TRANSLOCATION AND FUNCTION OF AKT IN THE MITOCHONDRIA
by
Keri April Mans
Gautam N. Bijur, COMMITTEE CHAIR
John Hablitz
Richard Jope
Anne Theibert
Scott Wilson
A DISSERTATION
Submitted to the graduate faculty of The University of Alabama at Birmingham,
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
BIRMINGHAM, ALABAMA
2010
TRANSLOCATION AND FUNCTION OF AKT IN THE MITOCHONDRIA
KERI APRIL MANS
DEPARTMENT OF NEUROBIOLOGY
ABSTRACT
The ubiquitously expressed kinase Akt is a known survival protein, and is
involved in multiple cell signaling cascades, notably the phosphatidylinositol 3-kinase
(PI3K) pathway. Active Akt translocates from the plasma membrane to different
subcellular
compartments,
including
mitochondria,
where
it
phosphorylates
compartment-specific substrates. The mechanism of this translocation and the specific
function of Akt within the mitochondria remains a mystery. The goals of this study were
to elucidate the mechanism by which Akt enters the mitochondria and examine the role of
Akt in mitochondrial function. Finally, the possibilities of using post-mortem human
brain tissue to study mitochondrial function in normal and diseased states were
considered.
Heat shock protein-90 (HSP90) is a major player in translocation of proteins into
mitochondria, and Akt is a known client protein of HSP90. Knockdown or
pharmacological inhibition of HSP90 caused decreases in mitochondrial Akt levels, as
well as inhibition of Akt protein translocation into isolated mitochondria. This indicates
HSP90 mediates mitochondrial accumulation of Akt. Furthermore, incubation of
constitutively active Akt with isolated mitochondria caused alterations in mitochondrial
morphology which are indicative of increased respiration.
ii
An ultrastructural study revealed widespread Akt labeling in neurons.
Surprisingly, mitochondria in neuronal processes (pMito) contained the highest levels of
Akt compared to mitochondria in the soma (cMito). Functional studies of the two
mitochondrial populations revealed that pMito produce ATP at higher rates than cMito,
and rates of ATP production in cortical mitochondria can be modulated via activation or
inhibition of PI3K signaling. These data support the notion that levels of Akt in
mitochondria can influence morphology and energy production.
To test if mitochondrial function could be examined in human brain, it was
necessary to determine if it is possible to obtain structurally and functionally intact
mitochondria from post-mortem brain. Studies in human and rodent models indicate that
functional mitochondria can be isolated from post-mortem brain up to 10 hours after
expiration, and these samples withstand cryopreservation for later testing.
These data provide novel insights into mitochondrial translocation and function of
Akt, as well as new frontiers for the study of mitochondria in human brain in health and
disease.
Keywords: Akt, Neuronal Mitochondria, HSP90, Mitochondrial Translocation, PI3K
iii
DEDICATION
Through life’s celebrations and struggles, my parents have carried their figurative
pom-poms with pride, my own personal cheerleading squad. My mom and dad have and
continue to support me in all my personal and career endeavors and encourage me to
press forward, never ceasing to remind me of the payoff at the end. Thus, it is with much
love and excitement that I dedicate this dissertation (at long last) to my parents and
biggest fans, Paula and Lee Peterson.
iv
ACKNOWLEDGMENTS
First, I extend my deepest thanks to my mentor, Dr. Gautam Bijur, who has
fostered my growth into a scientist. From side-by-side bench work to preparation of
papers and talks; from weekly meetings to cookouts at his home, he has been both a
mentor and a friend to me during my graduate career. I have deep respect for him as a
mentor and as a scientist, and look forward to further collaborations with his laboratory. I
would further like to acknowledge the members of my graduate committee, Drs. John
Hablitz, Richard Jope, Anne Theibert, and Scott Wilson, for guidance and support during
my graduate work. At each committee meeting, I could always expect to be challenged
by their questions and to receive positive feedback and new ideas for my project.
My parents (Lee and Paula Peterson), godparents (Tommy and Donna Serio), and
those that I consider family (Rich and Eileen Stoll) have my deepest gratitude for their
love and support during this journey. They have so much pride in everything I do, and I
appreciate them all for showing such interest in my goals. Both mine and Robert’s
families have been a limitless source of encouragement as the both of us have pursued
advanced degrees. Finally, but certainly not least, I thank my husband Robert, who brings
me flowers, says he likes my cooking, knows me like no one else ever could, and keeps
me around anyway.
Thank you all.
v
TABLE OF CONTENTS
Page
ABSTRACT........................................................................................................................ ii
DEDICATION................................................................................................................... iv
ACKNOWLEDGMENTS ...................................................................................................v
LIST OF FIGURES ......................................................................................................... viii
LIST OF TABLES...............................................................................................................x
INTRODUCTION ...............................................................................................................1
Akt (Protein Kinase B)..................................................................................................1
Activation Mechanism of Akt..............................................................................2
Role of Akt in Cell Survival ................................................................................3
Role of Akt in the Brain.......................................................................................4
Association between Akt and HSP90 ..................................................................6
Mitochondrial Function of Akt ............................................................................8
Mitochondria...............................................................................................................11
Evolutionary Origins of Mitochondria..............................................................11
Mitochondrial Structure ....................................................................................13
Mitochondrial Morphology...............................................................................13
Cellular Distribution of Mitochondria ..............................................................15
Mitochondrial Protein Translocation ................................................................17
Role of Mitochondria in Neurological Disorders .............................................19
Post Mortem Tissue: Direct Investigation of the Human Brain .................................22
Post-Mortem Tissue Viability...........................................................................22
The Study of Mitochondrial Function in the Human Brain..............................24
THE BASAL FLUX OF AKT IN THE MITOCHONDRIA IS MEDIATED
BY HEAT SHOCK PROTEIN-90.....................................................................................28
LOCALIZATION AND FUNCTION OF AKT IN BRAIN MITOCHONDRIA.............63
vi
MITOCHONDRIAL VIABILITY IN MOUSE AND
HUMAN POSTMORTEM BRAIN...................................................................................88
CONCLUSIONS..............................................................................................................121
LIST OF GENERAL REFERENCES .............................................................................134
APPENDICES
A
IACUC APPROVAL....................................................................................147
B
IRB APPROVAL..........................................................................................148
vii
LIST OF FIGURES
Figure
Page
INTRODUCTION
1
Hypothetical model of Akt activity within mitochondria ......................................26
2
Proposed mechanism of Akt translocation into mitochondria...............................27
THE BASAL FLUX OF AKT IN THE MITOCHONDRIA IS MEDIATED BY
HEAT SHOCK PROTEIN-90
1
HSP90 inhibitors decrease mitochondrial Akt levels
in SH-SY5Y and HEK293 cells.............................................................................55
2
siRNA knockdown of HSP90 expression results in decreased levels of
mitochondrial Akt ..................................................................................................57
3
HSP90 mediates the mitochondrial translocation of Akt.......................................58
4
HSP90 inhibitors do not block IGF-1 induced mitochondrial
Akt accumulation ...................................................................................................60
5
Akt affects mitochondrial morphology..................................................................61
LOCALIZATION AND FUNCTION OF AKT IN BRAIN MITOCHONDRIA
1
Localization of Akt and pSer473 in mouse cortex.................................................82
2
Confirmation of Akt levels in two populations of mitochondria...........................84
3
Modulation of mitochondrial Akt levels by activating or
inhibiting PI3K signaling.......................................................................................85
viii
MITOCHONDRIAL VIABILITY IN MOUSE AND HUMAN POSTMORTEM BRAIN
1
Effect of increasing PMIs on mitochondrial ΔΨmem .........................................112
2
Mitochondria ΔΨmem in human postmortem brain............................................114
3
Structurally intact mitochondria can be obtained from postmortem
human brain .........................................................................................................116
4
Cryopreservation for 1 week does not affect ΔΨmem.........................................117
DISCUSSION
1
Hypothesized model for mitochondrial Akt function in neurons ........................133
ix
LIST OF TABLES
Tables
Page
LOCALIZATION AND FUNCTION OF AKT IN BRIAIN MITOCHONDRIA
1
ATP Production rates in cMito and pMito under basal and
stimulated conditions .............................................................................................87
MITOCHONDRIAL VIABILITY IN MOUSE AND HUMAN POSTMORTEM
BRAIN
1
Demographic data of the human cases used ........................................................119
2
Rates of ATP production ....................................................................................120
x
INTRODUCTION
Akt (Protein Kinase B)
The origins of Akt research can be traced back to 1977 when Staal and colleagues
observed that a murine leukemia virus exhibited a high incidence of spontaneous
lymphoma (Staal 1977). This virus, named AKT8, caused malignant transformation in a
mink lung cell line and could induce focus formation in other cell lines (Staal 1987;
1988). These data suggested that the virus contained an oncogene which could play a role
in pathogenesis of malignancy, and this new oncogene was named akt (Staal 1987; 1988).
