University of Iowa
Iowa Research Online
Theses and Dissertations
2006
Structure and function of a mitochondrial PP2A
holoenzyme that regulates neuronal survival
Ruben Karim Dagda
University of Iowa
Copyright 2006 Ruben Karim Dagda
This dissertation is available at Iowa Research Online: http://ir.uiowa.edu/etd/84
Recommended Citation
Dagda, Ruben Karim. "Structure and function of a mitochondrial PP2A holoenzyme that regulates neuronal survival." PhD (Doctor of
Philosophy) thesis, University of Iowa, 2006.
http://ir.uiowa.edu/etd/84.
Follow this and additional works at: http://ir.uiowa.edu/etd
Part of the Pharmacology Commons
STRUCTURE AND FUNCTION OF A MITOCHONDRIAL PP2A HOLOENZYME
THAT REGULATES NEURONAL SURVIVAL
by
Ruben Karim Dagda
An Abstract
Of a thesis submitted in partial fulfillment
of the requirements for the Doctor of
Philosophy degree in Pharmacology
in the Graduate College of
The University of Iowa
May 2006
Thesis Supervisor: Assistant Professor Stefan Strack
1
ABSTRACT
Serine/threonine phosphatase 2A (PP2A) consists of an AC core dimer composed
of catalytic (C), structural (A) subunits complexed to a variable regulatory subunit
derived from three gene families (B, B’, B”). My dissertation work characterized the
structure and function of a neuron-specific splice variant of the Bβ regulatory gene
termed Bβ2. I found that the divergent N-terminus of Bβ2 does not affect phosphatase
activity or holoenzyme association but encodes a mitochondrial targeting signal.
Moreover, transient and stable expression of wild-type Bβ2 but not Bβ1, Bβ2 mutants
defective in mitochondrial targeting or a monomeric mutant unable to associate with the
holoenzyme, promotes apoptosis in neurons while knock-down of endogenous Bβ2 is
neuroprotective. Furthermore, I identified the mechanisms by which Bβ2 incorporates
the PP2A holoenzyme. By performing charge reversal mutagenesis in Bγ as a model for
B family regulatory subunits, I found that holoenzyme association requires multiple
electrostatic charges clustered in WD repeats 3 and 4 of the β-propeller. To identify
residues in Bβ2 important for mitochondrial association, I performed mutagenesis of the
divergent N-terminus of Bβ2 and identified basic and hydrophobic residues that are
critical for mitochondrial association. The variable N-terminal tail of Bβ2 is a cryptic
mitochondrial import sequence that promotes import of GFP, but not full-length Bβ2,
because its β-propeller domain resists the partial unfolding step necessary for
translocation. Lastly, I addressed the mechanism by which Bβ2 promotes apoptosis in
neurons. I found that overexpressing Bβ2 fragments mitochondria while RNAi of the
endogenous protein promotes mitochondrial fusion in neurons. Conversely, targeting
PKA, a well characterized prosurvival kinase, to the OMM by overexpressing A kinase
anchoring protein 121 (AKAP121) opposes the effects of the phosphatase by elongating
mitochondria. Furthermore, downregulating the endogenous AKAP121 by RNAi, or
inhibiting PKA at the OMM by overexpressing an inhibitor of PKA (OMM-PKI)
2
fragments mitochondria. The effects of OMM-targeted PP2A or PKA on survival require
remodeling of mitochondria, since blocking mitochondrial fission reversed the
proapoptotic effects of Bβ2 and OMM-PKI. My dissertation provides a novel mechanism
by which kinase/phosphatase signaling determines neuronal survival.
Abstract Approved: ____________________________________
Thesis Supervisor
____________________________________
Title and Department
____________________________________
Date
STRUCTURE AND FUNCTION OF A MITOCHONDRIAL PP2A HOLOENZYME
THAT REGULATES NEURONAL SURVIVAL
by
Ruben Karim Dagda
A thesis submitted in partial fulfillment
of the requirements for the Doctor of
Philosophy degree in Pharmacology
in the Graduate College of
The University of Iowa
May 2006
Thesis Supervisor: Assistant Professor Stefan Strack
Copyright by
Ruben Karim Dagda
2006
All Rights Reserved
Graduate College
The University of Iowa
Iowa City, Iowa
CERTIFICATE OF APPROVAL
_______________________
PH.D. THESIS
_______________
This is to certify that the Ph.D. thesis of
Ruben Karim Dagda
has been approved by the Examining Committee
for the thesis requirement for the Doctor of Philosophy
degree in Pharmacology at the May 2006 graduation.
Thesis Committee: ___________________________________
Stefan Strack, Thesis Supervisor
___________________________________
Mary Horne
___________________________________
Rory Fisher
___________________________________
Michael Knudson
___________________________________
Steven Green
To my dear wife Aidee for all her love and support throughout my graduate studies.
ii
“Get wisdom, get understanding; do not forget my words or deviate from them.
Do not forsake her, and she will protect you; love her and she will tend your needs. The
beginning of wisdom is the acquiring of wisdom and with all your acquiring, acquire
understanding. Extol her, and she will exalt you; embrace her and she will honor you. She
will set a garland of grace on your head and present you with a crown of splendor”
Proverbs 4:5-9
iii
ACKNOWLEDGMENTS
I will like to express my sincerest thanks to Dr. Stefan Strack for accepting me
into his laboratory, for his mentorship, counsel and support for four and a half years.
After graduating with an M.S. from the University of El Paso in Texas, I entered his
laboratory with little experience in molecular biology, tissue culture or neuroscience
techniques. Thanks to him, I can say today that I can perform high quality research in
molecular biology, think critically about experimental results, and have polished and
improved my writing skills. Overall, he has helped me to become a well-rounded scientist
throughout all my phases of my graduate career. I will also like to thank my graduate
committee members Dr. Mary Horne, Dr. Rory Fisher, Dr. Mike Knudson and Dr. Steven
Green for their valuable counsel, advice during graduate committee meetings and for
assisting me in my thesis writing. And of course, I will like to thank Dr. Gebhart for his
prompt assistance and guidance throughout my graduate career. The Department of
Pharmacology will miss his service, professionalism and leadership.
I will like to thank my parents Raul Dagda Medina and Agueda Saenzpardo for
their love, support, and counsel throughout my graduate career. Without their support, I
will not be where I am today. I will also like to thank my brother Raul Dagda for his
support and for staying with me in Iowa City for two grueling winters before going back
to Texas.
Also, I will never forget the vivid and wonderful moments that I had with my coworkers, colleagues and faculty of the Pharmacology department. Over the years, I have
witnessed the lab progressively grow from a few people to a competitive research
enterprise that attracts many graduate students and enjoys the mentorship provided by Dr.
Strack. I also had the privilege of enjoying the company of graduate students Amit Saraf,
Audrey Dickey, Mike Van Kanegan, and of post-doctoral associates Dr. Xinchang Zhou
and Dr. Ronald Merrill at work, at summer camping outings, and at get-togethers. I will
iv
like to specially thank Chris Barwacz, and Thomas Cribbs for their technical assistance in
subcloning constructs, and filling gaps in my research. I will like to thank Thomas
Monninger from the Central Microscopy Research Facility for instructing me on how to
use the confocal microscope and for his prompt assistance during technical difficulties.
The Phillibert lab for putting up with me in letting me use their microscope for many
hours. Also, I will like to thank other graduate students that assisted me throughout
different phases of my dissertation. One of these students is Vaibhavi Shah who helped
jump start my dissertation research during her lab rotation by subcloning GFP tagged
versions of all the Bβ isoforms. I will also like to thank Jim Denker for helping me
develop a dual-luciferase based assay for measuring neuronal survival. Jason Ulrich, a
productive graduate student with a lot of potential, for helping me refine the subjective
mitochondrial length assay and for unveiling the archrival of Bβ2, the survival kinase
PKA. This work was supported by NIH/NINDS (R01-NS043254) and NINDS/NRSA
Pre-doctoral fellowship (1 F31 NS49659-01).
v
ABSTRACT
Serine/threonine phosphatase 2A (PP2A) consists of an AC core dimer composed
of catalytic (C), structural (A) subunits complexed to a variable regulatory subunit
derived from three gene families (B, B’, B”). My dissertation work characterized the
structure and function of a neuron-specific splice variant of the Bβ regulatory gene
termed Bβ2. I found that the divergent N-terminus of Bβ2 does not affect phosphatase
activity or holoenzyme association but encodes a mitochondrial targeting signal.
Moreover, transient and stable expression of wild-type Bβ2 but not Bβ1, Bβ2 mutants
defective in mitochondrial targeting or a monomeric mutant unable to associate with the
holoenzyme, promotes apoptosis in neurons while knock-down of endogenous Bβ2 is
neuroprotective. Furthermore, I identified the mechanisms by which Bβ2 incorporates
the PP2A holoenzyme. By performing charge reversal mutagenesis in Bγ as a model for
B family regulatory subunits, I found that holoenzyme association requires multiple
electrostatic charges clustered in WD repeats 3 and 4 of the β-propeller. To identify
residues in Bβ2 important for mitochondrial association, I performed mutagenesis of the
divergent N-terminus of Bβ2 and identified basic and hydrophobic residues that are
critical for mitochondrial association. The variable N-terminal tail of Bβ2 is a cryptic
mitochondrial import sequence that promotes import of GFP, but not full-length Bβ2,
because its β-propeller domain resists the partial unfolding step necessary for
translocation. Lastly, I addressed the mechanism by which Bβ2 promotes apoptosis in
neurons. I found that overexpressing Bβ2 fragments mitochondria while RNAi of the
endogenous protein promotes mitochondrial fusion in neurons. Conversely, targeting
PKA, a well characterized prosurvival kinase, to the OMM by overexpressing A kinase
anchoring protein 121 (AKAP121) opposes the effects of the phosphatase by elongating
mitochondria. Furthermore, downregulating the endogenous AKAP121 by RNAi, or
inhibiting PKA at the OMM by overexpressing an inhibitor of PKA (OMM-PKI)
vi
fragments mitochondria. The effects of OMM-targeted PP2A or PKA on survival require
remodeling of mitochondria, since blocking mitochondrial fission reversed the
proapoptotic effects of Bβ2 and OMM-PKI. My dissertation provides a novel mechanism
by which kinase/phosphatase signaling determines neuronal survival.
vii
TABLE OF CONTENTS
LIST OF FIGURES .......................................................................................................... xii
LIST OF ABBREVIATIONS ...........................................................................................xv
CHAPTER I
INTRODUCTION.......................................................................................1
Classification of protein phosphatases..............................................................1
Classification of PSPs.......................................................................................2
Mechanisms of substrate recognition and catalytic regulation of PSPs ..........3
The PP2A class of PSPs ...................................................................................5
Protein phosphatases in brain function and disease..........................................7
Brain physiology .......................................................................................7
Brain disease..............................................................................................9
CHAPTER II A DEVELOPMENTALLY REGULATED, NEURON-SPECIFIC
SPLICE VARIANT OF THE VARIABLE SUBUNIT Bβ TARGETS
PROTEIN PHOSPHATASE 2A TO MITOCHONDRIA AND
MODULATES APOPTOSIS .........................................................................15
Abstract...........................................................................................................15
Introduction.....................................................................................................16
Types of cell death...................................................................................16
The intrinsic pathway of apoptosis..........................................................18
The role of PP2A in apoptosis.................................................................20
The B-family of regulatory subunits .......................................................21
PP2A inhibitors .......................................................................................22
Mitochondrial dysfunction in ischemic stroke and
neurodegenerative diseases .....................................................................23
Experimental procedure..................................................................................25
Generation of FLAG- and green fluorescent protein (GFP)-tagged
Bβ constructs and site-directed mutagenesis...........................................25
Generation of hairpins targeted against Bβ2 ...........................................26
Cell culture ..............................................................................................26
Immunocytochemistry of brain sections .................................................26
Generation of antibodies..........................................................................27
Generation of lentiviruses that overexpress Bβ1/2-GFP and Bβ2
targeted shRNAs......................................................................................28
Lentiviral infection of hippocampal neurons ..........................................29
Dual luciferase assays..............................................................................29
Immunoprecipitation and phosphatase activity assays............................29
Preparation of phosphatase substrates .....................................................30
Confocal imaging of GFP fusion proteins...............................................30
Mitochondrial translocation assay...........................................................31
Generation of tetracycline-inducible PC6-3 cell lines.............................31
Cell death assays......................................................................................32
Nuclear morphology assays in PC6-3 cells .............................................32
Annexin V staining..................................................................................33
Hippocampal survival assays ..................................................................33
Results.............................................................................................................34
Identification of a novel PP2A regulatory subunit ..................................34
viii
Characterization of Bβ2 expression ........................................................35
The Bβ2 N-terminus does not affect holoenzyme formation or
catalytic activity.......................................................................................37
The Bβ2 N-terminus encodes a mitochondrial localization signal .........38
Bβ2 promotes apoptosis ..........................................................................39
Bβ2 requires mitochondrial association and incorporation into the
PP2A heterotrimer to promote apoptosis ................................................43
Knock-down of the endogenous Bβ2 is neuroprotective ........................44
Discussion.......................................................................................................46
Complexity of Bβ gene expression .........................................................46
Structural implications.............................................................................47
PP2A in apoptosis....................................................................................49
CHAPTER III MECHANISMS OF PROTEIN PHOSPHATASE 2A
HOLOENZYME ASSEMBLY AND OUTER MITOCHONDRIAL
MEMBRANE TARGETING OF THE Bβ2 REGULATORY
SUBUNIT .......................................................................................................69
Abstract...........................................................................................................69
Introduction.....................................................................................................70
The β-propeller structure of WD repeat proteins ....................................70
Structure of the PP2A heterotrimer .........................................................71
Mechanisms of mitochondrial targeting ..................................................73
Mitochondrial import...............................................................................74
Experimental Procedure..................................................................................75
Structure modeling of Bγ.........................................................................75
Mutagenesis of Bγ ...................................................................................75
Mutagenesis and generation of Bβ2 fusion protein constructs ...............76
Cell culture ..............................................................................................77
Transfection of cDNAs............................................................................77
Antibodies................................................................................................78
Confocal microscopy...............................................................................78
Mitochondrial membrane potential assay................................................78
Apoptosis assays......................................................................................79
Immunoprecipitation of Bγ mutants ........................................................79
Proteolytic processing assay of Bβ2 mutants ..........................................80
Mitochondrial import assay.....................................................................80
Results.............................................................................................................81
Structure prediction of PP2A B-family regulatory subunits ...................81
Deletion mutagenesis of Bγ .....................................................................82
Charge-reversal mutagenesis...................................................................83
Identification of interacting residues .......................................................84
Determinants of mitochondrial localization of Bβ2 ................................85
Mitochondrial association determines proapoptotic activity...................87
The Bβ2 N terminus is a cryptic import signal .......................................87
Full-length Bβ2 is arrested at the OMM .................................................88
The β-propeller resists the unfolding step of import ...............................89
A prototypical import signal-β-propeller fusion protein
recapitulates OMM-targeting of Bβ2 ......................................................90
Translocase targeting is necessary, but not sufficient for apoptosis
induction by Bβ2 .....................................................................................91
Discussion.......................................................................................................92
Model of PP2A holoenzyme structure ....................................................92
Structural implications.............................................................................93
ix
Mitochondrial PP2A in apoptosis............................................................94
A novel OMM- targeting mechanism and its implications .....................95
CHAPTER IV REVERSIBLE PHOSPHORYLATION CONTROLS
MITOCHONDRIAL FISSION/FUSION AND NEURONAL
APOPTOSIS .................................................................................................112
Abstract.........................................................................................................112
Introduction...................................................................................................113
Mitochondrial fission/fusion .................................................................113
Reversible phosphorylation at the OMM in neuronal survival .............116
Mitochondrial PKA ...............................................................................116
Inhibitors of PKA ..................................................................................117
Role of OMM phosphatases in survival ................................................119
Experimental Procedures ..............................................................................120
Generation of plasmids..........................................................................120
Lentiviral infection of hippocampal neurons ........................................121
Apoptosis assays....................................................................................121
Subjective mitochondrial morphology assay.........................................122
Image analysis of mitochondrial morphology and mitochondrial
density....................................................................................................123
Statistical analysis .................................................................................123
Results...........................................................................................................124
PP2A/Bβ2 fragments mitochondria in neurons.....................................124
Catalytic inhibition of PP2A at the OMM fuses mitochondria .............126
PKA/AKAP121 opposes mitochondrial fragmentation ........................127
PKA/AKAP121 promotes survival .......................................................130
Mitochondrial restructuring is sufficient for neuronal survival
regulation...............................................................................................131
PP2A and PKA regulate mitochondrial morphology upstream of
apoptosis ................................................................................................133
Mitochondrial restructuring is required for the survival effects of
PP2A/PKA.............................................................................................133
Discussion.....................................................................................................134
Reversible phosphorylation alters mitochondrial morphology .............135
Reversible phosphorylation regulates survival through the MFF
machinery ..............................................................................................137
Substrates of PP2A/PKA among the MFF machinery ..........................140
Other potential physiological functions of PP2A/Bβ2 ..........................141
Possible mechanisms by which mitochondrial fragmentation may
promote apoptosis..................................................................................141
CHAPTER V CONCLUSION.......................................................................................161
The importance of subcellular targeting of PP2A in regulating neuronal
functions .......................................................................................................161
Role of PP2A/ Bβ2 in neurodegeneration ....................................................164
Bβ2 based therapies......................................................................................166
Implications of OMM PKA/PP2A mediated reversible phosphorylation
in modulating neuronal functions .................................................................167
Future directions ...........................................................................................169
REFERENCES ................................................................................................................173
x
APPENDIX- CURRICULUM VITAE............................................................................192
xi
LIST OF FIGURES
Figure 1
The phosphatome..............................................................................................13
Figure 2
PP2A heterotrimer model and possible holoenzyme combinations.. ...............14
Figure 3
Identification of a novel splice-variant of PP2A/Bβ.. ......................................54
Figure 4
The Bβ2 protein is expressed at lower levels than Bβ1.. .................................55
Figure 5
Immunocytochemistry of Bβ2 in the rat brain. ................................................56
Figure 6
In vitro characterization of PP2A holoenzymes containing Bβ isoforms. .......57
Figure 7 Mitochondrial targeting of Bβ2........................................................................58
Figure 8 Subcellular localization studies of Bβ11-32 and Bβ2 1-35-GFP by
confocal microscopy... .....................................................................................59
Figure 9
Cell stress induces mitochondrial translocation of Bβ2-GFP in PC6-3
cells. .................................................................................................................60
Figure 10 Cell stress induces mitochondrial translocation of Bβ2-GFP in
hippocampal neurons.. .....................................................................................61
Figure 11 Inducible expression of Bβ2 is proapoptotic....................................................62
Figure 12 Bcl-2 inhibits the cell death promoting activity of Bβ2. .................................63
Figure 13 Bβ2’s proapoptotic activity requires incorporation into the PP2A
holoenzyme in PC6-3 cells.. ............................................................................64
Figure 14 Bβ2 promotes apoptosis in hippocampal neurons and requires
incorporation into the PP2A heterotrimer and mitochondrial targeting.. ........65
Figure 15 RNA interference of Bβ2 is neuroprotective in primary hippocampal
neurons.............................................................................................................66
Figure 16 RNA interference of Bβ2 blocks apoptotic cell death of hippocampal
neurons.............................................................................................................67
Figure 17 Model figure that depicts the mechanism of apoptosis by Bβ2.. ....................68
Figure 18 Structure prediction of B-family regulatory subunits......................................97
Figure 19 Mapping the holoenzyme association domains of Bγ by deletion
mutagenesis. ....................................................................................................98
Figure 20 Identification of Bγ residues important for holoenzyme association. . ...........99
Figure 21 Summary of the effects of deletion and charge reversal mutagenesis
holoenzyme association.. ...............................................................................100
xii
Figure 22 Identification of interacting residues in Bγ and Aα.. ....................................101
Figure 23 Effects of N-terminal truncations on mitochondrial association of Bβ2.......102
Figure 24 Effects of N-terminal point mutations on mitochondrial association of
Bβ2.................................................................................................................103
Figure 25 The proapoptotic activity of Bβ2 requires mitochondrial association.. ........104
Figure 26 Proteolytic processing of Bβ2 N terminus fusion proteins. . ........................105
Figure 27 Submitochondrial localization of full-length Bβ2 and Bβ21-35-GFP. ...........106
Figure 28 The β-propeller of Bβ2 is a stop-transfer fold...............................................107
Figure 29 Unfolding-resistant translocase targeting of a model protein........................108
Figure 30 Translocase targeting is necessary but not sufficient for toxicity of
PP2A/Bβ2.. ....................................................................................................109
Figure 31 Model of the PP2A holoenzyme. ..................................................................110
Figure 32 Unfolding resistant translocase targeting of Bβ2. .........................................111
Figure 33 Overexpression of Bβ2 (but not Bβ1) fragments mitochondria in nondifferentiated and differentiated PC6-3 cells. ................................................144
Figure 34 PP2A/Bβ2 (w.t. but not mutants) fragments mitochondria in
hippocampal neurons. ....................................................................................145
Figure 35 Algorithmic quantification of mitochondrial morphology and
mitochondrial density.....................................................................................146
Figure 36 Inhibiting mitochondrial PP2A alters mitochondrial morphology in
PC6-3 cells.. ...................................................................................................147
Figure 37 Pharmacological activation of PKA elongates mitochondria in PC6-3
cells and in hippocampal neurons... ...............................................................148
Figure 38 Forskolin induces rapid mitochondrial elongation independent of
protein synthesis............................................................................................149
Figure 39 Overexpression of AKAP121 elongates mitochondria in PC6-3 cells..........150
Figure 40 Inhibiting PKA at the OMM fragments mitochondria in PC6-3 cells...........151
Figure 41 Overexpression of AKAP121 promotes fusion while RNAi fragments
mitochondria in hippocampal neurons...........................................................152
Figure 42 RNAi mediated knock-down of AKAP121 fragments mitochondria in
PC6-3 cells. ...................................................................................................153
Figure 43 Tools used for manipulating the MFF in hippocampal neurons.....................154
xiii
Figure 44 PKA at the OMM promotes survival while inhibition of PKA or RNAi
of AKAP121 promotes apoptosis in hippocampal neurons...........................155
Figure 45 The MFF machinery regulates mitochondrial remodeling in
hippocampal neurons. ....................................................................................156
Figure 46 Mitochondrial restructuring is sufficient for neuronal survival regulation. ...157
Figure 47 Mitochondrial restructuring and neuronal survival by PP2A/PKA is
upstream of cell survival and requires the MFF machinery.. ........................158
Figure 48 Mitochondrial restructuring is necessary but not sufficient for the
neuronal survival effects of AKAP121..........................................................159
Figure 49 Model summary: dynamic regulation of MFF and neuronal survival by
OMM-PP2A and PKA. ..................................................................................160
xiv
LIST OF ABBREVIATIONS
AKAP121 A kinase anchoring protein (of 121 kDa)
APP
Amyloid precursor protein
cDNA
Complementary deoxyribonucleic acid
CMV
Cytomegalovirus
CNS
Central nervous system
COX8
Cytochrome oxidase (8)
cAMP
Cyclic adenine monophosphate
DARP32
Dopamine and adenosine 3’5’-monophosphate regulated
phosphoprotein of 32 kDa
Dox
Doxycycline
DLP1
Dynamin like protein 1
DRP1
Dynamin related protein 1
DN-DRP1
GTPase deficient, dominant-negative Drp1
DIV
Days in vitro
FIS1
Fission protein 1
FIV
Feline immunodeficiency virus
GFP
Green fluorescent protein
GTP
Guanine triphosphate
HEPES
4-(2-Hydroxyethyl)piperazine-1ethanesulfonic acid
HS
Horse serum
HK1
Hexokinase 1
sHP
Single stranded hair pin
I2
Inhibitor 2
IP
Immunoprecipitate/immunoprecipitation
xv
HEAT
huntingtin, elongation factor, A subunit, and TOR kinase
IMM
Inner mitochondrial membrane
mAB
Monoclonal antibody
MFF
Mitochondrial fusion/fission
MRFP1
modified red fluorescent protein 1
MTS
3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4sulfophenyl)-2H- tetrazolium, inner salt)
NB/B27
Neurobasal B27 supplement media
NLS
Nuclear localization signal
OMM
Outer mitochondrial membrane
PfARP
Plasmodium falciparum aspartate rich protein
PC12
Pheochromocytoma clone 12
PC6-3
Pheochromocytoma clone 6-3
PKA
cAMP dependent protein kinase A
PKI
Inhibitor of protein kinase A
PP
Protein phosphatase
PP1
Protein phosphatase 1
PP2A
Protein phosphatase 2A
PP2B
Protein phosphatase 2B
PP2C
Protein phosphpatase 2C
PPP2R2B
Human Bβ gene
PTP
Protein tyrosine phosphatase
RNAi
RNA interference
ROS
Reactive oxygen species
SCA12
Spinocerebellar ataxia type 12
shRNA
short hairpin RNA
xvi
scFvs
single-chain Fv antibodies
TIM
Translocase of the inner membrane
TMRM
Tetramethylrhodamine methyl ester
TOM
Translocase of the outer membrane
VDAC
Voltage-dependent anion channel
WD
Tryptophan aspartate domain
WT
Wild-type
xvii
1
CHAPTER I
INTRODUCTION
Classification of protein phosphatases
Reversible protein phosphorylation is a key posttranslational regulatory
mechanism employed in many signal transduction cascades. Phosphorylation of serine,
threonine (Ser/Thr) or tyrosine (Tyr) residues is mediated by protein kinases while
dephosphorylation is carried out by protein phosphatases. While 516 kinases have been
identified in the human genome, only 143 protein phosphatase genes have been described
so far (~0.4% of all genes) and are derived from eight classes of protein phosphatases
(reviewed by [1-3]). Protein phosphatases are classified into three major groups based on
the presence of specific amino acids in their catalytic sites as aspartate-based, cysteinebased, and metal-based phosphatases. The metal-based protein phosphatases, also known
as protein Ser/Thr phosphatases (PSPs), specifically dephosphorylate phospho-serine and
threonine residues and are further classified into PPP and PPM families. The PPP family
include protein phosphatase 2A (PP2A), protein phosphatase-1 (PP1), PP2B
(calcineurin), PP4-PP7, while the PPM family consists of Mg2+-dependent PPases
including PP2C and pyruvate dehydrogenase phosphatase.
The Cys-based protein phosphatases are classified into four groups that include
the protein tyrosine phosphatases (PTPs), the dual-specificity phosphatases (DSPs), the
CDC25 group and the single member LMPTP group (reviewed by [2, 3], Fig. 1). All
groups of Cys-based protein phosphatases contain a well conserved (CX5R) motif. The
Asp-based protein phosphatases are characterized for containing two aspartates in their
catalytic motif (DXDXT). The Asp-based phosphatases dephosphorylate phosphoTyr/Ser residues and are related to the prokaryotic haloacid dehalogenase (HAD)
superfamily of hydrolases.
2
Classification of PSPs
Greater than 99% of cellular phosphoamino acids are phospho-serine and
threonines and are dephosphorylated by serine/threonine phosphatases (PSPs). PSPs
belong to the family of metallophosphoesterase enzymes that include DNA exonucleases
and polymerases. The catalytic domain of PSPs is structurally well conserved and is
composed of two α helices and β strands that bind metal ions Fe2+/3+, Zn2+, and Mn2+ at
well conserved amino acids in the loops between the secondary structures. PSPs employ a
sequential ordered mechanism where two metal ions bind the catalytic site followed by
the substrate prior to catalysis. The binding of metal ions increases the nucleophilic
character of water in the catalytic site and the electrophilicity of the phosphorus atom.
Catalysis then occurs by the direct attack of the water molecule on the phosphate ester
group of the substrate (reviewed by [3-5]).
PSPs are divided into two major families with little sequence homology: the PPP
and PPM. The PPP family consists of different holoenzyme groups based on their
catalytic specificity, metal cation requirements and sensitivity to inhibitors. The PPP
family includes PP2A, PP1, PP2B/calcineurin and a set of poorly characterized classes of
phosphatases termed PP4, PP5, PP6 and PP7. The PPM (PP2C) family consists of a
protein phosphatase that binds Mn2+ and the catalytic domain is structurally similar to the
PPP family. It is a monomeric enzyme and unlike the PPP family it requires binding to
Mg2+ for catalysis. PP2C has been shown to regulate cellular stress responses [6, 7]. PP1
is a very well characterized PSP. The catalytic subunit of PP1 arises from three genes
(α,β and γ) and forms dimers by interacting with a large array of regulatory subunits. The
catalytic subunits of PP1 may associate with regulatory subunits by binding to an RVXF
motif [8]. In the brain, regulatory subunits that interact with PP1 catalytic subunits
include the cytoskeletal proteins neurabin I, neurabin II, spinophilin to regulate
cytoskeletal dynamics in dendritic spines [9-12]. In neurons, PP1 regulates diverse
functions such as synaptic plasticity, dopamine signaling and cytoskeletal dynamics
3
(reviewed by [13-15]). PP2B, alternatively called calcineurin, is a Ca2+ /calmodulindependent phosphatase composed of two subunits termed CNA and the regulatory B
subunit denoted CNB. PP2B is a ubiquitous enzyme and is highly expressed in the brain.
The catalytic subunit of PP2B requires association with regulatory subunits and binding
to Ca2+ /calmodulin for catalytic activation.
Mechanisms of substrate recognition and catalytic
regulation of PSPs
It was once thought that PSPs are promiscuous enzymes since PP1 and PP2A
contribute up to 90% of the total Ser/Thr phosphatase activity in cells. However, PSPs are
not promiscuous enzymes but are regulated by binding to a large array of regulatory
subunits, activity modulators and chaperones. The binding of different regulatory
subunits to catalytic subunits impart enzyme specificity, subcellular targeting and
regulate catalytic activity of protein phosphatases. Over the years, investigators have
employed in vitro phosphatase assays to study the effects of regulatory subunits on
substrate specificity and activity of the catalytic subunit of PSPs. By performing in vitro
phosphatase assays, Price et al., 2000 demonstrated that the heterotrimeric form of PP2A
contains higher catalytic activity toward substrates compared to the AC core or the C
subunit alone [16]. Moreover, other in vitro phosphatase assays demonstrated that PP2A
holoenzymes that contain different regulatory subunits from different gene families (B
and B’) exhibit remarkable differences in activity towards substrates such as ERKs,
histone and casein [17, 18]. Overall, in vitro phosphatase studies have shown that PP2A
holoenzymes containing different regulatory subunits have distinct affinity and catalytic
activity of PP2A toward substrates.
The crystal structure of PP1 bound to myosin-targeting subunit of myosin
(MYPT1) has been solved and represents the only known PSP holoenzyme structure to
date. In its unbound form, the catalytic subunit of PP1 has low affinity toward myosin.
4
However, PP1 bound to MYPT1 increases the substrate affinity and catalytic activity
towards myosin which leads to smooth muscle relaxation [19]. MYPT1 binds PP1 at
three sites: 1) the N-terminal domain of MYPT1 forms an extended α-helix that precedes
the RVXF and interacts with a hydrophobic pocket of PP1, 2) the RVXF domain engages
six hydrophobic residues distal to the catalytic site of PP1 and 3) C-terminal ankyrin
repeats of MYPT1 interact with the C-terminal domain of PP1 involving residues Tyr
305 and Tyr 307. Overall, the final effect of the binding of MYPT1 on PP1 does not lead
to global changes in the tertiary structure of PP1 but expands the substrate binding
surface to include a long acidic groove on MYPT1 leading to an increased affinity and
specificity of the C subunit towards myosin. Thus, the PP1/MYPT1 structure
demonstrates that regulatory subunits confer substrate specificity of the catalytic subunits
and suggests the existence of a structural basis of substrate recognition in other PSPs
[20].
Mechanisms that regulate catalytic activity of PSPs include phosphorylation and
methylation of the C subunit. The catalytic activity of PP2A is regulated by methylation
of a conserved C-terminal Leu residue. Methylation of the C subunit at the C-terminal
Leu by methyltransferases stabilizes the PP2A holoenzyme by promoting the association
of regulatory subunits with the C subunit, which leads to increased activity of PP2A
towards substrates [21]. Conversely, phosphorylation of the C-terminus by kinase leads
to a remarkable decrease in PP2A activity. Guo et al., 1993 showed that
autophosphorylation-activated kinase in bovine extracts phosphorylated and inactivated
PP2A. This inhibitory activity of the kinase towards PP2A was increased in the presence
of microcystin suggesting that PP2A catalyzes an autodephosphorylation reaction
required for activation. It is not known whether phosphorylation causes conformational
changes in the C subunit or decreases the binding of the C subunit to regulatory subunits
[22].
5
The PP2A class of PSPs
The predominant form of PP2A is a heterotrimer composed of two subunits (A
and C) that may associate with a variable regulatory subunit. PP2A in conjunction with
PP1 dephosphorylate the majority of phospho-Ser/Thr in eukaryotes. PP2A is an essential
enzyme since pharmacological inhibition, RNA interference, or gene knockout of PP2A
is detrimental in eukaryotes [23-27]. The PP2A core dimer consists of a hook-shaped
scaffolding 65 kDa (A) subunit that binds to a well conserved 36 kDa catalytic (C)
subunit. The α4 protein, yeast homologue of Tap42 protein, has also been shown to
associate the C subunit to downregulate the activity of PP2A [28, 29]. The C and A
subunits are each encoded by two highly similar genes (Aα/β, Cα/β). The A and C
subunit isoforms contain a high degree of sequence conservation and exhibit subtle
differences in their tissue expression profiles. Although the α isoforms are significantly
more abundant than the β isoforms, the functional differences for each isoform have not
been established [30, 31]. Both Aα and Aβ are mutated in some forms of cancer
suggesting a role as tumor suppressors [32, 33]. The Aβ binds more weakly to the C and
regulatory subunits of PP2A compared to Aα [30]. Moreover, it been shown that proper
protein levels of Aα but not Aβ is critical for mammalian survival since Aβ was unable
to delay Aα RNAi induced cell death in PC12 cells [27]. The core AC dimer of PP2A
may associate to a third variable regulatory subunits that have been classified into three
multi-gene families termed PR55/B, PR61/B’, PR72/B” (Fig. 2). The association of the
core dimer with different regulatory subunits can form more than 48 distinct holoenzyme
complexes which adds to the level of complexity and fidelity of PP2A in regulating
signal transduction pathways (reviewed by [14, 34, 35]. Other AC core interacting
proteins include the SV40 small t antigen and the polyoma virus small and middle T
antigens. These proteins displace the regulatory subunits and subvert the tumor
suppressing activity of PP2A [36-38]. The PP2A core dimer has also been shown to
interact with cytoskeletal proteins such as WD repeat-containing proteins striatin, and
6
S/G(2) nuclear autoantigen (SGNA). However, it is currently debated whether SGNA and
striatin deserved to be classified as a proper subunit family of PP2A since a direct
interaction of these proteins with the AC dimer has not been demonstrated nor it is
known whether these proteins compete with PP2A regulatory subunits [39, 40].
The different families of regulatory subunits contain no obvious sequence or
predicted structural similarities but contain a loose consensus motif that associate to the
A subunit of PP2A [41]. There are four genes (Bα, Bβ, Bγ, Bδ) in the B-family of
regulatory subunits that give rise to proteins with a molecular mass of between 54-57 kDa
[42-46]. The regulatory subunits of the B family have a high degree of sequence
conservation at their C-terminus and contain N-terminal divergent tails that act as
subcellular targeting signals. An important feature of B family regulatory subunits is that
they are tryptophan-aspartate (WD) repeat containing proteins which are predicted to fold
into a β-propeller structure similar to the β subunit of G protein heterotrimers [47]. The
Bβ gene is the only gene member of the B-family of regulatory subunits that gives rise to
multiple splice variants. Although the B family of regulatory subunits was the first family
to be characterized (reviewed by [35]) little is known about their functions in the cell to
this date. Specific functions that have been ascribed to the B-family of regulatory
subunits include regulating cytoskeletal assembly, regulating the mitogen activated
protein (MAP) kinase pathway in neurons at the level of B-raf and by dephosphorylating
ERK, and regulating cell survival [17, 18, 48-51].
The B’ family of PP2A subunits are derived from five genes (B’α, B’β, B’γ, B’δ
and B’ε) and range in molecular masses from 54 kDa to 72 kDa [52-59]. B’α, B’β and
B’ε are predominantly localized in the cytosol while B’γ and B’δ are strictly nuclear
proteins [53]. The B’ family of regulatory subunits are phosphoproteins that have been
implicated in regulating the Wnt/β-catenin pathway, neuronal patterning during
embryonic development, synaptic plasticity functions such as LTP, survival, cellular
7
transformation, and bind to cyclin G2 to modulate the G/S transition of the cell cycle [26,
58-63].
The B” family of regulatory subunits comprises of three members that are
encoded by three genes PR48, PR59, and PR72/130. The PR72/130 subunits arise by
transcription initiation from the same gene. The B” regulatory subunits are nuclear
proteins that bind calcium ions via two EF hand structures [64]. Specific functions that
have been ascribed to the B” of regulatory subunits include regulating the G1-S cell cycle
transition [64-66]. The PR48 subunit was shown by yeast two hybrid screens to interact
with cdc6, a component of the DNA prereplication machinery, and may play a role in
regulating DNA replication [66].
Protein phosphatases in brain function and disease
Brain physiology
Neurons contain a vast array of signaling complexes localized at specialized
compartments such as axons and dendrites to regulate critical neuronal functions. These
signaling complexes are regulated by protein phosphorylation at serine/threonine residues
by protein kinases and are rapidly opposed by protein phosphatases. As previously
discussed above, protein phosphatases were initially though to be promiscuous enzymes,
and there is ever increasing evidence that demonstrates that substrate specificity of
protein phosphatases is regulated by regulatory subunits. While the PP1/MYPT1 crystal
structure suggests a structural basis of substrate recognition for PSPs, there is an
increasing amount of evidence that suggests that regulatory subunits regulate substrate
specificity by targeting PP2A to different subcellular compartments [17, 53, 67]. Thus,
the targeting of PSPs to specific subcellular sites enhances the fidelity and efficiency of
substrate dephosphorylation by C subunits to regulate signal transduction pathways (See
review by [68]).
8
In non-neuronal cells, PP1 regulates diverse functions such as glycogen
metabolism, survival, splicing of pre-mRNA, cell cycle progression and nuclear
assembly. In neurons, PP1 regulates a variety of critical functions that include synaptic
neurite outgrowth, synapse and dendritic morphology, glutaminergic synaptic
transmission, and synaptic plasticity (reviewed by [69]). By binding to spinophilin and
neurabin, PP1 dephosphorylates post-synaptic density (PSD) associated calcium
calmodulin kinase II (CaMKII) but not soluble CaMKII (dephosphorylated by PP2A), a
kinase involved in the regulation of neurotransmitter release, gene expression, and
synaptic plasticity. Furthermore, it was shown that peptides that disrupt PP1 localization
to the synapse enhanced and sustained long-term potentiation (reviewed by [70]). The
best characterized physiological function of PP1 in neurons is its involvement in
regulating dopamine signal transduction pathways in neostriatal neurons. Association of
dopamine to dopamine D1 receptors activates cAMP dependent protein kinase A (PKA)
which phosphorylates the PP1 inhibitor DARPP32. Phosphorylated DARPP32 then
associates with and inhibits PP1 which consequently leads to enhanced dopamine
signaling in neurons. Inhibition of PP1 by DARPP32 also alters other neuronal signaling
pathways such as NMDA receptor responses, dopamine modulation of AMPA-type
glutamate receptors and N- and P-type calcium channel responses [9, 13, 71, 72].
The regulation of neuronal functions by PP2A is mediated by the interaction of
the AC core dimer with regulatory subunits. The regulatory subunits of PP2A that are
exclusively expressed in brain are Bγ, a cytoskeletal associated regulatory subunit, and
the Bβ splice variants. PP2A holoenzymes that contain these neuronal regulatory subunits
have been shown to regulate cytoskeletal dynamics, survival, and promote neuronal
differentiation [17, 51, 67]. Unlike Bγ, and Bβ which are only expressed in brain, the
ubiquitous regulatory subunits Bα and Bδ dephosphorylate and inactivate ERKs in
neurons while B’ subunits have been shown to regulate survival by modulating Akt [18].
Furthermore, PP2A binds to neurofilaments where it dephosphorylates neurofilaments
9
NF-M and NF-L to regulate the stability of neurofilaments [73, 74]. PP2A has also been
shown to dephosphorylate microtubule associated proteins (MAP) such as MAP-2 and
tau at multiple serine residues, and dephosphorylation of MAPs by PP2A results in
increased assembly and stability of microtubules [49, 75-77].
Brain disease
Hyperphosphorylated tau, an axonal MAP, is the principal component of
neurofibrillary tangles, a pathological hallmark of Alzheimer’s disease. A PP2A
holoenzyme that contains the Bα regulatory subunit plays a role in the pathophysiology
of Alzheimer’s disease. PP2A/Bα has been shown to be the major phosphatase of tau
protein. Sontag et al., 2004 demonstrated that patients afflicted with Alzheimer’s disease
contained decreased expression levels of Bα and C subunits of PP2A in brain tissue
extracts suggesting that tau hyperphosphorylation may be consequence of decreased
PP2A activity. The carboxyl methylation of the C subunit of PP2A at a conserved Cterminal leucine mediated by methyl transferases is required for stability of the
holoenzyme [78, 79]. The instability of PP2A/Bα was attributed to a decrease in
expression levels of methyltransferases in Alzheimer’s disease patients and correlated
with the development of neurofibrillary tangles. [80-82].
PP2A is a regulator of tyrosine hydroxylase (TH), an enzyme involved in the
synthesis of dopamine. Phosphorylation of TH at serine residues (ie, Ser 19 and Ser40)
by PKA and ERK leads to increased synthesis of dopamine while dephosphorylation by
PP2A has the opposite effect (Review by [83]). Moreover, pharmacological inhibition of
PP2A by okadaic acid (OA) has been shown to lead to increased phosphorylation of TH
suggesting that PP2A dephosphorylates TH [84] Targeting the PP2A holoenzyme that
dephosphorylates TH can have therapeutic implications for treating Parkinson’s disease.
Parkinson’s disease is a common and gradual neurodegenerative disorder that leads to the
loss of nigrostriatal neurons leading to a substantial decline in the synthesis of dopamine
10
by tyrosine hydroxylase. Decreased dopamine synthesis in Parkinson’s patients leads to a
gradual decline in locomotor activity and coordination [85]. Thus, inhibiting or
knocking-down the yet unidentified PP2A holoenzyme responsible for mediating TH
dephosphorylation can be an alternative venue for treating PD.
PP2A has also been implicated in Angelman’s syndrome, a developmental
cognitive disorder characterized by epilepsy and severe mental retardation. Angelman’s
syndrome is the product of imprinting center mutations of the Ube3a gene that encodes
for an E6-AP ubiquitin ligase. A mouse model of Angelman’s disease containing a
mutation in the Ube3a gene showed decreases in NMDA receptor-dependent and
independent LTP. In addition, the mutant mice exhibited decreased CAMKII activity,
abnormal CAMKII autophosphorylation and impaired targeting of CAMKII to the
postsynaptic sites. Interestingly, autophosphorylation of CAMKII did not occur at the
autonomous site (Thr 286) but at Thr 305/306 of the calcium-calmodulin binding region
which renders the kinase inactive. The increased autophosphorylation of CAMKII at Thr
305/306 was correlated with a decreased in PP2A activity in whole hippocampal
homogenates in Ub3a mutant mice [86].
The Bβ regulatory gene of PP2A has been implicated in an autosomal dominant
neurodegenerative disease termed spinocerebellar ataxia type 12 (SCA12) which is
caused by a CAG repeat expansion located upstream of the presumed promoter region of
the human PPP2R2B gene. SCA12 was first described in a large pedigree of German and
Indian descent [87, 88]. Patients afflicted with these lethal disease exhibit lethargic
responses, bradykinesia, ataxic gait, cognitive decline and tremors [89]. Alternative
splicing of the Bβ gene gives rise to more than four isoforms of the regulatory subunit
that only differ in their 5’ UTR region and N-terminal tails. Since patients affected with
SCA12 develop widespread neuronal degeneration, proper expression of the Bβ gene
must be critical for neuronal survival. It is conceivable that the SCA12 mutation alters the
11
transcriptional regulation and splicing events of the Bβ gene may lead to altered
expression levels of each Bβ isoform.
My dissertation work characterized the structure and function of a neuron-specific
splice variant of the Bβ gene termed Bβ2. Bβ2 is widely expressed in many brain areas
and its expression is induced during postnatal development of the rat brain. By
employing confocal microscopy, immunofluorescence techniques and survival assays, the
subcellular localization and function of Bβ2 in PC6-3 cells and in rat dissociated
hippocampal neurons was elucidated. I determined that the divergent N-terminal tail of
Bβ2 is a subcellular targeting signal that targets PP2A to the mitochondria to promote
apoptosis. The proapoptotic activity of Bβ2 depends on a functional mitochondrial signal
and requires holoenzyme association since mutants of Bβ2 that are defective for binding
the A and C subunits of PP2A or for mitochondrial targeting failed to promote apoptosis
in neurons. In the same manner, RNA interference (RNAi) of endogenous Bβ2 promotes
neuronal survival against glutamate excitotoxicity and reactive oxygen species (ROS)
induced apoptosis. The third chapter of my thesis describes the mechanism by which the
B family of regulatory subunits associate the A and C subunits of PP2A. By employing
site directed and deletion mutagenesis of Bγ and immunoprecipitation/immunoblotting
techniques, I identified regions and residues in Bγ that are critical for holoenzyme
association. These studies provide a model by which the B-family regulatory subunits
associate with the A subunit and C subunits of PP2A. The PP2A model suggests that the
N-terminal tail is dispensable for holoenzyme association and point away from the
presumed β-propeller while multiple charge residues located in WD repeats 3 and 4
interact with charged residues in the A subunit. The second half of chapter 3 addresses
the mechanism of mitochondrial association of Bβ2. By using confocal microscopy
analysis of live neurons, I discovered that the N-terminus of Bβ2 fused to green
fluorescent protein (GFP)is as a cleavable mitochondrial import signal that interacts with
import receptors at the outer mitochondrial membrane (OMM) via basic and hydrophobic
12
residues. However, trypsin sensitivity assays of isolated mitochondrial fractions in PC6-3
cells determined that full-length Bβ2 is not imported to the matrix of mitochondria but it
is retained at the OMM by a novel mechanism termed unfolding resistant translocase
targeting which may have implications for other proteins. This mechanism entails a
mitochondrial import sequence that targets proteins to mitochondria but are retained at
the OMM by virtue of a rigid structure such as a β-propeller fold that resists the
unfolding step necessary for import.
In the fourth chapter of my thesis, I describe a mechanism by which PP2A/Bβ2
promotes apoptosis in neurons. By employing live confocal microscopy and survival
assays in PC6-3 cells and in hippocampal neurons, I discovered that Bβ2 promotes
apoptosis by fragmenting mitochondria. Conversely, catalytic inhibition of PP2A at the
OMM or RNAi of endogenous Bβ2 promotes mitochondrial elongation and prevents
apoptosis. In order, to identify the kinase that opposes PP2A/Bβ2, neurons were treated
with a variety of pharmacological activators of PKA and analyzed for changes in
mitochondrial morphology. I found that pharmacological activators of PKA, a well
characterized survival kinase, caused rapid fusion of mitochondria and this phenomenon
do not require protein synthesis. Moreover, I found that recruiting PKA to the OMM by
transiently transfecting A kinase anchoring protein 121 (AKAP121) causes mitochondrial
fusion and promotes survival in hippocampal neurons. Conversely, inhibiting PKA by
targeting the inhibitor of protein kinase A (PKI) to the OMM shortens mitochondria and
promotes apoptosis in neurons. Both OMM-PKA and PP2A/Bβ2 were found to regulate
survival via the mitochondrial fission/fusion (MFF) machinery since overexpression of
fusion protein modulators or RNAi of fission protein modulators inhibits mitochondrial
fragmentation and apoptosis by PP2A/Bβ2 and OMM-PKI. These findings suggest that
neuronal survival regulated by OMM reversible phosphorylation depends on remodeling
of mitochondrial morphology.
13
Figure 1 The phosphatome. There are approximately 143 protein phosphatases that
have been identified and catalogued into different groups based on sequence
similarity, sensitivity to catalytic inhibitors, catalytic mechanisms and divalent
metal ion requirements. The number of members and substrate residues for
each phosphatase group are indicated. Substrate residues are color coded
according to catalytic mechanisms. This figure was adapted with permission
from Merril et al., 2006 [3].
14
Figure 2 PP2A heterotrimer model and possible holoenzyme combinations. The
PP2A holoenzyme is composed of a hook-shaped scaffolding A subunit, a
catalytic subunit C, and a variable regulatory subunit. The AC core may
interact with three different families of variable regulatory subunits: B
(PR55), B’ (B56 or PR61), and B’’ (PR48, PR59, PR72/130). The X-ray
crystal structure of the A subunit [29] as well as model structures of C (based
on PP1/C) and B-family subunits based on Gβ [47] are depicted.
15
CHAPTER II
A DEVELOPMENTALLY REGULATED, NEURON-SPECIFIC SPLICE
VARIANT OF THE VARIABLE SUBUNIT Bβ TARGETS PROTEIN
PHOSPHATASE 2A TO MITOCHONDRIA AND MODULATES APOPTOSIS
Abstract
Heterotrimeric protein phosphatase 2A (PP2A) is a major Ser/Thr phosphatase
composed of catalytic, structural, and regulatory subunits. The Bβ regulatory subunit
gene is mutated in spinocerebellar ataxia type 12 (SCA12), and gives rise to several
neuron specific splice variants. Here we characterized Bβ2, a splice variant with a unique
N-terminal tail that is induced postnatally in the rat brain. At the mRNA and protein
levels, Bβ2 is widely expressed in most brain areas, and PP2A holoenzymes containing
Bβ2 are about 10-fold less abundant than those containing the Bβ1 (previously Bβ)
isoform. The divergent N-terminus of Bβ2 does not affect phosphatase activity, but
encodes a subcellular targeting signal. Bβ2, but not Bβ1, Bβ3, Bβ4 or an N-terminal
truncation mutant, is localized to mitochondria in PC12 cells and in hippocampal
neurons. Moreover, the Bβ2 N-terminal tail is sufficient to target green fluorescent
protein to this organelle.
Bβ2 is a proapoptotic regulatory subunit of PP2A, since inducible or transient
expression of Bβ2, but not Bβ1, Bγ, or Bβ2 mutants defective in holoenzyme formation
or mitochondrial targeting, accelerates apoptosis in PC12 cells and in hippocampal
neurons. Conversely, RNA interference (RNAi) mediated knock-down of endogenous
Bβ2 protects hippocampal neurons against glutamate excitotoxicity or rotenone toxicity.
Bβ2 regulates survival upstream of cytochrome c release or outer mitochondrial
membrane permeabilization since Bcl-2, an anti-apoptotic mitochondrial protein, blocks
apoptosis induction by Bβ2 in growth factor deprived PC12 cells. Thus, alternative
16
splicing of a mitochondrial localization signal generates a PP2A holoenzyme involved in
neuronal survival signaling.
Introduction
Types of cell death
Cell death is a complex process that is caused by a variety of factors such as
exposure to toxins, radiation, and injury or by disease. Different modes of cell death
have been described which include apoptosis, necrosis (oncosis), and autophagy.
Necrosis (oncosis) or accidental cell death is a passive and pathological process
that was first used by medical pathologists to describe morphological changes that occur
downstream of cell death. Since necrosis is the morphologic result of cell death, the
word oncosis is more widely accepted to describe a process that leads to cell death.
Oncosis was originally coined by von-Reckling hausen in 1910 to describe cell death by
swelling. Oncosis usually affects sheets of cells as seen during ischemic cell death, and is
an energy independent and unregulated process that generates cytoplasmic swelling,
swelling of organelles like the mitochondria and the endoplasmic reticulum, lost of
mitochondrial transmembrane potential, lost of membrane integrity and a rise in
intracellular calcium (reviewed by [90]). Ischemia cell death is one form of oncosis
which involves the failure of ionic pumps and leads to a rise of intracellular calcium. The
release of calcium stores during oncosis promotes cell death by activating downstream
cysteine proteases or cytosolic phospholipase A2’s which degrade organelles and
permeabilize the cell membrane respectively (reviewed by [91] [92, 93]). Unlike
apoptosis, oncosis does not lead to degradation of chromosomal DNA but chromatin
condensation is seen [94].
Autophagy, or also known as type II cell death is a highly regulated, evolutionary
conserved physiological mechanism for the degradation of cellular components such as
long lived proteins and cellular debris in autophagic vacuoles (reviewed by [94]).
17
Autophagy is essential for growth regulation, maintenance of cellular homeostasis and for
embryonic development [95, 96]. Autophagy is activated by intracellular and
extracellular stimuli that include nutrient depravation, bacterial infection, protein
aggregation, hormones, and by cellular debris. Cytoplasmic debris in dying cells is
sequestered by autophagosomes and this process is regulated by kinases, GTPases, lipids
and by ubiquitin-like conjugation systems [97, 98]. Autophagosomes then fuse with
lysosomes where contents are degraded [97]. Autophagy also plays a role in cell death
but the role of this process is not clear. Deregulation of this process contributes to the
pathophysiology of many diseases such as cancer, bacterial infection, and
neurodegenerative disorders such as Parkinson’s and Alzheimer’s disease [99]. For
instance, it has been demonstrated that NGF withdrawal of sympathetic neurons leads to
neurodegeneration by autophagy in a caspase independent manner [100].
Apoptosis, or also known as programmed cell death, is essential for pruning
excess cells during embryonic development, for removing virally infected cells, cancer
cells or damaged cells, and is a hallmark of many neurodegenerative diseases.
Programmed cell death is morphologically characterized by the formation of membrane
blebs, chromatin condensation, and cell shrinkage. Apoptosis can be divided in two main
categories: the extrinsic and intrinsic pathway of apoptosis (reviewed by [93, 101, 102]).
The extrinsic pathway of apoptosis usually involves a death ligand such as tumor
necrosis factor (TNF) that binds to a death receptor (P75), or a lipid such as ceramide that
elicits the release of intracellular calcium from calcium stores. Abnormal elevations of
intracellular calcium generate the formation of free radicals and leads to lipid
peroxidation of cell membranes. Elevated intracellular calcium leads to activation of
downstream “executioner” caspases or cysteine proteases such as calpain which then
cleave proteins such as Bcl-XL, the A subunit of PP2A and DNAases. The cleavage of
DNAases and endonucleases by caspases leads to their activation which ultimately
degrade chromosomal DNA, a hallmark of apoptosis (reviewed by [93, 101, 102]).
18
The intrinsic pathway of apoptosis
Mitochondria are not only “power houses” for generating ATP, but are also the
principle generators of reactive oxygen species (ROS), act as sinks for buffering calcium
and are the sites where many signal transduction pathways of apoptosis converge. The
intrinsic pathways of apoptosis are cell death signals that are triggered within the cell due
to an imbalance in cellular homeostasis and these pathways usually converged at the
mitochondria. There are a series of events that take place at the mitochondria before the
cell irreversibly commits to cell death. During toxic insult, pro-apoptotic Bcl-2 family
proteins are activated in the cytosol by posttranslational mechanisms and translocate to
the OMM to promote homodimerization of Bax, and Bak. Homodimerized Bax and Bak
insert into the mitochondrial lipid bilayer to form pore channels at the OMM, causing the
release of apoptogenic molecules such as cytochrome c and possibly SMAC/DIABLO.
Released cytochrome c then binds to the apoptosome activating factor-1 (Apaf-1) which
then interact and activate downstream “death executioners” such as caspase 9. Caspase 9
then activates downstream caspases 3 and 7 by proteolytic cleavage of their respective
precursor procaspases which cleave and activate DNAases which are the enzymes that
are ultimately responsible for degrading chromosomal DNA, a hallmark of apoptosis
(reviewed by [102, 103]).
Apoptosis is regulated by both anti-apoptotic and pro-apoptotic Bcl-2 family of
proteins. Members of the Bcl-2 family are classified according to the presence of Bcl-2
homology domains (BH1-4). Bcl-2 and Bcl-XL are anti-apoptotic proteins that contain all
four homology domains. In neurons, Bcl-2 and Bcl-XL are important for neuronal
development and for the maintenance of long lived post-mitotic neurons. The
proapoptotic Bcl-2 family members are further divided into two groups: the bax family
and the BH3 only containing family of proapoptotic proteins. The bax family of
proapoptotic proteins are BH1, BH2, and BH3 containing proteins that include Bax, Bak,
Bok while the BH3 only containing family of proapoptotic proteins include Bid, Bim,
19
Bad, Noxa, and PUMA. Of particular interest to neuroscientists are the proapoptotic
proteins Bim, Bax, Bak and Bid since they have been implicated in regulating neuronal
cell death. For instance, a splice variant of Bak, N-Bak, has been shown to promote
apoptosis of hippocampal, cortical and cerebellar granule cells by binding to and
inactivating Bcl-XL and by stimulating the pore forming activity of Bax at the
mitochondria [104]. Bim is a proapoptotic protein has been implicated in promoting
apoptosis of growth factor deprived sympathetic neurons. During growth factor
deprivation, Bim is phosphorylated and activated by c-Jun N-terminal kinases (JNK) at
Ser 65 and stimulates the pore forming activity of Bax at the mitochondria leading to the
release of cytochrome c [105]. Bid is a proapoptotic protein with intrinsic Bax-like pore
forming activity that plays a role in the death of neurons after ischemia. During ischemic
insult, Bid is cleaved by caspase 8 and the truncated form of Bid (t-Bid) translocates to
mitochondria to oligomerize with Bak and promote the release of cytochrome c by
forming pores at the surface of mitochondria [106-108].
Some pro- and anti-apoptotic Bcl-2 family proteins are postranslationally
regulated by reversible phosphorylation. While phosphorylation of proapoptotic Bcl-2
family members by multiple kinases usually promotes cell survival, PSPs that
dephosphorylate these proteins induce apoptosis. Several kinases, including protein
kinase B (PKB), protein kinase A (PKA), and ERK kinases phosphorylate and inactivate
Bad, a proapoptotic BH3 containing family member, at Ser -112, 136, 155 and 170 and
Bim at Ser -109 and Thre -110 while PSPs activate their proapoptotic activity by
dephosphorylation. Phosphorylation of Bad by these kinases leads to its association with
14-3-3 proteins which sequester the proapoptotic protein and prevent its translocation to
mitochondria. [109-113]. The proapoptotic activity of Bim is regulated by expression
levels in neurons by FOXO transcription factors. In sympathetic neurons, NGF dependent
phosphorylation of Bim at Ser-109 leads to a decrease in expression levels of the
20
proapoptotic protein while NGF withdrawal increases the expression levels and apoptotic
activity of Bim [114, 115].
The role of PP2A in apoptosis
PP2A has been shown to be both a negative and positive regulator of apoptosis.
These opposing roles are likely carried out by different PP2A holoenzymes. For instance,
treating cells with the PP2A inhibitor okadaic acid induces apoptosis in many cell types
suggesting that PP2A is essential for survival. In agreement with this view, silencing the
B’ subunits of PP2A, the scaffolding or the catalytic subunit of PP2A in Drosophila S2
cells by RNA interference causes a decrease in cell viability suggesting that PP2A
holoenzyme containing B’ subunits mediate survival [25, 26]. Moreover, the level at
which B’ subunits in Drosophila promote survival was mapped by Li et al., 2002.
Epistasis experiments demonstrated that apoptosis induced by RNAi mediated silencing
of the B’ subunits of PP2A requires the activation of caspases and places the survival
effects of PP2A/B’ holoenzymes upstream of p53 [62]. Recently, RNAi studies carried
out in mammalian cells by Strack et al., 2004 demonstrated that intact PP2A
heterotrimers containing the A, C subunits and all families of regulatory subunits are
critical for survival. In this study, he showed that RNAi mediated knock-down of the
scaffolding Aα subunit lead to a decrease in protein levels of C and B, B’ but not B”
regulatory subunits and a decline in cellular PP2A activity with detrimental consequences
in PC6-3 cells [27].
On the other hand, PP2A can promote apoptosis by dephosphorylating and
inactivating antiapoptotic proteins or by activating proapoptotic Bcl-2 family proteins.
For instance, phosphorylation of Bcl-2 at serine 70 by protein kinase C promotes survival
in interleukin-3 (IL-3) treated lymphoid cells. Conversely, dephosphorylation of Bcl-2 at
Ser70 is mediated by a mitochondrial isoform of PP2A containing the B’α regulatory
subunit which leads to an increased sensitivity of cells to ceramide induced apoptosis
21
[116, 117]. Serine phosphorylation of Bad (Ser 112, -136, -155, and -170) by survival
kinases such as PKA, Akt/PKB and ERK prevents its cell death promoting activity by
facilitating the sequestration of Bad in the cytoplasm by associating with 14-3-3 proteins
[109, 118]. Conversely, withdrawal of interleukin IL-3 in lymphoid cells promotes rapid
dephosphorylation of Bad by PP2A resulting in its dissociation from 14-3-3 proteins.
Liberated Bad then translocates to mitochondria to promote apoptosis [112, 119].
The B-family of regulatory subunits
Without taking into account the splice variants, the B-family of regulatory
subunits is comprised of four members (Bα-Bδ). The B-family of regulatory subunits
have a high degree of amino acid homology (~85%) and contain a C-terminal core region
predicted to fold into a β-propeller structure and divergent N-terminal tails that dictate
isoform specific functions (Fig. 3B). Supporting this view is the observation that Bfamily of regulatory subunits exhibit differences in spatial and temporal expression
particularly in the brain. For instance, Strack et al., 1998 demonstrated that the expression
levels of Bα and Bδ are constant while the expression levels of Bβ, and Bγ are
dramatically induced throughout postnatal development of the rat brain [67]. Moreover,
subcellular fractionation studies performed in brain homogenates showed that Bβ1
copurified with all subcellular fractions, Bα copurified with cytosolic and membrane
fractions while Bγ was enriched in the cytoskeletal fraction [67]. Determining the
targeting mechanisms of each member of the B-family of regulatory subunits will
increase our understanding as to how individual PP2A holoenzymes regulate specific
neuronal signal transduction pathways.
There has been a high interest in studying the function of the isoforms of the Bβ
gene (PPP2R2B in humans) of PP2A since a CAG repeat expansion in the presumed
promoter region of the PP2A regulatory subunit gene was identified in patients afflicted
with SCA12 (See Chapter 1). The observation that the SCA12 mutation is detrimental for
22
neurons suggest that proper regulation of the Bβ gene is critical for neuronal survival.
The Bβ gene gives rise to at least four alternatively spliced variants which include Bβ1Bβ4. Bβ1 and Bβ2 is expressed in lower and higher vertebrates, while Bβ3 and Bβ4
appears to be exclusively expressed in primates as judged by representation in EST
databases [120].
PP2A inhibitors
There is a vast repertoire of inhibitors of PP2A that have been identified to date
[121, 122]. These inhibitors are widely exploited as pharmacological tools to unveil novel
functional roles of PP2A and include the fatty acid derivatives okadaic acid (OA) and
calyculin A (CyA), the peptide toxins microcystin and nodularin, and the antitumor agent
tautomycin, a PP1/PP2A inhibitor [121]. OA and microcystin inhibit PSPs by binding to
two regions of the phosphatase, specifically at the hydrophobic groove adjacent to the
active site and at the acidic groove of the active site [123, 124]. More recently there has
been a high interest in identifying endogenous inhibitor proteins of PP2A. Endogenous
PP2A inhibitors have been grouped into different families which include two heat stable
inhibitors termed I1PP2A and I2PP2A , the SET proteins, template-activating factor (TAF),
putative human HLA class II associated protein (PHAP), and the nucleosome assembly
protein (NAP) [125-127]. An endogenous inhibitor of PP2A was isolated from
Plasmodium falciparum and was termed aspartate rich protein (PfARP). PfARP is a
cytosolic protein that contains a C-terminal domain rich in aspartate residues, and has
high homology to the SET/TAF family of proteins. Unlike OA and microcystin which
inhibit both PP2A, and PP1, PfARP potently and specifically inhibits PP2A [128]. The
physiological role of PfARP is not known to this date. The observation that PfARP is
localized to the cytoplasm and is highly specific for PP2A makes it an ideal molecular
tool to study the effects of PP2A inhibition at different subcellular compartments by
generating fusion proteins of PfARP that contain various targeting motifs
23
Mitochondrial dysfunction in ischemic stroke and
neurodegenerative diseases
Stroke along with neurodegenerative disorders such as Parkinson’s disease are
leading causes of mortality in the United States. Stroke is caused by the rupture, or
obstruction of cerebral arteries which leads to depletion of oxygen and nutrients and
widespread neurodegeneration. Blood loss caused by stroke leads to a depletion of ATP
levels and an excessive depolarization of the cell membrane due to an inability of ion
channels to maintain a proper balance of Na+, K+ ions, and Ca2+. This in turn leads to an
excessive release of excitatory neurotransmitters such as glutamate which bind to NMDA
and AMPA-type channels and promotes a massive influx of Ca2+ in neurons. At the core
of the infarct, a continual, massive influx of intracellular Ca2+ mediated by the activation
of glutamate receptors causes neurons to die mainly by necrosis by activating Ca2+dependent lipases and proteases while neurons in the penumbra retain sufficient levels of
ATP to undergo apoptotic cell death. (reviewed by [129, 130]). Glutamate excitotoxicity
has been shown to cause mitochondrial dysfunction in neurons. Mitochondria affected by
an abnormal accumulation of intraneuronal Ca2+ undergo swelling and/or fragmentation,
exhibit impaired mitochondrial functions such as a decline in transmembrane potential,
ATP synthesis, complex I activity and an increased production of ROS. An increase in
ROS production leads to lipid peroxidation of cell membranes and the downstream
activation of caspases and cysteine proteases which is ultimately detrimental to neurons
(reviewed by [131]).
Over the last decades, intensive research efforts have failed to develop an efficient
neuroprotective stroke therapy. Despite showing early promise in rodent ischemia
models, ionotropic glutamate receptor antagonists have failed in human clinical trials due
to the symptomatic effects associated with these drugs such as memory and learning
deficits, renal toxicity and epileptic episodes as a consequence of blocking AMPA and
NMDA related functions (reviewed by [132]). However, cell permeable inhibitors that
24
inhibit PSD95-NMDA interactions have been successful in reducing ischemia induced
brain damage and in circumventing secondary side effects [133]. Current efforts of
stroke intervention at the level of mitochondria include overexpression of antioxidant
enzymes (MnSOD, catalase, glutathione peroxidase), or antiapoptotic molecules (Bcl-2)
and developing efficient caspase inhibitors (caspase 3 and 9) ([134] [135] and reviewed
by [136]).
Parkinson’s disease is a slowly progressive, chronic, and lethal disease.
Mitochondrial dysfunction has been linked to PD. In PD, neurons exhibit impaired
mitochondrial functions such as decreased mitochondrial respiration, decreased complex
I activity, decreased ATP synthesis and an increased generation of ROS. Rotenone, a
pesticide commonly used as a chemical model of PD, reproduces many features of
mitochondrial dysfunction as seen in PD [137]. Treating organotypic cultures with low
doses of rotenone treatment leads to a reduction of complex I activity, altered Ca2+
homeostasis and an increased production of ROS which cause oxidative damage [138].
Furthermore, rotenone toxicity causes rapid swelling and fragmentation of mitochondria,
opening of the porin/VDAC transition pore, and the release of apoptogenic factors such
as cytochrome c which activate downstream caspases (ie, caspase 3), leading to activation
of DNAases which cleave chromosomal DNA ([139] and reviewed by [137]).
The studies presented in this chapter characterized a brain-specific splice variant
of the Bβ gene termed Bβ2. The divergent N-terminus of Bβ2 does not affect
phosphatase activity, but encodes a subcellular targeting signal. Bβ2, but not Bβ1, Bβ3,
Bβ4 or an N-terminal truncation mutant, is localized to mitochondria in PC12 cells and in
hippocampal neurons. We show that the Bβ2 N-terminal tail is sufficient to target GFP to
this organelle. Moreover, confocal analyses in hippocampal neurons and in PC6-3 cells
determined that Bβ2 translocates to mitochondria during cells stress to promote
apoptosis. Bβ2 is a proapoptotic regulatory subunit of PP2A, since inducible or transient
expression of Bβ2, but not Bβ1, Bγ, or Bβ2 mutants defective in holoenzyme formation
25
or mitochondrial targeting, accelerates apoptosis in PC12 cells and in hippocampal
neurons. Conversely, inhibiting the endogenous Bβ2 by RNA interference (RNAi)
strongly protects against glutamate excitotoxicity and rotenone induced toxicity. Our
results suggest that Bβ2 is an attractive target of stroke therapies and may be implicated
in neurodegenerative diseases such as Parkinson’s disease.
Experimental procedure
Generation of FLAG- and green fluorescent protein (GFP)tagged Bβ constructs and site-directed mutagenesis
Primers complementary to the N-terminus of Bβ1, Bβ2, Bβ3 and Bβ4 and to the
beginning of the common region (TEAD.) fitted with a HindIII cloning site in
conjunction with two nested reverse primers including sequence complementary to the
Bβ C-terminal end, the FLAG-epitope tag, and a SalI cloning site were used to PCR
amplify Bβ1, Bβ2, Bβ3, Bβ4 and Bβ∆N, respectively. PCR fragments were ligated into
pcDNA5/TO or pEGFP-N1 to generate fusion proteins with C-terminal FLAG- and
FLAG-GFP sequences, respectively. Bβ11-32-GFP and Bβ21-35-GFP were constructed by
excising C-terminal sequences from the full-length Bβ1/2-pEGFP-N1 plasmids by
EcoRI/XmaI digestion, filling in the overhangs with Klenow polymerase, and re-ligating
the plasmids. The Bβ2 RR168EE mutant was constructed by full-plasmid synthesis using
Pfu Ultra polymerase according to instructions for the QuikChange mutagenesis kit
(Stratagene). All constructs were fully sequenced at the University of Iowa DNA Facility.
Generation of hairpins targeted against Bβ2
Four hairpin shRNAs were generated to downregulate the expression of the
endogenous Bβ2 in hippocampal neurons. Two of these shRNA achieved >90%
silencing; Bβ203 is complementary to 5’-GAGACTGGTTTTTCATTTA-3’ in the 5’UTR region of Bβ2, and the Bβ201 shRNA is complementary to 5’-
26
TGCTTCTCTCGTTACCTGC-3’ in the variable N-terminal tail of Bβ2. Double-stranded
oligonucleotides encoding shRNAs were phosphorylated and ligated into pSUPER
vector. An shRNA containing a scrambled sequence (IR-CTRL) was used as a control for
specificity.
Cell culture
COS-M6 cells [140] were cultured in DMEM containing 10% fetal bovine serum
(FBS) and 4.5 g/L glucose and seeded into 6-well plates for transfection on the next day
at ~80% confluency. PC6-3 cells were cultured in RPMI 1640 containing 5% FBS and
10% horse serum (HS) in 2-well chambered slides (Nunc) at a density of 135,000
cells/well or in 12 well plates at a density of 90,000 cells/well.
Hippocampal neurons were cultured in the following manner. Hippocampi were
dissected and extracted from the brain of anesthetized embryonic (E18) rat pups using
sterile techniques. The collected hippocampi tissue was trypsinized in 0.25% trypsinEDTA, centrifuged at 1,000 RPM and resuspended and triturated in HEPES-buffered
saline. Dissociated tissue was removed by two sequential centrifugation and resuspension
steps in neurobasal/B27 supplement (NB/B27) media. Dissociated hippocampal neurons
were then plated in 24-well plates at a density of 60,000 neurons/ well with media
changes every 7 days of culture.
Immunocytochemistry of brain sections
Immunohistochemistry was performed to determine the Bβ2 brain expression
using the Bβ2 specific mAb (6E6) antibody. Adult male Lewis rats were transcardially
perfused with 4% paraformaldehyde in 0.1 M phosphate buffer following a saline rinse.
The perfusion was completed with 10% sucrose in 0.1 M phosphate buffer, followed by
brain removal and a 1 hour post-fix with 4% paraformaldehyde in 0.1 M phosphate
buffer. Brains were stored at 4°C in 30% sucrose in 0.1 M phosphate buffer until the
brains sank, and then, 40 µm sections were cut on a cryostat. Free-floating sections were
27
blocked in 2% normal donkey serum in TTBS for 1-2 hrs, followed by incubation of Bβ2
mAb (6E6) in 1% normal donkey serum in TTBS at 4 degrees overnight. Sections were
washed extensively in TTBS and mAb was bound with biotinylated sheep secondary
(Amersham Biosci., RPN1001) at 1:200 dilution (in TTBS) for 2 hrs. Following washes,
streptavidin-horseradish peroxidase (Amersham Biosci., RPN1051) was bound at 1:100
dilution (in TTBS) for 2 hrs and washed. Stain was developed using the DAB with
nickel staining kit (Vector laboratories, Burlingame, CA).
Generation of antibodies
A peptide derived from the N-terminus of Bβ2 (CFSRYLPYIFRPPNT) was
coupled to keyhole limpet hemocyanin via the sulfhydryl group of the N-terminal
cysteine, and polyclonal antibodies were generated in rabbits and affinity purified by
standard techniques [141].
Bβ2 specific monoclonal antibodies were generated by the Hybridoma Facility at
the University of Iowa. The first 26 amino acids excluding the initial Met of the Nterminus of Bβ2 was fused to glutathione S-transferase (GST) and subcloned into the
pGEX vector, transformed into BL-21 bacteria and the recombinant protein was purified
according to standard GST purification protocols [142]. The purified GST fusion protein
was subsequently crosslinked to a peptide containing the first 19 N-terminal residues of
Bβ2 via the peptide’s single C-terminal carboxyl group to increase representation of the
epitope. Two mice were immunized with the purified GST fusion protein and hybridomas
were generated from the spleen of the mice containing the highest antibody titers by
standard techniques [143]. 960 hybridoma supernatants were screened for Bβ2 specificity
by immunoblotting lysates of inducible PC6-3 clonal cell lines that overexpress FLAGtagged Bβ2 or Bβ1 as a control [17].
Monoclonal antibodies to the PP2A A subunit were a kind gift of Gernot Walter
(UCSD), and PP2A C subunit antibodies were purchased from Transduction
28
Laboratories. The adenine nucleotide translocase antibody was provided by Harmut
Wohlrab (Boston Biomedical Research Institute).
Generation of lentiviruses that overexpress Bβ1/2-GFP and
Bβ2 targeted shRNAs
The H1 promoter containing scrambled or Bβ2 specific shRNAs, and the cDNAs
encoding for Bβ1/2 –GFP were excised from pSUPER and pEGFP-N1 vectors
respectively by restriction digest and subsequently ligated into the lentiviral pVETL
shuttle vector (www.uiowa.edu/~gene) and transformed into DH-5α bacteria. At least 15
µg of extracted DNA from transformed DH-5α bacteria was submitted to the Gene
Transfer Vector Core facility at the University of Iowa for generation of lentiviruses. In
brief, 15µg of shuttle vector containing the DNA inserts were mixed with 4µg of a viral
backbone vector that expresses viral coat proteins and the DNA mix was transfected into
five plates of human embryonic kidney cells (HEK cells) by calcium phosphate methods
[144]. After seven days post-transfection, viral plaques in HEK cells were harvested by
collecting cells with a 5.0 ml sterile serological pipette. Collected cells were lysed by
freeze and thawing three times, followed by centrifugation at 3,500 rpm to remove the
cell debris. Viral particles were concentrated and purified from the supernatant by
standard CsCl gradient extraction/ultracentrifugation methods. The virus was then
dialyzed in 3% sucrose/PBS and diluted to 500 µl in 50%glycerol/BSA (1:1 dilution).
Lentiviral infection of hippocampal neurons
Hippocampal neurons (DIV10) grown on Poly-D-lysine coated 2-well chambered
slides at a density of 60,000 cells/well were incubated with FIV lentiviruses (1X106
pfu/mL for Bβ2-GFP, and 9X105 pfu/mL for Bβ1-GFP) that overexpress GFP fusion
proteins or shRNAs at 37°C. After 5-6 hours following transduction, hippocampal
neurons were washed once with fresh NB/B27 media to remove unbound viral particles
and further incubated in 37°C for 4-7 days to allow expression of GFP fusion proteins.
29
Dual luciferase assays
To assay for downregulation of proteins, hippocampal neurons seeded on 24 well
plates were co-transfected with proteins of interest fused to the Photinus luciferase
enzyme, with shRNAs that target the luciferase fusion protein of interest and with the
Renilla luciferase enzyme as a marker of transfection efficiency. After 3 days of
transfection, hippocampal neurons were washed once in PBS, and lysed with 1X passive
lysis buffer (Promega, DLA kit). Neurons were then scraped with a rubber policeman and
the cell lysate was then transferred into 1.5 ml centrifuge tubes. Photinus luciferase
activity was measured from cleared lysates (20,000 x g, 15 min) with a tube luminometer
according to the manufacturer’s instructions using the dual luciferase protocol and
normalized to the Renilla activity. Successful downregulation of proteins of interest by
shRNAs was indicated by a decrease in the ratio of Photinus/Renilla luciferase activity
compared to scrambled shRNA.
Immunoprecipitation and phosphatase activity assays
COS-M6 cells were transfected in 6-well plates using Lipofectamine 2000 (BD
Biosciences), and C-terminally FLAG-tagged Bβ subunits were immunoprecipitated with
FLAG antibody-agarose conjugates (Sigma) as described [47], except that the
immunoprecipitation buffer lacked protein phosphatase inhibitors. Aliquots of
immunoprecipitates were solubilized in SDS sample buffer for immunoblot analyses. For
PP2A activity assays, immunoprecipitates were stored at –20oC in 50% glycerol, 10%
ethyleneglycol, 20 mM Tris, pH 7.5, 5 mM dithiothreitol, 2 mM EDTA, and 0.1% Triton
X-100. 33P-labeled substrates (see below) were diluted to 0.2-0.5 mg/ml (2,000-10,000
cpm/µl) in 2 mg/ml BSA, 50 mM Tris pH 7.5, 2 mM EDTA, 2 mM EGTA, 2 mM
dithiothreitol, 1 mM benzamidine, 1 mg/ml leupeptin. Phosphatase reactions were started
by addition of 5 µl PP2A immunoprecipitates to 20 µl of diluted substrate, incubated for
30 min at 30oC with intermittent agitation on an Eppendorf shaking incubator, and
30
terminated by the addition of trichloroacetic acid (TCA) to a final concentration of 20%.
Following centrifugation at 22,000 x g, acid-soluble 33P-phosphate was quantified by
liquid scintillation counting. Less than 20% substrate dephosphorylation occurred under
our assay conditions, and activities were completely inhibited by 2.5 nM okadaic acid, a
specific inhibitor of PP2A at this concentration.
Preparation of phosphatase substrates
Partially dephosphorylated casein or bovine brain myelin basic protein (5 mg/ml,
Sigma) was phosphorylated by protein kinase A catalytic subunit (0.25 U/µl, Sigma) for
2-16 hours at 30oC in buffer containing 1 mM ATP, 100 µCi (γ-33P)ATP, 50 mM Tris pH
7.5, 10 mM MgCl2, 2 mM dithiothreitol, 1 mM EGTA, and 0.01% Triton X-100.
Phosphorylation reactions were stopped by addition to 20% TCA, followed by
centrifugation at 22,000 x g and successive washing of the pellet in 10% TCA, 70%
ethanol, and 100% acetone. Following the last wash, 33P-labeled substrates were
dissolved in 50 mM Tris pH 7.5 and stored aliquoted at –80oC.
Confocal imaging of GFP fusion proteins
PC6-3 cells or hippocampal neurons were seeded on collagen-coated or poly-Dlysine-coated, chambered No. 1 cover glasses (20 mm2 chamber, Nalge Nunc) and
transfected with 1 µg GFP fusion protein plasmids using Lipofectamine 2000 as
described [17]. Forty eight to 72 hours post-transfection, cells were imaged live on a
Zeiss LSM 510 laser-scanning confocal microscope at the Central Microscopy Facility of
the University of Iowa. In some experiments, MitoTracker Red CMXRos (Molecular
Probes, Portland, OR) or tetramethylrhodamine methyl ester (TMRM, Molecular Probes)
was added to 100 nM in order to stain mitochondria.
31
Mitochondrial translocation assay
Hippocampal neurons or PC6-3 cells were cultured in poly-D-lysine or collagen
coated 2-well chambered slides respectively (Nunc). PC6-3 cells were transfected with
1µg of BB2-GFP using Lipofectamine 2000 (BD Biosciences) and hippocampal neurons
were infected with lentiviruses that express Bβ1/2-GFP at a 1:200 dilution. After 3-7
days of transduction, PC6-3 cells were stressed with rotenone (10µM) or tunicamycin
(1.3µM) while hippocampal neurons were stressed with rotenone (1µM). Cells were
stained with 100 nM MitoTracker dye (Molecular Probes) and imaged “live” with a Zeiss
LSM 510 laser scanning confocal microscope at the University of Iowa Central
Microscopy Facility.
In order to quantify the colocalization of Bβ2-GFP with mitochondria, 7-10
randomly selected images each containing 4-8 transfected cells were analyzed blind to
the identity of the mutant. Each transfected cell was assigned a GFP/mitochondria
colocalization score from 0 to 4 (0 = mutual exclusion; 4 = perfect overlap [145]) The
colocalization scores were averaged per condition and converted to a ratio of the total
population.
Generation of tetracycline-inducible PC6-3 cell lines
Tetracycline-inducible (T-Rex system, BD Biosciences) PC6-3 cell lines stably
expressing FLAG-epitope tagged B-family PP2A regulatory subunits were generated as
described previously [51]. Between 40 and 60 blasticidine- and hygromycin-resistant
clones were expanded and tested for inducible expression by immunoblotting for the
FLAG epitope tag. In positive clones, maximum protein expression was achieved after 24
hour treatment with 1 µg/ml doxycycline.
Cell death assays
Tetracycline-inducible PC6-3 cell lines were seeded at 10,000 cells/well in
collagen-coated 96-well plates and grown for 72 hours in regular growth medium (10%
32
horse serum, 5% fetal bovine serum in RPMI-1640) in the presence of vehicle (0.1%
ethanol) or 1 µg/ml doxycycline. After two washes, serum-free RPMI-1640 ±
doxycycline was added and cell density was assayed by MTS tetrazolium reduction to
formazan according to the manufacturer’s instructions (CellTiter 96® AQueous
nonradioactive cell proliferation assay, Promega). Formazan production was quantified
after 3 hours by absorbance measurement at 490 nm using a 96-well plate reader. The
MTS assay was repeated after 24 hours in serum-free medium, and cell survival was
expressed as the ratio of the two measurements. Previous apoptosis studies with PC6-3
cells have documented excellent correlation between cell counts and metabolic activity as
assayed by tetrazolium salt reduction [146].
Nuclear morphology assays in PC6-3 cells
For nuclear morphology assays, native PC6-3 cells or tetracycline-inducible cells
were seeded at 200,000 cells/well of 20 mm2 on chambered cover glasses. Native PC6-3
cells were transiently transfected with 1 µg/chamber GFP-fusion protein plasmids using
Lipofectamine 2000 and cultured for 48 hours, whereas inducible cells were vehicle- or
doxycycline-treated for 72 hours prior to serum-deprivation. After 24 hours under serumfree conditions, cultures were fixed in 3.7% paraformaldehyde in phosphate- buffered
saline, incubated with the blue-fluorescent nuclear stain Hoechst 33342 at 1 µg/ml for 5
minutes and mounted on slides. Random microscopic fields (6-12 fields/culture, 50-200
cells/field) were captured on an epifluorescence microscope, and images were coded and
analyzed blind to the experimental condition. Cells with condensed, irregular, or
fragmented nuclei were scored as apoptotic.
Annexin V staining
For annexin V staining assays, PC6-3 cells were seeded on 20 mm2 two-well
chambered cover glasses at a density of 200,000 cells/well. Native PC6-3 cells were
transiently transfected with 1 µg/chamber GFP-fusion protein plasmids using
33
Lipofectamine 2000 and cultured for 48 hours. Cells were then growth factor deprived by
washing three times with RPMI 1640. After twenty four hours of growth factor
deprivation, cells were washed once with PBS and incubated with 1X Annexin V binding
buffer containing a 1:200 dilution of Annexin V conjugated to Cy3 for 5 min at R.T. in
the dark (US Biological, #A2296-30). Cells were then fixed in 3.9% paraformaldehyde
and washed twice in PBS. To analyze for nuclear morphology concurrently with Annexin
V staining, cells were incubated with Hoechst 33342 at 1 µg/ml for 5 minutes. Random
microscopic fields (6-12 fields/condition with 50-200 cells/field) were captured on an
epifluorescence microscope, and images were coded and analyzed blind to the
experimental condition. Annexin V positive cells that contain an intense red fluorescent
halo that lines the cell membrane were scored as apoptotic. At the same time, cells
containing condensed, irregular, or fragmented nuclei were scored as apoptotic.
Hippocampal survival assays
Rat hippocampal neurons (10 DIV) plated in 24 well plates were transfected with
GFP fusion proteins or with empty vector, scrambled hairpin and hairpins (hp) targeted
against endogenous Bβ2 and with β-galactosidase a marker of transfection. After at least
three days of transfection, hippocampal neurons were treated with 250 µM glutamate for
20 minutes followed by one wash of media or with 200-400 nM rotenone to induce
toxicity. After two days of toxic treatment, hippocampal neurons were fixed in 3.7% PF,
permeabilized in TTBS, immunostained for β-galactosidase using a rabbit anti-βgalactosidase monoclonal antibody at a 1:10,000 dilution (Molecular Probes, OR, USA)
or for GFP with an anti-GFP polyclonal antibody at a 1:2000 dilution dilution (Abcam,
Cambridge, UK) and for neurofilament using an mouse anti-neurofilament monoclonal
antibody at a 1:20 dilution (University of Iowa Hybridoma Facility). For glutamate
excitotoxicity and for some rotenone toxicity assays, hippocampal neurons that stained
for neurofilament and that retained an overall normal neuronal morphology were counted
34
as live. For apoptosis counts induced by rotenone toxicity, nuclei were counterstained
with Hoechst 33342 at 1 µg/ml for 5 minutes followed by two washed in TTBS.
Hippocampal neurons that contained a condensed (pyknotic), irregular, or fragmented
nuclei were scored as apoptotic.
Results
Identification of a novel PP2A regulatory subunit
A rat brain cDNA library was screened with a PP2A/Bβ cDNA probe in order to
identify novel isoforms of this brain-specific PP2A regulatory subunit. A clone was
isolated with an insert of 431 bp, of which the first 223 bases are novel and the last 208
bp are identical to the coding sequence for amino acid residues 23-91 of rat Bβ [147].
Conceptual translation of this partial cDNA predicts an isoform of Bβ with a novel 5’
UTR and N-terminal extension of 24 amino acids. The full-length cDNA for Bβ2 was
isolated by RT-PCR from rat brain RNA using primers flanking the coding sequence
(GenBank accession number AY251277).
Database searches with the unique rat Bβ2 sequence identified several human and
mouse ESTs with high degrees of sequence conservation at the nucleotide level and
100% amino acid identity in the coding region. The murine Bβ2 ortholog was recently
described and named Bβ.2 [120]. EST database searches also identified a Bβ2 ortholog
from rainbow trout (accession number CA376753) that has three conservative
substitutions in the N-terminal tail. No Bβ2 related sequences were found in other EST or
genome databases, suggesting that Bβ2 has evolved in the vertebrate subphylum.
The chromosomal organization of exons encoding human Bβ1 and Bβ2 was
determined by computer-aided alignment of the Bβ cDNAs with human and murine
genome databases, and is shown in fig. 3A. The gene structure of the human and murine
Bβ genes is highly conserved as has previously been noted [120]. The alternate N-termini
of Bβ1 and Bβ2 are encoded by exons separated by ~150 kb. Since the transcription start
35
site for the human Bβ1 mRNA is less than 600 nucleotides upstream of the initiation
codon [43], the Bβ2 transcript appears to be generated by use of an alternate promoter
upstream of exon 1.2, and splicing of exon 1.2 to the first common exon (Fig. 3A). Of
note, human EST BC031790 predicts an alternate Bβ mRNA in which exon 1.2 is fused
out-of-frame to exon 1.1 and the rest of the coding sequence. The existence of this
apparently incompletely spliced EST supports the notion that Bβ2 is generated by cis
splicing of a huge pre-mRNA (~500 kb) spanning the Bβ locus. Based on prior structure
modeling and site-directed mutagenesis of the related Bγ subunit [47], the variant Bβ2 Nterminus is predicted to extend from a presumed β-propeller core structure encoded by
common exons 2-9 (Fig. 3B).
Characterization of Bβ2 expression
Initial work on Bβ2 at the mRNA and protein level preceding my dissertation
was performed by Julie Zaucha and by Drs. Stefan Strack and Brian Wadzinsky’s lab at
Vanderbilt University. Ribonuclease protection assays using probes corresponding to the
divergent N-terminus revealed that Bβ2 is exclusively expressed in the brain but not in
other tissues [17]. Strack et al., 1998 previously reported that B-family regulatory
subunits exhibit distinct developmental expression profiles in the brain [67]. The
neuronal Bγ isoform is induced during postnatal brain development, whereas Bα and Bβ1
show constant expression and a slight postnatal decline in expression, respectively.