In 1991 three independent research teams published seminal findings which showed that:
the akt gene encoded a serine/threonine protein kinase related to protein kinase A (PKA)
and protein kinase C (PKC); the protein encoded by akt could be phosphorylated; and
that the akt gene product held significant homology to both PKA and PKC (reviewed by
Brazil and Hemmings, 2001). Thus, Akt is also referred to as Protein Kinase B.
Today, three known isoforms of Akt exist in mammalian cells: PKBα/Akt1,
PKBβ/Akt2, and PKBγ/Akt3 (Brazil 2001). There is 80% homology between the three,
and they contain similar phosphorylation (activation) sites (Santi 2009), but each has a
slightly different expression profile throughout the mammalian system. Knockout or
upregulation of each results in a distinct phenotype (Yang et al., 2004). For example,
Akt1 is expressed predominantly in the brain and knockout results in severe deficits in
brain growth as well as a significantly reduced body size (Yang 2004). Akt2 is most
1
highly expressed in insulin-sensitive tissues, and loss of Akt2 leads to a severely diabetic
phenotype (Cho 2001). Akt3 is highly expressed in the brain, and loss of this isoform is
known to result in an approximate 25% decline in brain mass (Easton 2005), while body
weight and glucose metabolism remain normal.
As a proto-oncoprotein, Akt was originally examined for its regulation of cellular
survival and proliferation pathways. However, Akt was considered an unrelated protooncogene product and protein kinase until it was demonstrated that Akt is activated by
growth factors in a phosphatidylinositol 3-kinase (PI3K) dependent manner (Whiteman
2002). Today, much more is known about Akt beyond its role in cell survival, and it has
emerged as a key regulator of many cellular processes, including gene expression and
metabolism.
Activation Mechanism of Akt
Akt is an abundant Ser/Thr protein kinase which is involved in multiple cell
signaling pathways. It can be activated by a number of signaling cascades, but perhaps
the most prominently studied pathway in the activation of Akt is the phosphatidylinositol
3-kinase (PI3K) signaling pathway. Stimulation of cell surface receptors with insulin or
growth factors, such as insulin-like growth factor-1 (IGF-1), results in the activation of
phosphatidylinositol 3-kinase and the formation of phosphatidylinositol (3,4,5)trisphosphate from phosphatidylinositol (4,5)-bisphosphate (Vanhaesebroeck 2000).
Phosphatidylinositol-dependent kinase-1 (PDK-1) and Akt are recruited to the plasma
membrane, where PDK1 directly phosphorylates Akt on its Thr308 site (Alessi, 1996;
Alessi, 1997; Fayard, 2005; Lessman, 2007). This results in a conformational change,
2
allowing phosphatidylinositol-dependent kinase-2 to phosphorylate Akt on its Ser 473
site (Lessman, 2007; Jacinto 2006). Phosphorylation at these sites results in the full
activation of Akt. Following activation, Akt rapidly translocates to other cellular
compartments, including the mitochondria where it phosphorylates compartment-specific
substrates (Andjelkovic 1997; Bijur 2003) to affect a multitude of cellular processes.
Role of Akt in Cell Survival
Akt is best known for its role as an anti-apoptotic, tumorigenic and cellular
survival protein. Several known Akt substrates are located throughout different parts of
the cell, many of which are involved in cell proliferation, differentiation, and survival.
Upon stimulation of the PI3K signaling pathway, Akt activity leads to inactivation of
multiple pro-apoptotic signaling proteins such as Bad, glycogen synthase kinase 3β
(GSK3β), caspase-9, and Bax (Scheid 2003; Kennedy 1999), as well as activation of prosurvival proteins such as Bcl-2 (Zhou 2000). Active Akt can also regulate transcription of
several genes involved in cell death and survival (Arciuch 2009). Akt has been shown to
directly inhibit mitochondrial swelling and release of cytochrome c from the
mitochondria in a caspase-independent mechanism (Kennedy 1999) through the
phosphorylation of the pro-apoptotic protein Bad, resulting in its sequestration (Arciuch
2009; Stiles 2009) and Bax, which inhibits its oligomerization to the mitochondria (Stiles
2009). It can also inhibit cell death through the direct phosphorylation of pro-caspase 9,
inhibiting its cleavage and subsequent apoptotic signaling (Trencia 2003). Furthermore,
inhibition of GSK3β by Akt leads to inability of GSK3β to carry out its pro-apoptotic
functions which include activation of pro-apoptotic proteins and inducing degradation of
anti-apoptotic proteins (Stiles 2009).
3
Role of Akt in the Brain
In addition to the role in general cell survival in multiple tissues, Akt has several
known functions in the developing and adult brain. It has been shown recently that in the
aging brain, hippocampal levels of activated Akt decrease while levels of proteins related
to cell death and Akt inactivation steadily increase (Jackson 2008). The loss of Akt in the
CA1 region of the hippocampus throughout the lifespan leads to reduced survival
signaling and increased propensity to neuronal insults in the aging brain (Jackson 2008).
In the developing brain, asymmetric distribution of Akt in developing neurons is a
requirement for the establishment of neuronal polarity, and Akt is selectively degraded in
dendrites to avoid development of multiple axons (Yan 2006; Arimura 2005).
Akt is well known to be involved in neuroprotection in the face of insults such as
excitotoxicity and ischemic stroke (Brywe 2005). For example, it has been shown in
oligodendrocyte progenitor cells that glutamate-induced excitotoxicity is a major
contributor to white matter loss in CNS demyelinating disorders (Pitt 2000, Matute 1997).
Given that oligodendrocyte loss leads to multiple CNS disorders including multiple
sclerosis, it is imperative to have an intact protection mechanism for these cells. Recent
studies have found that activation of the PI3-kinase pathway in oligodendrocytes prevents
glutamate-induced apoptotic signaling, thus promoting oligodendrocyte survival and the
continued protection of neurons (Ness 2004). It has also been shown that the herbal antiinflammatory drug baicalein is protective in ischemic stroke models by reducing
oxidative damage and infarct size and improving cognitive deficits (Lapchak 2007, Liu
2007). More recent work by Chao Liu and colleagues has suggested that the baicalein-
4
induced neuroprotection that occurs during middle cerebral artery occlusion ischemia is
modulated by the PI3K signaling pathway through the negative regulation of Bad as well
as positive regulation of anti-apoptotic proteins (Liu 2009). These aforementioned effects
are blocked by the addition of PI3-K inhibitors, suggesting that Akt plays an active role
in neuroprotection.
The role of Akt signaling, particularly in the brain, has been the subject of
numerous studies because of its involvement in psychiatric and neurological disease
pathology. Deregulation of the PI3K/Akt signaling cascade is postulated to play a direct
role in the pathogenesis of schizophrenia (Fayard 2005; Kalkman 2006). Akt is widely
known to phosphorylate GSK3β on the serine-9 residue, resulting in its inactivation (Ali
2001), which is notable because aberrant GSK3β activity in the brain has been associated
with the pathology of neurodegenerative and psychiatric disorders such as Alzheimer’s
disease, Parkinson’s disease, bipolar disorder, and schizophrenia (Hanger 1992; King
2001; Klein 1996; Kozlovsky 2000).
Akt has garnered interest not only from those interested in cellular survival
mechanisms, but also from scientists looking to examine signaling related to neuronal
plasticity. In fact, insulin signaling in the brain which activates Akt, has been reported to
induce changes in synaptic strength, which underlies the basis for all that is known about
learning and memory mechanisms (reviewed by van der Heide 2006). It has been shown
recently that exercise as well as high frequency stimulation of acute brain slices increases
the magnitude of long term potentiation in a PI3K-dependent manner and that acute brain
infusion of PI3K inhibitors causes a significant suppression of long term potentiation
(Bruel-Jungerman 2009, Sui 2008). Furthermore, in a rat model of adult neurogenesis
5
and synaptic plasticity, the PI3K pathway is activated during exercise, and the subsequent
increase in Akt activity has a significant role in mediating survival of newly formed
granule cells in the dentate gyrus of the hippocampus (Bruel-Jungerman 2009). Recent
work by Collingridge has shown that Akt is activated during long term potentiation and
through inhibitory phosphorylation of GSK3β, prevents the occurrence of long term
synaptic depression (Peineau 2007), suggesting yet another role for Akt in plasticity
mechanisms. From these experiments, one can conclude that Akt not only plays a
significant role in brain development and neuronal survival, but also is a major player in
neuronal plasticity.