Ribonuclease protection assays using a Bβ2-specific probe revealed that this isoform has
an expression pattern similar to Bγ, with near undetectable expression at birth rising to
adult levels by postnatal day 14 [17].
Our data indicate that Bβ2 mRNA is specifically found in mature brain, but do
not rule out a non-neuronal (e.g. glial) origin of expression. To address this issue,
multiple cell lines of neuronal (PC6-3, PC12, B104, SHSY5Y, Neuro2A), glial (C6,
Ng108), and other (COS, HEK293, NIH3T3, MCF7) origin were analyzed for
36
Bβ isoform expression by competitive RT-PCR. Whereas all neuronal cell lines tested
expressed Bβ1, none had detectable levels of Bβ2 (data not shown).
To analyze for protein expression levels of Bβ2 in the brain, polyclonal antibodies
were generated by immunizing rabbits with a peptide from the unique N-terminal tail of
Bβ2. The resulting antibody reacted specifically with heterologously expressed Bβ2 and
displayed no cross-reactivity with Bβ1 or other PP2A regulatory subunits. Bβ1 has been
previously shown to be detectable in total brain lysates ([67] and Fig. 4), antibody
detection of a protein with the size predicted for Bβ2 (52K) necessitated enrichment of
PP2A holoenzymes by microcystin-Sepharose affinity purification. COS cell lysates
expressing FLAG-epitope tagged Bβ splice variants were used as standards and
immunoblotted with Bβ isoform-specific and FLAG-directed antibodies to compare
detection strengths of Bβ1 and Bβ2 antibodies. Thus normalizing for antibody affinities
and titers, the relative abundance of Bβ1 and Bβ2-containing PP2A holoenzymes in rat
brain was estimated to be approximately 10:1 (Fig. 4). We also attempted to analyze the
subcellular distribution of Bβ2 in the rat brain by immunocytochemistry. Unfortunately,
the Bβ2 specific polyclonal antibody failed to detect Bβ2 over background stain by
immunocytochemistry. To this end, we generated a monoclonal antibody that specifically
recognizes the N-terminus of Bβ2 but not of Bβ1. The Bβ2 mAb (6E6) antibody
specifically recognized inducibly overexpressed Bβ2 but not Bβ1 in cell lysates of PC6-3
cells (Fig. 5A) By using this Bβ2 mAb, immunocytochemistry of rat brain sections was
performed by Dr. Merrill in our lab to analyze the subcellular distribution of Bβ2 in
neurons. In the cerebellum, he found that the Bβ2 monoclonal antibody stains the same
neuronal populations as previously reported for Bβ1. However, there were clear
differences in the subcellular distribution of the Bβ isoforms in that Bβ1 was localized in
the soma and in dendrites of Purkinje neurons while Bβ2 is localized in the soma but was
excluded from dendrites ([67], Fig. 5B). These results show that alternate splicing of the
Bβ gene gives rise to multiple isoforms with distinct subcellular localization in neurons.
37
Furthermore, Bβ2 was found to be expressed in all the areas of the hippocampus
including the CA1-CA3 regions, and in the dentate gyrus. Interestingly, the Bβ2 antibody
stained discrete patchy, punctate areas in the soma of neurons of the dentate gyrus and of
thalamic neurons suggesting that Bβ2 is targeted to an unidentified organelle/network
(Fig. 5C and D, and not shown for thalamic neurons).
The Bβ2 N-terminus does not affect holoenzyme formation
or catalytic activity
Previous structure-function studies indicated that the variable N-terminus of Bγ is
dispensable for binding to the A and C subunits [47]. A role of the Bγ N-terminus in
directing the PP2A holoenzyme to specific substrates in the MAP kinase pathway was
suggested by analyses of chimeras between this neuronal-specific regulatory protein and
the widely expressed Bα subunit [51]. To investigate whether the differentially spliced
N-termini of the Bβ isoforms play a role in formation or catalytic activity of the PP2A
holoenzyme, the cDNAs of four Bβ isoforms (Bβ1− Bβ4) where FLAG epitope tagged at
the C-terminus and transiently expressed in COS-M6 cells. A deletion mutant lacking the
divergent N-terminus, Bβ∆N, was also constructed and analyzed in parallel. The
ectopically-expressed proteins were immuno-isolated with anti-FLAG resin and analyzed
for association with endogenous A and C subunits by Western blot analysis. Bβ1, Bβ2,
Bβ3, Bβ4 and Bβ∆N could be expressed to similar levels and associated with equivalent
amounts of A and C subunits (Fig. 6A). Aliquots of Bβ1, Bβ2 and Bβ∆N
immunoprecipitates were then assayed for dephosphorylation of two model substrates,
myelin basic protein and casein phosphorylated in vitro by protein kinase A (Fig. 6B).
Myelin basic protein was a better substrate than the more acidic casein in these assays.
Importantly, the three PP2A heterotrimers had equivalent activities towards these
substrates. Therefore, we conclude that the divergent N-termini of the Bβ isoforms are
38
not involved in formation of the PP2A heterotrimer, and do not influence substrate
recognition, at least in these in vitro assays.
The Bβ2 N-terminus encodes a mitochondrial localization
signal
Bα, Bβ1, and Bγ regulatory subunits are found in different subcellular fractions
from brain and localize differentially to neuronal somata and processes [67]. We explored
a possible function of the N-termini of the Bβ isoforms in directing the protein to
different places in the cell by transiently transfecting the PC6-3 subline of neuronal PC12
cells with expression plasmids encoding Bβ1, Bβ2, Bβ3, Bβ4 and Bβ∆N tagged at the Cterminus with GFP. All GFP fusion proteins were not appreciably degraded and were
expressed at similar levels in PC6-3 cells. Furthermore, addition of the GFP moiety did
not compromise co-immunoprecipitation of Bβ with endogenous A and C subunits (data
not shown). Live confocal microscopy revealed that in every transfected cell, Bβ1-GFP,
Bβ3-GFP and Bβ4-GFP showed diffused localization in the cytoplasm, and clearly
excluded from the nucleus. Bβ2-GFP, on the other hand, was the only isoform that
showed a punctate localization in addition to diffuse cytoplasmic fluorescence (Fig. 7A,
middle top panel). The degree of punctate versus diffuse Bβ2-GFP fluorescence varied
between cells and transfections (10-50% cells with discernible punctae). The most
straightforward interpretation of these results is that the Bβ2 N-terminus is responsible
for localizing the protein to punctate structures in cells. Alternatively, the Bβ1 Nterminus may encode a cytoplasmic targeting signal that overrides a “default” address in
the common region of Bβ for punctate localization. The latter interpretation can be
discounted, since the subcellular distribution the N-terminal deletion mutant (Bβ∆NGFP) was indistinguishable from other GFP tagged Bβ isoforms (Fig. 7A).
The punctae labeled by Bβ2-GFP were identified as mitochondria in doublelabeling experiments with the red rosamine derivative dye MitoTracker, which
39
accumulates in actively respiring mitochondria (Fig. 7B). Bβ1-GFP, in contrast, appeared
deplete in areas with high densities of mitochondria.
To investigate whether the Bβ2 N-terminus is sufficient for targeting to
mitochondria, the first 35 amino acids of Bβ2, including the unique 24 residues and 11
residues shared with Ββ1 were fused to the N-terminus of GFP (Bβ21-35-GFP). The
corresponding N-terminal fusion of Bβ1 (Bβ11-32-GFP) served as a control. The Nterminus of Bβ2, but not Bβ1, was capable of targeting GFP to mitochondria in PC6-3
cells (Fig. 8, bottom right panel). In contrast to the full-length protein, Bβ21-35-GFP
showed a strikingly discrete mitochondrial localization in virtually every transfected cell,
with little if any diffuse fluorescence. It is conceivable that the common C-terminal
region of the Bβ splice variants associates with cytoplasmic proteins/structures, which
gives rise to the mixed diffuse/mitochondrial localization of full-length Bβ2-GFP.
Bβ2 promotes apoptosis
In addition to performing critical functions in biosynthesis and energy
metabolism, mitochondria are central to apoptotic signal transduction [148]. Since Bβ2GFP is localized in both cytosolic and mitochondrial compartments in healthy neurons,
we wanted to determine whether Bβ2 redistributes to mitochondria during apoptosis. The
PC6-3 subline of PC12 cells was established by Pittman and co-workers as a neuronal
apoptosis model that more closely resembles sympathetic neurons than the parental PC12
cell line [146]. Undifferentiated PC6-3 cells express primarily Bα, whereas nerve growth
factor-differentiated cells additionally express Bβ1 and Bγ [51]. Endogenous Bβ2
expression is undetectable by RT-PCR under either condition (data not shown).
PC6-3 cells were transfected with Bβ2-GFP followed by toxic insult with different
toxins and analyzed for colocalization of Bβ2-GFP with mitochondria by confocal
microscopy. In order to quantify colocalization of Bβ2-GFP with mitochondria,
randomly selected images each containing 4-8 transfected cells were analyzed blind to
40
the identity of the transfection condition. Each transfected cell was assigned a
GFP/mitochondria colocalization score using a scale from 0 to 4 where a score of 0 was
given to a cell showing mutual exclusion of GFP with mitochondria, a score of 1 was
given to a cell showing coincidental colocalization of GFP with mitochondria, scores of
2, 3 were assigned to cells showing increasing degrees of colocalization, and a score of 4
was given to a cell showing perfect overlap of GFP with mitochondria [145]. The
colocalization scores were averaged per condition and converted to a ratio of the total
transfected population. In healthy cells, only 10% of PC6-3 cells contained strong
colocalization of Bβ2-GFP with mitochondria. However, toxic insult of PC6-3 cells with
lethal doses of rotenone or tunicamycin, a protein glycosylation inhibitor, induced a
striking redistribution of Bβ2-GFP from the cytoplasm to mitochondria (Fig. 9A, and B).
A time course analysis revealed that mitochondrial translocation of Bβ2-GFP is rapid and
maximal (40%) after 24 hrs. of rotenone or tunicamycin treatment (Fig. 9C and D).
To study the subcellular localization of Bβ2 in primary neurons, we generated
lentiviruses that overexpress GFP fusion proteins of Bβ1/2. Lentiviruses efficiently infect
hippocampal neurons (~50%), and overexpressed Bβ1/2-GFP associate with endogenous
A and C subunits of PP2A (data not shown). In healthy neurons, Bβ2-GFP colocalized
with mitochondria in 40% of hippocampal neurons (Fig. 10A top right panel, and C).
However, toxic insult of hippocampal neurons with a lethal dose of rotenone (2µM)
induced a striking redistribution of Bβ2-GFP from the cytoplasm to mitochondria and
this phenomenon does not occur in neurons that express Bβ1-GFP (Fig. 10A, bottom
right panel). A time course analysis in hippocampal neurons revealed that maximal
mitochondrial translocation of Bβ2-GFP was achieved at 5 hrs. of rotenone treatment
(75%) (Fig. 10C). At the single cell level, we found a positive correlation of
mitochondrial translocation of Bβ2-GFP with apoptosis in PC6-3 cells and in
hippocampal neurons treated with rotenone. Furthermore, a time course analysis of
apoptosis and mitochondrial translocation demonstrated that translocation of Bβ2-GFP to
41
mitochondria (t1/2= 2.5 hr) precedes apoptosis induction (t1/2= 7.0 hr) in hippocampal
neurons. It is well documented that cells undergoing apoptotic cell death exhibit a
decrease in mitochondrial respiration (see review by [103]). We found that the
mitochondria of Bβ2 overexpressing hippocampal neurons did not stain well with the
mitochondrial respiration dependent TMRM dye compared to mitochondria of
untransfected neurons suggesting that Bβ2 may accelerate mitochondrial dysfunction
(Fig. 10B) Furthermore, overexpression of Bβ2 increased basal and rotenone induced
apoptosis compared to the untransfected population (Fig. 10C). Overall, these results
suggest that Bβ2 translocates to mitochondria to accelerate apoptosis (Fig. 9C; Fig.
10C).
The ability of Bβ2 to promote apoptosis was also tested in PC6-3 cells. To this
end, we generated a panel of stable, clonal PC6-3 cell lines that express Bβ1, Bβ2, or Bγ
under control of a tetracycline-inducible cytomegalovirus promoter [51].
Growth of the stable PC6-3 cell lines in doxycycline-containing medium led to
the induction of comparable levels of Bβ1, Bβ2 and Bγ, as detected with an antibody to
the FLAG epitope tag (Fig. 11A). Inducible expression of B-family regulatory subunits
was similar to endogenous Bα subunit, as visualized with a pan B-subunit antibody.
Levels of A and C subunits, as well as levels of members of the B’ regulatory subunit
family were unaltered following B subunit induction (Fig. 11A, and data not shown).
Growth rates were unaffected by doxycycline treatment in two independently isolated
Bβ2-expressing cell lines (data not shown). We also assayed cell viability in serumcontaining media, and found that Bβ2 induction is not toxic in PC6-3 cells (Fig. 11B).
Complete removal of serum kills 20-50% of PC6-3 cells within 24 hours as
assayed by tetrazolium salt reduction (Fig. 11C). In two different clonal cell lines,
inducible Bβ2 expression decreased survival by 30-40% assayed 24 hours after serum
withdrawal (Fig. 11C). Accelerated cell death was specific for this mitochondrialocalized Bβ splice variant, since Bβ1 or Bγ induction had moderate to no effect on cell
42
survival. Nuclear condensation and fragmentation are hallmarks of late-stage apoptosis.
We examined nuclear morphology after staining with a DNA dye to demonstrate that
Bβ2 decreases survival by promoting apoptosis. Inducible expression of Bβ2, but not
Bβ1, almost doubled the number of cells with apoptotic nuclei after 24 hours in serumfree media (Fig. 11D).
The effects of Bβ2 overexpression on survival was also tested by transient
transfection. To this end, we carried out apoptosis experiments in which various B-family
regulatory subunit constructs tagged at the C-terminus with GFP were transiently
transfected into PC6-3 cells. 24 hours following serum removal, GFP positive cells with
apoptotic nuclei were counted. In agreement with the data obtained from inducible cell
lines in PC6-3 cells, we found that transient expression of Bβ2, but not Bβ1, or Bγ,
increased the number of apoptotic cells compared to transfection with GFP alone (30%
vs. 5%, Fig. 13A). In hippocampal neurons, we found that transient expression of Bβ2GFP was sufficient to increase basal apoptosis compared to Bβ1-GFP, or GFP targeted to
the OMM via the α-helix transmembrane segments of MAS70 (OMM-GFP) [149],
(p<0.0001, Fig. 14).
The proapoptotic effects of Bβ2 were also confirmed by the Annexin V staining
assay. Annexin V is a protein that binds to phosphatidyl serine, a lipid of the plasma
membrane that normally faces the cytoplasmic side and is inverted to the exterior during
apoptosis. Annexin V and nuclear morphology assays revealed that overexpression of
Bβ2 increased apoptosis compared to Bβ1or GFP in growth factor deprived PC6-3 cells.
It is known that PP2A acts upstream of mitochondria during apoptosis but also acts
downstream as a substrate for caspases [117, 150]. To determine the level at which Bβ2
acts in apoptosis, PC6-3 cells were co-transfected with Bβ2-GFP and Bcl-2, a well
characterized prosurvival mitochondrial protein, followed by growth factor deprivation
for 24 hours. Annexin V staining and nuclear morphology assays determined that Bcl-2
efficiently abolished the proapoptotic activity of Bβ2 (Fig. 12) in serum starved PC6-3
43
cells suggesting that Bβ2 mediates apoptosis upstream of mitochondrial OMM
permeabilization and cytochrome c release.
Bβ2 requires mitochondrial association and incorporation
into the PP2A heterotrimer to promote apoptosis
It is conceivable that binding of the Bβ2 N-terminus to mitochondria has a
nonspecific toxic effect on cells. Arg165 and Arg166 of Bγ form critical salt bridges with
Glu100 and Glu101 of the Aα subunit [47]. To address this issue, we substituted the pair
of arginines in Bβ2 corresponding to Bγ with glutamates to generate a mutant
(RR168EE) that cannot associate with the AC dimer. Bβ2 RR168EE was able to
efficiently bind to an Aα subunit carrying the opposite charge reversal mutation
(EE100RR), demonstrating that the mutant Bβ2 protein folds normally (Fig. 13B). As
expected, we also failed to detect binding of PP2A A and C subunits to the mitochondriatargeting N-terminus of Ββ2 (Bβ21-35) (Fig. 13B). Neither monomeric Bβ2 point mutant
nor the Bβ2 N-terminus fused to GFP were able to promote apoptosis following growth
factor deprivation in transient transfection assays in PC6-3 cells (Fig. 13C, and D), even
though fluorescence levels were equivalent to (Bβ2 RR168EE, Fig. 13C) or much
greater (Bβ21-35) than full-length, wild-type Bβ2. In hippocampal neurons, transient
expression of the Bβ2 monomer did not promote apoptosis. In fact, the Bβ2 monomer
appeared to have a dominant negative effect in inducing a decrease in basal apoptosis
compared to neurons expressing OMM-GFP (Fig. 14).
An endogenous inhibitor of PP2A was isolated from Plasmodium falciparum
termed aspartate rich protein (PfARP). PfARP is a cytosolic protein that and contains a
C-terminal domain rich in aspartate residues, and has high homology to the SET/TAF
family of proteins. Unlike OA and microcystin which can inhibit both PP2A and PP1,
PfARP potently and specifically inhibits PP2A [128]. In order to determine the effects of
inhibiting the catalytic activity of mitochondrial PP2A on neuronal survival, we
44
generated a GFP fusion protein in which PfARP was redirected from the cytoplasm to the
mitochondria by N-terminally fusing the mitochondrial targeting domain of MAS70
(OMM-PfARP). Transient expression of PfARP alone increased basal apoptosis in
hippocampal neurons compared to OMM-GFP suggesting that PP2A activity is essential
for survival. This result agrees with previous studies which demonstrated that PP2A is an
essential enzyme since pharmacological inhibition or knocking out the gene that encodes
the C subunit of PP2A is detrimental [25, 151]. Unexpectedly, transient expression of
OMM-PfARP also increased basal apoptosis in hippocampal neurons (and to a lesser
extent compared to PfARP alone) compared to neurons expressing OMM-GFP
suggesting that PP2A activity at the mitochondria is critical for survival (Fig. 14). To
determine whether the proapoptotic activity of Bβ2 requires mitochondrial association,
we tested the ability of an N-terminal mutant of Bβ2 (R6A) that has impaired
mitochondrial targeting (Fig 24B) for promoting apoptosis. The R6A point mutant did
not promote apoptosis compared to wild-type Bβ2-GFP (Fig 14) suggesting that Bβ2
requires association with mitochondria to promote apoptosis.
Knock-down of the endogenous Bβ2 is neuroprotective
Since overexpression of Bβ2 is toxic in neurons, we wanted to determine whether
downregulating the endogenous protein by RNAi is neuroprotective. The effects of
downregulating endogenous Bβ2 on neuronal survival were analyzed in an in vitro model
of ischemia and Parkinson’s disease. To this end, hippocampal neurons were cotransfected with two Bβ2 specific shRNAs (which potently downregulate up to 90% of
expression of luciferase fusion protein of Bβ2) and β-galactosidase as a marker for
transfection (Fig. 15A). Three days post-transfection, neurons were stressed with
glutamate (250µM) for 20 min. followed by a media wash or with rotenone (50-200nM).
Two days following toxic insult, neurons were fixed and immunostained for
neurofilament and β-galactosidase. Glutamate treatment kills about 80% of neurons
45
transfected with an empty vector or a scrambled hairpin control, as evidenced by their
lack of neurofilament staining. However, neurons transfected with Bβ2 shRNA#1 stained
for neurofilament and retained a normal neuronal morphology (Fig. 15B). Quantifying
for the number of surviving neurons, we noticed that RNAi mediated knock-down of Bβ2
by shRNA #1 dramatically and significantly increased the number of surviving neurons
by a factor of 2.5 compared to hippocampal neurons transfected with empty vector or
scrambled shRNA controls following exposure to glutamate or rotenone (p<0.0001),
(Fig. 15C, and D). To test for specificity of knock-down, we determined the ability of a
second Bβ2 specific shRNA to promote survival against rotenone. While rotenone
treatment promotes apoptosis in neurons transfected with a scrambled shRNA, neurons
that express the second Bβ2 specific shRNA (Bβ2 #3 shRNA) retained a normal
dendritic network and had intact nuclei (Fig. 16A). Counting for the number of survival
neurons, we found that both Bβ2 specific shRNAs were equally efficient in decreasing
apoptosis followed by rotenone exposure compared to neurons transfected with a
scrambled shRNA (p<0.01), (Fig. 16B).
These data strongly support a model in which Bβ2 modulates neuronal survival
by targeting an active PP2A heterotrimer to dephosphorylate mitochondrial substrates
(Fig. 17).
Discussion
Protein kinases recognize their substrates via primary sequence motifs
surrounding phosphorylatable residues. In addition, the spatial constraint imposed by
tethering protein kinases to organelles and other subcellular structures further enhances
the fidelity of intracellular signaling via these enzymes [152]. In contrast, consensus
sequences appear to contribute little to substrate recognition by protein phosphatases
[153], and mechanisms for specific subcellular targeting of these enzymes remain
relatively unexplored. This study challenges the view that phosphatases lack specificity
46
by documenting the first example of a protein phosphatase subunit that contains a
mitochondrial targeting signal.
We show that the gene for the Bβ regulatory subunit of PP2A gives rise to an
alternative splice variant, termed Bβ2. Bβ2 mRNA and protein is highly expressed in
forebrain structures, and is induced during postnatal brain development. The unique Nterminal extension of Bβ2 is shown to be necessary and sufficient for targeting to
mitochondria. The functional consequence of mitochondrial localization appears to
modulate apoptosis upstream of Bcl-2, as demonstrated by transient and inducible
overexpression of Bβ2 in a neuronal cell line and by transient transfection in
hippocampal neurons.
Complexity of Bβ gene expression
The Bβ gene (PPP2R2B) is unique among genes encoding B-family PP2A
regulatory subunits in that it gives rise to multiple variants containing unique N-terminal
tails. In a recent report, Schmidt et al. reported the cloning of cDNAs for two novel
murine Bβ isoforms [120]. Bβ.2 is the murine ortholog of Bβ2, the subject of the present
report. Bβ.1, which has a distinct N-terminal tail, may be a murine-specific Bβ splice
form, because no orthologs are present in other EST databases and RT-PCR failed to
detect the presence of this isoform in rat brain (Strack, unpublished). The number of
potential Bβ isoforms appears to be even greater, since RT-PCR from human brain
samples combined with EST and genome database searches identified several other Bβ
gene transcripts that have unique 5’UTR and N-terminal sequences (Holmes and
Margolis, personal communication). Judging by the number of entries for Bβ2 in human
and mouse EST databases, Bβ2 appears to be the second most common Bβ isoform,
although it is much less abundant than Bβ1 at the protein level [17]. Furthermore,
immunocytochemistry using monoclonal antibodies against the regulatory subunits
demonstrated that both isoforms are expressed in the same populations of neurons in the
47
cerebellum and hindbrain. However, Bβ1and Bβ2 show remarkable differences in
subcellular localization supporting the premise that B-family regulatory subunits target
PP2A to distinct subcellular sites. It seems therefore likely that the remaining,
uncharacterized Bβ isoforms are expressed at extremely low levels or in only small
subsets of neurons.
Bβ has recently attracted the attention of the research community because of its
involvement in the neurodegenerative disorder SCA12. A CAG trinucleotide repeat
expansion immediately upstream of the transcription initiation site of Bβ1 was found to
be responsible for this disorder [87], which is a relatively common type of
spinocerebellar ataxia in India [154]. It will be important to address the effect of this
repeat expansion on mRNA and protein levels of not only Bβ1, but also other Bβ
isoforms, especially in light of our finding that Bβ2 overexpression promotes neuronal
apoptosis.
Structural implications
Members of the B-family of PP2A regulatory subunits are greater than 80%
identical at the amino acid level, with greatest sequence divergence at the N-terminus.
These proteins are predicted to adopt a toroidal, β-propeller structure that makes multiple
contacts with the AC dimer [47]. Pairs of conserved arginines that bind directly to a
adjacent glutamates in the A subunit, as well as other amino acids critical for holoenzyme
association map to WD repeats 3 and 4.in the mid portion of the B subunit molecule (Fig.
19; Fig. 20). Based on these data, we arrived at the model topology shown in figure 3B.
The divergent N-terminus is located opposite of the AC dimer interface, where it is free
to engage in macromolecular interactions that determine the subcellular localization of
the PP2A holoenzyme.
Consistent with this model and our previous mutagenesis data with Bγ [47], the
N-terminal tail can be deleted from Bβ without compromising its association to A and C
48
subunits (Fig. 6A). Furthermore, our in vitro phosphatase assays suggest that neither the
presence nor the identity of the N-terminal tail has any effect on phosphatase activities
towards two model substrates (Fig. 6B), arguing that the divergent residues are not
involved in substrate recognition. Instead, the N-terminus of Bβ2 was found to encode a
subcellular targeting module that can direct GFP to mitochondria as shown by
microscopy (Fig. 8). These results favor subcellular targeting mechanism for B
regulatory subunits of PP2A for dephosphorylating substrates. By analogy, we propose
that the differential localization of other B-family subunits [67] is also a function of their
N-terminal sequences.
There are two possibilities by which regulatory subunits may mediate substrate
recognition of PP2A. One possibility is that the binding of regulatory subunits to the AC
core induces an allosteric conformation in the catalytic subunit to increase the affinity
and specificity toward substrates. However, there is currently no evidence that supports
this model. In fact, the association of MYPT1 regulatory subunit to the C subunit of PP1
does not induce conformational change in the catalytic cleft when compared to the crystal
structures of the monomeric form of the catalytic subunit of PP1, PP1 bound to calyculin
A or to okadaic acid [20, 155-157]. Thus, we propose that regions within the presumed βpropeller of B family regulatory subunits not involved in holoenzyme association may
mediate substrate recognition by directly docking to substrates or by increasing the
substrate binding surface as seen for PP1/MYPT1 [20] (See Chapter V).
In vitro phosphatase assays have demonstrated a positive effect of B family
regulatory subunits in the desphosphorylation of substrates. For instance, in vitro
phosphatase assays have demonstrated that two PP2A holoenzymes containing Bα or Bδ
but not B’ regulatory subunits contains similar phosphatase activity towards ERKs [67].
Moreover, our in vitro results demonstrate that Bβ1 and Bβ2 contain similar phosphatase
activities towards phosphorylated myelin basic protein and casein. Thus, our in vitro
phosphatase results and previous results may support a role of the B subunit β-propeller
49
in mediating substrate recognition of the C subunit. The divergent N-termini further
restrict substrate specificity in vivo by mediating subcellular targeting.
PP2A in apoptosis
The pheochromocytoma PC12 cell line and its PC6-3 subline are established
model systems for neuronal differentiation and apoptosis studies [146, 158]. We have
generated stable, clonal PC6-3 lines that express neuronal PP2A regulatory subunits
under the control of a tetracycline-inducible promoter to investigate their involvement in
apoptosis signal transduction pathways. Inducible expression levels of neuronal
regulatory subunits are similar to the endogenous Bα subunit, and are not accompanied
by any changes in A, C, or other regulatory subunit levels (Fig. 11A). The lack of any
compensatory changes in combination with the known instability of monomeric B-family
regulatory subunits [47] suggests that the induced regulatory subunits complex to a pool
of free PP2A dimer in the cell [159]. Using this system, we show that inducible
expression of the mitochondria-targeted Bβ2 subunit potentiates neuronal death
following growth factor withdrawal without affecting cell viability in the presence of
serum (Fig. 11). It is important to point out that Bβ1, Bβ2, and Bγ induction levels in
PC6-3 cell lines are likely considerably higher than in native neurons, where Bα is the
most abundant B-family regulatory subunit [67]. The involvement of B-family regulatory
subunits in apoptosis signaling was also analyzed by transient transfection. Confirming
our inducible expression studies, we found that transient overexpression of Bβ2 but not
other regulatory subunits or GFP targeted to mitochondria promotes apoptosis in
hippocampal neurons and PC6-3 cells. Moreover, the proapoptotic activity of Bβ2
requires incorporation into the holoenzyme and a functional mitochondrial targeting
signal since a monomeric Bβ2 (RR168EE) or a mutant deficient for associating with
mitochondria (as shown for Bβ2 (R6A) in hippocampal neurons) failed to promote
50
apoptosis. In PC6-3 cells, co-expression of the mitochondrial survival protein Bcl-2
blocked Bβ2’s proapoptotic activity suggesting that PP2A/ Bβ2 modulates a
mitochondrial survival pathway upstream of OMM permeabilization and cytochrome c
release. Furthermore, the observation that Bβ2 relocalizes to the OMM during cell stress
suggests that PP2A/ Bβ2 is part of a positive feedback mechanism to accelerate apoptosis
in neurons.
The mechanism of mitochondrial translocation of Bβ2 remains to be elucidated.
In Chapter III, I present data that provides some insight into the mechanism of
mitochondrial translocation. The divergent N-terminus of Bβ2 is a cryptic mitochondrial
import sequence that associates with mitochondria via critical basic residues and stretches
of hydrophobic residues (Fig.23-24). However, Bβ2 is retained at the OMM via a
structure-based mechanism that involves a rigid β-propeller domain that resists the
unfolding step necessary for import. It is clear that the sole function of the divergent Nterminus of Bβ2 is to target the regulatory subunit to mitochondria since it is functionally
interchangeable with the non-homologous import sequence of COX8, and the outer
mitochondrial sequence of MAS70, which does not compromise the proapoptotic
function of Bβ2 (See chapter III, Fig.30). Interestingly, replacing the N-terminus of Bβ2
with that of MAS70 (MAS70- Bβ) shows a mixed cytosolic/mitochondrial localization
similar to wild-type Bβ2 while COX8- Bβ shows enhanced mitochondrial targeting (data
not shown). These observations suggest the N-terminus of Bβ2 contain residues that may
interact with cytosolic proteins/structures which keep the Bβ2 containing PP2A
holoenzyme in the cytosol. 14-3-3 proteins sequester and inactivate Bad in the cytosol by
binding to multiple phosphorylated serine residues in Bad (reviewed by [160, 161]). In
the presence of trophic support, it is conceivable that Bβ2 is inactivated by
phosphorylation and sequestered in the cytosol in a similar manner to Bad by interacting
with an unidentified 14-3-3 like protein. Supporting this notion are unpublished
observations have demonstrated that the N-terminus of Bβ2 can be phosphorylated by
51
PKA in vitro. It is conceivable that unidentified phosphorylated serine residues in the
divergent N-terminus of Bβ2 may be required for interacting with 14-3-3 proteins in the
cytosol. During cell stress, the N-terminus of Bβ2 may undergo desphosphorylation
causing the dissociation of Bβ2 from 14-3-3 proteins. The increase in net positive charge
of the N terminus may facilitate translocation of the proapoptotic regulatory subunit to
mitochondrial import receptors to promote apoptosis by dephosphorylating an unknown
substrate at the OMM (see chapter IV).
In order to demonstrate that the endogenous Bβ2 plays a role in neuronal survival,
the effects of knocking-down the endogenous protein by RNAi was analyzed in
hippocampal neurons. RNAi mediated downregulation of Bβ2 by two independent
shRNAs promoted significant neuroprotection against glutamate mediated excitotoxicity.
The observations that Bβ2 is a neuron specific regulatory subunit that participates in a
apoptotic signaling pathway downstream of glutamate receptors makes it an attractive
target of stroke therapies. It is conceivable to develop small anti-Bβ2 inhibitors (ie,
intrabodies) directed against the N-terminus of Bβ2 to prevent its mitochondrial
translocation and apoptotic activity (See Chapter V for details). Bβ2 translocation
inhibitors may be devoid of teratogenic and non-specific side effects in other tissues since
Bβ2 is expressed in the adult rat brain.
Moreover, we found that RNAi mediated downregulation of endogenous Bβ2 by
two shRNAs significantly and dramatically protects against rotenone induced apoptosis, a
chemical model of Parkinson’s disease. It is well documented that rotenone toxicity leads
to impaired mitochondrial functions as evidenced by a decrease in transmembrane
potential, ATP synthesis and complex I activity [148]. The observations that Bβ2
decreases the transmembrane potential and increases apoptosis in neurons treated with
rotenone raises the possibility that the regulatory subunit may play a role in the
pathogenesis of neurodegenerative diseases such as Parkinson’s disease (Fig. 10B; Fig.
14).
52
The decision between cell survival and apoptotic cell death depends on relative
expressions levels and phosphorylation states of pro- and antiapoptotic members of the
Bcl-2 family of proteins [148]. We hypothesize that targeting of PP2A to mitochondria
via the divergent N-terminus of Bβ2 tips the balance towards dephosphorylation,
facilitating the activation of proapoptotic proteins or the inactivation of antiapoptotic
proteins when cells are challenged by removal of survival factors.
Several lines of evidence support the idea that the balance of kinase and
phosphatase activities at the mitochondrial membrane is pivotal for survival signaling. In
a set of experiments complementary to the present study, Affaitati et al. showed that
inducible overexpression of the mitochondria-targeted A kinase anchoring protein
(AKAP) 121 promotes survival of PC12 cells [162]. The prosurvival effect of AKAP121
was suggested to involve enhanced phosphorylation by protein kinase A and cytosolic
sequestration of Bad, a proapototic Bcl-2 family protein. Although the substrates for Bβ2
are unknown, it is conceivable that PP2A may oppose the survival effects of PKA
targeted to the OMM via AKAP121. Significantly, a PP2A-like activity was implicated in
Bad dephosphorylation and apoptosis of lymphoid cells following interleukin-3 removal
[162]. In another set of studies with a leukemia-derived cell line, toxic concentrations of
the lipid second messenger ceramide were shown to activate PP2A, promote its
translocation to mitochondria, and cause dephosphorylation of the prosurvival protein
Bcl-2 at Ser70 [116, 117]. A member of the B’-family of PP2A regulatory subunits
highly expressed in non-neuronal tissues, B’α, was implicated in targeting PP2A to Bcl-2
in the latter studies, suggesting that induction of apoptosis may involve multiple PP2A
holoenzymes. Identifying the mitochondrial substrates of the Bβ2-containing PP2A
holoenzyme is a goal of ongoing experiments, and should provide further insights into the
function of Bβ2 in neurons and possibly the etiology of SCA12.
We can only speculate as to the reasons why the brain expresses a phosphatase
that promotes cell death. Apoptosis seen in adult organs is considered a cellular defense
53
mechanism against cancer, and Bβ2 expression may help explain why postmitotic
neurons rarely give rise to brain tumors. Our neuronal survival studies suggest that Bβ2
is part of a survival pathway for rapidly eliminating neurons that have sustained
mitochondrial damage induced by toxic insult. Although we have identified a functional
role of Bβ2 as a regulator of neuronal survival, we cannot rule out the possibility that the
regulatory subunit may have other physiological roles in the neuron (See chapter V).
Lastly, our neuronal survival data raises the possibility that preventing the mitochondrial
translocation of Bβ2 by developing small anti-Bβ2 inhibitors (ie, intrabodies) may be an
attractive avenue for the treatment of brain injuries, stroke and neurodegenerative
diseases such as Parkinson’s disease (See Chapter V).
54
Figure 3 Identification of a novel splice-variant of PP2A/Bβ. A: schematic
representation of Bβ1 and Bβ2 generated by alternative splicing of the Bβ
gene (PPP2R2B). Bβ2 transcription is driven by an alternate promoter
upstream of exon 1.2. This alternate exon is then ligated to the common exon
2, skipping exon 1.1 that encodes the Bβ1 N-terminal tail. Alternative splicing
of the Bβ gene gives rise to other splice variants that include Bβ3 and Bβ4
which are exclusively expressed in higher vertebrates containing shorter
divergent N-terminal tails than Bβ1/2. B: structure diagram of B-family
regulatory subunits of PP2A. B-family regulatory subunits contain seven WDrepeats (numbered, component β-strands are indicated by a-d and grouped by
shading) and are predicted to fold into a seven-blade β-propeller. The region
of the molecule that interacts with the AC dimer was previously delineated by
site-directed mutagenesis [47].
55
Figure 4 The Bβ2 protein is expressed at lower levels than Bβ1. Protein phosphatases
were affinity-purified from rat brain lysates with microcystin-sepharose either
in the absence (microcystin-Sepharose pellet, MCP) or in the presence of 5
µM free microcystin-LR (MCP control). Samples from brain lysate (20 µg),
MCP, MCP control and lysates from COS-M6 cells (5 µg) transiently
expressing either FLAG-epitope tagged Bβ1 or Bβ2 (FLAG-Bβ1/2) were
immunoblotted with antibodies directed against the divergent N-termini of the
corresponding Bβ isoform. An immunoreactive band migrating close to the
predicted molecular weight of Bβ2 (52 kD) can be detected in the MCP lane,
but not in the MCP control lane. Relative levels of Bβ1 and Bβ2 were
determined by densitometry of the band in the MCP lane. Calculations were
adjusted for antibody titers and affinities by probing duplicate FLAG-Bβ1/2
lanes with specific and common antibodies (FLAG epitope directed, not
shown). This experiment was performed by Dr. Stefan Strack.
56
Figure 5
Immunocytochemistry of Bβ2 in the rat brain. A: PC6-3 lysates expressing
FLAG-Bβ1 in lane 1 or FLAG-Bβ2 in lane 2 were probed with a Bβ2 specific
monoclonal antibody (mAb). The N-terminus of Bβ2 has been shown to be
proteolytically processed in mitochondria giving rise to N-terminal fragments
[163]. Shown in lane 2, the Bβ2 mAb immunoreacts with full-length Bβ2 as
well with N-terminal and C-terminal fragments. B-D: rat brain sections (40
µm) were stained for Bβ2 using a Bβ2 mAb and visualized by
immunoperoxidase staining. B: in the cerebellum (mo=molecular layer,
gr=granular layer, Pu=Purkinje layer). C: in the hippocampus and dentate
gyrus. D: a close-up of a brain section of the dentate gyrus immunostained by
the Bβ2 mAb. White arrows point to patchy areas in the soma of neurons
stained by the Bβ2 mAB. Overexpression experiments in PC6-3 cells were
performed by Dr. Stefan Strack while immunocytochemistry experiments
were performed by Dr. Ronal Merrill.
57
Figure 6
In vitro characterization of PP2A holoenzymes containing Bβ isoforms. A:
FLAG-epitope tagged Bβ1, Bβ2, Bβ3, Bβ4 or a truncation mutant lacking the
divergent N-terminus (Bβ∆N) were transiently expressed in COS-M6 cells;
transfections with empty vector served as controls. FLAG immunoprecipitates
were probed for transfected B subunits and endogenous A and C subunits. B:
PP2A holoenzymes containing the indicated FLAG-tagged Bβ subunits were
immuno-isolated as in panel A and assayed for activity towards exogenous,
33
P-labeled substrates (PKA-phosphorylated myelin basic protein, MBP;
PKA-phosphorylated casein). Subcloning of Bβ1-4, and Bβ∆N was performed
by Vaibhavi Shah and phosphatase activities of these constructs were assayed
by Dr. Stefan Strack.
58
Figure 7 Mitochondrial targeting of Bβ2. A: Bβ1, Bβ2, Bβ3, Bβ4 or the N-terminal
deletion mutant Bβ∆N were fused at the C-terminus to green fluorescent
protein (GFP) and transiently expressed in PC6-3 cells (a subline of PC12
cells) for fluorescence imaging of live cells. B: confocal images of live PC6-3
cells expressing Bβ1-, and Bβ2-GFP fusion proteins (green) stained with
MitoTracker dye to label mitochondria (red). Colocalization of signals is
apparent as yellow color in the red-green channel-merged images.
Scale bars = 10 µm. Confocal imaging of GFP fusion proteins were performed
in collaboration with Vaibhavi Shah.
59
Figure 8
Subcellular localization studies of Bβ11-32 and Bβ2 1-35-GFP by confocal
microscopy. The first 32 and 35 residues including the divergent N-termini
and 11 residues of the common domain of Bβ1 and Bβ2, respectively, were
fused to the N-terminus of GFP (Bβ11-32, Bβ21-35, see diagram) and analyzed
for colocalization with mitochondria as in Fig. 7 B. Scale bars = 10 µm.
The data concerning the mechanism of mitochondrial association of the Nterminal tail of Bβ2 is presented in figures 23-26 and explained in the
discussion section of chapter III.
60
Figure 9 Cell stress induces mitochondrial translocation of Bβ2-GFP in PC6-3
cells. A: representative images that demonstrate the effects of tunicamycin on
mitochondrial translocation of Bβ2-GFP in PC6-3 cells (PC12). In the top
panel, cells treated with DMSO vehicle control show diffuse localization of
Bβ2-GFP. In the bottom panel, cells treated with 1.3 µM tunicamycin (LD50)
for 3 hrs. show colocalization of Bβ2-GFP with mitochondria. Scale bars= 10
µM. B: representative images that demonstrate the effects of rotenone on
mitochondrial translocation of Bβ2-GFP. In the left panel, representative PC63 cells treated with DMSO vehicle control showed diffuse localization of
Bβ2-GFP. Shown for the right panel, PC6-3 cells treated with 10µM rotenone
for 2 hrs. show colocalization of Bβ2-GFP with mitochondria. Scale bars= 10
µM. C: quantification of the effects of tunicamycin on mitochondrial
translocation of Bβ2-GFP at two different time points. Cells that ranked 3 or 4
on a subjective scale from 0-4 [163] were scored positive for mitochondrial
localization of Bβ2. D: quantification of mitochondrial translocation and
apoptosis of Bβ2-GFP for different time points in PC6-3 cells treated with
rotenone. Cells that ranked 3 or 4 on a subjective scale from 0-4 were scored
positive for mitochondrial localization of Bβ2 and translocation averages were
converted to a ratio of the total population of transfected cells. To score for
apoptosis, cells that contained a pyknotic or fragmented nuclei were scored as
apoptotic and converted to a ratio of the total transfected population.
61
Figure 10 Cell stress induces mitochondrial translocation of Bβ2-GFP in
hippocampal neurons. A: primary hippocampal neurons (15 DIV) infected
with lentivirus (FIV) expressing Bβ1 or Bβ2-GFP were treated for 24 hr. with
DMSO vehicle control or 2 µM rotenone, followed by live confocal imaging
of GFP and mitochondria stained with TMRM. While Bβ2-GFP shows mainly
diffuse localization under basal conditions (top right panel), cell stress
increases colocalization of Bβ2-GFP with mitochondria (lower right panel).
Scale bars = 10 µm. B: a representative confocal image showing TMRM
stained mitochondria and DRAQ5 stained nuclei from untransfected and Bβ2GFP expressing hippocampal neurons. For comparison purposes, the GFP
channel was not overlayed in this image to demonstrate differences in
mitochondrial uptake of TMRM. Note that the two hippocampal neurons that
express Bβ2-GFP show a marked decrease in mitochondrial uptake of TMRM
compared to mitochondria of an untransfected neuron. C: quantification of
mitochondrial translocation and apoptotic acitivity of Bβ2-GFP for different
time points in hippocampal neurons treated with a lethal dose of 2 µM
rotenone. Bβ2 expressing neurons were scored for mitochondrial translocation
using subjective scale from 0-4 and translocation averages were converted to a
ratio of the total population of transfected neurons. To score for apoptosis,
neurons that contained a pyknotic or fragmented nuclei were scored as
apoptotic and converted to a ratio of the total number of neurons analyzed.
pVETL constructs and lentiviruses that overexpress Bβ1/2-GFP were
generated by Dr. Xinchang Zhou and Dr. Ronald Merrill.
62
Figure 11 Inducible expression of Bβ2 is proapoptotic. A: tetracycline-inducible PC63 cell lines that express either Bβ1, Bβ2 or Bγ as indicated were treated for 3
days in the presence of vehicle (-) or doxycycline (+ Dox) and analyzed for
inducible expression of FLAG-epitope tagged B subunits and other PP2A
subunits. Immunoblotting with a pan-specific B subunit antibody (pan-B)
shows that induction increases total B subunit levels by about two-fold. B: the
indicated PC6-3 lines that inducibly express Bβ2 were treated for 4 days in
the absence or presence of Doxycycline and analyzed for cell viability by
Trypan Blue dye exclusion. C: quantification of survival following serumdeprivation. The indicated cell lines were grown in serum-containing media
for 3 days in the absence or presence of doxycycline as indicated, and cell
numbers were determined 0 and 24 hours after serum-withdrawal using the
MTS cell proliferation assay. The bar graphs show percent survival ± S.E.M.
in quadruplicate wells of a representative experiment. D: quantification of
apoptosis by nuclear morphology. The indicated Bβ1- and Bβ2-expressing
cell lines were treated and serum-deprived for 24 hours as in panel C. Fixed
cells were stained with the nuclear dye Hoechst 33342 and classified as
apoptotic if they exhibited nuclear condensation or fragmentation. Data from a
representative experiment are shown as percent apoptotic cells ± S.E.M. in 6
randomly selected microscopic fields with 50-200 cells/field. Significant
differences to control (-Dox): *p<0.05, **p<0.0001. The PC6-3 cell lines that
inducibly overexpress Bβ2 (457, 463) and Bγ (275) were generated and
characterized by Dr. Stefan Strack.