Association between Akt and Heat Shock Protein-90
Heat Shock Protein-90 (HSP90) is one of the most abundant proteins in the cell,
and levels greatly increase following cell stress or heat shock conditions (McClellan
2007). HSP90 is an ATPase-driven molecular chaperone that is known for its activities in
recognizing unfolded proteins after cell stress, and with the aid of co-chaperones,
refolding these proteins (Kang 2007). It exists as an obligatory homodimer and each
subunit has three major domains: the N-terminal domain is essential for ATPase activity,
a requirement for substrate binding and release; the middle domain is typically known as
the substrate binding domain; the C-terminal domain is essential for homodimerization
(Pearl 2006).
HSP90, aside from its key role in recovery from cell stress, has multiple cellular
functions, including stabilization of cell signaling proteins, maturation of client proteins
after folding, and regulation of cell cycle progression (Kang et al. 2007, Meares 2004,
Sato 2000, McClellan 2007). Pharmacological inhibitors of HSP90 have provided much
6
insight into its function. For example, the coumarin-based compound novobiocin (NB)
inhibits HSP90 by targeting the C-terminal region of HSP90 adjacent to its
homodimerization domain (Marcu, 2000a). NB treatment has been associated with
destabilization and proteolysis of HSP90 client proteins (Marcu 2000b; Yun, 2004). In
addition, geldanamycin (GA), a benzoquinoid ansamycin antibiotic, inhibits HSP90 by
binding to its N-terminal ATPase domain (Whitesell, 1994; Panaretou, 1998). HSP90
has been shown to have an important role in trafficking proteins between subcellular
compartments, notably mitochondrial protein import (Fan 2006, Young 2003). The
import receptors for the translocase of the outer mitochondrial membrane (TOM) contain
specific binding sites for HSP90 and HSP70 (Young 2003), and recently it was reported
that inhibition of HSP90 with novobiocin and geldanamycin blocks the import of
preproteins into mitochondria in two distinct ways (Fan 2006). Novobiocin prohibits
initial binding of HSP90 to its client protein, while geldanamycin inhibits ATP hydrolysis,
preventing release of the client protein into the import pore of the mitochondrion (Fan
2006). Together, this information indicates that HSP90 plays a major role in the
translocation of proteins into mitochondria.
HSP90 is known to be responsible for the correct folding, stabilization, and
activity of many Ser/Thr protein kinases, and Akt is a well-known client protein of
HSP90 (Csermely 1998, Sato 2000). Binding of Akt to HSP90 provides stability by
protecting it from dephosphorylation (Sato 2000) and proteasomal degradation (Basso
2002), and by allowing it to phosphorylate downstream substrates. HSP90 binding to Akt
is required for Akt phosphorylation and subsequent activity, while apoptotic signaling is
enhanced when HSP90 is inhibited or becomes detached from Akt (Sato 2000).
7
Treatment of cells with geldanamycin affects Akt by inducing its dephosphorylation
(Fujita 2002; Xu 2003), and long treatments or high doses induce the degradation of Akt,
indicating that HSP90 is absolutely essential for stability and function of Akt. Thus, in
addition to phosphorylation; Akt activity is dependent on its interaction with HSP90.
Because HSP90 is known to bind Akt and to aid in chaperoning proteins to the
mitochondria, it is reasonable to hypothesize that HSP90 aids in mitochondrial
accumulation of Akt.
Mitochondrial Function of Akt
In 2003 it was shown by Bijur and Jope that Akt exists in the mitochondria, and
levels greatly increase following activation of the PI3K signaling pathway. Furthermore,
it was shown that Akt phosphorylates the β-subunit of ATP Synthase, also called
Complex V, of the mitochondrial electron transport chain. It has been suggested that
phosphorylation of the ATP synthase β-subunit by Ser/Thr kinases regulates rates of ATP
production in skeletal muscle as well as bacterial models (Hojlund 2003; Alekseeva
2009). Using this information, it can be hypothesized that Akt has some direct effect on
mitochondrial metabolism, possibly through enhancing rates of mitochondrial ATP
production (Figure 1). In fact, some investigators have elucidated effects of PI3K
signaling on mitochondrial function.
From activation of Akt at the cytoplasmic side of the plasma membrane, Akt has
some known effects on the mitochondria. Seminal work has been done using heart and
skeletal muscle tissue, because of the abundance of mitochondria and high energy
demands in these areas. Treatment of cardiac cells with IGF-1 leads to activation of Akt
8
and restores the electrochemical gradient in the mitochondria after depolarization, an
effect that is blocked by an inhibitor of PI3K signaling (Lai 2003). Furthermore, it has
been shown that ATP-dependent mitochondrial potassium channels are protective from
ischemic conditions in the heart, and that this protection is dependent upon Akt
translocation from the cytoplasm into the mitochondria (Ahmad 2006), though the exact
mechanism of protection has yet to be elucidated. Akt is tightly related to insulin, and in
skeletal muscle tissue with insulin resistance, mitochondrial number and size are reduced
and there are profound abnormalities in electron transport (Szendroedi 2008). Lastly,
work in cultured cardiomyocytes revealed that mitochondrial depolarization induced by
hydrogen peroxide treatment or calcium stress is prevented by the activation of Akt
(Miyamoto 2007).
A steady ATP supply is imperative for many tissues, and the brain is certainly no
exception. Experiments performed in cultured adult sensory neurons as well as human
muscle tissue showed that insulin infusion increases levels of ATP as well as rates of
mitochondrial ATP production and enhances the ΔΨmem (Stump 2003; Huang 2004).
Serum withdrawal is known to cause marked decreases in mitochondrial respiration in
multiple cell culture models using a variety of tissues, and treatment with small doses of
IGF-1 rescues deficits in respiration due to growth factor/serum withdrawal
(Unterluggauer 2008). Nerve growth factor signaling in the brain is also known to
enhance the ΔΨmem when applied locally to axons of cultured dorsal root ganglia cells
(Verburg 2008).
Akt has many effects on cell survival at the level of the mitochondrion. For
example, glucose deprivation is postulated to enhance apoptotic signaling by preventing
9
Akt-hexokinase interactions, thereby enhancing cleavage of the pro-apoptotic protein
BID and promoting the mitochondrial oligomerization of the pro-apoptotic protein BAX
(Majewski 2004). A series of studies have postulated and confirmed that activated Akt
leads to higher hexokinase levels, the enzyme that catalyzes the first committed reaction
of glycolysis (Gottlobb 2001; Majewski 2004). In fibroblasts, Akt is able to
phosphorylate mitochondrial hexokinase-II to inhibit apoptosis caused by BID cleavage,
BAX oligomerization, and subsequent cytochrome c release (Majewski 2004). Together,
the results of these studies indicate that Akt has an important role at the level of the
mitochondrion to not only affect mitochondrial function but also to promote cell survival.
Though these data support a role for the effect of Akt signaling on mitochondria,
surprisingly few studies have examined the effect of the mitochondrial pool of Akt on
organelle function. It has been suggested that intra-mitochondrial cycling of Akt regulates
cell cycle progression and cell fate (Antico 2009). Besides phosphorylation of the ATP
synthase β-subunit, mitochondrial Akt is also known to phosphorylate and inactivate
GSK3β, a reported deactivator of pyruvate dehydrogenase (Bijur 2003; Hoshi 1996). The
effect of this interaction inhibits GSK3β’s anti-apoptotic actions and theoretically
increases mitochondrial activity at the level of the tricarboxylic acid cycle (Pilegaard
2006) (Figure 1). Since Akt has known substrates within the mitochondria, it is feasible
that Akt may have a direct effect on mitochondrial function, and that these effects extend
to overall cellular health and function.
10
Mitochondria
Mitochondria are traditionally recognized for their ability to efficiently produce
large amounts of energy for cells to carry out vital functions. Only in the past several
years has our view of mitochondria as the “powerhouse of the cell” evolved into the
knowledge that mitochondria are intimately involved in cellular life, death and function.
Mitochondria are highly complex and compartmentalized organelles that are able to
change shape and move throughout the cell in response to varying cellular conditions.