63
Figure 12 Bcl-2 inhibits the cell death promoting activity of Bβ2. PC6-3 cells were
transiently transfected with the indicated GFP fusion proteins. Three days
post-transfection, PC6-3 cells were serum starved for 24 hr., and GFP positive
cells were scored for apoptosis by the nuclear morphology or by the Annexin
V staining assays. Note that the proapoptotic activity of BB2-GFP is blocked
in the presence of Bcl-2. (50-100 cells were each analyzed for each
transfection condition in this representative experiment).
64
Figure 13 Bβ2’s proapoptotic activity requires incorporation into the PP2A
holoenzyme in PC6-3 cells. A: C-terminal GFP fusion proteins of the
indicated B-family subunits or GFP alone were transiently expressed in PC6-3
cells. 24 hours following serum removal, apoptotic nuclei of GFP-positive
cells were quantified as in Fig. 8 (D). B: Bβ2 wild-type (w.t.), RR168EE
mutant (RE), or the first 35 residues (1-35) of Bβ2 tagged at the C-terminus
with the FLAG epitope and GFP were expressed in COS-M6 cells,
immunoprecipitated with FLAG antibodies, and immunoblotted for
association with endogenous PP2A A and C subunits. The position of
molecular weight markers (in kD) is indicated on the left. Arrowheads point to
full-length Bβ2 (closed) and Bβ21-35 (open) GFP fusion proteins. C:
representative microscopic fields (top: GFP fluorescence; bottom: Hoechst
33342 nuclear stain) of PC6-3 cells transiently expressing wild-type or
RR168EE mutant Bβ2 after 24 hours without serum. The arrows point to two
apoptotic cells expressing wild-type Bβ2-GFP. D: the GFP fusion proteins
characterized in B or GFP alone were transiently expressed in PC6-3 cells and
scored 24 hours after serum withdrawal for apoptosis-promoting activity as in
Fig. 8D. Significant differences to GFP control: **p<0.0001. The Bβ2
(RR168EE) mutant was generated and characterized by Thomas Cribbs.
65
Figure 14 Bβ2 promotes apoptosis in hippocampal neurons and requires
incorporation into the PP2A heterotrimer and mitochondrial targeting.
C-terminal GFP fusion proteins of the indicated B-family subunits, cytosolic
or mitochondrial targeted PfARP, or mitochondrial targeted GFP (OMMGFP) were transiently expressed in hippocampal neurons (15 DIV) for 5 days.
Neurons were then fixed in 3.7% PF, immunostained for GFP and for
neurofilament to identify transfected neurons. The nuclei of neurons were
counterstained with Hoechst dye. The percentage of transfected neurons
containing apoptotic nuclei was quantified for each transfection condition (#
of experiments indicated, 85-90 neurons each; PfARP=Plasmodium
falciparum aspartate rich protein (PP2A inhibitor)). Significant differences to
OMM-GFP except where indicated by brackets: *p<.05, **p<.001,
***p<.0001. The following constructs were generated by Chris Barwacz:
OMM-GFP, Bβ2-RR168EE, -R6A, PfARP and OMM-PfARP.
66
Figure 15 RNA interference of Bβ2 is neuroprotective in primary hippocampal
neurons. A: dissociated hippocampal neurons (15 DIV) were co-transfected
with a luciferase fusion protein of Bβ2, with pSUPER plasmids expressing a
scrambled shRNA or shRNAs targeting Bβ2 and with Renilla luciferase as a
marker of transfection efficiency. After 3 days post-transfection, knock-down
of luciferase fusion proteins were assayed by the dual luciferase assay. B:
dissociated hippocampal neurons (15 DIV) were co-transfected with Bβ2
specific shRNAs or with scrambled shRNA and with β-galactosidase as an
expression marker. After three days post-transfection, neurons were treated
with 250µM glutamate for 10 minutes or with rotenone (20-60 nM) to induce
toxicity. Two days following toxic insult, neurons were fixed in 3.7% PF and
immunostained with β-galactosidase and NF antibodies to identify transfected
neurons. Note the extensive neurite degeneration in glutamate treated control
neurons as shown by the absence of NF staining but not in Bβ2 shRNA#1
expressing neurons. C, and D: quantification of total surviving neurons from
quadruplicate transfected wells pooled (means±SEM) from the indicated
number experiments. Significant differences to vector or scrambled shRNA
controls: * p<0.01, **p<0.0001. The Bβ2 shRNAs #1 and #3 were generated
and characterized for their ability to knock-down exogenous luciferase fusion
protein of Bβ2 by Lisa Gomez and Thomas Cribbs. The scrambled shRNA
was generated by Thomas Cribbs.
67
Figure 16 RNA interference of Bβ2 blocks apoptotic cell death of hippocampal
neurons. Hippocampal neurons were transfected with the indicated shRNA
plasmids and GFP as an expression marker. 5 days post-transfection,
hippocampal neurons were treated with rotenone (400nM) or DMSO vehicle
control for 2 days. Following toxic treatment, hippocampal cells were then
fixed in 3.7% PF, immunostained for GFP and counterstained with a nuclear
dye (Hoechst 33342). A: representative microscopic fields of hippocampal
neurons transfected with the indicated shRNAs. The white arrows point to
apoptotic nuclei while yellow arrows point to intact nuclei. Note that neurons
that express scrambled shRNA but not Bβ2 shRNAs contain apoptotic nuclei
after rotenone treatment. Scale bars: 20µm. B: the percentage of transfected
neurons containing apoptotic nuclei was quantified for each transfection
condition and statistically compared to scrambled shRNA (mean±SEM from
at least 2 experiments, 100-200 neurons each). Significant differences to
scrambled shRNA in rotenone conditions: *p<.05, **p<.005
68
Figure 17 Model figure that depicts the mechanism of apoptosis by PP2A/Bβ2. Bβ2
exists in equilibrium with cytosolic and mitochondrial compartments. Upon
toxic insult, Bβ2 rapidly translocates to mitochondria to dephosphorylate an
unknown substrate to promote apoptosis. The apoptotic activity of Bβ2 is
inhibited by Bcl-2, a well characterized anti-apoptotic protein. Characteristics
of Bβ2 mediated apoptosis in neurons include inversion of cell membrane
lipids, a decline in mitochondrial transmembrane potential and nuclear
fragmentation.
69
CHAPTER III
MECHANISMS OF PROTEIN PHOSPHATASE 2A HOLOENZYME
ASSEMBLY AND OUTER MITOCHONDRIAL MEMBRANE
TARGETING OF THE Bβ2 REGULATORY SUBUNIT
Abstract
Heterotrimeric Ser/Thr protein phosphatase 2A (PP2A) is composed of catalytic
(C), structural (A), and regulatory subunits (B). B-family regulatory subunits contain
seven WD-repeats predicted to fold into a β-propeller structure. In order to understand
the mechanism of holoenzyme association of B-family regulatory subunits, we carried
out mutagenesis of Bγ and analyzed for the ability of mutants to associate with the A and
C subunits. We found that small internal deletions compromised holoenzyme association
while small N- and C-terminal Bγ deletions did not. Charge-reversal mutagenesis of Bγ
identified a cluster of conserved critical residues in WD repeats 3 and 4. Moreover, we
identified pair of arginine residues in Bγ that form salt bridges with glutamate residues in
the A subunit. Overall, these studies provide a model by which the N- terminal tails of
B-family regulatory subunits face away from the presumed β-propeller, while the middle
third of the β-propeller associate with the A subunit of PP2A via multiple electrostatic
interactions.
In agreement with the holoenzyme association studies performed in Bγ, the
divergent N-terminus of Bβ2, a neuron-specific proapoptotic splice variant of the Bβ
gene, was found to be dispensable for holoenzyme association. Moreover, holoenzyme
association studies of Bβ2 unveiled an additional role of the presumed β-propeller
domain as described below. Bβ2 is localized to the outer mitochondrial membrane
(OMM) by a novel mechanism, combining a cryptic mitochondrial import signal with a
structural arrest domain. Scanning mutagenesis demonstrates that basic and hydrophobic
residues mediate mitochondrial association and the proapoptotic activity of Bβ2. When
70
fused to green fluorescent protein (GFP), the N terminus of Bβ2 acts as a cleavable
mitochondrial import signal. Surprisingly, full-length Bβ2 is not detectably cleaved and
is retained at the OMM. Mutations that open Bβ2’s C-terminal β-propeller facilitate
mitochondrial import, indicating that this rigid fold acts as a stop-transfer domain by
resisting the partial unfolding step prerequisite for matrix translocation. Since hybrids of
prototypical import and β-propeller domains recapitulate this behavior, we predict the
existence of other similarly localized proteins and a selection against highly stable protein
folds in the mitochondrial matrix.
Introduction
The β-propeller structure of WD repeat proteins
The WD repeat domain consists of 40 to about 67 amino acids that often end with
WD (Trp-Asp) residues. The family of WD repeat proteins comprise over more than ~2%
of the yeast proteome and over 600 proteins in humans. They include RNA-processing
complexes, transcriptional regulators, regulators of mitochondrial fission and fusion
(MFF), regulators of vesicle formation and cytoskeletal assembly (reviewed by [164,
165]). WD repeat proteins whose 3D structures have been solved by X-ray
crystallography include Ski8p [166], Gβ1 subunit of heterotrimeric G proteins [167] and
the p40 subunit of the arp2/3 actin filament branching complex (p40-ARC, [168]). These
proteins contain α-helical N-terminal extensions and a C-terminal β-propeller structure.
The β-propeller folds of these WD repeat proteins form a rigid, closed circular or
“doughnut-shaped” structure composed of seven β propeller blades that are radially
arranged around the center. For other WD repeat proteins whose crystal structures have
not been solved, there is indirect experimental evidence that demonstrates that WD repeat
proteins contain a globular and compact structure consistent with a β-propeller fold. For
instance, in vitro synthesis of six WD repeat proteins using a rabbit reticulocyte system
were demonstrated to fold into compact globular structures and are resistant to
71
trypsinolysis [169] However, not all β-propellers are WD repeat proteins. Non-WD
repeat β-propellers may contain from four to up to sixteen β-propeller blades (as shown
for collagenase and methylamine dehydrogenase) (reviewed by [165]). Remarkably, the
β-propeller structures of WD repeat and non-WD repeat proteins are very similar in that
their β-propeller blades can be superimposed to each other without significant deviation
of their α-carbon atoms. In WD repeat proteins, each WD repeat contributes to the three
innermost β-strands of a β-propeller blade (a-c) and the outermost β-strand of the
adjacent β-propeller blade (d). All β-propeller structures have a mechanism for closing
the ring structure. In the case of Gβ1 and p40 ARC, the β-propeller structure is closed by
a “Velcro-snap” closure mechanism in that the first β-strand of WD repeat 1 “locks” in
place with the last β-strand of the last WD repeat [167]. In others, a disulfide bond that
links the N-terminal and C-terminal β-propeller blades is responsible for closing the
circular structure as seen for haemopexin, a four bladed β-propeller protein [170]. It has
been suggested that abolishing the closure mechanism can cause significant instability
and opening of the ring structure of Gβ1 [167]. Although no enzymatic activity has been
associated with the WD repeat motif, the β-propeller structures of WD repeat proteins
form stable platforms that are favorable for mediating protein-protein interactions and
can form large complexes such as seen for p40-ARC and Gβ1 [164, 165, 168].
Structure of the PP2A heterotrimer
The crystal structure of the scaffolding Aα subunit of PP2A has been solved [29],
confirming a previous model based on secondary structure prediction and mutagenesis
studies [171]. The A subunit is a hook-shaped protein that is comprised of 15 imperfect
repeats, each about 40 amino acids long. Each of these HEAT repeats (named after
proteins that contain them: huntingtin, elongation factor, A subunit, and TOR kinase)
consists of two antiparallel, amphipathic α-helices. Intra-repeat loops form a continuous
ridge along the inside of the hook that interact with catalytic and regulatory subunits of
72
PP2A. Regulatory subunits and viral antigens bind to the 10 N-terminal repeats, whereas
the catalytic subunit of PP2A binds via repeats 11-15 [171, 172]. It has been
demonstrated by Ruediger et al., 1999 that different residues located in HEAT repeats 110 differentially bind to distinct families of regulatory subunits of PP2A and viral tumor
antigens. For instance, two glutamate residues (Glu100, Glu101) are critical for binding
all regulatory subunit families and viral tumor antigens while an aspartate residue
(Asp177) is critical for binding B’ but not B, B” family of regulatory subunits or viral
tumor antigens [173]. The PP2A C subunit is thought to have a roughly globular
structure similar to the related PP1 C subunit [156, 174].
Mammalian B-family regulatory PP2A subunits (Bα-δ) display a high degree of
sequence conservation (> 80% amino acid identity). PP2A regulatory subunit families
have distinct primary and secondary structures since secondary structure prediction
analyses suggest that B-family regulatory subunits are almost entirely composed of βsheets and turns, whereas the B’ and B’’ subunits are mostly α-helical. Moreover,
structural modeling analyses predict that the C-terminal domain of B-family of regulatory
subunits adopt a β-propeller structure similar to Gβ1 while the N-terminus fold into an
extended α-helix. [47].
Knowing how regulatory subunits fold and interact with the core PP2A dimer has
therapeutic implications for generating small molecular drugs targeted against specific
PP2A holoenzymes implicated in neurodegenerative diseases such as SCA12 and
Parkinson’s disease (See chapter 1). In the first part of this chapter, I in collaboration
with other people in the lab performed deletion and site-directed mutagenesis of Bγ in
combination with structure modeling to identify domains and amino acids that are critical
for associating with the A and C subunits of PP2A. By complementary charge-reversal
mutagenesis, we showed that a pair of adjacent arginines in Bγ critically interact with a
pair of adjacent glutamates in the Aα subunit. The mutagenesis studies performed in Bγ
not only provides a model for PP2A holoenzyme association but may shed light into how
73
B-family regulatory subunits regulate the catalytic activity of PP2A and engage with
interacting protein/structures. More importantly, the holoenzyme association studies
performed in Bγ enabled me to predict regions that are critical for holoenzyme
association and for the apoptotic activity of Bβ2, and to understand why the β-propeller
is a rigid fold that resists the partial unfolding step necessary for mitochondrial import.
Mechanisms of mitochondrial targeting
Proteins can be anchored to mitochondria by a variety of mechanisms. A recent
review on mitochondrial targeting has classified this process as either dependent or
independent of the mitochondrial import machinery, also termed TIM/TOM complex
[175]. N-terminal extensions of preproteins that bind to the TIM/TOM complexes include
the N-terminal domains of MAS70 and cytochrome oxidase VIII (COX8). The noncleavable N-terminal targeting domain of MAS70 is a signal anchor sequence that
interacts with the TOM20/22 receptors and is laterally released to the lipid bilayer of the
OMM [149]. The first 31 amino acids of COX8 constitute a cleavable N-terminal
extension that interacts with the TIM/TOM complex and promotes import of COX8 into
the matrix of mitochondria [176]. Proteins that are targeted to mitochondria independent
of the TIM/TOM complex contain N-terminal extensions that may directly anchor to
proteins of the OMM. This is exemplified by the N-terminus of hexokinase I which binds
to the voltage dependent anion channel (VDAC) of the mitochondrial porin [177]. Fatty
acid modification of cysteine residues represents another mechanism by which proteins
anchor the OMM independent of the TIM/TOM complex. In the case of the small
GTPase protein Rab32, prenylation of two cysteines were shown to be required for
binding to the OMM [178]. Another mitochondrial targeting mechanism independent of
the TIM/TOM complex involves a C-terminal anchor domain that binds the lipid bilayer
of the OMM. This group of C-terminal anchored proteins include Bcl-2, TOM5 of the
import receptor, cytochrome b5 and OMP5 [179-182].
74
Mitochondrial import
The mitochondrial import machinery consists of outer and inner membrane
translocase complexes that are responsible for sorting nuclear encoded preproteins to
distinct compartments within the mitochondria. The TOM complex consists of at least
seven proteins which include the cytosolic components TOM20, TOM22, TOM5 and
TOM40 which forms the general import pore (GIP). Mitochondrial preproteins
containing cleavable mitochondrial targeting sequences first interact with TOM 20 and
TOM22 receptors. TOM22 interacts with basic residues while TOM20 interacts with
stretches of hydrophobic residues located at the N-terminal targeting sequences of
preproteins and guide preproteins to the β-barrel pore complex of the OMM (TOM40).
With the help of chaperone proteins (TIM9-10), preproteins destined for import are
partially unfolded, and are translocated across the narrow opening of the channel
(TOM40) to interact with components of the TIM complex. At this point, preproteins can
be destined either to the intermembrane space by interacting with the TIM22 complex or
shuttled directly to the matrix via TIM23. Preproteins that interact with the TIM23
complex are “pumped” into the matrix by a mechanism that requires heat shock protein
70 (mtHsp70), the transmembrane potential and expenditure of ATP. Once in the matrix,
the N-terminal presequences of preprotein are then cleaved by matrix processing
peptidases (MPPs) (reviewed by [175]).
In the second half of this chapter, we report that Bβ2 is localized to the OMM by
a previously undescribed mechanism. The N-terminal 26 residues of Bβ2 are sufficient to
import GFP into the mitochondrial matrix, with conserved basic and hydrophobic amino
acids playing critical roles. Full-length Bβ2 interacts with the mitochondrial import
complex, but is arrested at the OMM, unless its β-propeller domain is unraveled by small
deletions that disrupt the “Velcro” closure mechanism between the first and last βpropeller blade. A chimeric protein consisting of the matrix import sequence of
cytochrome oxidase subunit VIII (COX8) and the prototypical β-propeller of Gβ5
75
behaves identically, demonstrating that unfolding-resistant translocase targeting may be a
general mechanism by which proteins with thermodynamically stable tertiary structures
can be directed to the surface of mitochondria. Mitochondrial import of essential proteins
appears not be compromised by overexpression of proteins targeted in this manner,
judging by a lack of effect on mitochondrial membrane potential and cell viability. Our
results also constrain the types of tertiary folds that can be adopted by nuclear-encoded
proteins destined for the mitochondrial matrix.
Experimental Procedure
Structure modeling of Bγ
An amino acid alignment of B-family regulatory subunits from different phyla
was submitted to the 3D-PSSM protein fold recognition web server [183], which
generated a first round model based on the structure of the Gβ1 subunit of heterotrimeric
G proteins [167]. This model was globally and locally optimized for bond lengths, angles
and torsions of side chains using the steepest decent algorithm of the Swiss PDV Viewer
software[184]. In addition, breaks in the Cα trace were ligated with a cutoff value of 3.0
Å and missing hydrogen atoms were added to the model. Ribbon diagrams and surface
representations of the optimized B-subunit model were rendered, annotated and analyzed
using Rasmol and Swiss PDB Viewer software.
Mutagenesis of Bγ
The rat cDNA for Bγ was isolated by reverse transcriptase-polymerase chain
reaction (RT-PCR) from rat brain total RNA (Access RT-PCR kit, Promega, Madison,
WI), subcloned into a pcDNA3.1 mammalian expression vector under control of the
cytomegalovirus (CMV) promoter and Flag-epitope tagged at the N-terminus by PCR.
The Bγ ∆26-38 and ∆379-447 mutants were generated by restriction digest of the
wild-type plasmid with uniquely cutting restriction enzymes, followed by fill-in and
76
recircularization reactions. The Bγ ∆1-20 N-terminal truncation mutant was generated by
PCR amplification of the coding sequence with nested primers encoding the Flag-tag and
amino acids 21-25 of Bγ.
Generation of internal deletion mutants involved PCR-amplification of two halves
of the Bγ-cDNA containing plasmid, one extending from the 5’ end of the deletion to
approximately half-way around the plasmid, the other extending from the 3’ deletion
boundary to the same site in the vector backbone in the opposite direction. Reverse
primers annealing to sequences 5’ of the deletion and forward primers annealing to 3’
deletion boundaries additionally included a unique SacII site encoding a neutral “stuffer”
sequence (Ala, Ala-Ala, or Ala-Ala-Gly), and complementary forward and reverse
primers annealing to the plasmid backbone introduced a unique AscI site. PCR-generated
plasmid halves were digested with SacII and AscI and ligated to produce the complete
plasmid carrying the deletion.
The Bγ ∆434-447 and ∆440-447 C-terminal truncation mutants were generated by
site-directed mutagenesis of residues 434 and 440, respectively, to termination codons.
Site-directed mutagenesis was carried out by whole-plasmid synthesis with
complementary primers harboring mutations utilizing Pfu Turbo polymerase (Stratagene,
La Jolla, CA), followed by destruction of the template plasmid by digestion with DpnI.
All mutations were verified by automated sequencing. All Aα subunit plasmids have
been previously described [173].
Mutagenesis and generation of Bβ2 fusion protein
constructs
All deletion and amino acid substitution mutations in Bβ2 were generated by
polymerase chain reaction (PCR) using mutagenic primers harboring restriction sites.
PCR products were ligated into unique sites flanking the Bβ2 N terminus (residues 1-35)
in the pEGFP-N1 vector (BD Biosciences). Most of the mutations incorporated to the
77
Bβ21-35-GFP fusion protein were transferred into full-length Bβ2-GFP fusion proteins by
restriction digest and ligation. N-terminal fusions of GFP with the import sequence of
human COX8 (residues 1-31) [176] and the OMM anchor sequence of yeast MAS70
(residues 1-29) [149] were similarly generated by PCR using nested primers encoding the
targeting sequences.
The cDNA of the Gβ5 short isoform [185] was kindly provided by Rory Fisher
(University of Iowa). The full-length coding sequence of Gβ5, the Gβ5 coding sequence
lacking the last 8 amino acids (Gβ5∆C), and the β-propeller domain of Bβ2 (residues 28447) was inserted between COX81-31 and GFP following PCR with primers carrying
unique restriction sites to generate COX8-Gβ5-GFP, COX8-Gβ5∆C-GFP and COX8-BβGFP, respectively. MAS70-Bβ was created by insertion of Bβ228-447 between MAS701-29
and GFP. Internal (∆27-34) and C-terminal deletions (1-311, 1-437) of Bβ2 with and
without N-terminal alanine substitutions (K2A, IL18AA) were created similarly.
Cell culture
COS-M6 cells [140] were cultured in DMEM containing 10% fetal bovine serum
(FBS) and 4.5 g/L glucose and seeded into 6-well plates. PC6-3 cells were cultured in
RPMI 1640 containing 5% FBS and 10% horse serum (HS) in 2-well chambered slides at
a density of 135,000 cells/well or in 12 well plates at a density of 90,000 cells/well.
Transfection of cDNAs
COSM6 cells seeded into 6 well plates were transfected with 2 µg plasmid DNA
using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). For Bγ holoenzyme association
studies, Aα and Bγ subunits plasmids were co-transfected at 1:1 mass ratios. For live
confocal imaging or IP of N-terminal or full-length Bβ2-GFP mutants, PC6-3 cells
seeded on collagen coated 2-well chambered slides were transfected with 1 µg plasmid
DNAs mixed with 0.8% Lipofectamine 2000. Three days post-transfection, live confocal
imaging or IP of Bβ2 and Bγ mutants was performed as described below.
78
Antibodies
The “old” Bβ2 N terminus-directed polyclonal antibody was previously described
[17]. Rabbit anti-FLAG tag, mouse anti-FLAG tag (M2) conjugated to agarose, mouse
anti-EE tag, mouse anti-PP2A catalytic subunit and rabbit anti-GFP antibodies were
purchased from Affinity Bioreagents (Golden, CO), Sigma (St. Louis, MO), Babco
(Richmond, CA) and BD Biosciences (San Jose, CA), respectively.
Confocal microscopy
PC6-3 cells were cultured in 2-well chambered slides (Nunc) and transfected
with deletion and alanine-substitution mutants in the context of Bβ21-35-GFP using
Lipofectamine 2000 (BD Biosciences) as described [17]. After 2 days, cells were stained
with 100 nM MitoTracker dye (Molecular Probes) and imaged “live” with a Zeiss LSM
510 laser scanning confocal microscope at the University of Iowa Central Microscopy
Facility.
In order to quantify colocalization of N-terminal Bβ2 mutants with mitochondria,
7-10 randomly selected images each containing 4-8 transfected cells were analyzed blind
to the identity of the mutant. Each transfected cell was assigned a GFP/mitochondria
colocalization score from 0 to 4 (0 = mutual exclusion; 4 = perfect overlap [145]).
Mitochondrial membrane potential assay
PC6-3 cells cultured in 2-well chambered slides were loaded three days posttransfection with 50 nM of the potential-sensitive dye tetramethylrhodamine methyl ester
(TMRM, Molecular Probes) for 30 minutes at 37°C. Red and green channel confocal
images were acquired, and TMRM fluorescence in transfected cells was quantified using
NIH Image software, dividing the average pixel intensity of the mitochondria-containing
cytosol by the background intensity measured within the nucleus [186].
79
Apoptosis assays
Apoptosis of serum-starved PC6-3 cells was assessed by nuclear morphology as
previously described [17] with minor modifications. Two days post-transfection, cells
were washed three times with RPMI 1640 and cultured for an additional 24 h in the
absence of serum. In order to include floating cells in the analyses, the chambered slides
were centrifuged for 2 min at 12,000 x g prior to fixation of cells with 3.6%
paraformaldehyde and staining with 1 µg/ml of the DNA dye Hoechst 33342 (Sigma).
Immunoprecipitation of Bγ mutants
After 36-48 hours postransfection, COSM6 cells transfected with Bγ mutants
were rinsed once with PBS and lysed in 250 µl/well IP buffer (1% Triton X-100, 150 mM
NaCl, 20 mM Tris pH 7.5, 1 mM EDTA, 1 mM EGTA, 1 mM β-glycerolphosphate, 1
mM Na3V04, 1 mM Na4P2O7, 1 µM microcystin-LR, 1 mM phenylmethylsulfonyl
fluoride, 1 µg/ml leupeptin, 1 mM benzamidine) and sonicated for 2 seconds at low
intensity with a probe tip sonicator. Debris was pelleted (20,000 x g, 15 minutes) and
Flag-tagged Bγ subunits were immunoprecipitated from the cleared lysate with 6 µl antiFlag-tag antibody (M2) conjugated to agarose (Sigma-Aldrich, St. Louis, MO) by endover-end rotation at 4oC for 3-16 hrs. In some experiments, 200 µg/ml Flag-epitope
peptide was added to the cleared lysate as a specificity control. Immunoprecipitates were
washed with 6-8 ml IP lysis buffer and solubilized in SDS-sample buffer for immunoblot
analysis using the following antibodies: rabbit anti-Flag tag (Affinity Bioreagents,
Golden, CO), mouse anti-EE tag (Babco, Richmond, CA), mouse anti-PP2A catalytic
subunit (BD Pharmingen, San Diego, CA). Blots were processed for chemiluminescence
detection (Pierce SuperSignal, Pierce, New York, NY) and digital images were captured
on a Kodak Imaging station 440. Signal intensities were quantified by digital
densitometry using NIH image software (rsb.info.nih.gov/nih-image/).
80
Proteolytic processing assay of Bβ2 mutants
PC6-3 cells transiently transfected with FLAG-GFP fusion proteins using
Lipofectamine 2000 were lysed 3 d later in immunoprecipitation buffer containing 1%
(v/v) Triton X-100, 0.5% (w/v) deoxycholate, 150 mM NaCl, 20 mM Tris pH 7.5, 1 mM
EDTA, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 1 mM
benzamidine. Lysates were briefly sonicated with a probe tip sonicator, cleared by
centrifugation (20,000 x g, 15 min), and expressed proteins were immunoprecipitated
with agarose-conjugated anti FLAG-tag antibody followed by immunoblotting for GFP.
Digital densitometry was performed using NIH Image software, dividing the band
intensity of the major proteolytic fragment by total GFP immunoreactivity.
Mitochondrial import assay
Native PC6-3 cells or a PC6-3 cell line that inducibly overexpresses Bβ2 with a
C-terminal tandem affinity purification tag [187] were transfected with FLAG-GFP
fusion proteins, and, in the case of the inducible cell line, treated with 1 µg/ml of
doxycycline. Three days after transfection and/or induction, cells were disrupted by N2
cavitation (1200 psi, 15 min) in disruption buffer consisting of 250 mM sucrose, 20 mM
KCl, 10 mM HEPES pH 7.4, 2 mM EDTA, 2mM EGTA. Nuclei and unbroken cells were
removed by two consecutive centrifugations (5 min., 800 x g). The resulting supernatant
was centrifuged at high speed (15 min, 20,000 x g) to obtain a crude mitochondrial pellet,
which was washed once in disruption buffer. Resuspended mitochondrial fractions were
incubated with increasing concentrations of trypsin in the presence or absence of 1%
(v/v) Triton X-100 in disruption buffer for 25 minutes at 22°C with intermittent shaking.
The reactions were stopped by adding SDS-sample buffer. In some experiments, cell
homogenates were directly digested with trypsin ± Triton X-100 without prior isolation
of mitochondria. In this case, reactions were stopped by addition of immunoprecipitation
buffer supplemented with 0.25 mg/ml soybean trypsin inhibitor, and fusion proteins were
81
isolated with FLAG-tag directed antibodies conjugated to agarose. Mitochondrial
fractions or immunoprecipitates were immunoblotted for GFP.
Results
Structure prediction of PP2A B-family regulatory subunits
As has been noted previously [188] and as previously described in the
introduction of this chapter, B-family regulatory subunits contain several degenerate WD
repeats (4 to 7 depending on isoform and motif search threshold). WD (also called WD40
or Gβ) repeats are loosely defined, ~40 amino acid sequence motifs that often end with
the tryptophan-aspartate (WD) dipeptide [164]. The amino acid sequence of Bγ, a
representative member of the B subunit family, aligned by WD repeat motifs is shown in
Fig. 18A. Seven degenerate WD repeats are separated by regions of 13 to 46 residues in
length (c-d loops).
WD repeat containing proteins whose 3D structure has been solved to date are the
Gβ1 subunit of heterotrimeric G proteins [167], ski8p [166], and the p40 subunit of the
arp2/3 actin filament branching complex (p40-ARC, [168]). These proteins fold into a 7bladed β-propeller, a toroid structure in which 7 twisted, antiparallel β-sheets are radially
arranged around a common center. Each WD repeat contributes the outer (d) β-strand of
one propeller blade and the inner three β-strands (a-c) of the next propeller blade (Fig.
18B). This phase-shift of sequence and structural motifs allows for closure of the torus by
a “Velcro” mechanism [164]. Sequences preceding the first WD repeat and trailing the
last repeat may protrude from the core toroid (Fig. 18B).
Three web-based threading protein fold prediction algorithms (3D-PSSM[183];
FUGUE [189]]; 123D (http://genomic.sanger.ac.uk/123D/123D.html)) identified Gβ1 as
the closest structural homolog of PP2A B-family regulatory subunits, despite low
sequence similarity (~15% identity). The structure of B-family regulatory subunits was
modeled based on the Gβ1 crystal structure (see experimental procedures). A ribbon
82
diagram of this model shows the 7-bladed β-propeller fold characteristic of WD repeat
containing proteins (Fig. 18C). Since PP2A B-family regulatory subunits are larger than
Gβ1, portions of the larger loops connecting WD repeats are missing from the model.
Deletion mutagenesis of Bγ
In order to define regions and residues in B-family regulatory subunits critical for
association with the AC core dimer, deletion and site-directed mutagenesis of the coding
sequence of Bγ were carried out in COSM6 cells. For technical reasons, we chose to
mutagenise Bγ in our holoenzyme association studies since transient transfection of this
regulatory subunit gave the highest expression levels compared to other B-family
regulatory subunits in COSM6 cells (not shown). To this end, mutant N-terminal Flag
epitope tag Bγ cDNAs were transiently expressed in COS-M6 cells and co-expressed
with Aα subunit tagged with a C-terminal EE epitope [173]. Flag-Bγ mutants were
immunoprecipitated, and in vivo incorporation into the PP2A heterotrimer was assayed
by blotting Bγ immunoprecipitates for transfected Aα and endogenous C subunits. The
ability of Bγ mutants to associate with the core enzyme was quantified by densitometry
as the ratio of C to Bγ subunit bands in the same lane. In general, we found that mutating
Bγ affected Aα and C subunit binding to similar degrees, supporting the notion that
regulatory subunits interact with a structural unit of A and C subunits. A summary of the
effects of Bγ deletion and truncation mutants is demonstrated in Fig. 19A. B subunit
family members differ considerably in their first 20-30 residues. Deletion of the variable
20 N-terminal amino acids of Bγ (∆1-20) did not compromise binding to the A and C
subunit (Fig. 19B), consistent with a role of these residues in mediating isoform-specific
functions. This deletion extends into the predicted first (d) β-strand of WD repeat 1,
which, according to the crystal structures of Gβ1 and p40-ARC, is critical for closure of
the β-propeller core by interacting with the c-strand of WD repeat 7 (see Fig. 18B). It is
83
conceivable that the Flag epitope tag can substitute for the 5 residues deleted from WD
repeat 1; alternatively, the boundaries of this structural motif in Bγ may require revision.
In a similar manner, truncating the 8 amino acids of the C-terminus of Bγ that
follow the last WD repeat in Bγ (∆440-447) had no effect on holoenzyme association.
However, extending the truncation by just 6 amino acids (∆434-447) to include the
predicted c-strand of WD repeat 7 caused an almost complete loss of A and C subunit
binding. Thus, residues 434-439 are required for holoenzyme association, presumably
because they interact with N-terminal residues in WD repeat 1 that form the Velcro patch
required for the closure of the toroid structure of Bγ.
Four internal deletions throughout the Bγ protein ranging from 12 to 32 residues
in length completely abrogated co-immunoprecipitation of A and C subunits (3 to 7% of
wild-type); only the ∆381-401 deletion displayed close to wild-type binding activity (Fig.
19B). Three of these critical deletions (∆128-156, ∆259-270, ∆370-401) are predicted to
affect surface-exposed loops connecting WD repeats, whereas ∆26-38 deletes a portion of
WD repeat 1 predicted to be buried in the protein.
Charge-reversal mutagenesis
Site-directed mutagenesis was carried out to delineate specific sites of
holoenzyme interaction. All Bγ residues that were mutated are perfectly conserved in
other mammalian B-family isoforms, as well as their orthologs in worms, fruit flies, and
yeast. Carrying out similar mutagenesis experiments with the Aα subunit, we had
previously identified charged residues in HEAT repeats 3 (E100, E101) and 5 (R183)
important for binding to regulatory subunits and viral tumor antigens [173]. Hence, we
focused the Bγ mutagenesis on charge-reversal of basic and acidic residues with the goal
of identifying electrostatic interactions with the Aα subunit.
Three acidic-to-basic mutations of conserved residues in the N-terminal third of
Bγ (E66R, EE89RR, D112K) did not have an effect on holoenzyme association (Fig. 20).
84
In contrast, four Bγ mutants in WD repeat 3 (RR165EE, D184K, E186R, DD192RR) and
two mutants in the loop connecting WD repeat 3 and 4 (D212K, IK213EE) displayed
severely reduced binding to A and C subunits (between 2 and 12% of wild-type).
Mapping to WD repeat 4, Bγ mutant ED219RR incorporated into the PP2A holoenzyme
normally, whereas E223R was defective. Since Bγ ∆259-270 was binding-incompetent
(Fig. 19B), we tested the effect of mutating all acidic residues in this region. Bγ D259R
did not bind to the AC dimer, while Bγ EE266RR, E269R, and D270R had little or no
effect on the ability of Bγ to associate with A and C subunits. Lastly, the E343R mutation
in WD repeat 6 had an intermediate effect on Bγ’s ability to incorporate into the PP2A
heterotrimer (40% residual binding of the C subunit, Fig. 20).
The results of the deletion and site-directed mutagenesis experiments are
summarized in Fig. 21. Bγ mutants were classified as critical or non-critical depending on
the amount of co-immunoprecipitated C subunit (< 15% and >= 40% of wild-type,
respectively). Most critical amino acid substitutions cluster in the middle of the molecule
(165-259) encompassing WD repeats 3 and 4.
Identification of interacting residues
Evolutionarily conserved and surface-exposed, charged residues of Bγ were
speculated to interact with residues of opposite charge in Aα that were previously
identified as critical for regulatory subunit association (E100, E101, R183, [173]).
Consequently, charge-reversal mutants of the Aα subunit (EE100RR, R183E) were coexpressed with all opposite charge-reversal mutants of the Bγ subunit and tested for
complementation, i.e. restoration of holoenzyme assembly by co-immunoprecipitation.
None of the acidic-to-basic mutants of Bγ were able to associate with the basic-to-acidic
mutant Aα R183E (not shown). There are three potential reasons for this: 1) the
interacting residues in Bγ were not mutated, 2) incorporating mutations in Aα or Bγ,
while potentially affecting interacting residues, may cause structural perturbations that
85
may misalign other amino acids critical for interaction, 3) Aα R183 may not directly
interact with regulatory subunits.
However, when we paired Aα EE100RR with Bγ RR165EE, we observed binding
that was comparable to wild-type subunits (Fig. 22). Bγ RR165EE was unable to coimmunoprecipitate another binding-defective, acidic-to-basic Aα mutant, DW139RR (not
shown), demonstrating that the observed complementation is not a consequence of
altering the overall charge of the proteins.
The holoenzyme association studies performed in Bγ provided valuable insight in
to the mechanism by which B-family regulatory subunits associate with the A subunit
and C subunits of PP2A. Furthermore, the holoenzyme association studies performed in
Bγ enabled us to predict regions/residues of Bβ2 that are critical for binding the A and C
subunits of PP2A and for OMM retention. Structure and function analyses presented
below describe a mechanism by which the N-terminus of Bβ2 associates with
mitochondria, identify regions that are critical for the proapoptotic activity of the
regulatory subunit and updates a previous model that describes the mechanism by which
Bβ2 promotes apoptosis.
Determinants of mitochondrial localization of Bβ2
A fusion protein of the first 35 residues of Bβ2 to the N terminus of GFP (Bβ2135-GFP)
localizes to mitochondria [17]. To further delineate critical regions for
mitochondrial interaction, a systematic deletion and site-directed mutagenesis screen of
the Bβ2 N terminus was carried out. The first 26 amino acids of Bβ2, which include the
alternatively spliced region plus two amino acids shared with Bβ1, were sufficient to
accurately localize GFP to MitoTracker-labeled mitochondria of transiently transfected
PC6-3 cells. Representative confocal images are shown in Fig. 23A and Fig. 24A, and
quantification of GFP and MitoTracker colocalization from three experiments is graphed
in Fig. 23B and in Fig. 24B. Bβ21-19-GFP displayed partial mitochondrial targeting, while
86
an additional truncation by five amino acids resulted in localization indistinguishable
from GFP alone. Deleting residues 1-4 or 3-5 also abrogated colocalization with
mitochondria. (Fig. 23A, and B). These results suggest that targeting information is
distributed over the length of Bβ2’s divergent sequence.
An alignment of Bβ2 orthologs from rainbow trout (GenBank accession number
CA376753) and mammals (e.g., AY251277) was used as a guide for an alanine scan of
the N terminus (Fig. 24B). I found that individual mutations of two conserved basic
residues (K2A, R6A) greatly reduced the mitochondrial localization score of the GFPfusion proteins (Fig. 24A, and B). Neutralizing a more C-terminal arginine (R13A), on
the other hand, had relatively modest consequences (Fig. 24B). Significantly, alanine or
serine substitutions of the conserved Cys3 showed wild-type colocalization scores, ruling
out a possible palmitoylation of this residue as contributing to membrane anchoring.
Bβ2’s N terminus is relatively hydrophobic (10 of 24 residues). Replacing single
hydrophobic residues with alanine (F4A, Y7A, L8A, Y10A, I11A, F12A) had little if any
effect on mitochondrial localization (Fig. 24B). However, combining mutations of two or
three hydrophobic residues (YL7AA, YIF10AAA, IL18AA) strongly impaired
mitochondrial association (Fig. 24B). The T17A and LS19AA substitutions also
confounded targeting significantly. A triple alanine substitution of residues 3-5
(CFS3AAA) showed only marginally worse targeting than the individual alanine
substitutions, but deleting this region altogether (∆3-5) reduced localization to levels of
GFP alone (Fig. 23A, and B). The colocalization results were also confirmed by
subcellular fractionation, showing that mutating Lys2, but not Cys3 to Ala significantly
reduced co-purification of Bβ21-35-GFP with mitochondria (Fig. 26A). In aggregate,
these experiments reveal the importance of properly spaced basic residues (Lys2 and
Arg6), as well as polar and multiple hydrophobic residues.
87
Mitochondrial association determines proapoptotic
activity
We have previously shown that transient or inducible expression of Bβ2
potentiates apoptosis when PC6-3 cells are deprived of serum [17]. Bβ1, and a Bβ2
mutant unable to recruit the rest of the PP2A holoenzyme were inactive in this assay. To
assess whether the apoptotic activity of Bβ2 depends on its ability to localize to
mitochondria, we introduced several of the alanine substitutions described above into the
full-length Bβ2 coding sequence fused to the N terminus of GFP. All Bβ2 mutants could
be transiently expressed to levels similar to the endogenous Bα subunit in PC6-3 cells
(Fig. 25A). 24 hrs following serum withdrawal, nuclei of transfected cells were scored
for apoptotic morphology (Fig. 25B). When apoptotic sensitization of each mutant is
plotted against its mitochondrial localization score, a positive correlation is apparent
(R=0.74, Fig. 25B). In hippocampal neurons, only one mutant was tested to confirm the
premise that mitochondria localization is required for apoptosis induction by Bβ2. The
Bβ2 R6A mutant which had a low mitochondrial localization score showed reduced
proapoptotic activity in hippocampal neurons (Fig.14). Among the 10 mutants analyzed
in PC6-3 cells the only exception to this rule was Bβ2 CFS3AAA, which localized fairly
well to mitochondria, but had no detectable proapoptotic activity. Therefore,
mitochondrial association is prerequisite for the ability of Bβ2-containing PP2A
holoenzymes to promote apoptosis.
The Bβ2 N terminus is a cryptic import signal
Its net positive charge (+4), as well as the observation that basic and hydrophobic
residues are important for mitochondrial targeting led us to speculate that the N-terminal
tail of Bβ2 is a matrix import signal. Most nuclear-encoded proteins destined for the
mitochondrial matrix contain cleavable N terminal presequences characterized by basic
and hydrophobic residues [175, 190]. These presequences are initially recognized by two
88
receptor components of the TOM (translocase of outer membrane) complex, TOM22 and
TOM20, resulting in transfer of the preprotein through the pore-forming subunit TOM40
and the TIM (translocase of inner membrane) complex into the matrix, where
presequences are cleaved by the matrix processing peptidase (MPP) (Reviewed in the
introduction of this chapter and by [175]). Import is driven by the strongly negative
membrane potential of mitochondria acting on the positively charged presequence, and
by ATP hydrolysis catalyzed by matrix heat shock protein 70 (mtHSP70) to partially
unfold the imported protein (reviewed by [175]).
Indeed, the MITOPROT signal prediction algorithm [191] assigns a high score to
the Bβ2 N terminus and predicts cleavage by MPP after Pro15. Consistent with this
prediction, we noted that transiently expressed Bβ21-35-GFP migrates as a doublet of 36
and 34 kDa, and that the faster migrating band is enriched in a mitochondrial fraction.
Only the upper band that migrates close to the predicted size of the fusion protein is
recognized by an antibody raised against the extreme N terminus of Bβ2 (Fig. 26A).
Since the Bβ2 N terminus is devoid of internal translation start sites, these data strongly
suggest that the lower band is the product of N terminal cleavage by MPP. Arguing
against non-specific degradation by cytosolic proteases, we observed that only mutants of
Bβ21-35-GFP that scored high for mitochondrial colocalization underwent significant
proteolytic processing (Fig. 26A-C). Quantifying N-terminal proteolysis, we found a
strong correlation to mitochondrial targeting of the Bβ2 mutants (R=0.91, Fig. 26C).
Positive and negative controls for signal peptide processing by MPP included the import
signal of COX8, and the OMM anchor of MAS70, respectively (Fig. 26B, and C).
Full-length Bβ2 is arrested at the OMM
In contrast to Bβ21-35-GFP, full-length Bβ2 consistently migrates as a single
protein species that is immunoreactive with both N- and C-terminal directed antibodies
[17]. To resolve this conundrum, we carried out trypsin protection assays to determine
89
the submitochondrial localization of ectopically expressed fusion proteins. Mitochondrial
fractions were treated with increasing concentrations of trypsin in the presence or absence
of Triton X-100. The processed (34 kD) form of Bβ21-35-GFP is completely resistant to
trypsinolysis, unless mitochondria are lysed by the detergent. This result confirms that the
Bβ2 N terminus is a mitochondrial import signal. However, even without detergent
permeabilization, trypsin degrades the full-length Bβ2 protein as readily as OMManchored MAS70-GFP (Fig. 27). Thus, Bβ2 is targeted to mitochondria via an Nterminal import signal, but is somehow prevented from entering the matrix.
The β-propeller resists the unfolding step of import
OMM-anchored proteins commonly contain hydrophobic “stop-transfer”
sequences C terminal to an import signal, resulting in lateral release from the TOM
complex into the outer membrane [175]. A second, unusual form of anchoring was
recently reported for a population of amyloid precursor protein (APP), a protein mutated
in Alzheimer’s disease. APP arrests in the TIM/TOM complex by virtue of the
mitochondrial membrane potential repelling an acidic stop-transfer sequence [192].
To identify the stop-transfer signal in Bβ2, we created a series of C-terminal
truncations fused to the N terminus of GFP (Fig. 28A). Neither wild-type nor the
monomeric Bβ2 mutant RR168EE were detectably processed in transiently transfected
PC6-3 cells, indicating that PP2A holoenzyme association is not necessary for OMM
arrest. In contrast, deleting the two C-terminal β-propeller blades (Bβ21-311) or just the
last 10 amino acids (Bβ21-437) resulted in formation of a cleavage product with a size
consistent with N-terminal processing by MPP (Fig. 28B). The yield of these presumed
matrix-associated Bβ2 species was relatively low, which is consistent with large cytosolic
pools of the near full-length proteins ([17] and data not shown).
The last 10 residues of Bβ2 are neither particularly hydrophobic nor acidic.
However, according to a tertiary structure model based on Gβ [47], their deletion
90
removes the last β-strand of WD repeat 7, which interacts with the first β-strand of WD
repeat 1 to close the β-propeller by a so-called Velcro mechanism [164]. The import
channel formed by TOM40 can only accommodate the width of a single α-helix [175].
Combining these two observations, we surmised that the presumed structurally rigid βpropeller fold of Bβ2 may represent a stop-transfer domain, which resists the partial
unfolding required for translocation through the narrow import pore. To test this
hypothesis, we removed 8 amino acids including the Velcro patch of WD repeat 1
(Bβ2∆27-34, Fig. 28A). Similar N- and C-terminal Velcro patch deletions cause
dissociation of the related Bγ subunit from the PP2A/A and C subunit, consistent with
unraveling of the β-propeller torus [47]. Bβ2∆27-34 was processed to the same extent as
Bβ21-437 (Fig. 28B). The proteolysis of both Velcro patch mutants is specific and depends
on a functional import signal, because no cleavage products were observed for Bβ21-437
K2A and Bβ2∆27-34 IL18AA (Fig. 28B). These experiments show that an intact βpropeller structure is necessary for OMM retention of Bβ2.
A prototypical import signal-β-propeller fusion protein
recapitulates OMM-targeting of Bβ2
To further support this model and to generalize it to proteins other than Bβ2, we
attached the well-characterized mitochondrial import sequence of COX8 to a prototypical
β-propeller protein, the Gβ5 heterotrimer G protein subunit. Equivalent to Bβ21-437, we
also engineered a protein in which the Gβ5 β-propeller is destabilized by truncation of 8
residues (COX8-Gβ5∆C, Fig. 29A). Both GFP-tagged proteins faithfully localized to
mitochondria when expressed in PC6-3 cells (shown for COX8-Gβ5-GFP in Fig. 29B).
However, only the Velcro-disrupted β-propeller mutant showed cleavage of the COX8
import sequence, indicative of quantitative OMM retention of the intact β-propeller, and
~50% mitochondrial matrix localization of the destabilized mutant (Fig. 29C). The
processed, but not the unprocessed form of COX8-Gβ5∆C was resistant to trypsin
91
digestion of intact mitochondria, providing further evidence for import of the unraveled
β-propeller (Fig. 29D). These results demonstrate that β-propeller folds, and likely other
domains with similarly rigid tertiary structures, are refractory to translocation by the
TIM/TOM complex.
Translocase targeting is necessary, but not sufficient for
apoptosis induction by Bβ2
We previously showed that a Bβ2 mutant unable to recruit the PP2A holoenzyme
(RR168EE) is inactive in apoptosis assays [17]. An additional set of fusion proteins was
constructed to examine the relationship between OMM localization and apoptosis
induction (Fig. 30A). All Bβ2-based constructs were expressed to similar levels in PC6-3
cells; expression of the remaining, smaller fusion proteins was 3 to 5-fold higher (not
shown). The cryptic import sequence of Bβ2 was replaced with either the unrelated
COX8 import sequence, or the MAS70 OMM-anchor sequence. Like Bβ2, both proteins
co-immunoprecipitated with endogenous PP2A/A and C subunits, localized to
mitochondria, and migrated as single protein species, characteristic of OMM localization
(not shown). Bβ2 and COX8-Bβ sensitized cells to serum-starvation induced apoptosis
with equal potency, suggesting that the differentially spliced N terminus has no other
function than to target Bβ2 to the import complex (Fig. 30B). In contrast, PP2A
heterotrimers containing the transmembrane-anchored regulatory subunit (MAS70-Bβ)
displayed significantly reduced proapoptotic activity (Fig.30B), supporting a role for
reversible targeting of PP2A to OMM microdomains containing TIM/TOM complexes.
Physical obstruction of the TIM/TOM complex by a subpopulation of
Alzheimer’s disease APP has been proposed to compromise neuronal viability by
interfering with import of essential proteins [192]. To examine whether simply targeting
a β-propeller to the TIM/TOM complex could kill cells, serum starvation assays were
carried out with PC6-3 cells transfected with COX8-Gβ5 and control proteins. Despite
92
higher expression levels than Bβ2, the OMM-targeted β-propeller was completely inert in
these assays (Fig. 30B).
We lastly examined whether expression of OMM-targeted fusion proteins might
reduce mitochondrial respiration, as has been shown for APP [192]. Neither Bβ2, nor any
the artificial proteins affected mitochondrial membrane potential as determined by
TMRM fluorescence (Fig. 30C). Together, these results build upon a previous model in
which the cryptic import sequence and unfolding-resistant β-propeller domain of Bβ2
recruits PP2A to the OMM to dephosphorylate critical apoptosis regulators upstream of
Bcl-2 (Fig. 32). Unlike APP, the interaction of β-propellers with receptor components of
the TOM complex appears to be transient and readily reversible, as it is necessary but not
sufficient for the proapoptotic activity of Bβ2.
Discussion
Model of PP2A holoenzyme structure
Structure modeling and site-directed mutagenesis support the model of PP2A
holoenzyme structure depicted in Figure 31. In brief, the B-family of regulatory subunits
are predicted to adopt a β-propeller structure that is found in other proteins engaged in
multiple protein-protein interactions [167, 168]. The PP2A holoenzyme model predicts
that the Bγ N- and C-termini face away from the A subunit while multiple charged
residues located in WD repeats 3 and 4 and the intervening loop interact with residues of
the opposite charge in the A subunit. Supporting our model, the Bγ core (residues 21439) does not tolerate small internal deletions suggesting that the interaction of regulatory
subunits with the AC dimer requires the precise alignment of multiple interacting
residues. Multiple critical charged amino acids in Bγ were located upstream of the A
subunit-contacting residues R165 and R166, that cluster in WD repeats 3 and 4 and the
intervening loop. Furthermore, charge reversal complementation studies suggest that
PP2A holoenzyme association requires electrostatic interactions between adjacent
93
charged residues in the A subunit (E100 and E101 in Aα) and adjacent opposite charged
residues in B-family regulatory subunits (R165 and R166 in Bγ). Thus, we propose that
WD repeats 3 and 4 in B family regulatory subunits form the AC dimer interface which
makes extensive contacts with the intra-repeat loops of HEAT repeats 4-7 of the Aα
subunit, where many residues important for regulatory subunit binding are localized
[173]. The observation that the divergent N-terminal tail of Bγ is expendable for
holoenzyme association supports the view that the N-terminus dictates subcellular
localization of PP2A holoenzymes by interacting with specific anchoring proteins [67].
Overall, this report addresses the structural basis of PP2A holoenzyme function, but
requires ultimate verification and refinement by other methods such as X-ray
crystallography.
Structural implications
The holoenzyme association studies have structural implications for all B-family
of regulatory subunits considering that they are highly homologous (~85%). As seen for
Bγ, charge reversal complementation analyses also demonstrated that the pair of adjacent
acidic residues in the A subunit (EE100) interact with a pair of adjacent basic residues in
Bβ2 (RR168) (Fig. 13B; Fig. 22). The functional consequence of disrupting holoenzyme
association in B-regulatory subunits was exemplified in Bβ2. Survival assays performed
in hippocampal neurons and in PC12 cells demonstrated that Bβ2(RR168EE), a mutant
defective for incorporating the holoenzyme, failed to promote apoptosis (Fig.13D; Fig.
14). Furthermore, the divergent N-terminal tails of Bγ and Bβ isoforms were found to be
expendable for holoenzyme association (Fig. 6; Fig 19). The PP2A holoenzyme model
also predicts that the presumed β-propeller of B-family regulatory subunits require a
“Velcro” patch mechanism to close the rigid toroidal structure. Consistent with this view,
we found that small deletions in WD repeats 1 and 7 that disrupt the Velcro closure
mechanism which presumably opens the presumed β-propeller structure promotes the
94
import of β-propeller mutants but not of wild-type Bβ2 into the matrix of mitochondria
(Fig. 28B).
Mitochondrial PP2A in apoptosis
Bβ2 is a neuron-specific splice variant of the Bβ gene that translocates to
mitochondria upon cell stress to promote apoptosis while RNAi of the endogenous
protein is neuroprotective ([17], Fig. 15; Fig. 16). A comprehensive study on the
mechanism of mitochondrial association of Bβ2 will be instrumental for developing Bβ2
mitochondrial inhibitors (ie, intrabodies) as alternate treatments for stroke and
neurodegeneration. Residues required for mitochondrial localization (e.g. Lys2, Arg6) are
also critical for the proapoptotic activity of PP2A/Bβ2. This is a significant finding, since
a sizable fraction of the GFP-tagged protein is diffusely localized and could therefore
mediate dephosphorylation of cytosolic substrates [17]. Moreover, only the translocasetargeting function of the Bβ2 N terminus seems to be important, since it is functionally
interchangeable with the non-homologous import sequence of COX8. Somewhat
unexpectedly, the membrane-spanning sequence of MAS70 confers only limited
proapoptotic activity. A possible explanation for this finding is that Bβ2 requires some
degree of mobility around the TOM complex to dephosphorylate apoptosis-related
substrates, whereas MAS70-Bβ is fixed near the plane of the OMM.
Recent evidence suggests that a fraction of Alzheimer’s disease-related APP
obstructs the TIM/TOM complex to compromise mitochondrial respiration [192]. A poreblocking, stable translocation intermediate is formed, presumably because APP’s
cleavable import and acidic stop-transfer sequences are separated by more than 200
residues of largely hydrophilic character. In contrast to APP, the signal sequences of Bβ2
and model proteins with intact β-propellers show no evidence of cleavage, and even high
level expression of COX8-Gβ5 is not detrimental to cells. It is probable that a larger
separation between import sequence and β-propeller would expose the N terminus to the
95
matrix, resulting in MPP processing, irreversible blockade of the translocase, and
impaired mitochondrial function similar to APP.
In a model based on previous [17] and present results, the neuronal PP2A/Bβ2
holoenzyme exists in equilibrium between cytosol and mitochondrial surface. Its Nterminus transiently interacts with receptor components of the TOM complex, increasing
the local PP2A concentration to promote dephosphorylation of unidentified OMMassociated proteins, which in turn sensitizes neurons to proapoptotic insults (Fig. 32).
Other proapoptotic PP2A regulatory subunits appear to substitute for Bβ2 in nonneuronal cells. Ruvolo et al. [116, 117] reported that ceramide-induced
desphosphorylation of Bcl-2 in a lymphocyte cell line involves mitochondrial
redistribution of PP2A containing the B’α regulatory subunit. B’α is structurally
unrelated to Bβ2, has no recognizable import signal, and is therefore likely localized by a
different mechanism.
A novel OMM- targeting mechanism and its implications
Partial unfolding of proteins destined for the mitochondrial matrix is catalyzed by
mtHSP70, a component of the presequence-associated motor (PAM) complex at the
matrix-side of the TIM complex. MtHSP70 is thought to unfold preproteins by a ratchet
mechanism [193, 194]. Facilitated by thermal “breathing” of the preprotein’s tertiary
structure, its positively charged N terminus is electrostatically pulled into the matrix,
where the partially linearized polypeptide is trapped by mtHsp70 and prevented from
sliding back into the translocase pore. Assisted by chaperones, the imported protein then
refolds into its original conformation.
Conformational stability is a well-established factor governing the efficiency of
mitochondrial import. Fusion of various import signals to dihydrofolate reductase
(DHFR) promotes efficient import of this cytosolic enzyme into the mitochondrial
matrix. The DFHR inhibitor methotrexate stabilizes the enzyme’s conformation so that
96
the fusion protein arrests in the TIM/TOM complex, an experimental trick commonly
exploited in translocase structure/function studies [175, 195]. To our knowledge, Bβ2 is
the first non-artificial protein identified that employs a structure-based mechanism for
OMM localization, and the β-propeller constitutes the first fold shown to resist import in
a ligand-independent manner. We predict that additional examples of proteins targeted in
this fashion will be discovered, utilizing β-propellers or other rigid tertiary structures as
stop-transfer domains.
These results also imply that the diversity of folds that can be adopted by nuclearencoded matrix proteins is constrained by the translocase. That is, β-strand rich proteins
with unfolding energies in the range of β-propellers (> 10 kcal*mol–1) should be
underrepresented in the mitochondrial matrix, because they are poor translocase
substrates. Indeed, only 2 of the 751 proteins (0.25%) constituting the mitochondrial
proteome of Saccharomyces cerevisiae [196] are predicted β-propellers, including one
membrane protein (YHR186C/KOG1). This contrasts with a ~1% representation of βpropellers in the entire yeast proteome (http://bmerc-www.bu.edu/wdrepeat/yeast.html).
97
Figure 18 Structure prediction of B-family regulatory subunits. A: the amino acid
sequence of Bγ was aligned according to boundaries of the seven WD repeats
and component β-strands (d, a-c) provided by the Pfam web application
(http://pfam.wustl.edu, [197]). Sequence conservation of WD repeats is
indicated by gray and black shading. B: schematic of β-strand arrangement of
the β-propeller fold, highlighting the phase-shift of WD repeats (identified by
shading) and propeller blades. C: ribbon view of B subunit model based on
the Gβ1 crystal structure; note that large loops connecting WD repeats (c-d
loops) were not completely modeled. The molecular model of Bγ was initially
generated by Dr. Stefan Strack and subsequently refined and optimized by
Ruben K. Dagda.
98
Figure 19 Mapping the holoenzyme association domains of Bγ by deletion
mutagenesis. Wild-type (w.t.) Flag-tagged Bγ, the diagrammed deletion
mutants (A) or vector alone were transiently expressed in COS-M6 cells,
immunoprecipitated (IP), and tested for association with co-transfected Aα
subunit (EE epitope-tagged) and endogenous C subunit by immunoblotting.
A representative immunoblot is shown in (B); the broad band below the Aα
subunit band is immunoglobulin heavy chain. Co-immunoprecipitated C
subunit (C coIP) was quantified as the ratio of C to Bγ subunit in each lane,
and is listed relative to wild-type Bγ below the blot (average of 2 to 4
independent experiments). All constructs were generated by Thomas Cribbs,
Chris Barwacz, Stefan Strack and Raul Dagda. The representative
immunoblot (B) was performed by Ruben K. Dagda.
99
Figure 20 Identification of Bγ residues important for holoenzyme association. Wildtype (w.t.) Flag-tagged Bγ, the indicated site-directed mutants, or empty
vector were expressed into COS-M6 cells, and tested for association with
transfected Aα (EE-tagged) and endogenous C subunits by coimmunoprecipitation (coIP). Percent binding of the C subunit was quantified
as in Fig. 19 and is shown as the average of 2 to 5 experiments. With the
exception of Bγ -EE89RR-, -RR165EE, -D112K, -D184K-, and E223R, all
constructs were generated by Thomas Cribbs, Chris Barwacz, Ruben Dagda
and Raul Dagda. The representative immunoblots were performed by Ruben
K. Dagda.
100
Figure 21 Summary of the effects of deletion and charge reversal mutagenesis on
holoenzyme association. Summary of Bγ deletion and site-directed
mutagenesis results. Critical mutations displaying < 15% wild-type C subunit
binding activity are indicated on the top of the domain diagram; non-critical
mutations (>= 40% wild-type binding) on the bottom.
101
Figure 22 Identification of interacting residues in Bγ and Aα. Wild-type (w.t.) or
mutant Flag-tagged Bγ and EE-tagged Aα subunits were co-expressed in the
indicated combinations in COS-M6 cells, and tested for association by Flagimmunoprecipitation, followed by immunoblotting for PP2A subunits. The
experiment was performed by Dr. Stefan Strack.
102
Figure 23 Effects of N-terminal truncations on mitochondrial association of Bβ2.
A: representative confocal images of selected Bβ2 N terminus-GFP fusion
proteins (green) expressed in PC6-3 cells after labeling with the mitochondrial
dye MitoTracker (mito). Truncations were created in the context of Bβ21-35GFP. Yellow in the larger, merged panels indicates colocalization. Scale
bars=10 µm.Throughout the figure, green fonts and graph coloring represents
strong mitochondrial localization (score > 2), while red colors are reserved for
critical mutations (score < 2). B: summary quantification of mitochondrial
(mito.) localization of the indicated Bβ2 sequences fused to GFP (means ±
SEM of 3 or more experiments, >50 cells/experiment). All constructs were
generated by Chris Barwacz.
103
Figure 24 Effects of N-terminal point mutations on mitochondrial association of
Bβ2. A: representative confocal images of selected Bβ2 N terminus-GFP
fusion proteins (green) expressed in PC6-3 cells after labeling with the
mitochondrial dye MitoTracker (mito). Point mutations were created in the
context of Bβ21-35-GFP. Yellow in the larger, merged panels indicates
colocalization. Scale bars=10 µm.Throughout the figure, green fonts and
graph coloring represents strong mitochondrial localization (score > 2), while
red colors are reserved for critical mutations (score < 2). B: summary of
mitochondrial colocalization scores of the listed mutations in the context of
Bβ21-35- or Bβ21-26-GFP (means ± SEM of 1 to 6 experiments as indicated,
>50 cells/experiment). With the exception of Bβ21-35-GFP, all constructs
were generated by Thomas Cribbs and Chris Barwacz.
104
Figure 25 The proapoptotic activity of Bβ2 requires mitochondrial association. A:
cell lysates of PC6-3 cells transfected with the indicated C-terminally GFP
tagged constructs were immunoblotted with a B-pan antibody that recognizes
all B family of regulatory subunits. The ratio of GFP fusion protein to
endogenous Bα protein immunoreactivity was quantified using Image J and
indicated for each lane. The black arrow points to the GFP fusion proteins
while the white arrow points to endogenous Bα. B: Positive correlation
(R=0.74) between mitochondrial localization (data from Figure panel B) and
proapoptotic activity. The indicated mutations were introduced into full-length
Bβ2 fused to GFP and analyzed for sensitization of apoptosis induced by 24 h
serum withdrawal. Percent sensitization is plotted relative to GFP alone
(typically 5% apoptotic nuclei) and wild-type (w.t.) Bβ2 (~30% apoptosis)
(means ± SEM from 3 to 6 apoptosis assays, >50 cells/assay). All constructs
were generated by Chris Barwacz and Thomas Cribbs, and midi-preps were
generated by Ruben K. Dagda.
105
Figure 26 Proteolytic processing of Bβ2 N terminus fusion proteins. A: cytosolic
(cyt) and mitochondrial fractions (mito) of PC6-3 cells transiently expressing
Bβ21-35-GFP wild-type (w.t.) and substitution mutants were probed with the
indicated antibodies. The position of the 31 kDa molecular weight marker is
shown. B: the indicated fusion proteins were immunoprecipitated from total
cell lysates and probed for GFP. As in panel A, solid and open arrow heads
point to full-length and processed forms of Bβ2 N terminus fusion proteins,
respectively. C: positive correlation (R=0.93) between mitochondrial
localization (data from figure 24B) and proteolysis of Bβ2 fusion proteins.
Cleavage was quantified by densitometry of GFP blots (see panel B) and
expressed relative to Bβ21-35-GFP wild-type (means ± SEM from 3 to 4
independent experiments). The COX8-GFP and MAS70-Bβ constructs were
generated by Chris Barwacz. The experiment shown in the top panel was
performed by Chris Barwacz.
106
Figure 27 Submitochondrial localization of full-length Bβ2 and Bβ21-35-GFP.
Mitochondrial fractions of PC6-3 cells expressing the indicated proteins were
incubated ± 1% Triton X-100 and the listed concentrations of trypsin,
followed by probing with antibodies directed to GFP (top two panels) and the
FLAG tag (Bβ2). The processed form of Bβ21-35-GFP is protected from
trypsinolysis of intact mitochondria, whereas full-length Bβ2 and the OMM
protein MAS70-GFP are degraded. The asterisk indicates a non-specific band
also detected in untransfected cells; the position of molecular weight markers
is shown on the right. The doxicycline inducible PC6-3 cell line that
overexpresses Bβ2-TAP was generated by Thomas Cribbs and Ruben K.
Dagda. The proteolysis experiment was performed bt Dr. Stefan Strack.
107
Figure 28 The β-propeller of Bβ2 is a stop-transfer fold. A: schematic of β-propellermutant Bβ2 fusion proteins. Numbered boxes (1-7) represent WD repeats, and
the brackets indicate intact or disrupted closures of the toroidal structure. B:
the listed Bβ2-GFP fusion proteins were immunoprecipitated from total PC63 cell lysates via the FLAG tag and immunoblotted for GFP. All Velcro
closure mutants are cleaved. Black arrows point to full-length β-propellermutant Bβ2 fusion proteins while the white arrows point to processed form of
β-propeller-mutant Bβ2 fusion proteins. The Bβ2 –(1-437 (K2A)) and ∆27-34
(IL18AA) constructs were generated by Audrey Dickey.
108
Figure 29 Unfolding-resistant translocase targeting of a model protein. A: schematic
of a mitochondria-targeted G protein β subunit (COX8-Gβ5) and its Cterminal truncation of 8 residues that unravels the β-propeller (COX8Gβ5∆C). WD repeats are numbered 1-7. B: confocal image of PC6-3 cell
showing colocalization of COX8-Gβ5 (GFP) and MitoTracker-labeled
mitochondria (mito). Scale bar=10 µm. C: transfected fusion proteins were
immunoisolated via their FLAG epitope and immunoblotted for GFP. The
position of the 66 kDa marker is shown. D: PC6-3 cells expressing the COX8Gβ5∆C were disrupted by N2 cavitation, and homogenates were incubated ±
100 µg/ml trypsin, ± 1% Triton X-100 as indicated. The fusion protein was
immunoisolated from solubilized extracts via its FLAG tag and
immunoblotted for GFP. As in panel C, solid and open arrowheads point to
full-length and processed COX8-Gβ5∆C, respectively. The COX8- Gβ5 and
COX8-Gβ5∆C constructs were generated by Chris Barwacz.
109
Figure 30 Translocase targeting is necessary but not sufficient for toxicity of
PP2A/Bβ2. A: schematic of fusion proteins tested for effects on cell viability
and mitochondrial function. B: PC6-3 transfected with the indicated fusion
proteins were tested for sensitization to serum deprivation-induced apoptosis
(see Fig. 25B) (means ± SEM of 3 to 8 assays as labeled, >50 cells/assay).
Significant differences by Student’s t-test: *, p<0.001 between Bβ2/COX8Bβ and MAS70-Bβ, p<0.0001 compared to GFP. C: PC6-3 cells expressing
the listed GFP constructs, untransfected cells, and detached cells (“floaters”)
were loaded with the mitochondrial (mito.) membrane potential-reporting dye
TMRM and analyzed by confocal microscopy (cytosolic to nuclear
fluorescence ratio, means ± SD of two experiments with 20-30 cells each).
The COX8-Bβ and MAS70-Bβ constructs were generated by Chris Barwacz.
110
Figure 31 Model of the PP2A holoenzyme. A space-filling representation of the
structure of the Aα subunit [29] is arranged with model structures of the C
and Bγ subunits (based on the PP1 catalytic subunit [174], and Gβ1 [167],
respectively, see text). Residues whose mutation disrupts subunit association
(critical) are indicated in black; interacting residues are colored green. R183
and W257 are highlighted as representative of several critical residues in
HEAT repeats 5 and 7 of the Aα subunit [173]. Some critical Bγ residues are
not shown because they are either absent from the model (K214, D259) or are
buried (D192).
111
Figure 32 Unfolding resistant translocase targeting of Bβ2. Figure model depicting
the mechanism of OMM targeting and apoptosis sensitization by PP2A/Bβ2.
Bβ2 exists in equilibrium with cytosolic and mitochondrial compartments.
Upon cell stress, Bβ2 translocates to mitochondria via its N-terminus which
binds to mitochondrial import receptors. However, Bβ2 does not undergo
import but is retained at the OMM via its presumed rigid β-propeller domain
that resists the unfolding step required for import. Once targeted to the OMM,
PP2A/Bβ2 dephosphorylates an unknown substrate at the OMM to promote
apoptosis upstream of Bcl-2.
112
CHAPTER IV
REVERSIBLE PHOSPHORYLATION CONTROLS
MITOCHONDRIAL FISSION/FUSION AND NEURONAL
APOPTOSIS
Abstract
Mitochondria are dynamic organelles that continuously undergo fission
(fragmentation) and fusion (elongation) and these processes are regulated by large
dynamin-like GTPases. While some evidence has linked fission to apoptosis in nonneuronal cells, it is not clear to what extent neuronal survival requires mitochondrial
fission/fusion (MFF). In this report we describe novel mechanisms by which Bβ2, a
neuron specific splice variant of the Bβ gene that is targeted to the OMM to promote
apoptosis[163], and PKA, a well characterized prosurvival kinase, at the OMM regulate
neuronal survival. Implicating a role of PP2A as regulator of MFF, confocal microscopy
analyses of live neurons revealed that overexpression of Bβ2 fragments mitochondria
while RNAi mediated knock-down or catalytic inhibition of endogenous PP2A/Bβ2 by
targeting an endogenous inhibitor of PP2A, Plasmodium falciparum aspartate rich protein
(PfARP), to mitochondria promoted fusion. Implicating PKA as a positive regulator of
MFF, agents that activate PKA, or recruiting PKA to the OMM by overexpressing A
kinase anchoring protein (AKAP121) promoted mitochondrial fusion and neuronal
survival. Conversely, PKI targeted to the OMM (OMM-PKI), knocking-down
endogenous AKAP121 by RNAi, or inhibiting PKA with an inhibitor of PKA (PKI)
targeted to the OMM fragmented mitochondria and promoted apoptosis in hippocampal
neurons. Moreover, our survival assays demonstrate that the MFF machinery is required
for survival since proteins that promote fission such as dynamin like protein 1 (DLP1)
promotes apoptosis while dominant negative (DN) DLP1 (K38A) or RNAi knock-down
of FIS1 fuses mitochondria and promotes survival. Epistasis experiments were then
113
carried out to determine whether the survival effects of OMM-PKA and PP2A/Bβ2
require MFF. RNAi mediated knock-down of FIS1 or co-expression of DN-DLP1 were
unable to rescue neurons from apoptosis induced by RNAi mediated knock-down of
endogenous AKAP121 suggesting that other AKAP121 interacting proteins recruited to
the OMM are required for basal survival. However, overexpression of Bcl-2, DN-DLP1,
or RNAi mediated knock-down of FIS1 blocked mitochondrial fission and apoptosis
induced by Bβ2 and OMM-PKI suggesting that OMM-PKA and PP2A/Bβ2 require MFF
and act upstream of apoptosis to regulate survival. Overall, this is the first report that
demonstrates that mitochondrial morphology and neuronal survival are dynamically
regulated by reversible phosphorylation.
Introduction
Mitochondrial fission/fusion
Mitochondria are not static but are highly dynamic structures that exhibit
motility, divide and undergo drastic changes in morphology in response to intracellular
and extracellular signals such as calcium and growth factors. Mitochondria continuously
undergo fusion (elongation) and fission (fragmentation) and the balance of these two
processes ultimately determines the ability of mitochondria to acquire a spherical,
elongated morphology, or form a continuous network (reviewed by [198-202]).
Mitochondrial morphology differs among cell types. For instance, it has been shown that
mitochondria in cardiac myocytes and hepatocytes are elongated while mitochondria in
the cell body of neurons are generally more fragmented [203]. Physiological
consequences associated with mitochondrial fission include decreased mitochondrial
respiration and increased sensitivity to apoptosis. On the other hand, physiological
consequences associated with mitochondrial fusion include stabilizing the mitochondrial
genome, delaying cell aging and increasing mitochondrial respiration (See review by
([200]). Mitochondrial fragmentation is an obligatory step for apoptosis [202, 204].
114
Apoptotic insult promotes fragmentation of mitochondria and this process coincides with
the translocation of Bax to mitochondrial scission sites and precedes the release
apoptogenic factors such as cytochrome c, and caspase activation [205]. Fission is an
early event during apoptosis since the anti-apoptotic Bcl-2 protein prevents apoptotic cell
death but does not block mitochondrial fission [206, 207]. Mitochondrial fusion has been
linked to promoting survival and delaying the onset of apoptosis in eukaryotes [198,
208]. Thus, the balance of mitochondrial fission and fusion can ultimately determine
whether a cell survives toxic insult. However, the positive correlation of mitochondrial
fusion and survival has been recently challenged by one report that suggests that
mitochondrial fusion is required for apoptotic pathways that rely on mitochondrial
mediated calcium wave propagation. Szabadkai et al., 2004 reported that mitochondrial
fission sensitized cells to staurosporine or etoposide induced apoptosis but protected cells
against ceramide induced apoptosis by interfering with the network dependent calcium
wave propagation [209].
Mitochondrial morphology is regulated by a group of large GTPases. A mediator
of fission is the cytosolic dynamin related protein (DRP1) or dynamin like protein 1
(DLP1) which translocates to mitochondria and binds with the membrane adaptor protein
FIS1 at the OMM. DLP1 requires GTPase activity to promote fission since a mutation of
a conserved nucleotide binding residue (K38A) inhibits hydrolysis of GTP and blocks its
fission promoting activity. Driving fission by overexpression of wild-type (W.T.) DLP1
promotes apoptosis while overexpressing DN-DLP1 prevents apoptosis. One model has
proposed that DLP1 promotes fission by assembling into large oligomers that wrap
around mitochondria to form a ring which ultimately constricts and pinches mitochondria
into two fragments [202].
On the other side of the coin, three large GTPases have been identified to catalyze
mitochondrial fusion. They are mitofusins 1 and 2 (Mfn1/2) and optic atrophy 1 (Opa1)
proteins. Mitofusins are large GTPases that promote fusion of outer mitochondrial
115
membranes (See review by ([200]). Opa1 is localized at the inner membrane space and
catalyzes the fusion of inner mitochondrial membranes. Mitofusins promotes fusion of
mitochondria in a process similar to SNARE proteins during membrane docking of
synaptic vesicles. Fusion mediated by mitofusins is a GTPase dependent process by
which the coiled-coil domain of one mitofusin clamps with the coiled-coil domain of
another mitofusin to bring mitochondria in close proximity [202]. Mfn1/2 are essential
for embryonic development since knock-out of either protein is lethal in mice [210].
While both Mfn1 and Mfn2 can homodimerize to promote fusion, Mfn2 is functionally
distinct from Mfn1 in that it mediates long-range signaling functions in the cell [211213].
There is increasing evidence that links abnormal remodeling of mitochondrial
morphology to neurodegeneration. Genetic mutations in Opa1 cause excessive fission
and underlies neurodegeneration of retinal ganglion neurons in dominant optic
neuropathy, the most common form of hereditary blindness [214]. Genetic mutations
mapped at the GTPase and coiled-coil domains of Mfn2 and in ganglioside-induced
differentiation associated protein 1 (GDAP), a fission promoting protein, are associated
with Charcot-Marie-Tooth disease (CMT), the most common form of hereditary
peripheral neuropathy [215, 216]. These findings suggest that a proper balance of
mitochondrial morphology is critical for neuronal survival. Although a lot of research has
been focused on studying the physiological implications of mitochondrial remodeling in
non-neuronal cells, little work has been done in neurons. Mitochondrial fission is
necessary for the transport of mitochondria to dendrites to stimulate the development of
dendritic spines in hippocampal neurons [217]. In addition, hypomorphic mutations of
DLP1 in flies deplete the synapses of mitochondria which cause a decrease in
neurotransmission due to an impaired ability to mobilize reserve synaptic vesicle pools
[218].
116
Reversible phosphorylation at the OMM in neuronal
survival
Phosphatases and kinases are present in all the compartments of the
mitochondrion and their activities regulate diverse mitochondrial functions and long
range signaling pathways. For instance, kinases such as PKA, and JNK at the OMM
regulate apoptosis while PKCδ at the IMM regulates mitochondrial metabolism by
phosphorylating /inactivating the pyruvate dehydrogenase complex (Reviewed by [219].
There is currently a high interest in studying the role of OMM reversible
phosphorylation in mediating neuronal survival. Mitochondrial kinases have been
characterized to be both positive and negative regulators of apoptosis. For instance,
PINK1, a PTEN induced kinase implicated in a familial form of Parkinson’s disease, is a
mitochondrial kinase that has been shown to prevent neuronal apoptosis. On the other
hand, a Parkinson’s disease bearing mutation or a kinase inactive mutation in PINK1 fails
to promote neuronal survival [220, 221]. Translocation of phosphorylated ERK1/2 to the
nucleus promotes neuronal survival by activating gene transcription programs of
immediate early genes (IEGs) to promote neuronal survival [222]. However, ERK also
plays a detrimental role in neurons. It has been found that an aberrant localization of
phosphorylated ERK1/2 at the mitochondria may contribute to neurodegeneration
possibly by inducing an increase in mitochondrial turnover and mitochondrial
dysfunction (reviewed by[223]). Other kinases at the OMM that have been shown to
stimulate apoptosis include GSK3β, JNK and others (reviewed by [219]). Perhaps the
best characterized prosurvival signaling pathway at the mitochondria to date is the role of
OMM targeted PKA in stimulating survival as described below.
Mitochondrial PKA
In neurons, PKA regulates most cAMP dependent functions such as
neurotransmitter release, gene transcription, survival and synaptic plasticity. The PKA
117
holoenzyme consists of two catalytic (C) subunits bound to two regulatory (R) subunits.
The binding of two cAMP molecules to the R subunits promote the dissociation of the C
subunits to allow phosphorylation of substrates at Ser/Thr residues. There are two classes
of R subunits (RI, RII) with two isoforms each (α, β) that exhibit distinct tissue
expression and subcellular compartmentalization [224]. The RI and RII subunits are
widely expressed in most regions of the brain. At the subcellular level RI is usually
cytoplasmic while the RII subunits are preferentially localized to the ER, Golgi and cell
membrane [225].
Inhibitors of PKA
The effects of PKA are opposed by small thermostable proteins known as
inhibitors of proteins kinase (PKI). Three isoforms of PKI have been characterized
(PKIα-PKIγ) with each exhibiting cell-type and tissue specific expression patterns. PKI
potently inhibits the activity of free catalytic subunits of PKA and has also been
implicated in exporting PKA from the nucleus. Furthermore, by inhibiting PKA and by
suppressing the expression of right side specific genes, PKIα has been shown to play a
role in organizing the left-right axis formation during embryonic development [226]. In
the brain, PKI has been shown to downregulate neurotransmission, and synaptic
plasticity. Purified peptides of PKI have been previously employed as molecular tools to
analyze the effects of inhibiting PKA in primary neurons (Reviewed by [227]).
The targeting of PKA to distinct subcellular compartments is mediated by
molecular scaffolds termed A-kinase anchoring proteins (AKAPs). AKAPs target PKA
holoenzymes to distinct subcellular sites to regulate a variety of physiological processes
that include survival, metabolism and synaptic plasticity [228-230]. To date, three OMMtargeted AKAPs that have been identified are D-AKAP1 (AKAP84/121), D-AKAP2, and
Rab32. Of the three, AKAP121 is the best characterized mitochondrial AKAP [178, 231,
232, 233]. AKAP121 is expressed in all tissues and the expression of this mitochondrial
118
AKAP in the brain is low. To date, eight splice variants of AKAP121 have been
identified and contain divergent N-terminal and C-terminal domains. The N1 splice
variant contains an alternatively spliced N-terminal domain that directs AKAP121 to the
ER. The N0 splice variant of AKAP121 lacks the ER targeting domain and exposes an Nterminal mitochondrial targeting domain of 30 amino acids that redirects the protein to
the mitochondria [231]. The mechanism of mitochondrial association of AKAP121
remains unresolved to this date. AKAP121 has been speculated to associate the OMM by
binding to porin/VDAC receptor since the N-terminal targeting domain contains a weak
homology to the mitochondrial targeting domain of hexokinase I [231]. Another group
proposed that the N0 splice variant of AKAP121 localizes to mitochondria by interacting
with β-tubulin [234]. Another proposed alternative mechanism is that AKAP121 may
anchor the OMM via an N-terminal transmembrane spanning α-helix. AKAP121 shares
homology to the transmembrane α-helix of MAS70 and is predicted to be a single-pass
transmembrane protein (THHMM, DAS algorithms, unpublished observations).
In addition to binding PKA, AKAP121 also recruits mRNA via a C-terminal KH
domain, PTPD1 and Src kinase to the OMM [162, 235, 236]. AKAP121 mediated
targeting of mRNA to the OMM leads to increased protein translation of mitochondrial
proteins while targeting PTPD1 and PKA to the OMM increases oxidative respiration
and survival respectively[236, 237].
There are two reports that have shed light into understanding the mechanism by
which OMM-PKA promotes neuronal survival. PKA is constitutively targeted to the
mitochondria by mitochondrial AKAPs. Prosurvival stimuli elicit an increase in
intracellular concentration cAMP, which bind and activates PKA by promoting
dissociation of the R subunits and consequently releasing the C subunits to phosphorylate
the proapoptotic Bad protein. Phosphorylated Bad is inactivated and sequestered by 14-33 proteins in the cytosol [111]. PKA has been shown to be essential for survival since
119
mutating the PKA binding site of AKAP121 depletes the mitochondria of endogenous
PKA, leads to impaired mitochondrial metabolism and sensitizes cells to apoptosis [162].
Role of OMM phosphatases in survival
There is currently scarce evidence regarding the role of OMM phosphatases in
neuronal survival. To date, relatively few phosphatases have been identified in the
mitochondria. At the OMM, Ser/Thr specific phosphatases such as PP1 and PP2A
regulate nitric oxide synthase and apoptosis respectively while protein tyrosine
phosphatase D1 (PTPD1) stimulates mitochondrial metabolism [17, 117, 219, 238]. A
PP2A holoenzyme, termed spermine stimulated phosphatase, with specific activity
towards the pyruvate dehydrogenase complex was one of the first mitochondrial
phosphatases to be characterized [239]. Only two mitochondrial PP2A holoenzymes have
been implicated in regulating apoptosis. In non-neuronal cells, a PP2A holoenzyme
containing the B’α regulatory subunit translocates to mitochondria to promote ceramide
induced apoptosis by dephosphorylating/inactivating the anti-apoptotic Bcl-2
protein[116, 117].
We have previously characterized a neuron specific proapoptotic PP2A
holoenzyme, PP2A/Bβ2 that translocates to the OMM by virtue of a cryptic
mitochondrial import signal. Furthermore, the proapoptotic activity of PP2A/Bβ2 was
shown to require a functional import signal and incorporation into the PP2A holoenzyme
[17, 163]. Here we describe a mechanism by which PP2A/Bβ2 promotes apoptosis that
involves mitochondrial fission and identified OMM-PKA as the kinase that opposes the
effects of the phosphatase in hippocampal neurons.
120
Experimental Procedures
Generation of plasmids
shRNAs targeting DLP1, AKAP121 and FIS1 were generated in the following
manner. Four shRNAs (IR-AKAP121 shRNA1-4) were generated to target AKAP121 by
annealing to bps 778 -888, 1333-1353, 1064-1084 and 738-758 of the rat coding
sequence respectively. To target FIS1, four shRNAs (IR-FIS1 shRNA1-4) were generated
to anneal at sites 86-105, 165-183, 315-333, 421-439 of the rat cDNA respectively. To
target rat DLP1, three shRNAs (IR-DLP1 shRNA 1-3) were generated to anneal at sites
421-439, 927-945, 2105-2120 respectively. The shRNAs were inserted into the H1
promoter by PCR amplification of the H1 promoter of the pSUPER vector using a
forward primer fitted with a BamH1 restriction site that anneals upstream of the H1
promoter and a reverse primer containing the shRNA and a HindIII site. The amplified
PCR products were digested with HindIII and BamHI restriction enzymes and
subsequently ligated into HindIII/BglII digested pCDNA3.1 vector that expresses the βgalactosidase protein.
In order to construct a GFP fusion protein of AKAP121, the core domain of
AKAP121 (aa 1-525) was amplified by RT-PCR from rat brain mRNA using forward
primer fitted with a HindIII restriction site and reverse primer fitted with a SalI restriction
site. The PCR product was digested with HindIII and SalI restriction enzymes and
subsequently ligated into HindIII/SalI digested pEGFP-N1 vector. Subsequently, the
AKAP121 GFP fusion construct was used as a template to generate a PKA binding
deficient mutant of AKAP121 (I310P, L316P) by two rounds of PCR in the following
manner. A nested forward primer containing a HindIII site and a reverse primer that
contains mutagenic sequences (I310P, L316P) were used to amplify a 5’ fragment of
AKAP121 from AKAP121-pEGP-N1 vector by PCR. The resulting PCR product was gel
purified, and used as a forward primer in conjunction with a reverse primer containing a
121
SalI site that anneals to the extreme C-terminus of AKAP121 to amplify the remaining 3’
portion of AKAP121. The PCR product was gel purified, digested with HindIII and SalI
restriction enzymes and ligated into HindIII/SalI digested pEGFP-N1 vector.
The MFF modulators wild-type and dominant-negative (DN) DLP1 (K38A) were
generously provided by Yisang Yoon from the University of Rochester.
GFP fusion proteins of PKA and PKI targeted to the OMM via the porin/VDAC
sequence of hexokinase I (OMM-PKA, OMM-PKI), and nuclear targeted PKA (NLSPKA) were generously provided by Steven Green from the University of Iowa [240].
Untagged Bcl-2 was a generous gift from Mike Knudson from the University of Iowa
Lentiviral infection of hippocampal neurons
Hippocampal neurons (DIV10) seeded on Poly-D-lysine coated 2-well
chambered slides (20 mm2 chamber, Nalge Nunc) at a density of 60,000 cells/well were
incubated with FIV lentiviruses (1X105-1X106 plaque forming units/mL) that
overexpress GFP fusion proteins or shRNAs at a 1:200 dilution at 37°C. After 5-6 hours
of incubation, hippocampal neurons were washed once with fresh NB/B27 media to
remove unbound viral particles.
Apoptosis assays
E18 hippocampal neurons were plated in 24 well plates and transfected at 10 DIV
with GFP fusion proteins, with scrambled control or with shRNAs targeting MFF
modulators. Four to five days after transfection, hippocampal neurons were treated with
methanol (1:1000) as a vehicle control or with 400nM rotenone to induce toxicity. Two
days following the insult, hippocampal neurons were fixed in 3.7% paraformaldehyde,
permeabilized in Tris-buffered saline containing 0.1% Triton X-100 (TTBS) and
immunostained for GFP using a rabbit anti-GFP polyclonal antibody at a 1:2,000 dilution
(Abcam, Cambridge, UK), and for neurofilament using an anti-neurofilament monoclonal
antibody at a 1:20 dilution (U. of Iowa Hybridoma Core Facility). The nuclei of neurons
122
were counterstained with the UV filter sensitive Hoechst dye (33342) at 1 µg/ml for 5
minutes, followed by two additional washes of TTBS. Hippocampal neurons containing
condensed, irregular, or fragmented nuclei were scored as apoptotic.