Outside their role in ATP production, mitochondria can buffer calcium to regulate
neurotransmission and cell signaling events, regulate cell cycle progression, and
participate in cellular and apoptotic signaling. In addition, proteins that were previously
thought to be cytoplasmic and nuclear, such as Akt and GSK3β, are currently emerging
as significant players in the world of mitochondrial function.
Evolutionary Origins of Mitochondria and Development of Mitochondrial Research
The accepted theory of the inception and evolution of the mitochondrion is today
known as the Endosymbiotic Theory (described by IE Scheffler in Mitochondria, 2nd
edition; Gray 1993). Millions of years ago, a primitive prokaryote capable of oxidative
phosphorylation was taken up by a glycolytic anaerobic proto-eukaryotic cell. The protoeukaryote had DNA compartmentalized in a nucleus and some other primitive
compartmentalization while the prokaryote was more similar to an archaebacterium with
no nucleus. The symbiotic relationship that occurred led to the ability of the host to
efficiently perform oxidative phosphorylation with the aid of the archaebacterium, or
“protomitochondrion”. The protomitochondrion could, in turn, rely on the host cell for
11
transcription and protein synthesis, thus many of the mitochondrial DNA genes were
transferred to the nucleus. Though this process is known to have occurred over millions
of years, today’s eukaryotic cell contains mitochondria with only plasmid DNA encoding
very few genes. Mitochondria now rely on cellular machinery for transcription and
translation to make the majority of proteins needed for oxidative phosphorylation and
ATP synthesis.
As the mitochondrion itself evolved over millions of years, mitochondrial
research and our knowledge of this organelle as it contributes to metabolism has evolved
over the past 120 years (described by IE Scheffler in Mitochondria, 2nd edition). In 1890,
mitochondria were first glimpsed in cells of higher organisms by Altman, who referred to
them as “bioblasts” which were elemental living units that formed bacteria like colonies
in the host cytoplasm. Lehninger and Warburg in 1912 were the first to suggest that these
granular subcellular structures were associated with respiration. The discovery of the
tricarboxylic acid cycle and ATP in the 1930’s led to the common nickname
“powerhouse of the cell”. It was not until the 1960’s, however, that the five complexes
of the electron transport chain and the flow of electrons throughout the mitochondrion
were characterized. Since these seminal discoveries, mitochondrial research has leapt
forward significantly and it is now known that mitochondria play large roles in almost
every cellular function from mitosis, cell cycle, survival, death, and homeostatic
signaling.
Mitochondrial Structure
12
Mitochondria are complex, double membrane organelles made up of five major
constituents (reviewed by I Scheffler in Mitochondria, 2nd Edition): the outer membrane
(OM), inner membrane (IM), the inter-membrane space (IMS), cristae, and the matrix.
The OM contains docking sites for signaling proteins such as those involved in cell fate,
as well as major receptors for protein trafficking into the organelle. The IMS is the fluidfilled space between the OM and IM, while the IM contains all of the complexes of the
electron transport chain (ETC) involved in respiration and oxidative phosphorylation.
Complexes I, III, and IV of the ETC serve as proton pumps which fuel the flow of
electrons and continually pump protons from the matrix into the IMS, creating a charge
separation across the inner membrane known as the ΔΨmem (Nicholls 2000). Protons
flow back into the mitochondrial matrix to fuel Complex V, or ATP Synthase, which
phosphorylates ADP into ATP, the energy currency of the cell (Nicholls 2000). The inner
membrane is continuous with multiple invaginations that completely fill the space, or
matrix, of the organelle. These folds are referred to as cristae, and are able to change
shape as the mitochondrion adapts to cell signaling events.
Mitochondrial Morphology
Mitochondria are highly dynamic organelles that continuously change shape,
divide, and move throughout the cell in response to various stimuli (Okamoto 2005).
Mitochondrial remodeling is required for function, and processes such as fission, fusion,
condensation, and matrix volume changes are all examples of mitochondrial remodeling.
Fission is a division process that is required for dividing and differentiating cells, coping
with energy demands, and toxin response, while the purpose of mitochondrial fusion is to
13
maintain a tubular branched mitochondrial network within the cell, and optimal
mitochondrial motility and function (Bossy-Wetsel 2003). The processes of fusion and
fission are highly complex, act on the two mitochondrial membranes separately but
coordinately, and therefore must be maintained in a delicate balance (Scorrano 2005). For
example, loss of Drp1 or hFis1, both proteins required for mitochondrial fission, has been
shown to prevent fragmentation during apoptosis (Parone 2006), but under basal
conditions the loss of either protein results in large increases in fusion, loss of
mitochondrial membrane potential, decreases in ATP production, decreased calcium
buffering, and loss of mitochondrial DNA during mitochondrial replication (Parone
2008). Conversely, deletion or mutation of mitofusin proteins, which are required for
mitochondrial fusion and formation of mitochondrial networks, results in fragmentation
of mitochondria, premature release of cytochrome c, and increased apoptosis (Scorrano
2005).
In the 1960’s Charles Hackenbrock described two main configurations of
mitochondria and signaling that leads to adoption of these configurations. Under
conditions of cell stress, apoptosis, large changes in osmotic pressure, lacking nutrition or
loss of oxidizable substrates, mitochondrial condensation often occurs. This is a phase in
which the cristae are compressed against the outer membrane of the organelle, causing
large empty vacuoles to appear in the middle of the mitochondrion (Hackenbrock 1966).
The condensed phase of mitochondria indicates an unhealthy, non-respiring
mitochondrion, and apoptotic signaling (Rasola 2007). Alternately, the orthodox phase is
one in which the mitochondrial cristae fill up the entire matrix space, indicating increased
respiration, cellular survival signaling, and a healthy organelle (Rasola 2007). In the
14
orthodox phase, the cristae are tightly compacted and held together by the mitochondrial
protein OPA1, sequestering cytochrome c, inhibiting its release and subsequent apoptotic
signaling (Olichon 2002; Gottlieb 2003; Frezza 2006). In 2003 it was shown that ΔΨmem
is tightly coupled to the state of matrix morphology and mitochondrial function (Gottlieb).
Polarizing compounds, such as oxidizable substrates and complex activators, cause a
buildup of protons in the inner membrane space, and this leads to the orthodox
morphology (Gottlieb, 2003). Alternatively, depolarizing compounds such as the
mitochondrial complex I inhibitor rotenone, retain protons in the matrix and leads to the
condensed morphology (Gottlieb, 2003). Furthermore, an altered mitochondrial
membrane potential can cause decreases in ATP production because of the lack of
available protons in the IMS for ATP synthase utilization, as well as failure of protein
translocation, and interference with other mitochondrial functions. Changes in
mitochondrial morphology and membrane potential take place in an energy-dependent
and homeostatic fashion to ensure that the mitochondrion functions efficiently, and
contributes to overall cellular function.
Cellular Distribution of Mitochondria
In the past, mitochondria were thought of as static organelles that simply reside in
one location within a cell churning out energy, a notion that could not be farther from the
truth. Mitochondria in all cell types are known to change position and location in
response to the cellular environment (Chang 2006). Much work has been done recently
regarding mitochondrial trafficking and motility in neurons and yeast, which are highly
polarized cells and particularly sensitive to deficits in mitochondrial movement and
15
distribution (Frederick 2007). Many studies have shown that mitochondria have a specific
distribution in neurons, and in the earliest electron microscopy explorations of the
nervous system, synapses were often identified by the high concentration of mitochondria
(Palay 1956; Hollenbeck 2005). It has been shown by many researchers that
mitochondrial biogenesis occurs in the neuronal cell body, in close proximity to the
cellular transcriptional and translational machinery (Miller 2004; Chang 2006).
Subsequently, nascent mitochondria are transported to different areas of the cell (Miller
2004; Chang 2006). It is now known that mitochondria are able to move directionally
along axons in response to cellular signals to areas of high energy demand (i.e. synapses,
axonal growth cones, and Nodes of Ranvier) through the association with cellular motor
proteins, and are retained in their respective locations through interaction with
cytoskeletal components and scaffolding proteins (Chada 2003; Hollenbeck 2005; Chang
2006; Frederick 2007). Mitochondria with a high membrane potential produce more ATP
and are rapidly transported toward axonal and dendritic terminals where the energy
demand is high (Miller 2004). Suppression of this transport by disease processes or
mitochondrial toxins is known to abolish synaptic potentiation in slice preparations (Tong
2007). Mitochondria that are more depolarized, indicating aging or pharmacological
inhibition, are transported back to the cell body (Miller 2004). Furthermore, under
circumstances of decreased synaptic activity and decreased energy demand, such as long
term synaptic depression, mitochondria are retrograde transported back to the soma
(Chang 2006).