To assay for apoptosis and mitochondrial morphology in the same neuron at the
same time by confocal microscopy, hippocampal neurons seeded on 2-well poly-D-lysine
coated slides were stained with 10 µM of the infra-red sensitive DRAQ5 dye to visualize
the nucleus (Alexis Biochemical, San Diego, CA) and with 100 nm of
tetramethylrhodamine methyl ester (TMRM) to stain the mitochondria (Molecular
Probes, OR, USA).
Subjective mitochondrial morphology assay
PC6-3 cells or hippocampal neurons were seeded on collagen-coated or poly-Dlysine-coated, chambered No. 1 cover glasses (20 mm2 chamber, Nalge Nunc)
respectively and transfected with 1 µg GFP fusion protein plasmids using Lipofectamine
2000. For some experiments, the effects of different pharmacological agents on
mitochondrial morphology was elucidated by treating native PC6-3 cells or hippocampal
neurons with different concentrations of forskolin (MP Biomedicals, CA, USA), nerve
growth factor (NGF) (Upstate Biotechnologies, NY, USA) or with cpt-AMP (Sigma,
MO, USA) for 18 hrs. Twenty four to 96 hours post-transfection, cells were imaged live
using a Zeiss LSM 510 laser-scanning confocal microscope at the Central Microscopy
Facility of the University of Iowa. To visualize mitochondria, cells were treated with 100
nM of TMRM. (Molecular Probes, Portland, OR) and incubated for at least 30 min. at
37°C. At least 15 random confocal images that cut through the mid section of neuronal
somata per transfection condition were collected and processed for analysis of
mitochondrial morphology. Mitochondrial morphology was quantified using a subjective
scale of 0 to 3 where a score of 0 was assigned to a cell containing mostly fragmented
123
mitochondria and increasing scores were assigned to neurons containing elongated
mitochondria by comparison to a set of reference images.
Image analysis of mitochondrial morphology and
mitochondrial density
Mitochondrial morphology was quantified using Image J 1.34 software (National
Institute of Health., US). Confocal images of neuronal somatic mitochondria were
converted to black/white images and adjusted to a threshold value of pixelation. The
aspect ratio of individual mitochondria that contained a minimum size of 7 pixels was
determined by measuring the lengths of minor and major axis and averaged for all the
mitochondria within each neuron. The average mitochondrial aspect ratios for the total
population of neurons/transfection condition were calculated and tabulated using Excel
software.
Mitochondrial density in each transfected neuron was calculated using Image J
1.34 software. Images were processed in the same manner as described above except that
mitochondrial density was calculated as the percentage of the area occupied by
mitochondria relative to the area of the soma of the neuron. In order to measure for the
area of the soma, the entire soma of the neurons was painted with black color using the
“fill” function of the software and the total area occupied by black pixels neuron was
measured with Image J. The average mitochondrial density for the total population of
neurons/transfection condition was calculated and tabulated using the Excel software.
Statistical analysis
The means and standard error measurements (SEM) from at least three
independent experiments were analyzed for significance by using the unpaired, two-tailed
Student t test function of the Excel software and a p value less than 0.05 was considered
statistically significant.
124
Results
PP2A/Bβ2 fragments mitochondria in neurons
We have previously characterized a neuron specific PP2A holoenzyme termed
PP2A/Bβ2 that rapidly translocates to the OMM by virtue of a cryptic N-terminal
targeting signal to promote apoptosis. A functional mitochondrial targeting signal and
association to the PP2A heterotrimer are required for the proapoptotic activity of Bβ2
[17, 163]. Moreover, silencing the endogenous phosphatase is neuroprotective against
glutamate and rotenone toxicity. We initially hypothesized that PP2A/Bβ2 may sensitize
neurons to apoptosis by impairing mitochondrial metabolism or by dephosphorylating
and activating Bcl-2 family of proapoptotic proteins. However, after a series of
preliminary experiments in PC6-3 cells, we failed to demonstrate an effect of Bβ2
overexpression on mitochondrial metabolism or on the phosphorylation status of Bad,
Bcl-2 or Bcl-XL, another survival regulator identified as a substrate of PP2A (data not
shown). However, we noticed an interesting effect of Bβ2 on the morphology of
mitochondria in neurons. Live confocal microscopy revealed that mitochondria of PC6-3
cells, hippocampal neurons, or dorsal root ganglion neurons appeared more punctate or
fragmented compared to the mitochondria of Bβ1-GFP OMM-GFP expressing cells (Fig.
33A, Fig. 34A, and not shown for DRGs). These results implicated Bβ2 as a regulator of
mitochondrial morphology. However, the effects of Bβ2 on mitochondrial morphology
were initially determined by qualitative analysis of high resolution confocal images of
somatic mitochondria of neurons. Thus in order to reliably and routinely quantify
changes in mitochondrial morphology in neurons, we developed the subjective
mitochondrial length assay as described below.
The subjective mitochondrial length assay consists of randomly taking confocal
sections that cut through the soma of each hippocampal neuron or cell body of a PC6-3
cell for least 15 cells per condition. The images were blindly scored by at least three
125
independent observers and visually scored for mitochondrial length using scale from 0 to
3 compared to a set of reference images. A score of 0 was assigned to a neuron that
contain mostly fragmented or spherical shaped mitochondria while more fused
mitochondria were incrementally assigned higher scores. The mitochondrial length scores
were averaged for the total number of neurons analyzed per transfection condition. We
found that the subjective mitochondrial length assay is reliable in that differences of
mitochondrial length score averages for the same transfection condition determined by
different observers were small (SEM= ±0.20). Using the subjective mitochondrial length
assay, we determined that Bβ2-GFP expression significantly fragments mitochondria of
PC6-3 cells or hippocampal neurons compared to Bβ1, or GFP targeted to the OMM
respectively (p<0.0001) (Fig. 33B; Fig. 34B). Interestingly, overexpression of Bβ2
apparently does not fragment mitochondria in glia when compared to glia that transiently
express Bβ1 (data not shown).Our subjective quantification method was further verified
by algorithmic methods using Image J. By measuring the axial ratios of individual
mitochondria in the soma of neurons, Image J computed a significant decrease in
mitochondrial ellipticity for Bβ2 compared to neurons that express Bβ1 (p<0.05), (Fig.
35B). The effects of Bβ2 on mitochondrial fission may be underestimated by Image J
since overlapping fragmented mitochondria tended to be detected as a single
mitochondrion. Thus, algorithmic quantification requires high resolution images
containing well separated mitochondria.
Furthermore, we found that the mitochondrial fragmenting activity of Bβ2
requires association to the A and C subunits of PP2A since transient expression of
Bβ2(RR168EE), a mutant that does not bind the holoenzyme, was unable to fragment
mitochondria. Surprisingly, Bβ2(RR168EE) had an apparent dominant negative
phenotype in that the mitochondria of neurons that express this construct appeared more
fused compared to the mitochondria of neurons that transiently express OMM-GFP
(p<0.003), (Fig. 34B). To determine whether endogenous Bβ2 promotes fission,
126
transfecting two shRNAs that target the 5’-UTR and the divergent N-terminal domain
(Bβ2 shRNA#1 and shRNA #3) of Bβ2 significantly elongated mitochondria compared
to mitochondria of neurons that express a scrambled shRNA (p<0.001), (Fig. 34A, and
B).
Catalytic inhibition of PP2A at the OMM fuses
mitochondria
To further show that mitochondrial fission requires the catalytic activity of Bβ2,
an endogenous inhibitor of PP2A termed PfARP was redirected from the cytosol to
mitochondria by N-terminally fusing the protein with either the divergent N-terminus of
Bβ2 (Bβ2 1-35) or with the mitochondrial targeting domain of MAS70(MAS701-36) [149]
(Fig. 36A). Both fusion proteins colocalized with mitochondria of PC6-3 cells and
hippocampal neurons (Fig. 35A, right panel; Fig. 36B, top panel). Bβ2 1-35-GFPPfARP effectively inhibits phosphatase activity of PP2A since transient overexpression
of this construct in COSM6 cells increases the phosphorylation of Cdk/MAP kinase
substrates on Thr-Pro residues (Fig. 36C). In agreement with Bβ2 1-35 being an import
signal, we found that Bβ2 1-35-PfARP but not MAS701-36-PfARP is proteolytically
processed by MPPs in mitochondria (Fig. 36D). Interestingly, PC6-3 cells that transiently
expressed Bβ2 1-35-PfARP but not MAS701-36-PfARP (OMM-PfARP) displayed
mitochondria that contained a swollen/disk-shaped morphology (Fig. 36B). OMMPfARP did not cause swelling of mitochondria in PC6-3 cells transfected for two days but
some swelling was evident after three days of transient expression (data not shown).
Unlike PC6-3 cells, a distinct mitochondrial phenotype was observed in hippocampal
neurons in that expression of both OMM-GFP-PfARP and Bβ2 1-35-GFP-PfARP
dramatically elongated mitochondria compared to OMM-GFP control (Fig. 35A, and not
shown for Bβ2 1-35-PfARP). In fact, some OMM-PfARP expressing neurons contained
127
abnormally long mitochondria that spanned the entire soma and extended into the
dendrites (not shown).
Similar to PC6-3 cells, transient overexpression of OMM-GFP-PfARP or Bβ2 135-GFP-PfARP swelled the mitochondria of glia (not shown). One possible explanation
for the distinct phenotypes in mitochondrial morphology is that mitochondrial fusion in
hippocampal neurons may be a consequence of inhibiting PP2A/Bβ2 by OMM-PfARP.
Since PC6-3 cells and glia do not express Bβ2, it is possible that distinct, unidentified
PP2A holoenzymes that are inhibited by OMM-PfARP are required for maintaining a
normal mitochondrial morphology.
It is conceivable that mitochondrial elongation induced by OMM-PfARP may be
a consequence of mitochondrial biogenesis (growth) as opposed to MFF events. Unlike
MFF, mitochondrial biogenesis increases mitochondrial density in cells (reviewed by
[219]). In order to rule out this possibility, we analyzed for changes in mitochondrial
density in neurons. Looking at mitochondrial density we did not detect an increase, but in
fact there was a small decrease in mitochondrial density in OMM-GFP-PfARP
expressing neurons compared to OMM-GFP. This small decrease in mitochondrial
density may be an artifact of the image analysis or may be a consequence of fused
mitochondria occupying less space in the soma since OMM-PfARP overexpression tends
to clump mitochondria in some neurons (not shown). Along the same lines, we also
quantified for possible changes in mitochondrial density induced by PP2A/Bβ2.
Although transient expression of Bβ2 significantly fragments mitochondria it does not
alter the mitochondrial density in hippocampal neurons consistent with MFF (Fig. 35B
and C).
PKA/AKAP121 opposes mitochondrial fragmentation
In order to identify the kinase that opposes the mitochondria-fragmenting activity
of PP2A/Bβ2, PC6-3 cells and hippocampal neurons were treated with pharmacological
128
PKA activators (cpt-cAMP, forskolin) or with NGF for PC6-3 cells, and analyzed for
changes in mitochondrial morphology. After 18 hours of treatment, a significant increase
in mitochondrial fusion was observed in PC6-3 cells and hippocampal neurons treated
with PKA activators (cpt-cAMP and forskolin) in PC6-3 cells, suggesting that PKA is the
kinase that promotes fusion. Treating PC6-3 cells with NGF also promoted mitochondrial
fusion compared to PC6-3 cells treated with DMSO. However, mitochondrial fusion
mediated by NGF may be a consequence of morphological changes induced by neuronal
differentiation and not by activating PKA since NGF does not activate PKA related
signaling pathways (Fig. 37A, and B). PC6-3 cells have been shown to survive NGF
withdrawal mediated cell death after treatment with 10µM forskolin for 24 hrs. In our
assays, we see similar effects on mitochondrial fusion by forskolin at the optimal survival
concentration (10µM) and at a concentration that does not promote survival (2µM) of
sympathetic neurons and PC12 cells against NGF withdrawal cell death, suggesting that
mitochondrial fusion does not depend on the survival state of cells [241] (Fig. 37B).
Furthermore, a time course analysis of mitochondrial morphology in PC6-3 cells revealed
that the effects of PKA activation are rapid since mitochondrial fusion induced by
forskolin is maximal within 30 min. and persists for over 16 hrs. after treatment (Fig.
37C, Fig. 38A and B). Unlike MFF which relies on preexisting components of the MFF
machinery to regulate mitochondrial morphology, mitochondrial biogenesis requires de
novo protein synthesis. To rule out an effect of PKA on mitochondrial biogenesis, PC6-3
cells were pretreated with cyclohexamide, a protein synthesis inhibitor, followed by
treatment with PKA activators and analyzed for changes in mitochondrial morphology by
confocal microscopy. Cyclohexamide did not inhibit mitochondrial fusion induced by 30
min. or 60 min. of forskolin treatment compared to cells treated with PKA activators
alone (Fig. 38A and B). To further support the hypothesis that mitochondrial PKA
modulates mitochondrial fusion, I analyzed the effects of recruiting PKA to the OMM by
AKAP121. Expression of AKAP121-GFP (1-526 of the N0 splice variant with C-
129
terminal GFP) significantly increased fusion in PC6-3 cells and in hippocampal neurons
compared to cells expressing a GFP targeted to the OMM by the mitochondrial targeting
sequence of MAS70 (OMM-GFP) or in untransfected cells (Fig. 39A, and B; Fig. 41A
and B). Cells expressing moderate levels of AKAP121 displayed an elongated phenotype
of mitochondrial morphology (Fig. 39, middle panel). However, PC6-3 cells expressing
high levels of AKAP121 appeared to achieve a single fused mitochondrial network or a
“superfused” phenotype of mitochondrial morphology (Fig. 39A, bottom panel). In
hippocampal neurons, AKAP121 overexpression elongated mitochondria but did not
produce the “superfused” phenotype seen in PC6-3 cells (Fig. 41A). In addition to
recruiting PKA, AKAP121 is a scaffolding protein that also recruits other molecules such
as PTPD1, Src kinase and mRNA to the OMM. AKAP121 binds the RI and RII subunits
of PKA via multiple hydrophobic residues localized within an extended amphipathic
helix (I310, L316) [242]. Substituting these hydrophobic residues to proline adds a kink
to the secondary structure and disrupts binding of AKAP121 to PKA. In order to
determine whether AKAP121 promotes fusion by recruiting PKA, a PKA binding
deficient mutant AKAP121 (I310P, L316P) was transfected in hippocampal neurons and
analyzed for mitochondrial morphology. AKAP121 (I310P, L316P) was unable to
elongate mitochondria suggesting that PKA and not other AKAP121 binding proteins
mediate fusion (Fig. 41B). Further implicating PKA as a mediator of fusion, just
targeting PKA to the OMM was sufficient to fuse mitochondria in hippocampal neurons
(Fig. 41A, and B). However, while some hippocampal neurons contained long threadlike mitochondria, other neurons expressing high levels of OMM-PKA appeared to
clump mitochondria or produced the “superfused” mitochondrial morphology phenotype
as seen in PC6-3 cells that overexpress AKAP121 (Fig. 41A). To determine whether
endogenous PKA mediates mitochondrial fusion, PC6-3 cells and hippocampal neurons
were transfected with a GFP fusion protein of PKI that contains the mitochondrial
targeting sequence of hexokinase I (OMM-PKI) and its effects on mitochondrial
130
morphology were analyzed by confocal microscopy. Inhibiting PKA at the mitochondria
by OMM-PKI prominently fragmented mitochondria in PC6-3 cells and in hippocampal
neurons, and reduced the mitochondrial fusion activity of forskolin in PC6-3 cells (Fig.
40A, and B; Fig. 41A, and B). These effects are a consequence of inhibiting PKA at the
OMM since PKI targeted to the nucleus (NLS-PKI) or untargeted PKI (cytosol) in PC6-3
cells does not fragment mitochondria compared to OMM-GFP (Fig. 40A, and B). In
order to determine whether endogenous AKAP121 regulates MFF, the two most potent of
four shRNAs (shown to downregulate at least 50% of exogenous AKAP121 expression,
Fig. 42A, Fig. 43C) were transfected in hippocampal neurons and PC6-3 cells and
analyzed for effects on mitochondrial morphology by confocal microscopy. Transient
transfection of both shRNAs robustly fragmented mitochondria in PC6-3 cells and in
hippocampal neurons (Fig. 41B; Fig. 42B, and C). It is possible that mitochondrial
fragmentation may occur as a consequence of a toxic effect of shRNAs in PC6-3 cells.
Arguing against a non-specific toxic effect, RNA mediated knock-down of AKAP121 did
not affect labeling of mitochondria by the potentiometric sensitive TMRM dye in PC6-3
cells (Fig. 42C). In aggregate, these results suggest that PKA is the kinase responsible for
promoting mitochondrial fusion in neurons.
PKA/AKAP121 promotes survival
PKA at the mitochondria delays cell death of spiral ganglion neurons in culture, in
PC6-3 cells and in hippocampal neurons [111, 162, 243]. In order to determine the extent
to which mitochondrial PKA promotes survival, hippocampal neurons were transfected
with various GFP fusion constructs of PKA and PKI and treated with 400nM rotenone
(achieves 50% apoptosis) for two days and analyzed for survival. Overexpression of
AKAP121-GFP but not the PKA binding incompetent AKAP121 mutant (I310P, L316P)
significantly decreased basal (p<0.05) and ROS mediated apoptosis (p<0.005) compared
to neurons transiently expressing OMM-GFP (Fig. 44B). Conversely, inhibiting
131
endogenous mitochondrial PKA by overexpressing OMM-PKI significantly increased
basal apoptosis (p<0.05) compared to neurons transiently expressing OMM-GFP (Fig.
44B). To determine whether endogenous AKAP121 is critical for survival, hippocampal
neurons were transfected with two AKAP121 shRNAs (AKAP121 shRNA#1 and
shRNA#4) and quantified for apoptosis. We noticed that transfection of both AKAP121
shRNAs is sufficient to increase apoptosis in the absence of toxicity (Fig. 44A and B).
Moreover, RNAi mediated knock-down of endogenous AKAP121 by two shRNAs
significantly increased rotenone induced apoptosis compared to neurons expressing a
scrambled shRNA control (p<0.05), (Fig. 44B). Unexpectedly, OMM-PKA was not as
efficient in protecting neurons against rotenone toxicity compared to AKAP121. In fact,
OMM-PKA overexpression unexpectedly decreased basal viability in hippocampal
neurons (Fig. 44B). One possible explanation is that excessive mitochondrial fusion by
OMM-PKA-GFP can be as detrimental as excessive mitochondrial fragmentation. In
aggregate, these results demonstrate that OMM-localized PP2A and -PKA holoenzymes
regulate mitochondrial morphology and neuronal survival.
Mitochondrial restructuring is sufficient for neuronal
survival regulation
There is some evidence that shows that MFF modulators can either drive
apoptosis by promoting mitochondrial fission or delay apoptosis by promoting
mitochondrial fusion [217, 218]. To determine whether mitochondrial remodeling
mediated by the MFF machinery is sufficient to regulate neuronal survival, we
transfected hippocampal neurons with different MFF modulators or with shRNAs that
target endogenous FIS1 and DLP1 and analyzed for dual effects on mitochondrial
morphology and survival. We found that transient expression of WT DLP1 disrupted the
mitochondrial network in hippocampal neurons while transient expression of DN-DLP1
fused mitochondria compared to OMM-GFP (Fig. 45A, and B). To determine whether
132
downregulating endogenous fission promoting proteins elongates mitochondria in
hippocampal neurons, one DLP1 specific shRNA and two FIS1 specific shRNAs which
potently downregulates at least 70% of exogenous expression of luciferase fusion
proteins DLP1 or FIS1 (Fig. 43A, B and D) were tested for effects on mitochondrial
morphology in hippocampal neurons. RNAi mediated knock-down of endogenous DLP1
or FIS1 dramatically fused mitochondria in hippocampal neurons compared to neurons
expressing OMM-GFP (Fig. 45A, and B). Furthermore, RNAi of FIS1 with one shRNA
reduced mitochondrial fragmentation induced by rotenone toxicity compared to neurons
transiently expressing OMM-GFP (Fig. 45A, and B). In agreement with previous reports,
these results demonstrate the MFF machinery regulates mitochondria morphology in
hippocampal neurons [217, 218]. Unpublished observations suggest that Bcl-xL, a Bcl-2
family protein, may regulate the rate of fission and fusion in neurons [244]. In order to
determine whether other Bcl-2 family members regulate MFF, we co-transfected
hippocampal neurons with Bcl-2 and GFP as a transfection marker and analyzed for
effects on mitochondrial morphology by confocal microscopy. In our assays, we did not
detect an effect of the antiapoptotic Bcl-2 protein on mitochondrial morphology in
hippocampal neurons under basal conditions or following toxic insult with rotenone (Fig.
45A, and B).
Since the MFF machinery modulates mitochondrial morphology in hippocampal
neurons, we wanted to determine whether MFF is required for neuronal survival. We
found that mitochondrial fragmentation induced by DLP1-GFP overexpression promoted
a four-fold increase in basal apoptosis in hippocampal neurons compared to neurons
transiently expressing OMM-GFP (p<0.02) and modestly increased rotenone mediated
apoptosis (p<0.05) (Fig. 46). Conversely, elongating mitochondria by transiently
expressing DN-DLP1 or by RNAi mediated knock-down of endogenous FIS1
significantly increased neuronal survival against rotenone toxicity compared to neurons
expressing OMM-GFP. In fact, neuronal survival induced by these constructs is
133
comparable to survival induced by overexpression of Bcl-2 (Fig. 46). Overall, these
results suggest that MFF is sufficient for regulating survival of neurons.
PP2A and PKA regulate mitochondrial morphology
upstream of apoptosis
Since mitochondrial fragmentation is obligatory and is intimately associated with
apoptosis, it is conceivable that mitochondrial fragmentation induced by PP2A/Bβ2 or by
PKA inhibition is downstream of apoptosis. To address this possibility, hippocampal
neurons were co-transfected with Bβ2-GFP or OMM-PKI and with Bcl-2 to block
apoptosis and analyzed for effects on mitochondrial morphology. As previously
demonstrated (Fig. 34B; Fig. 41A and B), co-transfecting Bβ2-GFP or OMM-PKI with a
scrambled shRNA robustly fragments mitochondria compared to OMM-GFP control
(Fig. 47A- C). However, co-expression of Bcl-2 significantly blocked basal apoptosis
induced by Bβ2 (p<0.0001), (Fig. 47D), but failed to block the mitochondrial fission
activity of PP2A/Bβ2 (Fig. 47 B) Likewise, co-expression of Bcl-2 significantly blocked
basal apoptosis induced by OMM-PKI (p<0.0001), (Fig. 47E) but failed to abolish its
mitochondrial fragmenting activity (Fig. 47A, and C). Furthermore, preliminary
experiments demonstrated that co-expression of Bcl-2 failed to block mitochondrial
fragmentation induced by knock-down of endogenous AKAP121 with two shRNAs in
neurons (Fig. 48A). Overall, these results suggest that mitochondrial fragmentation
induced by PP2A/Bβ2, OMM-PKI and by AKAP121 mediated knock-down is upstream
of cell death.
Mitochondrial restructuring is required for the survival
effects of PP2A/PKA
In order to determine whether mitochondrial remodeling is required for the
survival effects of PP2A/Bβ2 and OMM-PKA, hippocampal neurons were co-transfected
with phosphatase/kinase proteins and with MFF modulators for five days, and analyzed
134
for mitochondrial morphology by confocal microscopy or for apoptosis by the nuclear
morphology assay. Unexpectedly, we found that transfecting shRNAs that target FIS1 or
co-expressing DN-DLP1 was unable to rescue neurons from AKAP121 RNAi mediated
apoptosis (Fig. 48B). However, co-expressing Bcl-2 completely rescued neurons from
AKAP121 RNAi mediated apoptosis (Fig. 48B). It is conceivable that the inability of
MFF modulators to prevent apoptosis induced by RNAi mediated knock-down of
AKAP121 is due to a lack of an effect on mitochondrial fusion. However, preliminary
results show that co-expression of DN-DLP1 or FIS1 specific shRNAs inhibit
mitochondrial fragmentation induced by RNAi mediated knock-down of endogenous
AKAP121 (Fig. 48A). Overall, these results suggest that promoting mitochondrial fusion
does not rescue neurons from the absence of endogenous AKAP121. In addition to PKA,
these results suggest that other macromolecules- Src1, PTPD1 or MnSOD mRNArecruited by AKAP121 at the OMM or that ER targeted AKAP121 (AKAP121 shRNAs
target ER and mitochondrial targeted AKAP121) may be required for basal neuronal
survival independent of MFF.
In contrast, transient expression of DN-DLP1 or RNAi mediated knock-down of
endogenous FIS1 by two shRNAs blocked the apoptotic and mitochondrial fission
activity of Bβ2 (Fig. 47A, B and D). Likewise, mitochondrial fission and apoptosis
induced by overexpression of OMM-PKI was blocked by co-expression of DN-DLP1 or
by RNAi mediated knock-down of endogenous FIS1 (Fig. 47A, C and E). Overall, these
results suggest that reversible phosphorylation at the OMM dynamically regulates
mitochondrial morphology and neuronal survival via the MFF machinery.
Discussion
Mitochondrial dysfunction is associated with variety of neurodegenerative
diseases and with hypoxic-ischemic injury [131, 192, 245, 246]. In Parkinson’s disease,
substantia nigra neurons exhibit a decline in mitochondrial respiration and an increased
135
susceptibility to ROS induced oxidative damage [246-248]. The apoptotic pathways that
lead to mitochondrial dysfunction have begun to be addressed in different models of
neurodegenerative diseases. Reversible phosphorylation is a key posttranslational
mechanism that neurons employ to activate or inactivate apoptosis signaling pathways.
The regulation of some neuronal survival pathways that coalesce at the mitochondria is
attributed to reversible phosphorylation of Bcl-2 family proteins.
Here we describe a novel neuronal survival pathway that involves reversible
phosphorylation at the OMM and the MFF machinery. A neuron specific regulatory
subunit of the Bβ gene, mutated in SCA12, targets PP2A to the OMM to promote
apoptosis [17, 163]. Furthermore, the endogenous PP2A/Bβ2 regulates neuronal survival
since RNAi mediated silencing of the phosphatase dramatically protects neurons in an in
vitro model of ischemia and Parkinson’s disease. In summary, we identified a novel
mechanism by which PP2A/Bβ2 promotes neuronal apoptosis which requires
mitochondrial fragmentation mediated by fission promoting proteins. Furthermore, we
identified PKA as the opposing kinase that antagonizes mitochondrial fission and
apoptosis induced by PP2A/ Bβ2. We found that neuroprotection induced by PKA
targeted to the OMM by AKAP121 is a consequence of mitochondrial fusion via the
MFF machinery. In summary, we provide a model that suggests that OMM-PKA and
PP2A/Bβ2 are enzymes that participate in an apoptotic signaling pathway that requires
the MFF machinery (Model shown in Fig. 49).
Reversible phosphorylation alters mitochondrial
morphology
Previous studies have demonstrated that MFF modulates survival in non-neuronal
cells [198]. Our studies are the first to provide evidence that remodeling the morphology
of mitochondria is sufficient to regulate survival in neurons, and that MFF is regulated by
Ser/Thr kinases and phosphatases. Supporting this view is the observation that
136
overexpression of a PP2A holoenzyme targeted to the OMM promotes mitochondrial
fission upstream of Bcl-2. Furthermore, mitochondrial fragmentation induced by PP2A/
Bβ2 requires catalytic activity and incorporation into the holoenzyme since catalytic
inhibition of PP2A or transient expression of a monomeric mutant of Bβ2 promoted
mitochondrial fusion. Interestingly, PP2A inhibition by OMM-PfARP in glia or in PC6-3
cells confers mitochondria with a disc shaped or swollen morphology. Since PP2A/Bβ2
is not expressed in glia or in PC6-3 cells, this result suggests that other PP2A
holoenzymes at the OMM are critical for maintaining a proper mitochondrial morphology
(Fig. 36B, shown for PC6-3 cells). Moreover, overexpression of Bβ2 was unable to
disrupt the mitochondrial network of glia. It is conceivable that glia contain high
endogenous levels of OMM-PKA which may counteract the fission activity of exogenous
Bβ2 or that glia lack an unidentified neuron specific component of the MFF machinery
that is required by Bβ2 to promote fission.
Little is known regarding the signal transduction cascades that activate
mitochondrial fusion or fission. We found that activating a cAMP dependent signal
transduction pathway promotes mitochondrial fusion. Furthermore, the observations that
pharmacological activation of PKA causes mitochondrial fusion independent of protein
synthesis, does not alter total mitochondrial density, and occurs within minutes of
forskolin treatment are consistent with rapid phosphorylation of preexisting components
of the MFF machinery (Fig. 35B; Fig. 37B; Fig. 38B). Moreover, these phenomena
appears to be specific for mitochondrial targeted PKA since overexpression of AKAP121
or OMM-PKI but not nuclear targeted, cytosolic PKI or OMM-GFP promotes
mitochondrial fusion in PC6-3 cells (Fig. 39 A and B; Fig. 40A and B). There is some
evidence that provide insight into the signaling pathways or molecules that activate the
fission activity of DLP1. For instance, signals that trigger a rise in cytosolic Ca2+
stimulate the fission activity of DLP1. It has been demonstrated that treating cells with
Ca2+ ionophores promotes the redistribution of DLP1 from the cytosol to mitochondria to
137
cause fission. This suggests that Ca2+ may be an important regulator of mitochondrial
morphology [249]. One group reported that the Bax mediated release of DDP/TIMM8a,
an apoptogenic molecule, from the intermembrane space of mitochondria binds DLP1 in
the cytosol and targets the fission promoting protein to mitochondrial scission sites to
promote mitochondrial fragmentation [250]. Furthermore, little is known regarding the
mechanism by which DLP1 anchors the mitochondria. One group reported that that
DLP1 is a substrate of SumoI. SumoI was shown to interact with DLP1 in mitochondrial
scission sites to sumoylate the fission promoting protein. Sumoylated DLP1 is resistant to
degradation by the ubiquitin proteosome pathway and has a higher affinity for binding
the OMM where it concentrates at scission sites to promote fission. Thus, appropriate
anchoring of DLP1 to the OMM is required for fission and apoptosis induction [251]. In
neurons, our studies raise the possibility that the fission activity of DLP1 is stimulated by
dephosphorylation at the OMM mediated by PP2A/Bβ2. Thus, reversible
phosphorylation at the OMM mediated by PP2A/Bβ2 and OMM-PKA provides an
additional level of regulation of the MFF machinery.
Reversible phosphorylation regulates
survival through the MFF machinery
How does altering the balance of fission/fusion by reversible phosphorylation
affect survival? Our survival assays suggest that MFF is sufficient to regulate neuronal
survival and that the MFF machinery in turn is regulated by OMM mediated reversible
phosphorylation. Supporting this notion is the observation that overexpression of DLP1
or Bβ2 promotes mitochondrial fission and leads to a decrease in basal neuronal survival
whereas co-expressing DN-DLP1 inhibits the proapoptotic activity of Bβ2 (Fig. 34B;
Fig. 45B; Fig. 46; Fig. 47 A, B and D). It is conceivable that PP2A/ Bβ2 stimulates the
fission activity of DLP1 or FIS1, or may inhibit the fusion activity of Mfn1/2. We
propose that PP2A/Bβ2 sensitize neurons to apoptosis by stimulating the fission activity
138
of DLP1 by dephosphorylation. In the presence of toxic insult, PP2A/Bβ2 rapidly
translocates to the OMM and concentrates the catalytic activity of PP2A at sites of
mitochondrial scission where it dephosphorylates and activates the mitochondrial scission
activity of DLP1 (See model in Fig. 49). It has been reported that DLP1 rapidly
translocates to scission sites and associates with Bax proteins during the early stages of
apoptosis [205]. Consistent with this view, we found that mitochondrial translocation of
Bβ2 is rapid (t1/2=2.5 hr) and precedes apoptosis induction in hippocampal neurons
(t1/2=7.0 hr) (Fig.10). Thus, we propose that Bβ2 may shuttle PP2A to sites of
mitochondrial scission to interact and stimulate the fission activity of DLP1 by
dephosphorylation.
We also report novel mechanisms by which AKAP121 promotes survival in
neurons. Prior to our studies, other investigators have demonstrated that AKAP121
promotes neuronal survival by recruiting PKA to the OMM where it phosphorylates and
inactivates Bad at Ser (155) in the cytosol [111, 162]. However, although these studies
show that PKA at the OMM is necessary for survival, they failed to show that Bad
phosphorylation at Ser (155) is related to survival. Thus, the current model on AKAP121
mediated neuronal survival can be amplified given that the data on Bad phosphorylation
is not very strong. Adding to the current model, our studies suggest that the prosurvival
effect of AKAP121 is attributed to the mitochondrial fusion activity of OMM-PKA (Fig.
41B; Fig. 44B; Fig. 47E). We have found that abolishing PKA activity at the
mitochondria by expressing OMM-PKI fragments mitochondria and promotes apoptosis
presumably due to the unopposed activity of PP2A/Bβ2 at the OMM. Further implicating
PKA in survival, transient expression of a PKA binding deficient mutant of AKAP121
(I310, L316) does not confer survival against rotenone toxicity compared to W.T.
AKAP121 (Fig. 44B). However, the observation that AKAP121(I310P, L316P) does not
increase rotenone induced apoptosis is inconsistent with a proapoptotic effect of this
mutant as previously reported by Affaitati et al., 2003 in PC12 cells [162]. Unlike PC12
139
cells, it is likely that AKAP121 (I310P, L316P) in our studies does not compete for
endogenous AKAP121 in neurons and is simply inactive for mitochondrial fusion and
survival.
We also propose that a proper balance of MFF is critical for neuronal survival.
Excessive fusion (as mediated by DLP1 and PP2A/Bβ2) or fission can be equally
detrimental to neurons. Genetic mutations in GDAP1, a fission promoting protein, cause
a recessive form of peripheral neuropathy called Charcot-Marie-Tooth disease [215].
Consistent with this observation, our mitochondrial morphology and survival assays
demonstrated that excessive fusion promoted by OMM-PKA increases basal apoptosis
and mildly rescues neurons against rotenone toxicity (Fig. 44). Thus, a proper balance of
MFF is crucial for neuronal survival.
AKAP121 is vital to neuronal survival since our survival assays demonstrate that
silencing AKAP121 by RNAi significantly increases basal and ROS induced apoptosis
(p<0.005), (Fig. 44A and B). However, apoptosis by RNAi mediated knock-down of
AKAP121 could not be reversed by overexpressing DN-DLP1 or by RNAi mediated
silencing of endogenous FIS1 (Fig. 48B). In addition to recruiting PKA, AKAP121 also
recruits mRNAs, PTPD1 and Src kinases at the OMM [235-237]. Our results suggest that
AKAP121 interacting macromolecules other than PKA or that the ER-targeted AKAP121
(the N1 splice variant of AKAP121 is also targeted by AKAP121 specific shRNAs) are
essential for basal neuronal survival and that mitochondrial fusion is necessary but not
sufficient for the prosurvival effects of AKAP121. Another explanation is that AKAP121
mediated fusion of mitochondria via PKA may not protect against all forms of toxic
insults in neurons. In support of this notion are unpublished observations that
demonstrate that overexpressing the catalytic subunit of PKA promotes mitochondrial
fusion in spiral ganglion neurons but does not protect against cell death induced by coexpressing a constitutively active mutant of Bad (S112A) that cannot be phosphorylated
by PKA at this site [243] . However, there is little support regarding the contribution of
140
Bad in promoting apoptosis in hippocampal neurons. For instance, induction of seizurelike activity by removing kainuric acid in the media or glutamate mediated excitotoxicity
in hippocampal neurons promotes PP2B mediated desphosphorylation and translocation
of Bad, and complex formation with Bcl-XL. However, manipulations of this apoptotic
signaling pathway (treating neurons with inhibitors of calcineurin) confer partial or no
significant neuroprotection in hippocampal neurons indicating that t-Bid and caspases but
not Bad may be the major instigator of apoptosis in hippocampal neurons [252, 253].
Thus, the mechanism by which AKAP121 promotes survival of hippocampal neurons
involves fusion of mitochondria and possibly the targeting of other AKAP121 interacting
macromolecules to the OMM. Changes in Bad -phosphorylation, if they occur, may not
be relevant in our assays of hippocampal neuron survival.
Future experiments will be performed to determine the extent by which the
interacting partners of AKAP121 promote neuronal survival. For instance, deletion
mutagenesis of W.T. AKAP121 will be performed to truncate the binding sites of Src,
PTPD1 and mRNA analyzed for the effects of AKAP121 deletion mutants on basal
survival in hippocampal neurons.
Substrates of PP2A/PKA among the MFF machinery
The discovery that OMM-PKA and PP2A/Bβ2 regulate fusion and fission suggest
that MFF proteins may be potential substrates of these opposing enzymes. Furthermore,
the observation that PKA activation by forskolin elicits mitochondrial fusion suggests
that either the rate fission is decreased or rate of fusion is increased (Fig. 37C; Fig. 38B).
Dynamin in the synaptic terminal promotes endocytosis by constricting the membrane
stalk of the invaginating clathrin-coated vesicles. This process is regulated by reversible
serine phosphorylation where Cdk5 phosphorylates and inactivates while calcineurin
(PP2B) dephosphorylates and activates dynamin [254]. One attractive hypothesis is that
mitochondrial fission may be regulated by reversible phosphorylation of DLP1 in a
141
manner similar to dynamin. Amino acid alignment of dynamin family proteins reveals
two putative PKA sites preceding and following the GTPase effector domain of DLP1.
Supporting this concept, unpublished observations in our lab demonstrate that forskolin
increases the phosphorylation of endogenous DLP1 in PC6-3 cells metabolically labeled
with 32P suggesting that the fission promoting GTPase is a substrate of PKA. It is
conceivable that PKA at the OMM phosphorylates and inactivates the fission activity of
DLP1 by serine phosphorylation. To test this hypothesis, it will be critical to determine
whether PKA recruited to the OMM by AKAP121 mediates phosphorylation and
inactivation of DLP1. Conversely, we observed that Bβ2 translocates to the OMM to
promote fission and apoptosis. In a similar manner to PP2B mediated dephosphorylation
of dynamin, we propose that PP2A/Bβ2 may dephosphorylate DLP1 at serine residues to
activate its fission promoting activity. However, future experiments will be required to
verify this hypothesis.
Other potential physiological functions of PP2A/Bβ2
Mitochondrial fission mediated by DLP1 was found to be necessary for the
transport of mitochondria into dendrites to stimulate the development of dendritic spines.
At the postsynaptic compartment, the fission activity of DLP1 promotes the transport of
mitochondria into synaptic terminals which was found to be required for replenishing the
reserve pool of neurotransmitter vesicles [217, 218]. Although PP2A/ Bβ2 in these
studies have been shown to be a regulator of mitochondrial morphology and survival, is
very likely that Bβ2 may play additional physiological roles such as promoting the
development of dendritic spines or replenishing the reserve pool of neurotransmitter
vesicles. Consistent with this view, the expression profile of Bβ2 during postnatal
development of the rat brain shows a dramatic increase in expression that peaks at the
second week which coincides with synapse formation and stabilization in rodents.
142
Possible mechanisms by which mitochondrial
fragmentation may promote apoptosis
Fission is intimately linked to apoptosis since inhibiting mitochondrial fission
promotes survival against toxic insult. How does mitochondrial fragmentation facilitate
apoptosis? During apoptosis, proapoptotic Bax proteins translocate to sites of
mitochondrial scission to permeabilize the OMM and promote the release of apoptogenic
factors such as cytochrome c which in turn activate downstream effector caspases.
Furthermore, the translocation of Bax coincides with the translocation of DLP1 during
apoptosis [205]. Mitochondrial fission has been shown to be required for Bax mediated
permeabilization of mitochondria and release of cytochrome c [206]. Conversely, it has
been shown by Neuspiel et al., 2003 that mitochondrial fusion mediated by Mfn1/2
mitochondria inhibits Bax mediated permeabilization of mitochondria and the release of
cytochrome c during apoptosis (reviewed by [213]). Thus, remodeling of mitochondria is
required for Bax mediated release of apoptogenic factors. Another mechanism by which
mitochondrial fission contributes the release of cytochrome c is by altering the
morphology of the cristae of mitochondria. It has been demonstrated that fission
increases the gap where the cristae meets the OMM and these regions in mitochondria are
“hot spots” where the C-terminal tails of proapoptotic proteins such as Bid and Bik
anchor to further widen the gap, facilitate the release of cytochrome c and consequently
activate downstream caspases (reviewed by [213]).
Another mechanism by which fission may promote apoptosis is by facilitating the
opening of the permeability transition pore. It has been demonstrated that fusion of
mitochondria by Mfn1/2 reduces the susceptibility of cells to ROS mediated
mitochondrial depolarization and opening of the permeability transition pore which leads
to the release of apoptogenic factors such as cytochrome c [213]. Although fission has
not been proven to cause the opposite effect, it is likely that fragmenting mitochondria
143
leads to an increased sensitivity of cells to ROS mediated depolarization and opening of
the permeability transition pore.
Mitochondrial interconnectivity can influence the propagation of cytosolic
calcium waves [186]. Mitochondrial fragmentation mediated by DLP1 has been shown to
protect against ceramide induced apoptosis by interfering with the network dependent
calcium wave propagation [209]. The latter observation suggests that apoptosis mediated
by ceramide requires fusion as opposed to fission as seen in other forms of apoptosis. In
neuronal cells, unpublished results performed by Dr Yuriy Usachev at the University of
Iowa suggest that Bβ2, a mitochondrial fission promoting protein, increases the calcium
uptake in mitochondria as determined by Fura-2 measurements of glutamate induced
Ca2+ transients in hippocampal neurons. The latter result suggests that PP2A/Bβ2 may
sensitize neurons to apoptosis by increasing the Ca2+ uptake of mitochondria which has
been shown to be required for glutamate induced excitotoxicity. However, additional
experiments will be carried out to determine whether Bβ2 mediated mitochondrial Ca2+
uptake requires mitochondrial fission as outlined in chapter V.
144
Figure 33 Overexpression of Bβ2 (but not Bβ1) fragments mitochondria in nondifferentiated and differentiated PC6-3 cells. A: PC6-3 were transfected
with Bβ1- and Bβ2-GFP for three days. In some experiments, cells were
treated with NGF (10ng/ml) for 18 hrs. To analyze for mitochondrial
morphology, cells were loaded with 100nM TMRM and imaged by confocal
microscopy (Z plane confocal sections, mito=TMRM stain). Scale bar=10 µm.
B: quantification of mitochondrial morphology was scored using a subjective
scale (0-3) in naïve and in differentiated PC6-3 cells transfected with the
indicated proteins (mean±SEM from 35-45 neurons each from at least 3
experiments) .Significant differences to Bβ1: *p<0.05
145
Figure 34 PP2A/Bβ2 (w.t. but not mutants) fragments mitochondria in hippocampal
neurons. A: hippocampal neurons (15DIV) were transfected with the
indicated cDNAs or infected with lentiviruses that express Bβ1-, Bβ2-GFP,
scrambled or Bβ2 specific shRNAs for five days. GFP and TMRM stained
mitochondria were imaged by confocal microscopy. Scale bar=10 µm. B:
quantification of mitochondrial morphology was assessed blindly with a
subjective scale (0-3) in hippocampal neurons (# of experiments indicated, 2530 neurons each) PfARP= Plasmodium falciparum aspartate rich protein.
Significant differences to OMM-GFP for overexpression experiments or to
scrambled shRNA for RNAi experiments: *p<0.05, **p<0.0001. The
lentivirus that overexpresses Bβ2 shRNA#1 was generated by Dr. Ronald
Merrill.
146
Figure 35 Algorithmic quantification of mitochondrial morphology and
mitochondrial density. A: representative confocal images of somatic
mitochondria from hippocampal neurons (15 DIV) expressing OMM-PfARP
(MAS701-29-GFP-PfARP) or OMM-GFP (MAS701-29-GFP). Scale bar=10 µm.
B and C: hippocampal neurons were transfected with the indicated GFP
fusion constructs for at least three days. Mitochondrial morphology and
mitochondrial density was quantified algorithmically by using an Image J
macro (mean±SEM from at least 17 neurons each quantified from at least 2
experiments).Significant differences to Bβ1 or OMM-GFP: *p<0.05
147
Figure 36 Inhibiting mitochondrial PP2A alters mitochondrial morphology in PC63 cells. A: domain diagram of Bβ21-35-GFP-PfARP and MAS701-29-GFPPfARP(OMM-PfARP), an endogenous inhibitor of PP2A targeted to
mitochondria via the divergent N-terminus of Bβ2 or via the VDAC/porin
sequence of hexokinase I. B: representative confocal images of PC6-3 cell
transiently expressing Bβ21-35-GFP-PfARP or MAS701-29-GFP-PfARP. Scale
bar=10 µm. Note that Bβ21-35-GFP-PfARP causes mitochondrial swelling
while mitochondria of cells expressing MAS70-GFP-PfARP and
untransfected cells display a normal mitochondrial morphology (red). C:
Bβ21-35-GFP-PfARP but not Bβ21-35-GFP (control) causes an increase in
phosphorylation on Thr-Pro residues of high molecular weight proteins (160200kDa). This experiment was carried out by Dr. Stefan Strack. D: Bβ21-35GFP-PfARP but not MAS701-29-GFP-PfARP is proteolytically processed in
mitochondria suggesting that Bβ21-35-GFP-PfARP is targeted to import
receptors (solid arrow=full-length protein, open arrow=processed protein).
148
Figure 37 Pharmacological activation of PKA elongates mitochondria in PC6-3 cells
and in hippocampal neurons. A: representative images of MitoTracker
stained mitochondria of PC6-3 cells or hippocampal neurons treated with
2µM forskolin or with DMSO as a vehicle control. Hippocampal neurons
were identified by looking for their raised, phase bright cell bodies using
differential interference contrast (DIC). Scale bar=10 µm. B: quantification of
mitochondrial length of PC6-3 cells treated with the indicated
pharmacological activators of PKA or with NGF and of hippocampal neurons
treated with forskolin (mean±SEM from at least 2 experiments, 15-35 neurons
each). All length increases are significant (p<0.05).
149
Figure 38 Forskolin induces rapid mitochondrial elongation independent of protein
synthesis. A: representative images of TMRM stained mitochondria of a
PC6-3 cell captured at the same confocal plane in the absence or presence of
50µM forskolin for 30 min. Note the dramatic increase in mitochondrial
fusion after forskolin treatment at the indicated zoomed-in region of interest.
B: PC6-3 cells were treated with DMSO, or with 2µM forskolin in the
presence or absence of 100µg/ml cyclohexamide as indicated and scored for
mitochondrial length after 30 and 60 min of forskolin treatment (mean±SEM
from a representative experiment of 25-35 neurons each). All length increases
are significant (p<0.0001).
150
Figure 39 Overexpression of AKAP121 elongates mitochondria in PC6-3 cells.
A: representative confocal sections of TMRM-stained (red) mitochondria of
PC6-3 cells transfected with OMM targeted GFP (OMM-GFP), and
AKAP121-GFP. Scale bar=10 µm. Note that AKAP121 (middle panel) but
not OMM-GFP causes robust mitochondrial fusion and its effects is increased
with higher expression levels causing mitochondrial clumping (bottom panel).
B: quantification of mitochondrial length of untransfected and of PC6-3 cells
transiently expressing OMM-GFP, or AKAP121-GFP (mean±SEM from at
least 2 experiments, 15-35 neurons each). Significant differences to control
(untransfected or OMM-GFP): *p<0.05. The AKAP121 construct was
generated by Jason Ulrich. Confocal analyses of mitochondria in PC6-3 cells
was performed by Jason Ulrich.
151
Figure 40 Inhibiting PKA at the OMM fragments mitochondria in PC6-3 cells.
A: representative confocal sections of TMRM-stained (red) mitochondria of
PC6-3 cells transiently expressing OMM targeted GFP proteins (OMM-GFP,
hexo-I-GFP), nuclear targeted PKI (NLS-PKI) and PKI targeted to the OMM
(OMM-PKI). Scale bar=10 µm. Note that OMM-PKI but not OMM-GFP or
NLS-PKI causes robust mitochondrial fragmentation. . B: quantification of
mitochondrial length of untransfected or of PC6-3 cells transiently expressing
OMM-GFP, NLS-PKI, or OMM-PKI with or without forskolin (2 µM)
treatment (mean±SEM from the number of experiments indicated per
condition, 15-35 neurons each). *p<0.05, **p<.005
152
Figure 41 Overexpression of AKAP121 promotes fusion while RNAi fragments
mitochondria in hippocampal neurons. A: representative confocal images
of TMRM-stained mitochondria (red) of hippocampal neurons transfected
with the indicated GFP fusion proteins or with shRNA plasmids directed
against AKAP121. Note that OMM-PKA causes clumping of mitochondria.
Scale bar=10 µm. B: quantification of mitochondrial length of hippocampal
neurons transiently expressing the indicated GFP fusion proteins or shRNAs
plasmids (mean±SEM from the number of experiment indicated, 15-35
neurons each). Significant differences to OMM-GFP for overexpression
experiments or to scrambled shRNA for RNAi experiments: *p<.05,
**p<.0001. The AKAP121 (I310P, L316P) construct was generously
generated by Thomas Cribbs.
153
Figure 42 RNAi mediated knock-down of AKAP121 fragments mitochondria in
PC6-3 cells. A: PC6-3 cells were co-transfected at 0.2:1 and 1:1 plasmid
ratios of AKAP121-luciferase fusion proteins and pSUPER plasmids
expressing shRNAs targeting luciferase (fluc) as a positive control, or
AKAP121 and assayed for knock-down by the dual luciferase assay. B:
Representative confocal images of TMRM stained mitochondria (red) of
PC6-3 cells transfected with control or AKAP121 specific shRNAs (identified
by GFP fluorescence as a marker for transfection). Scale bar=10 µm.
C: quantification of mitochondrial length in PC6-3 cells expressing scrambled
control or AKAP121 specific shRNAs (mean±SEM of 30-40 neurons each
from a representative experiment). Significant differences to scrambled
shRNA: *p<0.01. All experiments were done in collaboration with Jason
Ulrich.
154
Figure 43 Tools used for manipulating the MFF in hippocampal neurons.
A, B: COSM6 cells were co-transfected with a FIS1-luciferase fusion protein
or DLP1-GFP and with pSUPER plasmids expressing shRNAs targeting
luciferase (fluc), FIS1 or DLP1 proteins at the indicated plasmid ratios, and
with Renilla luciferase as a marker for transfection efficiency. After 3 days
post-transfection, knock-down of luciferase fusion proteins were assayed by
the dual luciferase assay or by immunoblotting for GFP (FIS1-GFP). C and
D: hippocampal neurons were co-transfected with DLP1- or AKAP121luciferase fusion proteins and with pSUPER plasmids expressing a scrambled
shRNA or shRNAs targeted against AKAP121 and DLP1 proteins and
assayed for knock-down by the dual luciferase assay. All experiments were
performed by Dr. Stefan Strack and Thomas Cribbs.
155
Figure 44 PKA at the OMM promotes survival while inhibition of PKA or RNAi
of AKAP121 promotes apoptosis in hippocampal neurons. Hippocampal
neurons were transfected with the indicated GFP fusion proteins or shRNA
plasmids. After 5 d. post-transfection, hippocampal neurons were treated with
rotenone (400nM) or DMSO vehicle control for 2 d. Hippocampal cells were
then fixed in 3.7% PF, immunostained for GFP and counterstained with a
nuclear dye (Hoescht 33342). A: representative microscopic fields of
hippocampal neurons transfected with scrambled or AKAP121 shRNAs and
GFP as a transfection marker. Note that neurons that express AKAP121
shRNAs but not scrambled shRNA degenerate and undergo apoptosis. The
white arrows point to apoptotic nuclei while yellow arrows point to intact
nuclei. Scale bars: 20 µm. B: The percentage of transfected neurons
containing apoptotic nuclei was quantified for each transfection condition.
(mean±SEM from at least 2 experiments, 100-200 neurons each). Significant
differences to OMM-GFP for overexpression experiments or to scrambled
shRNA for RNAi experiments: *p<.05, **p<.005
156
Figure 45 The MFF machinery regulates mitochondrial remodeling in hippocampal
neurons. Hippocampal neurons were transfected with GFP tagged MFF
modulators, Bcl-2 or with the indicated shRNAs plasmids. After 5 d. posttransfection, hippocampal neurons were treated with rotenone (400nM) or
DMSO vehicle control for 2 d and analyzed for mitochondrial morphology
(identified by GFP fluorescence as a marker for transfection). A:
representative confocal images of TMRM stained mitochondria (red) of
hippocampal neurons transfected with the indicated plasmids or shRNAs.
Scale bar=10 µm. B: quantification of the effects of overexpressing or
knocking-down different MFF modulators on mitochondrial length in
hippocampal neurons. (mean±SEM from at least 1-4 experiments, 15-35
neurons each). Significant differences to OMM-GFP: *p<.01, **p<.0001. The
pSUPER plasmids containing the shRNAs that target FIS1 and DLP1 were
generated and tested for knock-down by Thomas Cribbs.
157
Figure 46 Mitochondrial restructuring is sufficient for neuronal survival regulation.
Hippocampal neurons were transfected with the indicated GFP tagged MFF
modulators, Bcl-2 or with shRNA plasmids directed against endogenous FIS1
protein. After 5 d. post-transfection, hippocampal neurons were treated with
rotenone (400nM) or with DMSO vehicle control for 2 d. Hippocampal cells
were then fixed in 3.7% PF, immunostained for GFP and counterstained with
a nuclear dye (Hoescht 33342). The percentage of transfected neurons
containing apoptotic nuclei was quantified for each transfection condition.
(mean±SEM of at least 2 experiments, 100-200 neurons each). Significant
differences to OMM-GFP: *<p.05, **p<.005
158
Figure 47 Mitochondrial restructuring and neuronal survival by PP2A/PKA is
upstream of cell survival and requires the MFF machinery.
A: Representative confocal images of TMRM stained mitochondria (red) of
hippocampal neurons co-transfected with the indicated GFP fusion proteins, ±
shRNA plasmids, or ± Bcl-2 and identified by GFP fluorescence as a marker
for transfection. Scale bar=10 µm. B, C: quantification of mitochondrial
length scores in hippocampal neurons transfected with the indicated plasmids
and shRNAs. D, E: hippocampal neurons were transfected with the indicated
plasmids and shRNAs. 5 d. postransfection, hippocampal neurons were fixed
in 3.7% PF, immunostained for GFP, counterstained with a nuclear dye
(Hoescht 33342). The percentage of transfected neurons containing apoptotic
nuclei was quantified for each transfection condition. (mean±SEM of at least
2 experiments, 100-200 neurons each) *p<.05, **p<.005, ***p<.0005.
159
Figure 48 Mitochondrial restructuring is necessary but not sufficient for the
neuronal survival effects of AKAP121.
A. quantification of mitochondrial length in hippocampal neurons transfected
with shRNAs targeting AKAP121 in combination with other shRNAs or GFP
fusion proteins and identified by GFP fluorescence as a marker for
transfection. B: quantification of apoptosis in hippocampal neurons
transfected with plasmids and shRNAs as indicated above. 5 d.
postransfection, hippocampal neurons were fixed in 3.7% PF, immunostained
for GFP and counterstained with a nuclear dye (Hoescht 33342). The
percentage of transfected neurons containing apoptotic nuclei was quantified
for each transfection condition. Significant differences to scrambled shRNA in
A or as indicted by brackets in B: *p<.05, **p<.005
160
Figure 49 Model summary: dynamic regulation of MFF and neuronal survival by
OMM-PP2A and PKA. A neuron specific splice variant of the regulatory
subunit Bβ gene termed Bβ2 promotes translocation of PP2A from the cytosol
to mitochondrial import receptors via its divergent N-terminus to promote
fission and apoptosis. PKA anchored at the OMM via AKAP121 opposes the
effects of PP2A/ Bβ2 by preventing mitochondrial fragmentation and
stimulating neuronal survival. Other molecules recruited by AKAP121 such
as PTPTD, Src1 and mRNA at the OMM may be essential for basal neuronal
survival independent of the MFF machinery.
161
CHAPTER V
CONCLUSION
The importance of subcellular targeting of PP2A in
regulating neuronal functions
Neurons contain specialized compartments such as dendrites and axons with each
harboring unique signal transduction pathways that mediate diverse functions such as
synaptogenesis, survival and synaptic plasticity. Reversible phosphorylation is a
posttranslational mechanism mediated by kinases and phosphatases. It was once believed
that phosphatases promiscuously dephosphorylate substrates since they rely on fewer
catalytic subunits than kinases (~140 phosphatases compared to 518 kinases identified in
the human genome). How phosphatases and kinases are able to mediate reversible
phosphorylation of a vast amount of substrates to regulate signal transduction pathway in
different subcellular compartments have been a subject of high interest. Neurons rely on
the compartmentalization of select pools of phosphatases and kinases to regulate diverse
neuronal functions. The compartmentalization of kinase/phosphatase activity in neurons
is mediated by anchoring proteins, which associate with and target kinases/phosphatases
to distinct subcellular compartments to regulate diverse neuronal functions such as
survival, glutamate receptor activation, dendritic morphology and calcium homeostasis.
For instance, AKAP79/150 forms a large signaling complex that consists of two kinases
(PKA, PKC) and a phosphatase (PP2B) at the plasma membrane by associating with
membrane associated guanylyl kinase (MAGUK) to regulate multiple cell membrane
receptors that include AMPA, NMDA and the β-adrenergic receptor (reviewed by [255]).
The specificity of PP2A is conferred by association with a large array of
regulatory subunits. Supporting this notion is the discovery that regulatory subunits of
PP2A show differences in tissue expression and in subcellular localization [67, 120]. In
chapter 2 of my thesis I present evidence that supports the view that the substrate
162
specificity of PP2A relies on subcellular location. By using deletion mutagenesis and
holoenzyme association assays, we showed that truncating the divergent N-termini of Bβ
isorforms does not compromise binding to A and C subunits of PP2A suggesting that the
divergent N-terminus does not play a role in substrate recognition.
There are different possibilities by which regulatory subunits may mediate
substrate recognition of PP2A. One possibility is that the association of regulatory
subunits to the AC core induces an allosteric conformation in the catalytic subunit to
increase the affinity and specificity of the catalytic subunit toward substrates. However,
there is currently no evidence that supports this model. In fact, the association of MYPT1
regulatory subunit to the C subunit does not induce a change in conformation of the
catalytic cleft when compared to the crystal structures of monomeric PP1, or PP1 bound
to okadaic acid, microcystin, or calyculin A [20, 155-157]. It is most likely that
regulatory subunits of PP2A may directly dock to substrates to mediate substrate
recognition. The association of MYPT1 to the catalytic subunit of PP1 does not lead to
global changes in the tertiary structure but expands the substrate binding surface to
include a long acidic groove on MYPT1 leading to an increased affinity for basic residues
in myosin. Thus, the PP1/MYPT1 crystal structure suggests a structural basis of substrate
recognition by PSPs [20]. In vitro phosphatase assays performed by different
investigators have highlighted the importance of regulatory subunits in mediating
substrate recognition of PP2A. For instance, Kamibayashi et al., 1994 demonstrated by
performing in vitro phosphatase assays that PP2A holoenzymes containing different
regulatory subunits (Bα, Bβ, B’β) show differences in catalytic activity and specificity
toward substrates such as phosphorylated histone [256]. Other in vitro phosphatase assays
have demonstrated that two PP2A holoenzymes containing Bα or Bδ but not B’
regulatory subunits mediate dephosphorylation and inactivation of ERKs. Moreover, the
Bβ isoforms were shown to contain similar catalytic activities in vitro toward
phosphorylated casein and myelin basic protein [17, 18]. The Bα regulatory subunit has
163
been show to play a critical role in mediating substrate specificity of PP2A. A PP2A
holoenzyme containing the Bα subunit has been shown to specifically dephosphorylate
substrates phosphorylated by p34cdc2/cyclin B kinase, a kinase that regulates the G2-M
cell cycle [257, 258]. In vivo, it was shown that P element insertions that mapped in the
PR55 gene leads to mitotic and severe developmental defects in Drosophila [259].
Overall, ours and previous results support a role of the B subunit β-propeller in directly
docking to substrates or extending the substrate binding surface of the C subunit. The
divergent N-termini further restrict substrate specificity in vivo by mediating subcellular
targeting.
There is evidence that supports a role of B’ and B’’ family of regulatory subunits
in mediating substrate recognition by directly docking to substrates. For instance, B’ε
was demonstrated to interact with APC by yeast two hybrid screens. PP2A/ B’ε interacts
with APC to dephosphorylate components of the APC signaling complex leading to
inhibition of beta-catenin expression [58]. Furthermore, PR48 is a nuclear targeted B’’
family regulatory subunit that has been shown to interact with the N-terminal domain of
cdc6, a component of the prereplication complex, by yeast two hybrid screens.
PP2A/PR48 associates and dephosphorylates cdc6 at serine residues to modulate DNA
replication in human cells [66]. Moreover, subcellular targeting of B’ family of
regulatory subunits may also play a role in mediating substrate specificity of PP2A. For
instance, B’ subunits that localize in the cytosol (B’β, B’ε) regulate the Wnt/β catenin
pathway and LTP, in mitochondria (B’α) to regulate survival or in the nucleus (B’γ1) to
reorganize subnuclear structures [58, 63, 117, 260, 261].
Adding to the diversity of mechanisms of catalytic regulation of PP2A, B’
regulatory subunits are regulated by serine phosphorylation [53, 260]. For instance, B’
subunits specifically dephosphorylates and inactivates ERK. However, ERK in turn can
regulate B’ subunits by serine phosphorylation. IEX-1, an immediate early gene product
induced by growth factors, forms a complex with B’ subunits and phosphorylated ERK
164
to mediate the ERK dependent phosphorylation of B’ subunits at serine residues (ie, Ser
327 for B’γ ) promoting dissociation of B’ subunits from the PP2A holoenzyme [262].
Thus, other factors such as phosphorylation in addition to subunit composition of PP2A
heterotrimers impact the catalytic activity of PP2A.
Role of PP2A/ Bβ2 in neurodegeneration
PP2A can be both a positive and negative regulator of cell survival. While PP2A
mediated dephosphorylation of Bad and Bcl-2 promotes apoptosis, other PP2A
holoenzymes containing B’ subunits identified in Drosophila are prosurvival [26, 62,
113]. Recent evidence shows that the regulatory subunits target PP2A to different
substrates to activate apoptotic signaling cascades. The B’α subunit promotes PP2A
mediated dephosphorylation/inactivation of Bcl-2 in mitochondria and sensitizes
lymphocytes to ceramide induced apoptosis [117]. Prior to my dissertation work, there
was no knowledge regarding a PP2A holoenzyme that regulates neuronal survival. In
chapter II, I characterized a PP2A holoenzyme containing a splice variant of the Bβ gene
that regulates survival in neurons. Since the Bβ gene gives rise to multiple splice variants
with differences in subcellular localization, I wanted to determine whether the divergent
N-termini of the Bβ isoforms encode subcellular targeting signals. By performing
confocal microscopy in neurons, I demonstrated that Bβ2 but not other Bβ isoforms is
localized to both cytosolic and mitochondrial compartments in healthy neurons.
However, I found that cells stress induced the rapid redistribution of Bβ2 from the
cytosol to mitochondria. Implicating Bβ2 as a regulator of neuronal survival, I
determined that overexpression of Bβ2 in hippocampal neurons by transient, and
lentiviral expression promotes basal and rotenone induced apoptosis while knock-down
of endogenous Bβ2 is neuroprotective in vitro models of ischemia and Parkinson’s
disease. Furthermore, Bcl-2 co-expression blocked the apoptotic activity of Bβ2
165
suggesting that PP2A/Bβ2 acts upstream of OMM permeabilization and cytochrome c
release to regulate survival.
SCA12 is an autosomal dominant neurodegenerative disease caused by a mutation
in one of the presumed promoters of the Bβ gene (PPP2R2B). Immediately upstream of
the Bβ1 initiation site, a CAG repeat is expanded from a normal length of 10-14 repeats
to up to 90 repeats. SCA12 patients show widespread neurodegeneration and atrophy of
cortical layers and the cerebellum which leads to the development of ataxia, loss of gait
and posture, tremors and cognitive decline. Among CAG repeat expansions diseases,
SCA12 is unusual in that the mutation maps to a non-coding sequence [87]. To this date,
the molecular mechanism of pathogenesis for SCA12 has not been elucidated (reviewed
by [263]). It is conceivable that the SCA12 mutation may alter promoter usage of
upstream promoters or splicing of upstream exons leading to increased expression of one
or more isoforms. Supporting this view, unpublished observations by Russ Margolis
from the University of Hopkins demonstrated that the SCA12 mutation leads to an
increase in promoter activity as shown by a luciferase based reporter system suggesting
that increased Bβ1 expression may be involved in SCA12 pathogenesis. (reviewed by
[263]). It is not known whether the SCA12 mutation alters expression levels of Bβ2 (or
any of the other Bβ splice variants). Bβ2 is generated by cis splicing of a huge premRNA (~500 kb) spanning the Bβ locus. We hypothesize that the SCA12 mutation may
be a cis-acting mutation, which may affect the use of alternate splice sites. It is
conceivable that the CAG repeat expansion may induce changes in the secondary
structure of the Bβ2 pre-mRNA which may interfere with the ability of the spliceosome
complex to splice and ligate exon 1.1 to exons 2-9. This in turn leads to a favorable
splicing and fusion of exon 1.2 to exons 2-9 and an increase in the generation of
Bβ2 mRNA. Thus, an increase in ratio of Bβ2:Bβ1 may be detrimental to neurons and
may contribute to the pathogenesis of SCA12.
166
Furthermore, we argue against an effect of increased Bβ1 expression on SCA12
pathogenesis since we have shown that transient, stable or lentiviral mediated
overexpression of Bβ1 leads to a modest or no effect on apoptosis in neurons. To this
end, we favor the view that Bβ2 contributes to the pathogenesis of SCA12 since our
survival assays convincingly demonstrate that Bβ2 but not Bβ1 promotes apoptosis in
neurons in vitro. Thus, a functional increase of Bβ2 may contribute to the pathogenesis of
SCA12 by causing late onset neurodegeneration
Bβ2 based therapies
There is currently no effective therapy of stroke. Despite showing early promise
in rodent ischemia models, ionotropic glutamate receptor antagonists have failed in
human clinical trials [132]. On the other hand, preventing ischemic neuronal death
downstream of glutamate receptors should significantly widen the treatment window,
perhaps to several days after the stroke. Bβ2 is an attractive target of stroke therapies
since its expression is restricted to brain and since mitochondrial dysfunction is
protracted. It is important to point out that developing mitochondrial targeted inhibitors
of PP2A may not be a viable neuroprotective therapy since our in vitro survival assays in
hippocampal neurons demonstrated that inhibiting mitochondrial PP2A using a
mitochondrial targeted inhibitor of PP2A (OMM-PfARP) is toxic. However, the
observation that PP2A/Bβ2 relocalizes to the OMM via its N-terminus to promote
apoptosis supports the concept that translocation inhibitors of Bβ2 may have the potential
to be employed as neuroprotective agents (antiapoptotic drugs). Currently, there are a few
antiapoptotic drugs that target apoptotic signal transduction pathways at or downstream
of mitochondria under clinical development with the potential for treating
neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease (reviewed by
[136, 264]). A rational approach is to develop intracellular single-chain antibodies
(scFvs) specific for the N-terminus of Bβ2 which should promote neuroprotection by
167
blocking the translocation of PP2A/Bβ2 to mitochondria. Recently, antibody engineering
technology has been used to develop intrabodies which may have the potential to be
employed as therapeutic drugs for treating various disorders that include cancer and
Huntington’s disease [265, 266]. Our lab has recently developed hybridomas that express
Bβ2 specific monoclonal antibodies. By using mRNA from hybridomas that express Bβ2
antibodies, it will be possible to amplify the variable segments of immunoglobin heavy
(VH) and light chains (VL) by RT-PCR. The amplified VL and VH PCR products will be
used generate a phage display library that produces Bβ2 directed scFvs. Plaque colonies
that produce Bβ2 directed scFvs will be identified for their ability to bind to purified
recombinant Bβ21-27-GST. Bβ2 directed scFvs that confer neuroprotection of
hippocampal neurons against different toxic insults in vitro can be evaluated as a
prototype for therapies of stroke and neurodegenerative diseases. We propose that Bβ2
translocation inhibitors should be devoid of teratogenic and non-specific side effects in
other tissues since Bβ2 is expressed in the adult brain and is not expressed in embryos.
Implications of OMM PKA/PP2A mediated reversible
phosphorylation in modulating neuronal functions
As previously reviewed in chapter II, mitochondria are not merely ‘powerhouses’
of the cell but are also high capacity calcium buffers, modulators of apoptosis signal
transduction pathways, main source and sinks of ROS production, and sites of multiple
metabolic pathways. In addition, mitochondria are not static but are dynamic structures
that display motility and have the capacity to acquire a spherical, elongated morphology,
or form a continuous network (reviewed by [204]). The role of OMM reversible
phosphorylation in modulating mitochondrial functions such as apoptosis signaling has
been a topic of high interest over the past twenty years. Prior to my dissertation, little
was known regarding the mechanism by which OMM-reversible phosphorylation
modulates neuronal survival.
168
In chapter IV, we identified a novel mechanism by which PP2A/Bβ2 promotes
neuronal apoptosis. PP2A/Bβ2 overexpression sensitizes neurons to apoptosis by
promoting mitochondrial fragmentation in a process that requires the MFF machinery.
Furthermore, we identified PKA at the OMM as the kinase responsible for antagonizing
fission and apoptosis induced by PP2A/Bβ2. Thus, OMM-PKA and PP2A/Bβ2 are
enzymes that reciprocally regulate neuronal survival via the MFF machinery.
It is clear that OMM-reversible phosphorylation by kinase/phosphatase signaling
may regulate other neuronal functions through the MFF machinery besides survival.
Supporting this notion are the observations that mitochondrial fragmentation induced by
DLP1 is necessary for the transport of mitochondria to axonal terminals and dendritic
spines which has functional implications on synaptogenesis and neurotransmission [217,
218].
Dendritic spines are dynamic structures composed of spine heads and necks that
have the capacity to change shape. For instance, the spine head of dendritic spines can
acquire a stubby, mushroom or elongated appearance [267]. The induction of LTP by
high frequency stimulation in hippocampal neurons increases the size and number of
dendritic spines [268]. During the last decade, there has been a lot of research that has
unveiled the molecular mechanisms that regulate synaptic terminal remodeling. The
formation, remodeling and maintenance of dendritic spines are regulated by a variety of
factors that include hormonal fluctuations, changes in temperature, and by long term
potentiation (LTP) (reviewed by [269]). Most recently, it has been shown that dendritic
mitochondria are essential for remodeling and for the development of dendritic spines in
hippocampal neurons. Mitochondrial fission of somatic mitochondria mediated by DLP1
drives the transport of mitochondria into dendrites to stimulate the development of
dendritic spines in hippocampal neurons. Conversely, promoting mitochondrial fusion by
overexpressing DN-DLP1 depletes dendritic spines of mitochondria and retards
synaptogenesis in hippocampal neurons [217].
169
At the postsynaptic terminal, mitochondrial ATP is essential for functional
refilling/mobilization of the readily releasable pool of neurotransmitter vesicles and for
reestablishing the membrane polarity after neurotransmission (reviewed by ([270] [271]).
Hypomorphic mutations of DLP1 in flies deplete the synapses of mitochondria which
lead to a decreased neurotransmission due to an impaired ability to mobilize reserve
synaptic vesicle pools. These findings suggest that mitochondria at the postsynaptic
terminals are essential for regulating the ATP-dependent replenishment of the releasable
pool of synaptic vesicles [218].
While our data supports a role of Bβ2 in regulating apoptosis by promoting
mitochondrial fission in neurons, I propose that a physiological function of PP2A/Bβ2 is
to mediate the transport of mitochondria from the soma into the dendrites to regulate
synaptogenesis. It is noteworthy that the developmental expression profile of Bβ2 does
not match that of any known apoptosis regulator but coincides with CAMKII expression
and with the consolidation of synapses supporting a role of Bβ2 in regulating
synaptogenesis. Conversely, PKA at the OMM may retract/collapse synapses by
depleting synapses of mitochondria and mitochondrial ATP that is required for
synaptogenesis. Thus, by regulating synaptogenesis, PP2A/Bβ2 may have functional
implications in higher brain functions such as learning and memory by consolidating
synapses and thereby reforming/rewiring synaptic circuits.
Future directions
There are different lines of studies that can arise from my dissertation that can help
unveil other physiological roles of f PP2A/ Bβ2 or AKAP121 in neurons. Here are some
potential projects that can be pursued in the future:
1.
As previously mentioned, knocking-down endogenous Bβ2 is
neuroprotective in vitro in models of ischemia and Parkinson’s disease suggesting a
role of the regulatory protein in neurodegeneration. As a proof of concept, it is
170
critical to prove that knocking-down Bβ2 in vivo is neuroprotective. To this end, we
have generated “straight” knockout mice in which the divergent N-terminal
mitochondrial sequence of Bβ2 has been replaced with the β-galactosidase cDNA.
We have shown that homozygous and heterozygous knock-out mice are viable
without obvious behavioral or morphological abnormalities. Furthermore, Northern
blot, RT-PCR and Western blot analyses have confirmed the mRNA and protein of
Bβ2 have been replaced with β-galactosidase whereas Bβ1 is not affected. To follow
up on the effects of knocking-down endogenous Bβ2 on neuronal survival,
homozygous, heterozygous and wild-type mice will be challenged by focal cerebral
ischemia (MCAO) or by chemical insults by stereotactically injecting low doses of
rotenone or 3-nitropropionic acid . The effects of knocking down endogenous Bβ2
on survival will be assessed by performing Nissl or Fluor-Jade staining of
transcardially perfused brain sections to quantify the lesion. A decrease in Nissl or
Fluor-Jade staining of Bβ2 knockout mice compared to heterozygous or wild-type
mice will suggest that Bβ2 promotes neurodegeneration in vivo.
2.
It will be critical to determine whether the AKAP121-, Bβ2-, and FIS1shRNAs efficiently knock-down the expression of the respective targeted
endogenous protein. However, proving the ability of shRNAs in knocking down their
respective endogenous proteins may pose a technical challenge since endogenous
Bβ2 is expressed at low levels and there is also a lack of reliable commercial FIS1 or
AKAP121 specific antibodies. However, the observation that two independent
shRNAs targeting AKAP121 or FIS1 have similar effects on survival and
mitochondrial morphology in hippocampal neurons rules out off-target effects and
probably suggests that the endogenous protein is targeted and knocked down by their
respective shRNAs. It will be critical to test the ability of Bβ2 shRNAs to knock
down the endogenous protein since the effects of knocking down the endogenous
regulatory subunit in survival will be tested in vivo in a the MCAO model and in a
171
model of Parkinson’s disease. To this end, dissociated hippocampal neurons (DIV10)
will be co-transfected with Bβ2 shRNAs and with β-galactosidase for three days,
fixed in 3.7% PF and immunostained for Bβ2 using the Bβ2 mAb and for βgalactosidase to identify transfected neurons. At the single cell level, a significant
decrease in Bβ2 specific immunostaining in hippocampal neurons transfected with
Bβ2 shRNAs compared to IR-CTRL transfected neurons will be indicative of
successful downregulation of endogenous Bβ2. It is possible that Bβ2 is expressed at
such low levels that it will not be possible to detect subtle differences in knock down
by Bβ2 shRNAs. To this end, the hippocampi of rat brains will be stereotactically
injected with lentiviruses that overexpress the Bβ2 shRNAs #1 or with a lentivirus
that express IR-CTRL as a control. After a week post-injection, hippocampi will then
be extracted from rat brains and a microcystin pull-down will be performed to enrich
PP2A holoenzymes and probed for Bβ2 immunoblotting with the Bβ2 specific mAb.
A decrease in the immunoreactivity of Bβ2 in hippocampi of rats injected with the
Bβ2 shRNA#1 lentivirus compared to hippocampi of rats injected with the IR-CTRL
expressing lentivrus will be indicative of successful downregulation by Bβ2 specific
shRNAs in vivo.
3.
As stated above, the fission of somatic mitochondria mediated by DLP1
drives the transport of mitochondria into dendrites which stimulates the development
of dendritic spines in hippocampal neurons [217]. Since Bβ2 drives fission of
somatic mitochondria in neurons, we propose that a physiological function of Bβ2 is
to stimulate the development of dendritic spines. Supporting this view, preliminary
results in our lab that demonstrates that overexpression of Bβ2 increases the density
of mitochondria in dendrites compared to hippocampal neurons transiently
expressing Bβ1 while knock-down of Bβ2 has the opposite effect (not shown).
Furthermore, we will assess the effects of targeting PKA to the OMM via AKAP121
or of PKA inhibition via OMM-PKI on the distribution of dendritic mitochondria. In
172
order to determine an effect PP2A/Bβ2 and OMM-PKA on the development of
dendritic spines, future experiments will determine whether transient overexpression
of Bβ2 increases the maturation of synapses by immunolabeling for PSD95, a
postsynaptic localized protein, and count the number of PSD95 punctae. At the same
time, electrophysiological recordings in hippocampal neurons by patch clamp
techniques will determine whether Bβ2 increases the number of functional synapses
by quantifying for changes in the frequency of miniature excitatory postsynaptic
currents (EPSPs) over time.
4.
In collaboration with Yuriy Usachev at the University of Iowa, Fura-2
recordings that measure the FCCP dependent rise in cytosolic [Ca2+] in hippocampal
neurons have demonstrated that overexpression of Bβ2 but not Bβ1 increases the
depolarization induced Ca2+ uptake into mitochondria. An excess uptake of calcium
during glutamate excitotoxicity has been shown to cause mitochondrial dysfunction
and leads to the development of an irreversible increase in cytosolic [Ca2+] termed
delayed Ca2+ deregulation [272-274]. By measuring for [Ca2+] transients in
hippocampal neurons treated with 200µM glutamate, it will be intriguing to know
whether Bβ2 overexpression sensitizes neurons to apoptosis by increasing the
probability or accelerating the development of delayed [Ca2+] deregulation in
hippocampal neurons. Conversely, it will be interesting to determine whether PKA
recruited at the OMM by AKAP121 subverts the development of delayed [Ca2+]
deregulation during glutamate excitoxicity.
173
REFERENCES
1.
Dunn, D.A., Mining the human "kinome". Drug Discov Today, 2002. 7(22): p.
1121-3.
2.
Manning, G., et al., The protein kinase complement of the human genome.
Science, 2002. 298(5600): p. 1912-34.
3.
Merrill, R.A. and S. Strack, Protein kinases and phosphatases, in Pain: gene,
synapse, disease, M. Zhu and G.F. Gebhart, Editors. 2006. p. In press.
4.
Fjeld, C.C. and J.M. Denu, Kinetic analysis of human serine/threonine protein
phosphatase 2Calpha. J Biol Chem, 1999. 274(29): p. 20336-43.
5.
Williams, N.H., Models for biological phosphoryl transfer. Biochim Biophys
Acta, 2004. 1697(1-2): p. 279-87.
6.
Kaap, S., et al., Apoptosis by 6-O-palmitoyl-L-ascorbic acid coincides with JNKphosphorylation and inhibition of Mg2+-dependent phosphatase activity.
Biochem Pharmacol, 2004. 67(5): p. 919-26.
7.
Nguyen, A.N. and K. Shiozaki, Heat-shock-induced activation of stress MAP
kinase is regulated by threonine- and tyrosine-specific phosphatases. Genes Dev,
1999. 13(13): p. 1653-63.
8.
Wakula, P., et al., Degeneracy and function of the ubiquitous RVXF motif that
mediates binding to protein phosphatase-1. J Biol Chem, 2003. 278(21): p.
18817-23.
9.
MacMillan, L.B., et al., Brain actin-associated protein phosphatase 1
holoenzymes containing spinophilin, neurabin, and selected catalytic subunit
isoforms. J Biol Chem, 1999. 274(50): p. 35845-54.
10.
McAvoy, T., et al., Regulation of neurabin I interaction with protein phosphatase
1 by phosphorylation. Biochemistry, 1999. 38(39): p. 12943-9.
11.
Oliver, C.J., et al., Targeting protein phosphatase 1 (PP1) to the actin
cytoskeleton: the neurabin I/PP1 complex regulates cell morphology. Mol Cell
Biol, 2002. 22(13): p. 4690-701.
12.
Colbran, R.J., et al., Association of brain protein phosphatase 1 with cytoskeletal
targeting/regulatory subunits. J Neurochem, 1997. 69(3): p. 920-9.
13.
Munton, R.P., S. Vizi, and I.M. Mansuy, The role of protein phosphatase-1 in the
modulation of synaptic and structural plasticity. FEBS Lett, 2004. 567(1): p. 1218.
14.
Mumby, M.C. and G. Walter, Protein serine/threonine phosphatases: structure,
regulation, and functions in cell growth. Physiol Rev, 1993. 73(4): p. 673-99.
15.
Cohen, P., The structure and regulation of protein phosphatases. Ann Rev
Biochem, 1989. 58: p. 453-508.
174
16.
Price, N.E. and M.C. Mumby, Effects of regulatory subunits on the kinetics of
protein phosphatase 2A. Biochemistry, 2000. 39(37): p. 11312-8.
17.
Dagda, R.K., et al., A developmentally regulated, neuron-specific splice variant of
the variable subunit B-beta targets protein phosphatase 2A to mitochondria and
modulates apoptosis. J Biol Chem, 2003. 278(27): p. 24976-24985.
18.
Van Kanegan, M.J., et al., Distinct protein phosphatase 2A heterotrimers
modulate growth factor signaling to extracellular signal-regulated kinases and
Akt. J Biol Chem, 2005. 280(43): p. 36029-36.
19.
Rajashree, R., B.C. Blunt, and P.A. Hofmann, Modulation of myosin phosphatase
targeting subunit and protein phosphatase 1 in the heart. Am J Physiol Heart Circ
Physiol, 2005. 289(4): p. H1736-43.
20.
Terrak, M., et al., Structural basis of protein phosphatase 1 regulation. Nature,
2004. 429(6993): p. 780-4.
21.
Xie, H. and S. Clarke, Protein phosphatase 2A is reversibly modified by methyl
esterification at its C-terminal leucine residue in bovine brain. J Biol Chem,
1994. 269(3): p. 1981-4.
22.
Guo, H. and Z. Damuni, Autophosphorylation-activated protein kinase
phosphorylates and inactivates protein phosphatase 2A. Proc Natl Acad Sci U S
A, 1993. 90(6): p. 2500-4.
23.
Cohen, P., C.F. Holmes, and Y. Tsukitani, Okadaic acid: a new probe for the
study of cellular regulation. Trends Biochem Scien, 1990. 15(3): p. 98-102.
24.
Carelli, V., F.N. Ross-Cisneros, and A.A. Sadun, Optic nerve degeneration and
mitochondrial dysfunction: genetic and acquired optic neuropathies. Neurochem
Int, 2002. 40(6): p. 573-84.
25.
Gotz, J., et al., Delayed embryonic lethality in mice lacking protein phosphatase
2A catalytic subunit Calpha. Proc Nat Acad Sci U S A, 1998. 95(21): p. 12370-5.
26.
Silverstein, A.M., et al., Actions of PP2A on the MAP kinase pathway and
apoptosis are mediated by distinct regulatory subunits. Proc Nat Acad Sci U S A,
2002. 99(7): p. 4221-6.
27.
Strack, S., J.T. Cribbs, and L. Gomez, Critical role for protein phosphatase 2A
heterotrimers in mammalian cell survival. J Biol Chem, 2004. 279(46): p. 477329.
28.
Nanahoshi, M., et al., Alpha4 protein as a common regulator of type 2A-related
serine/threonine protein phosphatases. FEBS Lett, 1999. 446(1): p. 108-12.
29.
Groves, M.R., et al., The structure of the protein phosphatase 2A PR65/A subunit
reveals the conformation of its 15 tandemly repeated HEAT motifs. Cell, 1999.
96(1): p. 99-110.
175
30.
Zhou, J., et al., Characterization of the Aalpha and Abeta subunit isoforms of
protein phosphatase 2A: differences in expression, subunit interaction, and
evolution. Biochem J, 2003. 369(Pt 2): p. 387-98.
31.
Van Hoof, C., et al., Molecular cloning and developmental regulation of
expression of two isoforms of the catalytic subunit of protein phosphatase 2A from
Xenopus laevis. Biochem Biophys Res Commun, 1995. 215(2): p. 666-73.
32.
Wang, S.S., et al., Alterations of the PPP2R1B gene in human lung and colon
cancer. Science, 1998. 282(5387): p. 284-7.
33.
Colella, S., et al., Reduced expression of the Aalpha subunit of protein
phosphatase 2A in human gliomas in the absence of mutations in the Aalpha and
Abeta subunit genes. Int J Can, 2001. 93(6): p. 798-804.
34.
Lechward, K., et al., Protein phosphatase 2A: variety of forms and diversity of
functions. Acta Biochim Pol, 2001. 48(4): p. 921-33.
35.
Janssens, V. and J. Goris, Protein phosphatase 2A: a highly regulated family of
serine/threonine phosphatases implicated in cell growth and signalling. Biochem
J, 2001. 353(Pt 3): p. 417-39.
36.
Walter, G., et al., Association of protein phosphatase 2A with polyoma virus
medium tumor antigen. Proc Natl Acad Sci U S A, 1990. 87(7): p. 2521-5.
37.
Pallas, D.C., et al., Polyoma small and middle T antigens and SV40 small t
antigen form stable complexes with protein phosphatase 2A. Cell, 1990. 60(1): p.
167-76.
38.
Sontag, E., et al., The interaction of SV40 small tumor antigen with protein
phosphatase 2A stimulates the map kinase pathway and induces cell proliferation.
Cell, 1993. 75(5): p. 887-97.
39.
Yu, X.X., et al., Methylation of the protein phosphatase 2A catalytic subunit is
essential for association of Balpha regulatory subunit but not SG2NA, striatin, or
polyomavirus middle tumor antigen. Mol Biol Cell, 2001. 12(1): p. 185-99.
40.
Moreno, C.S., et al., WD40 repeat proteins striatin and S/G(2) nuclear
autoantigen are members of a novel family of calmodulin-binding proteins that
associate with protein phosphatase 2A. J Biol Chem, 2000. 275(8): p. 5257-63.
41.
Li, X. and D.M. Virshup, Two conserved domains in regulatory B subunits
mediate binding to the A subunit of protein phosphatase 2A. Eur J Biochem, 2002.
269(2): p. 546-52.
42.
Healy, A.M., et al., CDC55, a Saccharomyces cerevisiae gene involved in cellular
morphogenesis: identification, characterization, and homology to the B subunit of
mammalian type 2A protein phosphatase. Mol Cell Biol, 1991. 11(11): p. 576780.
43.
Mayer, R.E., et al., Structure of the 55-kDa regulatory subunit of protein
phosphatase 2A: evidence for a neuronal-specific isoform. Biochemistry, 1991.
30(15): p. 3589-97.
176
44.
Pallas, D.C., et al., The third subunit of protein phosphatase 2A (PP2A), a 55kilodalton protein which is apparently substituted for by T antigens in complexes
with the 36- and 63-kilodalton PP2A subunits, bears little resemblance to T
antigens. J Virol, 1992. 66(2): p. 886-893.
45.
Zolnierowicz, S., et al., Diversity in the regulatory B-subunits of protein
phosphatase 2A: identification of a novel isoform highly expressed in brain.
Biochemistry, 1994. 33(39): p. 11858-67.
46.
Strack, S., et al., Cloning and characterization of B delta, a novel regulatory
subunit of protein phosphatase 2A. FEBS Lett, 1999. 460(3): p. 462-6.
47.
Strack, S., et al., Protein phosphatase 2A holoenzyme assembly. Identification of
contacts between B-family regulatory and scaffolding A subunits. J. Biol. Chem.,
2002. 277: p. 20750-20755.
48.
Sontag, E., et al., A novel pool of protein phosphatase 2A is associated with
microtubules and is regulated during the cell cycle. J Cell Biol, 1995. 128(6): p.
1131-44.
49.
Sontag, E., et al., Regulation of the phosphorylation state and microtubulebinding activity of Tau by protein phosphatase 2A. Neuron, 1996. 17(6): p. 12017.
50.
Turowski, P., et al., Vimentin dephosphorylation by protein phosphatase 2A is
modulated by the targeting subunit B55. Mol Biol Cell, 1999. 10(6): p. 19972015.
51.
Strack, S., Overexpression of the protein phosphatase 2A regulatory subunit
B[gamma] promotes neuronal differentiation by activating the MAP kinase
cascade. J Biol Chem, 2002. 277(44): p. 41525-41532.
52.
McCright, B. and D.M. Virshup, Identification of a new family of protein
phosphatase 2A regulatory subunits. J Biol Chem, 1995. 270(44): p. 26123-8.
53.
McCright, B., et al., The B56 family of protein phosphatase 2A (PP2A) regulatory
subunits encodes differentiation-induced phosphoproteins that target PP2A to
both nucleus and cytoplasm. J Biol Chem, 1996. 271(36): p. 22081-9.
54.
Csortos, C., et al., High complexity in the expression of the B' subunit of protein
phosphatase 2A0. Evidence for the existence of at least seven novel isoforms. J
Biol Chem, 1996. 271(5): p. 2578-88.
55.
Tanabe, O., et al., Molecular cloning of a 74-kDa regulatory subunit (B" or delta)
of human protein phosphatase 2A. FEBS Lett, 1996. 379(1): p. 107-11.