The specific cellular and mitochondrial signals that regulate mitochondrial
movement are still under investigation. It has, however, been reported that nerve growth
16
factor (NGF) signaling, which leads to activation of Akt, regulates motility and docking
of axonal mitochondria (Kimura 1994; Chada 2004). Specifically, NGF-coated beads
taken up into axons through the NGF receptor caused focal accumulation of mitochondria
in areas of NGF signaling (Chada 2004). Mitochondria which enter the proximity of NGF
beads in axons remain anchored there, whereas mitochondria entering proximity of
control beads pass through the area without stopping, effects that are dependent upon
intact cytoskeleton protein docks, such as presence of F-actin (Chada 2004). Furthermore,
disruption of the PI3K signaling pathway by LY204002 completely and specifically
eliminates NGF effects of mitochondrial behavior in neurons (Chada 2003). From these
data, there seems to be another effect of PI3K signaling, specifically Akt activity, on the
behavior and function of mitochondria.
Mitochondrial Protein Translocation
It is well-known that while mitochondria contain their own DNA (mtDNA) for
the synthesis of RNAs and other proteins needed for oxidative phosphorylation
(Anderson et. al. 1981; described by I. E. Scheffler in Mitochondria, 2nd edition), most of
the 1000 proteins found in the mitochondria are coded by the nucleus, synthesized on
ribosomes in the cytoplasm, and imported into the mitochondria through large complexes
in the mitochondrial membranes (Koehler 2004). This is a complicated event, because the
translocation complexes through which proteins are imported must translocate to the
mitochondria and assemble with the correct subunit composition and orientation before
protein translocation from the cytoplasm can begin. Multiple proteins have been
identified to play key roles in targeting and import of proteins to the mitochondrion.
Targeting of proteins to the mitochondria is accomplished by binding to cytoplasmic
17
molecular chaperones such as HSP90 or HSP70 or the specialized mitochondrial import
stimulating factor (MSF) (Takahashi 1998; Young 2003; Fan 2006). The major
mitochondrial complexes involved in protein import are the translocase of the outer
mitochondrial membrane (TOM) and the translocase of the inner mitochondrial
membrane (TIM) (Pfanner 1997; Neupert 1997). The TOM complex is made up of
multiple subunits, some of which extend into the cytoplasm and serve as specialized
receptors for recognition of incoming proteins (Pfanner 1997; Neupert 1997). The current
understanding of these receptors is that the TOM20 and TOM22 receptors preferentially
recognize proteins with a cleavable mitochondrial targeting presequence while the
TOM70 receptor recognizes and prevents aggregation of mature proteins with an internal
mitochondrial targeting sequence (Yamamoto 2009). Cytoplasmic chaperones bind the
preprotein and deliver it to the TOM receptors. Docking of HSP90 or HSP70 to these
receptors occurs by virtue of a tetratricopeptide repeat domain on the receptor similar to
the domains found on cytoplasmic co-chaperones of HSP90 and HSP70 (Young, 2003;
Yano 2004). After interacting with the precursor protein receptors on the TOM complex,
proteins are sent through TOM40, the import pore for the TOM complex (Pfanner 1997;
Becker 2005) (Figure 2). The other TOM subunits exist to position the protein for
interaction with the central pore, to modulate TOM complex assembly and dissociation,
and to prepare the protein for further mitochondrial trafficking (Dietmeier 1997). Once
the precursor protein has traversed the outer mitochondrial membrane, it must be further
sorted to its mitochondrial destination. In this step of the translocation process, proteins
can be inserted into the outer membrane or the inner membrane, or they can be further
sorted to the intermembrane space or matrix. Proteins that are to be inserted into the outer
18
membrane are chaperoned by small soluble proteins in the intermembrane space,
designated the small translocases of the inner mitochondrial membrane (Tiny TIMs) and
the outer-membrane bound sorting and assembly machinery (SAM) (Baker 2007).
Proteins that are to be inserted into the inner membrane or traverse the inner membrane
into the matrix are first handled by the translocases of the inner mitochondrial membrane
(TIM). The TIM complex is similar in structure and function to the TOM complex, but it
also sorts proteins by their final destination, whether it is the matrix or insertion into the
inner membrane (Bolender 2008).
Multiple reports have indicated that the cytoplasmic molecular chaperone HSP90
plays a substantial role in targeting preproteins to the Tom70 receptor on the OM (Young
2003; Fan 2006). Furthermore, Akt does not, to our knowledge, contain a cleavable
mitochondrial targeting presequence, making it an ideal candidate for interaction with the
TOM70 receptor. Considering that Akt is a known client protein of HSP90, it stands to
reason that HSP90 mediates mitochondrial accumulation of Akt by binding it in the
cytoplasm, targeting it to the TOM70 receptor and releasing it to the TOM complex for
further mitochondrial import and sorting (Figure 2).
Role of Mitochondria in Neurological Disorders
The role of mitochondria in neurodegenerative and neuropsychiatric disorders has
been gaining interest over the past several years. Mitochondrial malfunction components
have been suggested for a plethora of human diseases; hence the question becomes one of
cause and effect. In the case of a mutation in a specific mitochondrial gene or
mitochondrial protein, the mitochondrion is the origin of the disease. In the case of other
types of cellular problems, the mutation may be within the nuclear DNA, and the
19
mitochondrion becomes negatively affected in the process. Whether mitochondrial
function is the cause or an effect of a given disease state, it is well accepted this organelle
plays a role in the pathogenesis of many types of diseases, especially in the central
nervous system. Indeed, neurodegenerative disorders have been the focus of much
mitochondrial research. Alzheimer’s disease, Parkinson’s disease, Huntington’s disease,
multiple sclerosis, and supranuclear palsy are just a few of the progressive
neurodegenerative diseases that have been shown include some type of mitochondrial
dysfunction (Castellani 2002; Wang 2009; Vila 2008; Browne 2008; Heales 1999; Park
2001). The exact problem with the mitochondrion itself varies widely, as does the onset
of each disease. Mitochondrial dysfunction in neurodegenerative disorders can be due to
increased reactive oxygen species production and oxidative stress, loss of mitochondrial
membrane potential, mutations in mitochondrial enzymes, imbalance of the fission and
fusion processes, impairment in mitochondrial axonal trafficking to synapses, and
functional deficits in specific mitochondrial proteins (Heales 1999; Park 2001; Wang
2009;
Orr
2008;
Huttemann
2008).
Deficits
in
mitochondrial
function
in
neurodegenerative diseases are thought to partially underlie progressive loss in synaptic
plasticity and eventually lead to neuron loss and further multi-system malfunctions
(Mattson 2003).
To further support a role for mitochondria in neurological diseases, studies have
shown that ischemia that occurs during stroke causes up-regulation of reactive oxygen
species production. This leads to oxidative stress, upregulation of apoptotic signaling, and
lasting mitochondrial damage in the form of respiratory dysfunction (Niizuma 2009;
Anderson 1999). Ischemia lasting more than a few minutes often leads to permanent
20
tissue damage; mitochondrial damage and mitochondrial-mediated apoptotic signaling
are thought to underlie the massive cell death that occurs after cerebral ischemia (Sims
2001; Niizuma 2009).
Another area of mitochondrial research as it relates to overall brain function has
been centered on mood disorders. This is still a rather new endeavor, but it has been
suggested that canonical mitochondrial diseases also show a psychiatric component (Shao
2008). Furthermore, it has been shown that mice containing neuron-specific mutations or
deletions in the mitochondrial DNA proofreading enzyme polymerase-γ have an
accumulation of mitochondrial DNA (mtDNA) mutations and a mood disorder phenotype
similar to bipolar disorder that can be successfully treated with lithium (Kasahara 2006).
Also, mtDNA mutations have been shown to exist in patients with bipolar disorder, major
depressive disorder, and schizophrenia (Rollins 2009). In bipolar disorder research, it has
been suggested that mitochondrial calcium buffering is adversely affected due to
increased basal intracellular calcium levels in patients with bipolar disorder, leading to
impaired neurotransmitter release and loss of synaptic signaling (Quiroz 2008). In
schizophrenia, electron microscopy studies of post-mortem patient brain have suggested
that there are fewer mitochondria concentrated at synaptic terminals compared to control
patients (Kung 1999; Shao 2008).
Considering the available data, it appears that mitochondrial function plays an
important role not only in homeostatic stabilization of neuronal signaling, but may also
underlie many neurodegenerative and psychiatric disorders. Since Akt and GSK3β
deregulation have been shown to contribute to pathogenesis of mood disorders, and the
insulin signaling pathway has become a potential therapeutic target (Jope 2002;
21
Prickaerts 2006; Lovestone 2007), it stands to reason that the mitochondrial pool of Akt
may play some role in metabolic deficits that occur in stroke, mood disorders, and
neurodegeneration.
Post-Mortem Tissue: Direct Investigation of the Human Brain
The study of human disease, both neurological and non-neurological, was
hampered for many years by the belief that irreversible degradation of the body’s tissues
occurs before the tissue can be properly examined in a scientific setting. However, with
increasing knowledge of brain circuitry and advances in technology and methodology for
scientific exploration, studies of the postmortem human brain have become an essential
element to understanding the neurobiology of neurodegenerative and psychiatric
disorders.
Post-Mortem Tissue Viability
For decades it has been accepted that the brain is irreparably damaged following a
5-10 minute interruption of blood and oxygen supply (Verwer 2002b). Moments after
death, the brain is placed in a condition of asphyxia, tissues are deprived of oxygen, and
mitochondrial ATP generation ceases. Among the early consequences of decreased ATP
production is the destruction of cellular membranes, release of (normally sequestered)
degrading enzymes, and cellular digestion (Reviewed in “Chemical and Ultrastructural
Aspects of Decomposition”, by Gill-King, H. in “Forensic Taphonomy: The postmortem
fate of human remains” Edited by William D. Haglund and Marcella H. Sorg, 1997, CRC
Press LLC, Boca Raton FL).
22
Post-mortem tissue has been used to study a wide variety of human diseases in
many tissues of the human body. For example, mitochondrial pathology has been
examined in the striatum of post-mortem control and schizophrenic brain, which
indicated altered brain anatomy and decreased mitochondrial numbers in synapses of
caudate and putamen (Lewis 2002; Kung 1999). However, most studies of postmortem
human brain that have been conducted have been strictly observational.
In order for tissue from postmortem human brains to be a potential source of
information about human disease, postmortem bodies should be handled in a consistent
manner. For example, when the body is transferred to a 2oC refrigerator within one hour
of death, brain integrity and cell morphology remains remarkably intact up to 18 hours
post-mortem (Chandana 2009). Of course, this preferred situation for body storage is not
always the case. Cause of death, premortem medical interventions such as surgeries or
medications, illegal drug use, season and time of death, time interval to body discovery,
storage temperature of postmortem tissue, and total postmortem interval (PMI) are all
factors which must be considered (Lewis 2002). When comparing postmortem tissue
from two groups, the design of the postmortem study requires specialized attention to
characterization of the tissue and the potential confounding factors stated above (Lewis
2002).
Recent findings have solidified the idea that postmortem human brains can be a
potential source of data despite complications with tissue procurement and storage. It was
shown in 2002 that postmortem human motor cortex tissue from adult neurological
patients and controls within 8 hours after death could be sliced and organotypically
cultured for up to 78 days (Verwer 2002 a, b). Furthermore, cells within the cultured
23
brain slices could be experimentally manipulated to induce gene expression as well as
alter activity of certain enzymes (Verwer 2002 a, b). Together, the available data suggest
that there is a vast amount yet to be learned about basal brain function and deficits that
occur during neurological disease from studies of postmortem human brain tissue.
The Study of Mitochondrial Function in Human Brain
Postmortem human brains, when available, are potentially an abundant source of
mitochondrial material. Previous dogma, however, held that functional assessments could
not be carried out in mitochondria from post-mortem tissue. This is because of the
common belief there is a rapid deterioration of the tissue following death and all
metabolic activity is irreversibly arrested.
To understand the role of mitochondria in the human central nervous system in
the past, investigators have utilized frozen postmortem brain tissues to analyze
mitochondrial enzymatic activities (Devi 2008), mitochondrial protein levels (Park 2001),
and mitochondrial DNA (Alam 1997; Vila 2008). However, some mitochondrial
functional indices provide insight into brain mitochondrial activity and cannot be
conducted in frozen tissue samples. These include measurements of overall indicators of
mitochondrial activity such as ΔΨmem, ATP production, calcium buffering capacity, and
respiration, which together give an overall assessment of mitochondrial health and
activity. However, these analyses can only be conducted in structurally intact and
functional mitochondria. Furthermore, if these measurements are undertaken in isolated
mitochondria, large quantities of relatively pure mitochondrial preparations are required.
These key features; mitochondrial purity, structural integrity, and abundance, are
24
common barriers which can encumber brain mitochondrial research. An additional
hindrance for research on human brain mitochondria is the extremely limited source of
human brain tissues in general.
Following studies designed to examine the possibility the functional mitochondria
can be isolated from post-mortem human brain, the next logical step would be to extend
this project to examine specific proteins within the brain, namely Akt. From a
researcher’s standpoint, data from postmortem human studies can provide a novel and
direct approach for the study of Akt function in the human brain under basal and diseased
conditions.
Based on this background information, the overall purpose of this project is to
provide insight into the mechanism of mitochondrial accumulation of Akt, as well as the
role of mitochondrial Akt in the overall function of the organelle. To achieve these goals,
the translocation and function of Akt in isolated mitochondria is explored, as is the
hypothesis that functional mitochondria can be isolated from post-mortem brain for
further testing of mitochondria in health and disease.
25
ATP
Synthase β
PO4
ATP
Akt
GSK3β
PDH
PO4
Fig 1. Hypothetical model of Akt activity within mitochondria. Akt is known to
phosphorylate the β-subunit of ATP Synthase, or Complex V, of the mitochondrial
electron transport chain. It is hypothesized that this action, combined with an intact
ΔΨmem, causes enhanced rates of ATP production. Mitochondrial Akt also
phosphorylates and inhibits mitochondrial GSK3β. This releases GSK3β inhibition on the
E1α subunit of pyruvate dehydrogenase, theoretically increasing rates of energy
production at the level of the tricarboxylic acid (Kreb’s) cycle.
26
HSP90
Akt
HSP90
HSP90
20
70
TOM40
Akt
TIM Complex
Akt
Akt
mitochondrion
cytoplasm
Fig 2. Proposed mechanism of Akt translocation into mitochondria. Cytoplasmic HSP90
binds Akt and provides stability. At a basal rate, HSP90 delivers Akt to the TOM70
receptor on the mitochondrial OM. As HSP90 hydrolyzes ATP, Akt is released from
HSP90 and guided through the TOM40 pore and further sorted into the IM, OM, and
matrix by the TIM complex.
27
THE BASAL FLUX OF AKT IN THE MITOCHONDRIA IS MEDIATED
BY HEAT SHOCK PROTEIN-90
by
Keri A. Barksdale and Gautam N. Bijur
Journal of Neurochemistry 108(5), 1289-99
Copyright 2009
by
Elservier B. V.
Used with permission
Format adapted for dissertation
28
Abstract
Akt is a known client protein of heat shock protein-90 (HSP90). We have found
that HSP90 is responsible for Akt accumulation in the mitochondria in unstimulated cells.
Treatment of SH-SY5Y neuroblastoma cells and HEK293 cells with the HSP90
inhibitors novobiocin and geldanamycin caused substantial decreases in the level of Akt
in the mitochondria without affecting the level of Akt in the cytosol. Moreover, intracerebroventricular injection of novobiocin into mice brains decreased Akt levels in
cortical mitochondria. Knockdown of HSP90 expression with siRNA also caused a
significant decrease in Akt levels in the mitochondria without affecting total Akt levels.
Using a mitochondrial import assay it was found that Akt is transported into the
mitochondria. Furthermore, it was found that the mitochondrial import of Akt was
independent of Akt activation as both an unmodified Akt and a constitutively active
mutant Akt both readily accumulated in the mitochondria in an HSP90-dependent manner.
Interestingly, incubation of isolated mitochondria with constitutively active Akt caused
visible alterations in mitochondrial morphology, including pronounced remodeling of the
mitochondrial matrix. This effect was blocked when Akt was mostly excluded from the
mitochondria with novobiocin treatment. These results indicate that the level of Akt in
the mitochondria is dependent on HSP90 chaperoning activity and that Akt import can
cause dynamic changes in mitochondrial configuration.