56.
Tehrani, M.A., M.C. Mumby, and C. Kamibayashi, Identification of a novel
protein phosphatase 2A regulatory subunit highly expressed in muscle. J Biol
Chem, 1996. 271(9): p. 5164-70.
57.
Zolnierowicz, S., et al., The variable subunit associated with protein phosphatase
2A0 defines a novel multimember family of regulatory subunits. Biochem J, 1996.
317(Pt 1): p. 187-94.
177
58.
Seeling, J.M., et al., Regulation of beta-catenin signaling by the B56 subunit of
protein phosphatase 2A. Science, 1999. 283(5410): p. 2089-91.
59.
Yamamoto, H., et al., Inhibition of the Wnt signaling pathway by the PR61
subunit of protein phosphatase 2A. J Biol Chem, 2001. 276(29): p. 26875-82.
60.
Chen, W., et al., Identification of specific PP2A complexes involved in human cell
transformation.[see comment]. Cancer Cell, 2004. 5(2): p. 127-36.
61.
Okamoto, K., et al., p53-dependent association between cyclin G and the B'
subunit of protein phosphatase 2A. Mol Cell Biol, 1996. 16(11): p. 6593-602.
62.
Li, X., et al., B56-associated protein phosphatase 2A is required for survival and
protects from apoptosis in Drosophila melanogaster. Mol Cell Biol, 2002. 22(11):
p. 3674-84.
63.
Yang, J., et al., PP2A:B56epsilon is required for Wnt/beta-catenin signaling
during embryonic development. Development, 2003. 130(23): p. 5569-78.
64.
Janssens, V., et al., Identification and Functional Analysis of Two Ca2+-binding
EF-hand Motifs in the B"/PR72 Subunit of Protein Phosphatase 2A. J Biol Chem,
2003. 278(12): p. 10697-10706.
65.
Voorhoeve, P.M., E.M. Hijmans, and R. Bernards, Functional interaction
between a novel protein phosphatase 2A regulatory subunit, PR59, and the
retinoblastoma-related p107 protein. Oncogene, 1999. 18(2): p. 515-24.
66.
Yan, Z., et al., PR48, a novel regulatory subunit of protein phosphatase 2A,
interacts with Cdc6 and modulates DNA replication in human cells. Mol Cell Biol
2000. 20(3): p. 1021-9.
67.
Strack, S., et al., Brain protein phosphatase 2A: developmental regulation and
distinct cellular and subcellular localization by B subunits. Journal of Comp
Neurol, 1998. 392(4): p. 515-27.
68.
Price, N.E. and M.C. Mumby, Brain protein serine/threonine phosphatases. Curr
Opin in Neurobiol, 1999. 9(3): p. 336-42.
69.
Cohen, P., Protein phosphatase 1- targeted in many directions. J Cell Sci, 2002.
115(2): p. 241-256.
70.
Colbran, R.J., Protein phosphatases and calcium/calmodulin-dependent protein
kinase II-dependent synaptic plasticity. J Neurosci, 2004. 24(39): p. 8404-8409.
71.
Huang, H.B., et al., Characterization of the inhibition of protein phosphatase-1 by
DARPP-32 and inhibitor-2. J Biol Chem, 1999. 274(12): p. 7870-8.
72.
Blank, T., et al., The phosphoprotein DARPP-32 mediates cAMP-dependent
potentiation of striatal N-methyl-D-aspartate responses. Proc Natl Acad Sci U S
A, 1997. 94(26): p. 14859-64.
73.
Saito, T., et al., Neurofilament-associated protein phosphatase 2A: its possible
role in preserving neurofilaments in filamentous states. Biochemistry, 1995.
34(22): p. 7376-84.
178
74.
Strack, S., et al., Protein serine/threonine phosphatase 1 and 2A associate with
and dephosphorylate neurofilaments. Brain Res Mol Brain Res, 1997. 49(1-2): p.
15-28.
75.
Merrick, S.E., J.Q. Trojanowski, and V.M. Lee, Selective destruction of stable
microtubules and axons by inhibitors of protein serine/threonine phosphatases in
cultured human neurons. J Neurosci, 1997. 17(15): p. 5726-37.
76.
Sontag, E., et al., Molecular interactions among protein phosphatase 2A, tau, and
microtubules. Implications for the regulation of tau phosphorylation and the
development of tauopathies. J Biol Chem, 1999. 274(36): p. 25490-8.
77.
Alexa, A., et al., The phosphorylation sate of threonine-220, a uniquely
phosphatase-sensitive protein kinase A site in microtubule-associated protein
MAP2C, regulates microtubule binding and stability. Biochem, 2002. 41(41): p.
12427-35.
78.
Tolstykh, T., et al., Carboxyl methylation regulates phosphoprotein phosphatase
2A by controlling the association of regulatory B subunits. Embo J, 2000. 19(21):
p. 5682-91.
79.
Wu, J., et al., Carboxyl methylation of the phosphoprotein phosphatase 2A
catalytic subunit promotes its functional association with regulatory subunits in
vivo. Embo J, 2000. 19(21): p. 5672-81.
80.
Trojanowski, J.Q. and V.M. Lee, Phosphorylation of paired helical filament tau
in Alzheimer's disease neurofibrillary lesions: focusing on phosphatases. Faseb J,
1995. 9(15): p. 1570-6.
81.
Sontag, E., et al., Altered expression levels of the protein phosphatase 2A
ABalphaC enzyme are associated with Alzheimer disease pathology. J
Neuropathol Exp Neurol, 2004. 63(4): p. 287-301.
82.
Sontag, E., et al., Downregulation of protein phosphatase 2A carboxyl
methylation and methyltransferase may contribute to Alzheimer disease
pathogenesis. J Neuropathol Exp Neurol, 2004. 63(10): p. 1080-91.
83.
Nagatsu, T., Tyrosine hydroxylase: human isoforms, structure and regulation in
physiology and pathology. Essays Biochem, 1995. 30: p. 15-35.
84.
Leal, R.B., et al., Tyrosine hydroxylase dephosphorylation by protein phosphatase
2A in bovine adrenal chromaffin cells. Neurochem Res, 2002. 27(3): p. 207-13.
85.
Bosboom, J.L. and E. Wolters, Psychotic symptoms in Parkinson's disease:
pathophysiology and management. Expert Opin Drug Saf, 2004. 3(3): p. 209-20.
86.
Weeber, E.J., et al., Derangements of hippocampal calcium/calmodulindependent protein kinase II in a mouse model for Angelman mental retardation
syndrome. J Neurosci, 2003. 23(7): p. 2634-44.
87.
Holmes, S.E., et al., Expansion of a novel CAG trinucleotide repeat in the 5'
region of PPP2R2B is associated with SCA12 [letter]. Nat Genet, 1999. 23(4): p.
391-2.
179
88.
Fujigasaki, H., et al., SCA12 is a rare locus for autosomal dominant cerebellar
ataxia: a study of an Indian family. Ann Neurol, 2001. 49(1): p. 117-21.
89.
Holmes, S.E., E. O'Hearn, and R.L. Margolis, Why is SCA12 different from other
SCAs? Cytogenet Genome Res, 2003. 100(1-4): p. 189-97.
90.
Majno, G. and I. Joris, Apoptosis, oncosis, and necrosis. An overview of cell
death. Am J Pathol, 1995. 146(1): p. 3-15.
91.
Liu, X., T. Van Vleet, and R.G. Schnellmann, The role of calpain in oncotic cell
death. Annu Rev Pharmacol Toxicol, 2004. 44: p. 349-70.
92.
Trump, B.F. and I.K. Berezesky, The role of altered [Ca2+]i regulatin in
apoptosis, oncosis, and necrosis. Biochim Biophys Acta, 1996. 1313: p. 173-8.
93.
Kanduc, D., et al., Cell death: apoptosis versus necrosis. Int. J. Oncol., 2002. 21:
p. 165-170.
94.
Trump, B.F., et al., The pathways of cell death: oncosis, apoptosis, and necrosis.
Toxicol. Pathol., 1997. 25: p. 82-88.
95.
Yorimitsu, T. and D.J. Klionsky, Autophagy: molecular machinery for self-eating.
Cell Death Differ, 2005. 12 Suppl 2: p. 1542-52.
96.
Baehrecke, E.H., Autophagy: dual roles in life and death? Nat Rev Mol Cell Biol,
2005. 6(6): p. 505-10.
97.
Gozuacik, D. and A. Kimchi, Autophagy as a cell death and tumor suppressor
mechanism. Oncogene, 2004. 23: p. 2891-2906.
98.
Ohsumi, Y., Molecular dissection of autophagy: two ubiquitin-like systems. Nat
Rev Mol Cell Biol, 2001. 2: p. 211-16.
99.
Bursch, W. and A. Ellinger, Autophagy--a basic mechanism and a potential role
for neurodegeneration. Folia Neuropathol, 2005. 43(4): p. 297-310.
100.
Xue, L., G.C. Fletcher, and A.M. Tolkovsky, Autophagy is activated by apoptotic
signalling in sympathetic neurons: an alternative mechanism of death execution.
Mol Cell Neurosci, 1999. 14(3): p. 180-98.
101.
Farber, E., Programmed cell death: necrosis versus apoptosis. Mod Pathol, 1994.
7: p. 605-609.
102.
Ashe, E. and M. Berry, Apoptotic signaling cascades. Prog in
Neuropsychopharmacol Biol Psychiatry, 2003. 27(2): p. 199-214.
103.
Jin, Z. and W.S. El-Deiry, Overview of cell death signaling pathways. Cancer
Biol Ther, 2005. 4(2): p. 139-63.
104.
Takuma, U., K. Yoshito, and R. Morrison, Neurons exclusively express N-Bak, a
BH3 domain-only Bak isoform that promotes neuronal apoptosis. J Biol Chem,
2005. 280(10): p. 9065-73.
180
105.
Putcha, G.V., et al., JNK-mediated BIM phosphorylation potentiates BAXdependent apoptosis. Neuron, 2003. 38(6): p. 899-914.
106.
Wei, M.C., et al., tBID, a membrane-targeted death ligand, oligomerizes BAK to
release cytochrome c. Genes Dev, 2000. 14(16): p. 2060-71.
107.
Yin, X.M., et al., Bid-mediated mitochondrial pathway is critical to ischemic
neuronal apoptosis and focal cerebral ischemia. J Biol Chem, 2002. 277(44): p.
42074-81.
108.
Plesnila, N., et al., BID mediates neuronal cell death after oxygen/ glucose
deprivation and focal cerebral ischemia. Proc Natl Acad Sci U S A, 2001. 98(26):
p. 15318-23.
109.
Dramsi, S., et al., Identification of a novel phosphorylation site, Ser-170, as a
regulator of bad pro-apoptotic activity. J Biol Chem, 2002. 277(8): p. 6399-405.
110.
Biswas, S.C. and L.A. Greene, Nerve growth factor (NGF) down-regulates the
Bcl-2 homology 3 (BH3) domain-only protein Bim and suppresses its
proapoptotic activity by phosphorylation. J Biol Chem, 2002. 277(51): p. 495116.
111.
Harada, H., et al., Phosphorylation and inactivation of BAD by mitochondriaanchored protein kinase A. Mol Cell, 1999. 3(4): p. 413-22.
112.
Chiang, C.W., et al., Protein phosphatase 2A activates the proapoptotic function
of BAD in interleukin- 3-dependent lymphoid cells by a mechanism requiring 143-3 dissociation. Blood, 2001. 97(5): p. 1289-97.
113.
Klumpp, S. and J. Krieglstein, Serine/threonine protein phosphatases in
apoptosis. Curr Opin Pharmacol, 2002. 2(4): p. 458-62.
114.
Gilley, J., P.J. Coffer, and J. Ham, FOXO transcripton factors directly activate
bim gene expression and promote apoptosis in sympathetic neurons. J Cell Biol,
2003. 162(4): p. 613-622.
115.
Putcha, G.V., et al., Induction of BIM, a proapoptotic BH3-only BCL-2 family
member, is critical for neuronal apoptosis. Neuron, 2001. 29(3): p. 615-28.
116.
Ruvolo, P.P., et al., Ceramide induces Bcl2 dephosphorylation via a mechanism
involving mitochondrial PP2A. J Biol Chem, 1999. 274(29): p. 20296-300.
117.
Ruvolo, P.P., et al., A functional role for the B56 alpha-subunit of protein
phosphatase 2A in ceramide-mediated regulation of Bcl2 phosphorylation status
and function. J Biol Chem, 2002. 277(25): p. 22847-52.
118.
Masters, S.C., et al., Survival-promoting functions of 14-3-3 proteins. Biochem
Soc Trans, 2002. 30(4): p. 360-5.
119.
Chiang, C.W., et al., Protein phosphatase 2A dephosphorylation of phosphoserine
112 plays the gatekeeper role for BAD-mediated apoptosis. Mol Cell Biol, 2003.
23(18): p. 6350-62.
181
120.
Schmidt, K., et al., Diversity, developmental regulation and distribution of murine
PR55/B subunits of protein phosphatase 2A. Eur J Neurosci, 2002. 16(11): p.
2039-2048.
121.
Herzig, S. and J. Neumann, Effects of Serine/Threonine protein phosphatases on
ion channels in excitable membranes. Physiol Rev, 2000. 80(1): p. 173-210.
122.
Fujiki, H. and M. Suganuma, Tumor promotion by inhibitors of protein
phosphatase 1 and 2A: the okadaic acid class of compounds. Adv Cancer Res,
1993. 61: p. 143-154.
123.
Maynes, J.T., et al., Crystal structure of the tumor-promoter okadaic acid bound
to protein phosphatase-1. J Biol Chem, 2001. 276(47): p. 44078-82.
124.
Kita, A., et al., Crystal structure of the complex between calyculin A and the
catalytic subunit of protein phosphatase 1. Structure, 2002. 10(5): p. 715-724.
125.
Li, M. and Z. Damuni, I1PP2A and I2PP2A. Two potent protein phosphatase 2Aspecific inhibitor proteins. Methods Mol Biol, 1998. 93: p. 59-66.
126.
Saito, S., et al., Functional domains of template-activating factor-I as a protein
phosphatase 2A inhibitor. Biochem Biophys Res Commun, 1999. 259: p. 471-5.
127.
Fing, T.M., et al., Localization of the gene encoding the putative human HLA
class II associated protein (PHAPI) to chromosome 1q22.3-q23 by fluorescence
in situ hybridization. Genomics, 1995. 29(309-310.).
128.
Dobson, S., et al., Characterization of a unique aspartate-rich protein of the
SET/TAF-family in the human malaria parasite, Plasmodium falciparum, which
inhibits protein phosphatase 2A. Mol Biochem Parasitol, 2003. 126(2): p. 239-50.
129.
Onteniente, B., et al., The mechanisms of cell death in focal cerebral ischemia
highlight neuroprotective perspectives by anti-caspase therapy. Biochem
Pharmacol, 2003. 66(8): p. 1643-9.
130.
Reynolds, I.J. and T.G. Hastings, Glutamate induces the production of reactive
oxygen species in cultured forebrain neurons following NMDA receptor
activation. J Neurosci, 1995. 15(5 Pt 1): p. 3318-27.
131.
Fiskum, G., Mitochondrial participation in ischemic and traumatic neural cell
death. J Neurotrauma, 2000. 17(10): p. 843-55.
132.
Akins, P.T. and R.P. Atkinson, Glutamate AMPA receptor antagonist treatment
for ischaemic stroke. Curr Med Res Opin, 2002. 18 Suppl 2: p. s9-13.
133.
Aarts, M., et al., Treatment of ischemic brain damage by perturbing NMDA
receptor- PSD-95 protein interactions. Science, 2002. 298(5594): p. 846-50.
134.
Sun, Y., et al., Adeno-associated virus-mediated delivery of BCL-w gene improves
outcome after transient focal cerebral ischemia. Gene Therapy, 2003. 10(2): p.
115-22.
135.
Yenari, M.A., et al., Gene therapy and hypothermia for stroke treatment. Ann N
Y Acad Scien, 2003. 993: p. 54-68; discussion 79-81.
182
136.
Waldmeier, P.C., Prospects for antiapoptotic drug therapy of neurodegenerative
diseases. Prog Neuropsychopharmacol Biol Psychiatry, 2003. 27(2): p. 303-21.
137.
Bove, J., et al., Toxin-induced models of Parkinson's disease. NeuroRx, 2005.
2(3): p. 484-94.
138.
Testa, C.M., T.B. Sherer, and J.T. Greenamyre, Rotenone induces oxidative stress
and dopaminergic neuron damage in organotypic substantia nigra cultures. Brain
Res Mol Brain Res, 2005. 134(1): p. 109-18.
139.
Alam, A. and W.J. Schmidt, Rotenone destroys dopaminergic neurons and
induces parkinsonian symptoms in rats. Behav Brain Res, 2002. 136.
140.
Horowitz, M., C.L. Cepko, and P.A. Sharp, Expression of chimeric genes in the
early region of SV40. J Mol Appl Genet, 1983. 2(2): p. 147-59.
141.
Harlow, E. and D.P. Lane, Antibodies: a laboratory manual. 1988, Cold Spring
Harbor: Cold Spring Harbor Laboratory.
142.
Frangioni, J.V. and B.G. Neel, Solubilization and purification of enzymatically
active glutathione S-transferase (pGEX) fusion proteins. Anal Biochem, 1993.
210(1): p. 179-87.
143.
Galfre, G. and C. Milstein, Preparation of monoclonal anibodies: strategies and
procedures. Methods Enzymol, 1981(73): p. 3-46.
144.
Jordan, M., A. Schallham, and F.W. Wurm, Transfecting mammalina cells:
optimization of critical parameters affecting calcium phospphate precipitate
formation. Nucleic Acid Res, 1996(24): p. 596-601.
145.
Strack, S., R.B. McNeill, and R.J. Colbran, Mechanism and regulation of
calcium/calmodulin-dependent protein kinase II targeting to the NR2B subunit of
the N-methyl-D-aspartate receptor. J Biol Chem, 2000. 275(31): p. 23798-806.
146.
Pittman, R.N., et al., A system for characterizing cellular and molecular events in
programmed neuronal cell death. J Neurosci, 1993. 13(9): p. 3669-80.
147.
Hatano, Y., et al., Expression of PP2A B regulatory subunit beta isotype in rat
testis. FEBS Lett, 1993. 324(1): p. 71-5.
148.
Wang, X., The expanding role of mitochondria in apoptosis. Genes &
Development, 2001. 15(22): p. 2922-33.
149.
McBride, H.M., et al., A signal-anchor sequence selective for the mitochondrial
outer membrane. J Cell Biol, 1992. 119(6): p. 1451-7.
150.
Santoro, M.F., et al., Regulation of protein phosphatase 2A activity by caspase-3
during apoptosis. J Biol Chem, 1998. 273(21): p. 13119-28.
151.
Gehringer, M.M., Microcystin-LR and okadaic acid-induced cellular effects: a
dualistic response. FEBS Lett, 2004. 557(1-3): p. 1-8.
183
152.
Faux, M.C. and J.D. Scott, Molecular glue: kinase anchoring and scaffold
proteins. Cell, 1996. 85(1): p. 9-12.
153.
Kennelly, P.J. and E.G. Krebs, Consensus sequences as substrate specificity
determinants for protein kinases and protein phosphatases. J Biol Chem, 1991.
266(24): p. 15555-8.
154.
Srivastava, A.K., et al., Molecular and clinical correlation in five Indian families
with spinocerebellar ataxia 12. Ann Neurol, 2001. 50(6): p. 796-800.
155.
Maynes, J.T., et al., Crystal structure of the tumor-promoter okadaic acid bound
to protein phosphatase-1. J Biol Chem, 2001. 276(47): p. 44078-82.
156.
Egloff, M.P., et al., Crystal structure of the catalytic subunit of human protein
phosphatase 1 and its complex with tungstate. J Mol Biol, 1995. 254(5): p. 94259.
157.
Kita, A., et al., Crystal structure of the complex between calyculin A and the
catalytic subunit of protein phosphatase 1. Structure, 2002. 10(5): p. 715-24.
158.
Greene, L.A., et al., PC12 pheochromocytoma cells: culture, nerve growth factor
treatment, and experimental exploitation. Methods Enzymol, 1987. 147: p. 20716.
159.
Kremmer, E., et al., Separation of PP2A core enzyme and holoenzyme with
monoclonal antibodies against the regulatory A subunit: abundant expression of
both forms in cells. Mol Cell Biol, 1997. 17(3): p. 1692-701.
160.
Muslin, A.J., et al., Interaction of 14-3-3 with signaling proteins is mediated by
the recognition of phosphoserine. Cell, 1996. 84(6): p. 889-97.
161.
Fu, H., R.R. Subramanian, and S.C. Masters, 14-3-3 proteins: structure, function,
and regulation. Annu Rev Pharmacol Toxicol, 2000. 40: p. 617-47.
162.
Affaitati, A., et al., Essential role of A-kinase anchor protein 121 for cAMP
signaling to mitochondria. J Biol Chem, 2003. 278(6): p. 4286-94.
163.
Dagda, R.K., et al., Unfolding-resistant translocase targeting: a novel mechanism
for outer mitochondrial membrane localization exemplified by the Bbeta2
regulatory subunit of protein phosphatase 2A. J Biol Chem, 2005. 280(29): p.
27375-82.
164.
Smith, T.F., et al., The WD repeat: a common architecture for diverse functions.
Trends Biochem. Sci., 1999. 24(5): p. 181-5.
165.
Neer, E.J., et al., The ancient regulatory-protein family of WD-repeat proteins.
Nature, 1994. 371(6495): p. 297-300.
166.
Cheng, Z., et al., Crystal structure of Ski8p, a WD-repeat protein with dual roles
in mRNA metabolism and meiotic recombination. Protein Sci, 2004. 13(10): p.
2673-84.
167.
Lambright, D.G., et al., The 2.0 A crystal structure of a heterotrimeric G protein.
Nature, 1996. 379(6563): p. 311-9.
184
168.
Robinson, R.C., et al., Crystal structure of Arp2/3 complex. Science, 2001.
294(5547): p. 1679-84.
169.
Garcia-Higuera, I., et al., Folding of proteins with WD-repeats: comparison of six
members of the WD-repeat superfamily to the G protein beta subunit.
Biochemistry, 1996. 35(44): p. 13985-94.
170.
Faber, H.R., et al., 1.8 A crystal structure of the C-terminal domain of rabbit
serum haemopexin. Structure, 1995. 3(6): p. 551-9.
171.
Ruediger, R., et al., Molecular model of the A subunit of protein phosphatase 2A:
interaction with other subunits and tumor antigens. J Virol, 1994. 68(1): p. 123-9.
172.
Ruediger, R., et al., Identification of binding sites on the regulatory A subunit of
protein phosphatase 2A for the catalytic C subunit and for tumor antigens of
simian virus 40 and polyomavirus. Mol Cell Biol, 1992. 12(11): p. 4872-82.
173.
Ruediger, R., K. Fields, and G. Walter, Binding specificity of protein phosphatase
2A core enzyme for regulatory B subunits and T antigens. J Virol, 1999. 73(1): p.
839-42.
174.
Goldberg, J., et al., Three-dimensional structure of the catalytic subunit of protein
serine/threonine phosphatase-1. Nature, 1995. 376(6543): p. 745-53.
175.
Wiedemann, N., A.E. Frazier, and N. Pfanner, The protein import machinery of
mitochondria. J Biol Chem, 2004. 279(15): p. 14473-6.
176.
Van Kuilenburg, A.B., et al., Human heart cytochrome c oxidase subunit VIII.
Purification and determination of the complete amino acid sequence. FEBS Lett,
1988. 240(1-2): p. 127-32.
177.
Weiler, U., et al., The regulation of mitochondrial-bound hexokinases in the liver.
Biochem Med, 1985. 33(2): p. 223-35.
178.
Alto, N.M., J. Soderling, and J.D. Scott, Rab32 is an A-kinase anchoring protein
and participates in mitochondrial dynamics. J Cell Biol, 2002. 158(4): p. 659-68.
179.
Kuroda, R., et al., Charged amino acids at the carboxyl-terminal portions
determine the intracellular locations of two isoforms of cytochrome b5. J Biol
Chem, 1998. 273(47): p. 31097-102.
180.
Nemoto, Y. and P. De Camilli, Recruitment of an alternatively spliced form of
synaptojanin 2 to mitochondria by the interaction with the PDZ domain of a
mitochondrial outer membrane protein. Embo J, 1999. 18(11): p. 2991-3006.
181.
Nguyen, M., et al., Targeting of Bcl-2 to the mitochondrial outer membrane by a
COOH-terminal signal anchor sequence. J Biol Chem, 1993. 268(34): p. 252658.
182.
Nakamura, Y., et al., Targeting and assembly of rat mitochondrial translocase of
outer membrane 22 (TOM22) into the TOM complex. J Biol Chem, 2004.
279(20): p. 21223-32.
185
183.
Kelley, L.A., R.M. MacCallum, and M.J. Sternberg, Enhanced genome
annotation using structural profiles in the program 3D-PSSM. J Mol Biol, 2000.
299(2): p. 499-520.
184.
Guex, N. and M.C. Peitsch, SWISS-MODEL and the Swiss-PdbViewer: an
environment for comparative protein modeling. Electrophoresis, 1997. 18(15): p.
2714-23.
185.
Watson, A.J., A. Katz, and M.I. Simon, A fifth member of the mammalian Gprotein beta-subunit family. Expression in brain and activation of the beta 2
isotype of phospholipase C. J Biol Chem, 1994. 269(35): p. 22150-6.
186.
Frieden, M., et al., Ca(2+) homeostasis during mitochondrial fragmentation and
perinuclear clustering induced by hFis1. J Biol Chem, 2004. 279(21): p. 2270414.
187.
Rigaut, G., et al., A generic protein purification method for protein complex
characterization and proteome exploration. Nat Biotechnol, 1999. 17(10): p.
1030-2.
188.
Griswold-Prenner, I., et al., Physical and functional interactions between type I
transforming growth factor beta receptors and Balpha, a WD-40 repeat subunit of
phosphatase 2A. Mol Cell Biol, 1998. 18(11): p. 6595-604.
189.
Shi, J., T.L. Blundell, and K. Mizuguchi, FUGUE: sequence-structure homology
recognition using environment- specific substitution tables and structuredependent gap penalties. J Mol Biol, 2001. 310(1): p. 243-57.
190.
Omura, T., Mitochondria-targeting sequence, a multi-role sorting sequence
recognized at all steps of protein import into mitochondria. J Biochem (Tokyo),
1998. 123(6): p. 1010-6.
191.
Claros, M.G. and P. Vincens, Computational method to predict mitochondrially
imported proteins and their targeting sequences. Eur J Biochem, 1996. 241(3): p.
779-86.
192.
Anandatheerthavarada, H.K., et al., Mitochondrial targeting and a novel
transmembrane arrest of Alzheimer's amyloid precursor protein impairs
mitochondrial function in neuronal cells. J Cell Biol, 2003. 161(1): p. 41-54.
193.
Okamoto, K., et al., The protein import motor of mitochondria: a targeted
molecular ratchet driving unfolding and translocation. Embo J, 2002. 21(14): p.
3659-71.
194.
Voisine, C., et al., The protein import motor of mitochondria: unfolding and
trapping of preproteins are distinct and separable functions of matrix Hsp70.
Cell, 1999. 97(5): p. 565-74.
195.
Eilers, M. and G. Shatz, Nature, 1986. 322: p. 228-232.
196.
Rehling, P., et al., Protein insertion into the mitochondrial inner membrane by a
twin-pore translocase. Science, 2003. 299(5613): p. 1747-51.
186
197.
Bateman, A., et al., The Pfam protein families database. Nucleic Acids Res.,
2002. 30(1): p. 276-80.
198.
Lee, Y.-j., et al., Roles of the Mammalian Mitochondrial Fission and Fusion
Mediators Fis1, Drp1, and Opa1 in Apoptosis. Mol . Biol. Cell, 2004. 15(11): p.
5001-5011.
199.
Chen, H. and D.C. Chan, Mitochondrial dynamics in mammals. Curr Top Dev
Biol, 2004. 59: p. 119-44.
200.
Westermann, B., Mitochondrial membrane fusion. Biochim Biophys Acta, 2003.
1641(2-3): p. 195-202.
201.
Yoon, Y., Sharpening the scissors: mitochondrial fission with aid. Cell Biochem
Biophys, 2004. 41(2): p. 193-206.
202.
Youle, R.J. and M. Karbowski, Mitochondrial fission in apoptosis. Nat Rev Mol
Cell Biol, 2005. 6(8): p. 657-63.
203.
Collins, T.J., et al., Mitochondria are morphologically and functionally
heterogeneous within cells. Embo J, 2002. 21(7): p. 1616-27.
204.
Perfettini, J.L., T. Roumier, and G. Kroemer, Mitochondrial fusion and fission in
the control of apoptosis. Trends Cell Biol, 2005. 15(4): p. 179-83.
205.
Karbowski, M., et al., Spatial and temporal association of Bax with mitochondrial
fission sites, Drp1, and Mfn2 during apoptosis. J Cell Biol, 2002. 159(6): p. 9318.
206.
Frank, S., et al., The role of dynamin-related protein 1, a mediator of
mitochondrial fission, in apoptosis. Dev Cell, 2001. 1(4): p. 515-25.
207.
James, D.I., et al., hFis1, a novel component of the mammalian mitochondrial
fission machinery. J Biol Chem, 2003. 278(38): p. 36373-9.
208.
Karbowski, M. and R.J. Youle, Dynamics of mitochondrial morphology in healthy
cells and during apoptosis. Cell Death Differ, 2003. 10(8): p. 870-80.
209.
Szabadkai, G., et al., Drp-1-dependent division of the mitochondrial network
blocks intraorganellar Ca2+ waves and protects against Ca2+-mediated
apoptosis. Mol Cell, 2004. 16(1): p. 59-68.
210.
Chen, H., et al., Mitofusins Mfn1 and Mfn2 coordinately regulate mitochondrial
fusion and are essential for embryonic development. J Cell Biol, 2003. 160(2): p.
189-200.
211.
Ishihara, N., Y. Eura, and K. Mihara, Mitofusin 1 and 2 play distinct roles in
mitochondrial fusion reactions via GTPase activity. J Cell Sci, 2004. 117(Pt 26):
p. 6535-46.
212.
Eura, Y., et al., Two mitofusin proteins, mammalian homologues of FZO, with
distinct functions are both required for mitochondrial fusion. J Biochem (Tokyo),
2003. 134(3): p. 333-44.
187
213.
Neuspiel, M., et al., Activated mitofusin 2 signals mitochondrial fusion, interferes
with Bax activation, and reduces susceptibility to radical induced depolarization.
J Biol Chem, 2005. 280(26): p. 25060-70.
214.
Votruba, M., Molecular genetic basis of primary inherited optic neuropathies.
Eye, 2004. 18(11): p. 1126-32.
215.
Pedrola, L., et al., GDAP1, the protein causing Charcot-Marie-Tooth disease type
4A, is expressed in neurons and is associated with mitochondria. Hum Mol
Genet, 2005. 14(8): p. 1087-94.
216.
Kijima, K., et al., Mitochondrial GTPase mitofusin 2 mutation in Charcot-MarieTooth neuropathy type 2A. Hum Genet, 2005. 116(1-2): p. 23-7.
217.
Li, Z., et al., The importance of dendritic mitochondria in the morphogenesis and
plasticity of spines and synapses. Cell, 2004. 119(6): p. 873-87.
218.
Verstreken, P., et al., Synaptic mitochondria are critical for mobilization of
reserve pool vesicles at Drosophila neuromuscular junctions. Neuron, 2005.
47(3): p. 365-78.
219.
Pagliarini, D.J. and J.E. Dixon, Mitochondrial modulation: reversible
phosphorylation takes center stage? Trends Biochem Sci, 2006. 31(1): p. 26-34.
220.
Petit, A., et al., Wild-type PINK1 prevents basal and induced neuronal apoptosis,
a protective effect abrogated by Parkinson disease-related mutations. J Biol
Chem, 2005. 280(40): p. 34025-32.
221.
Rogaeva, E., et al., Analysis of the PINK1 gene in a large cohort of cases with
Parkinson disease. Arch Neurol, 2004. 61(12): p. 1898-904.
222.
Huang, E.J. and L.F. Reichardt, Neurotrophins: roles in neuronal development
and function. Ann Rev Neurosci, 2001. 24: p. 677-736.
223.
Chu, C.T., et al., Oxidative neuronal injury. The dark side of ERK1/2. Eur J
Biochem, 2004. 271(11): p. 2060-6.
224.
Bauman, A.L., A.S. Goehring, and J.D. Scott, Orchestration of synaptic plasticity
through AKAP signaling complexes. Neuropharmacol, 2004. 46(3): p. 299-310.
225.
Cadd, G. and G.S. McKnight, Distinct patterns of cAMP-dependent protein
kinase gene expression in mouse brain. Neuron, 1989. 3(1): p. 71-9.
226.
Kawakami, M. and N. Nakanishi, The role of an endogenous PKA inhibitor,
PKIalpha, in organizing left-right axis formation. Development, 2001. 128(13): p.
2509-15.
227.
Dalton, H. and W. Dewey, Protein kinase inhnibitor peptide (PKI): A family of
endogenous neuropeptides that modulate neuronal cAMP-dependent protein
kinase function. Neuropeptides, 2006. 40(1): p. 23-24.
228.
Tasken, K. and E.M. Aandahl, Localized effects of cAMP mediated by distinct
routes of protein kinase A. Physiol Rev, 2004. 84(1): p. 137-67.
188
229.
Wong, W. and J.D. Scott, AKAP signalling complexes: focal points in space and
time. Nat Rev Mol Cell Biol, 2004. 5(12): p. 959-70.
230.
dell'Acqua, M. NMDA receptor-PP2B pathways implicated in LTD induced loss
of AKAP79/150 and anchored PKA-RII from synapses. in FASEB summer
research conference on protein phosphatases. 2004. Snowmass Village, CO.
231.
Huang, L.J., et al., NH2-Terminal targeting motifs direct dual specificity Akinase-anchoring protein 1 (D-AKAP1) to either mitochondria or endoplasmic
reticulum. J Cell Biol, 1999. 145(5): p. 951-9.
232.
Huang, L.J., et al., D-AKAP2, a novel protein kinase A anchoring protein with a
putative RGS domain. Proc Natl Acad Sci U S A, 1997. 94(21): p. 11184-9.
233.
Wang, L., et al., Cloning and mitochondrial localization of full-length D-AKAP2,
a protein kinase A anchoring protein. Proc Natl Acad Sci U S A, 2001. 98(6): p.
3220-5.
234.
Cardone, L., et al., A-kinase anchor protein 84/121 are targeted to mitochondria
and mitotic spindles by overlapping amino-terminal motifs. J Mol Biol, 2002.
320(3): p. 663-75.
235.
Cardone, L., et al., Mitochondrial AKAP121 binds and targets protein tyrosine
phosphatase D1, a novel positive regulator of src signaling. Mol Cell Biol, 2004.
24(11): p. 4613-26.
236.
Ginsberg, M.D., et al., PKA-dependent binding of mRNA to the mitochondrial
AKAP121 protein. J Mol Biol, 2003. 327(4): p. 885-97.
237.
Livigni, A., et al., Mitochondrial AKAP121 Links cAMP and src Signaling to
Oxidative Metabolism. Mol Biol Cell, 2006. 17(1): p. 263-71.
238.
Ghafourifar, P. and E. Cadenas, Mitochondrial nitric oxide synthase. Trends
Pharmacol Sci, 2005. 26: p. 190-195.
239.
Damuni, Z. and L.J. Reed, Purification and characterization of a divalent cationindependent, spermine-stimulated protein phosphatase from bovine kidney
mitochondria. J Biol Chem, 1987. 262(11): p. 5133-8.
240.
Bok, J., et al., An extranuclear locus of cAMP-dependent protein kinase action is
necessary and sufficient for promotion of spiral ganglion neuronal survival by
cAMP. J Neurosci, 2003. 123(3): p. 777-787.
241.
Rukenstein, A., R.E. Rydel, and L.A. Greene, Multiple agents rescue PC12 cells
from serum-free cell death by translation- and transcription-independent
mechanisms. J Neurosci, 1991. 11(8): p. 2552-63.
242.
Newlon, M.G., et al., A novel mechanism of PKA anchoring revealed by solution
structures of anchoring complexes. Embo J, 2001. 20(7): p. 1651-62.
243.
Huang, J. and S.H. Green. cAMP-dependent proetein kinase A (PKA) promotes
spiral ganglion neuron (SGN) survival through outer miochondrial membrane
(OMM)signaling. in Association for Research in Otolaryngology 29th Midwinter
Research Meeting. 2006. Baltimore, MD.
189
244.
Jones, A.F., et al. Bcl-xL modulates synaptic transmission in cultured
hippocampal neurons. in Society for Neuroscience Conference. 2005. Washington
D. C.
245.
Carelli, V., F.N. Ross-Cisneros, and A.A. Sadun, Mitochondrial dysfunction as a
cause of optic neuropathies. Prog Retin Eye Res, 2004. 23(1): p. 53-89.
246.
Floor, E. and M.G. Wetzel, Increased protein oxidation in human substantia
nigra pars compacta in comparison with basal ganglia and prefrontal cortex
measured with an improved dinitrophenylhydrazine assay. J Neurochem, 1998.
70: p. 268-275.
247.
Dexter, D.T., et al., Basal lipid peroxidation in substantia nigra is increased in
Parkinson's disease. J Neurochem, 1989. 52: p. 381-389.
248.
Pearce, R.K., et al., Alterations in the distribution of glutathione in the substantia
nigra in Parkinson's disease. J Neural Transm, 1997. 104: p. 661-667.
249.
Breckenridge, D.G., et al., Caspase cleavage product of BAP31 induces
mitochondrial fission through endoplasmic reticulum calcium signals, enhancing
cytochrome c release to the cytosol. J Cell Biol, 2003. 160(7): p. 1115-27.
250.
Arnoult, D., et al., Bax/Bak-dependent release of DDP/TIMM8a promotes Drp1mediated mitochondrial fission and mitoptosis during programmed cell death.
Curr Biol, 2005. 15(23): p. 2112-8.
251.
Harder, Z., R. Zunino, and H. McBride, Sumo1 conjugates mitochondrial
substrates and participates in mitochondrial fission. Curr Biol, 2004. 14(4): p.
340-5.
252.
Meller, R., et al., Seizure-like activity leads to the release of BAD from 14-3-3
protein and cell death in hippocampal neurons in vitro. Cell Death Differ, 2003.
10(5): p. 539-47.
253.
Wang, H.G., et al., Ca2+-induced apoptosis through calcineurin
dephosphorylation of BAD. Science, 1999. 284(5412): p. 339-43.
254.
Smillie, K.J. and M.A. Cousin, Dynamin I phsphorylation and the control of
synaptic vesicle endocytosis. Biochem Soc Symp, 2005: p. 87-89.
255.
Carnegie, G.K. and J.D. Scott, A-kinase anchoring proteins and neuronal
signaling mechanisms. Genes Develop, 2003. 17: p. 1557-1568.
256.
Kamibayashi, C., et al., Comparison of heterotrimeric protein phosphatase 2A
containing different B subunits. J Biol Chem, 1994. 269(31): p. 20139-48.
257.
Ferrigno, P., T.A. Langan, and P. Cohen, Protein phosphatase 2 A1 is the major
enzyme in vertebrate cell extracts that dephosphorylates several physiological
substrates of cyclin-depndent protein kinases. Mol Biol Cell, 1993. 4: p. 669-677.
258.
Agostinis, P., et al., Specificity of the polycation-stimulated (type-2A) and ATP,
Mg-dependent (type-1) protein phosphatases towards substrates phosphorylated
by p34cdc2 kinase. Eur. J. Biochem., 1992. 205: p. 241-248.
190
259.
Mayer-Jaekel, R.E., et al., Drosophila mutants in the 55 kDa regulatory subunit of
protein phosphatase 2A show strongly reduced ability to dephosphorylate
substrates of p34cdc2. J Cell Sci, 1994. 107(Pt 9): p. 2609-16.
260.
Fukunaga, K., et al., Decreased protein phosphatase 2A activity in hippocampal
long-term potentiation. J Neurochem, 2000. 74(2): p. 807-17.
261.
Gigena, M.S., et al., A B56 regulatory subunit of protein phosphatase 2A localizes
to nuclear speckles in cardiomyocytes. Am J Physiol Heart Circ Physiol, 2005.
289(1): p. H285-94.
262.
Letourneux, C., G. Rocher, and F. Porteu, B56-containing PP2A dephosphorylate
ERK and their activity is controlled by the early gene IEX-1 and ERK. Embo J,
2006. 25(4): p. 727-38.
263.
Holmes, S.E., et al., Spinocerebellar ataxia type 12, in Genetic Instabilities and
Neurological Diseases, R.D. Wells and A. T., Editors. 2006, Elzeiver-Academic
Press: San Diego. p. In Press.
264.
Dessolin, J., et al., Selective targeting of synthetic antioxidants to mitochondria:
towards a mitochondrial medicine for neurodegenerative diseases? Eur J
Pharmacol, 2002. 447(2-3): p. 155-61.
265.
Alvarez, R.D., et al., A cancer gene therapy approach utilizing an anti-erbB-2
single-chain antibody-encoding adenovirus (AD21): a phase I trial. Clin Cancer
Res, 2000. 6(8): p. 3081-7.
266.
Lecerf, J.M., et al., Human single-chain Fv intrabodies counteract in situ
huntingtin aggregation in cellular models of Huntington's disease. Proc Natl
Acad Sci U S A, 2001. 98(8): p. 4764-9.
267.
Wong, W.T. and R.O. Wong, Changing specificity of neurotransmitter regulation
of rapid dendritic remodeling during synaptogenesis. Nat Neurosci, 2001. 4(4): p.
351-2.
268.
Desmond, N.L. and W.B. Levy, Changes in the postsynaptic density with longterm potentiation in the dentate gyrus. J Comp Neurol, 1986. 253(4): p. 476-82.
269.
Ethell, I.M. and E.B. Pasquale, Molecular mechanisms of dendritic spine
development and remodeling. Prog Neurobiol, 2005. 75(3): p. 161-205.
270.
Gonzalez-Aguilar, F., Electrical and chemical synaptic transmission as an
interacting system. Med Hypotheses, 2000. 54(1): p. 40-6.
271.
Heidelberger, R., ATP is required at an early step in compensatory endocytosis in
synaptic terminals. J Neurosci, 2001. 21(17): p. 6467-74.
272.
Starkov, A.A., C. Chinopoulos, and G. Fiskum, Mitochondrial calcium and
oxidative stress as mediators of ischemic brain injury. Cell Calcium, 2004. 36(34): p. 257-64.
191
273.
Chen, Q.X., et al., Secondary activation of a cation conductance is responsible for
NMDA toxicity in acutely isolated hippocampal neurons. J Neurosci, 1997.
17(11): p. 4032-6.
274.
Randall, R.D. and S.A. Thayer, Glutamate-induced calcium transient triggers
delayed calcium overload and neurotoxicity in rat hippocampal neurons. J
Neurosci, 1992. 12(5): p. 1882-95.
192
APPENDIX- CURRICULUM VITAE
Ruben K. Dagda was born on April 27, 1978 in El Paso, Texas. He is the third son
of Raul Dagda Medina and Agueda Saenzpardo Prieto. He graduated from Montwood
High School, El Paso, Texas, in the spring of 1995 and entered The University of Texas
at El Paso in the fall of 1995. During his undergraduate studies, he was an active
participant in various organizations that included the American Society for Microbiology,
the National Honor Society and the Golden Key National Honor Society. He graduated
with a Bachelor’s of Science in microbiology on December of 1998 with an Honor’s
certificate.
He then entered graduate school on spring of 1999 in pursuit of a Master’s of
Science in Biology. He worked as a research assistant for Dr. Eppie Rael under the
SCORE Program for 2 years. Upon graduating with an M.S. degree, he entered the
doctoral program in Pharmacology at the University of Iowa and pursued his Ph.D. in
Pharmacology in the laboratory of Dr. Stefan Strack. During his predoctoral studies, he
presented his dissertation work at numerous prestigious national conferences and local
conferences which included the Experimental Biology 2002 Conference, the Society for
Neuroscience Conferences of 2003 and 2005, the FASEB Phosphatase Conference in
2004, the American National Society for Chicanos and Native Americans meetings in
September 2005, and at the Research Week Conferences of 2003 and 2004. In July of
2004, he successfully obtained a NINDS F31 Predoctoral Fellowship Grant (NS049659)
to fund his final two years of dissertation work. Upon graduating with a Ph.D. in
Pharmacology at the University of Iowa, he will pursue a postdoctoral associate program
at the University of Pittsburgh Medical Center in the laboratory of Dr. Charleen Chu.
This thesis was typed by Ruben Karim Dagda.