29
Introduction
The serine/threonine kinase Akt, also referred to as protein kinase B, is closely
associated with growth factor signaling. Stimulation of cell surface receptors with insulin
or growth factors, such as insulin-like growth factor-1 (IGF-1), results in the activation of
phosphatidylinositol 3-kinase and phosphatidylinositol dependent kinases, the latter
which directly phosphorylates Akt on its Thr308 site, and possibly on its Ser473 site
(Alessi, 1996; Alessi, 1997; Fayard, 2005). Phosphorylation at these sites results in the
full activation of Akt. Akt activation occurs at the cytoplasmic face of the plasma
membrane, however subsequent to its activation Akt moves from the plasma membrane
and translocates into different subcellular compartments including the cytosol, the
nucleus (Andjelkovic, 1997), and mitochondria (Bijur, 2003), where it can phosphorylate
compartment-specific substrates. The translocation of Akt into the mitochondria is rapid,
occurring in quick succession to its activation. Although mitochondrial import of Akt is
blocked by a dissipation of the mitochondrial transmembrane potential (Bijur, 2003), the
molecular mechanism underlying Akt mitochondrial transport is unknown.
Akt activity is also affected by its binding to heat shock protein-90 (HSP90) (Sato,
2000). Akt is a well-known client protein of HSP90. Pharmacological inhibitors of
HSP90 have provided much insight into how HSP90 functions. For example, the
coumarin based compound novobiocin (NB) inhibits HSP90 by targeting the C-terminal
region of HSP90 adjacent to its homodimerization domain (Marcu, 2000a). NB treatment
has been associated with destabilization and proteolysis of HSP90 client proteins (Marcu
2000b; Yun, 2004). In addition, geldanamycin (GA) is a benzoquinoid ansamycin
antibiotic which inhibits HSP90 by binding to its N-terminal ATPase domain (Whitesell,
30
1994; Panaretou, 1998). Treatment of cells with geldanamycin affects Akt by inducing its
dephosphorylation (Fujita 2002; Xu, 2003), and also induces the degradation of Akt (Kim,
2003), indicating that HSP90 is essential for stability and function of Akt. Thus, in
addition to phosphorylation, Akt activity is dependent on its interaction with HSP90.
Recently it was reported that inhibition of HSP90 with NB and GA can block the
import of proteins into mitochondria (Fan, 2006), indicating that HSP90 plays a major
role in the translocation of proteins into the mitochondrion. The import of proteins into
the mitochondria is a dynamic energy-dependent process relying on the concerted efforts
of several translocase proteins resident within the mitochondria, a pore complex through
which the proteins are inserted, and chaperone proteins which facilitate the import
process (Pfanner, 1990; Komiya, 1997). Although many mechanistic aspects of
mitochondrial import of proteins are unknown, some features have been elucidated. Two
key chaperone proteins responsible for protein import into the mitochondria are heat
shock proteins-90 and -70 (HSP90 and HSP70, respectively) (Young, 2003; Young,
2004). Before proteins are imported into mitochondria (preproteins) they are bound to
chaperones in the cytoplasm, and notably, HSP90 was recently reported to play a major
role in the mitochondrial protein import process (Fan, 2006). In the initial process of
mitochondrial import, HSP90 (with its preprotein cargo) binds to one of the
mitochondrial surface receptors on the translocase of the outer mitochondrial membrane
(TOM) complex, the TOM70 receptor (Yano, 2004; Fan, 2006). Docking of the HSP90
to TOM70 occurs by virtue of a tetratricopeptide repeat domain on the receptor similar to
the domains found on the co-chaperones of HSP90 (Young, 2003; Yano 2004). Due to
the close juxtaposition of HSP90 to the TOM complex, the preprotein can make direct
31
contact with the TOM70 receptor. Subsequently, the preprotein is imported through the
outer mitochondrial membrane via the TOM40 import pore (Gabriel, 2003; Chacinska,
2004). Once a preprotein has entered the mitochondria it is further assimilated into
specific mitochondrial subcompartments by numerous transport proteins, most notably
the translocases of the inner membrane, commonly known as TIMs (Kerscher, 1997;
Kutik, 2007). In addition, transport of proteins into the deeper compartments of the
mitochondria, the inner membrane and the matrix, requires the electrochemical gradient
across the inner membrane (Martin, 1991; Geissler, 2000). Thus, HSP90, in conjunction
with the TOM and TIM proteins, function in the efficient translocation of proteins into
the mitochondria and its subcompartments.
Given that Akt can translocate into the mitochondria, and that Akt is a well
known client protein of HSP90, it was hypothesized that Akt levels in the mitochondria
might be affected by HSP90 inhibitors. The present report describes how inhibition of
HSP90 activity or decreased HSP90 expression causes marked decreases in the levels of
Akt in the mitochondria and inhibits the translocation of Akt into the mitochondria. In
addition, increased Akt signaling in the mitochondria is shown to cause remodeling of the
mitochondrial cristae, indicating that Akt activity has profound affects on mitochondrial
morphology.
Materials and methods
Cell Culture and Treatments
SH-SY5Y human neuroblastoma cells were grown in continuous culture RPMI
media containing 10% horse serum, 5% Fetal Clone II (Hyclone, Logan, UT), 2 mM L-
32
glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin. HEK293 human
embryonic kidney cells were grown in F12/DMEM containing 10% fetal bovine serum
(Invitrogen), 2 mM L-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin.
For serum withdrawal, adherent cells were rinsed twice with serum-free media
supplemented with 2 mM L-glutamine, 100 units/ml penicillin, and 100 μg/ml
streptomycin. Prior to all treatments cultured cells were washed once with serum-free
RPMI media (for SH-SY5Y cells) or F12/DMEM (for HEK293 cells) containing 2 mM
L-glutamine, 100 units/mL penicillin, and 100 ug/mL streptomycin. Cells were
maintained in serum-free media overnight prior to treatments with 625 μM novobiocin
(NB) (Alexis Biochemicals), and 5 μM geldanamycin (GD) (AG Scientific). Cells were
either subjected to a 60-minute time course with NB, or a 90 minute GA time course.
Where indicated, cells were pretreated with NB or GA for 45 minutes prior to a 15
minute treatment of 50 ng/mL insulin-like growth factor-1 (IGF-1) (US Biological).
siRNA HSP90 Knockdown
Adherent HEK293 cells were transfected with 2 µg of plasmid from the pGBHSP90 siRNA vector mix obtained from Biovision Incorporated (Mountainview, CA).
Since this plasmid mix confers G418 resistance, stably transfected cells were grown in
continuous culture media containing 225 µg/ml G418. Control cell lines were stably
transfected with an empty pcDNA vector conferring G418 resistance. The cells were
continuously cultured in serum-containing F12/DMEM media, described above.
33
Mitochondrial Purification
Mitochondria were isolated essentially as described in Bijur and Jope (2003) with
modifications as follows. After the cells were harvested and plasma membranes disrupted
by nitrogen cavitation, the cell homogenate was centrifuged once at 700 x g for 5 min to
remove unbroken cells and nuclei. The crude mitochondria-containing supernatant was
transferred to a new tube and centrifuged for 30 min at 17,500 x g to collect a
mitochondrial pellet. The supernatant was saved as the cytosolic sample, and the pellet
was washed twice with cavitation buffer followed by a 15 min centrifugation at 17,500 x
g. Mitochondrial extracts were obtained by incubating mitochondria for 30 min at 4oC in
lysis buffer (20 mM Tris, pH 7.5, 0.2% Nonidet P-40, 150 mM NaCl, 2 mM EDTA, 2
mM EGTA, and protease and phosphatase inhibitors). The mitochondrial extract was
clarified by centrifugation at 20,800 x g for 10 min. Protein concentrations were
determined by the bicinchoninic acid method (Pierce). To verify complete separation of
the mitochondrial and cytosolic fraction, the two fractions were immunoblotted with βtubulin, an abundant cytosolic protein, and pyruvate dehydrogenase, a mitochondrial
protein.
Immunoblotting
Immunoblotting was performed as described previously (Bijur, 2003). Proteins
separated by SDS-PAGE were transferred onto nitrocellulose membranes. The
membranes were blotted with antibodies to Akt (Sigma, Cell Signaling), HSP90 (BD
Transduction), HSP70, HSC70, and HSP40 (BDTransduction), phospho-Ser21/9-
34
GSK3α/β, phospho-Ser473-Akt, phospho-Akt-substrates (Cell Signaling), pyruvate
dehydrogenase , or β-tubulin (Sigma).
Mitochondrial Import Assay
Methods for the mitochondrial import assay were adapted from Fan, et al (2006)
with minor modifications. Cell-free translation of wtAkt and DDAkt proteins were
performed using the TNT-coupled transcription/translation rabbit reticulocyte lysate
system (Promega), according to manufacturer’s directions. Briefly, T7 Polymerase
(Promega) and 35S-Methionine (Amersham) were mixed with reticulocyte lysate, transfer
RNAs, reaction buffer, RNAsin (Promega), and 1 µg of wtAkt or DDAkt plasmid
containing the T7 promoter. The total reaction volume was adjusted to 50 µl with
nuclease free water. Reactions were incubated for ninety minutes at 30oC then adjusted to
250 mM sucrose.
Mitochondria from HEK293 cells were isolated as described and aliquoted
equally in import buffer (20 mM HEPES-KOH pH 7.5, 5 mM MgOAc2, 80 mM KOAc,
and 250 mM Sucrose) containing 2 mM ATP, 0.4 mM ADP, 1 mM dithiothreitol, and 10
mM succinate. Mitochondria were then added to the translated proteins such that the
volume of mitochondria made up 40% of the reaction volume and the exogenous protein
constituted 35% of the total reaction volume. Novobiocin was added to the translation
product for a final concentration of 625 μM before the mitochondria were added. The
translocation reaction was then allowed to incubate for 45 min at 30oC with intermittent
gentle mixing. Mitochondria were collected and half of each sample was digested for 10
minutes on ice with 250 ug/mL proteinase K to remove proteins on the outside of the
mitochondria. The digested samples were then treated with 2 mM phenylmethylsulfonyl
35
fluorine (PMSF) for 10 minutes on ice to terminate digestion. Half of each sample was
left untreated. Mitochondrial lysates were obtained and mitochondrial proteins were
separated by SDS-PAGE. The gels were dried and the radioactive Akt bands were
visualized with a phosphorimager.
Intra-cerebroventricular (ICV) Injections
All procedures were in accordance with the Institutional Animal Care and Use
Committee (IACUC) guidelines. 6-8 week old male C57BL6 mice were anesthetized
using 100mg/kg ketamine and 10 mg/kg xylazine. Novobiocin (312.5 mM in 1 μL total
volume) was bilaterally injected into ventricles, to achieve a final concentration of 625
μM assuming a 1mL cerebrospinal fluid volume, using a 10 μL Hamilton syringe (0.8
mm posterior and 1.6 mm left and right of Bregma). Control mice were injected with an
equivalent volume of sterile 0.9 % saline solution. Following a 2 hour recovery, mice
were rapidly decapitated and cortical mitochondria and cytosolic fractions were isolated
and immunoblotted for Akt.
Akt Activity Assay
All reagents for the Akt activity assay, except translocation components, were
obtained from Cell Signaling. Cell-free translations of wtAkt and DDAkt proteins were
performed, as described above, and each reaction was adjusted to 200ul total volume with
1X Cell Lysis Buffer. 20ul immobilized Akt antibody slurry was added, and the mixture
was incubated on an orbital rocker overnight at 4oC. The immobilized antibody pellets
were washed two times with 1X cell lysis buffer and two times with 1X kinase buffer.
Antibody pellets were incubated with 1X kinase buffer supplemented with 1ug GSK3
fusion protein and 0.2 mM ATP for 30 minutes at 30oC. Reactions were terminated with
36
SDS sample buffer and boiled 5 minutes. Membranes were immunoblotted for phosphoGSK3α/β reactivity.
Transmission Electron Microscopy
Adherent HEK293 cells were cultured as described and either serum-starved
overnight or maintained in serum media, where indicated. Functional mitochondria were
isolated using nitrogen cavitation and incubated with DD-Akt protein in the presence or
absence of novobiocin for 45 minutes at 30oC. Control reactions were incubated with
reticulocyte lysates containing no translated protein. Following incubation, mitochondria
were pelleted and fixed overnight at room temperature in 2% paraformaldehyde/2.5%
glutaraldehyde in 0.1M sodium cacodylate buffer. The pellet was washed extensively
with sodium cacodylate buffer, and post-fixed with 1% osmium tetroxide in sodium
cacodylate buffer. Pellets were then washed for one hour with phosphate-buffered saline
(PBS). The cell pellets were dehydrated with 10 minute incubations of 50% EtOH, 70%
EtOH, and 95% EtOH, followed by four 15-minute incubations in 100% EtOH. Pellets
were washed twice for 10 minutes with 100% propylene oxide, and placed into a 50:50
mixture of propylene oxide and embedding resin (Embed 812, Electron Microscopy
Sciences, Fort Washington, PA) for 12-18 hours. The tissue was transferred to 100%
embedding media for one hour, changed to fresh embedding media for an additional hour,
and incubated in an oven at 60oC overnight to polymerize. Following polymerization, the
resin blocks were sectioned 1-2 μm thick with a diamond knife using an ultramicrotome.
The sections were stained with toluidine blue to use as a reference for thin sectioning.
Thin sections were made using a diamond knife (Diatome, Electron Microscopy Sciences,
Fort Washington, PA) at 70-100 nm, and the sections were placed on copper mesh grids.
37
After drying, sections were stained with heavy metals uranyl acetate and lead citrate for
contrast. Grids were allowed to dry, and then viewed on a FEI Tecnai Twin 120kv TEM
(FEI, Hillsboro, OR). Digital images were taken with an AMT CCD (Advanced
Microscopy Techniques) camera and the images were saved on a computer memory
device. Two blinded observers who were unaware of the treatments were given several
representative images of mitochondria with varying configurations as the criterion for
quantitation to prevent the inclusion of fragmented mitochondria or debris in the
quantitation. The grayscale settings were kept the same for all the images. The observers
were told to manually and accurately circumscribe 30 or more objects on the computer
screen using the ArcSoft photoimaging software, only those objects in the experiments
that fit the criteria. The optical density of each object was then automatically tabulated
and entered into a statistical program to determine means, standard errors, and statistical
significance.
Results
Inhibition of HSP90 decreases mitochondrial Akt levels.
Previously it was reported that a pool of Akt exists in the mitochondria, the level
of which can flux greatly (Bijur, 2003). Two well characterized HSP90 inhibitors, NB
and GA, have previously been shown to block the import of HSP90 cargo proteins into
the mitochondria (Fan 2006), indicating that HSP90 plays a key role in the process of
mitochondrial protein import. The goal here was to test if HSP90 inhibitors would affect
the levels of Akt in the mitochondria under normal growth conditions. Akt exists mostly
in the cytoplasm (Andjelkovic, 1997; Borgatti, 2000), thus it was necessary to confirm
38
that there was no cytosolic contamination in the mitochondrial fractions. The purity of the
mitochondrial fraction used to measure Akt levels was assessed by blotting for the
abundant cytosolic protein β-tubulin, which could not be detected in the mitochondrial
fraction (Fig 1A). The mitochondrial protein pyruvate dehydrogenase was clearly evident
in the mitochondrial fraction and not in the cytosolic fraction, indicating complete
separation of these two fractions. The existence of HSP90 on the mitochondria was also
confirmed. Isolated mitochondria were separated into two separate aliquots and one
aliquot was digested with proteinase K to strip proteins on the mitochondrial surface.
Internal mitochondrial proteins are inaccessible to proteinase K activity and therefore
remain intact. Figure 1B shows an abundant amount of HSP90 on the mitochondrial
surface, and digestion with proteinase K removed most of this HSP90. A small amount of
HSP90 was found to exist within the mitochondria. The level of ATP synthase-β which
exists on the inner mitochondrial membrane was unaffected by proteinase K digestion.
Thus, HSP90 resides on the mitochondrial surface.
The effect of NB and GA treatments on Akt mitochondrial levels was tested in
unstimulated cells. NB is a coumarin-type compound that affects the C-terminal region of
HSP90 (Marcu, 2000a). NB treatment of SH-SY5Y neuroblastoma cells caused a timedependent reduction in the level of Akt in the mitochondria (Fig 1C). After 15 min of
treatment with 0.625 mM NB there was a 42% reduction of Akt in the mitochondria
compared to control untreated cells. After 45 min of NB treatment, the level of Akt in the
mitochondria was reduced by approximately 60% of control, and further reductions in
Akt levels in the mitochondria were not seen. HSP90 inhibition results in the degradation
of its client proteins including Akt, however at the concentration and the time points of
39
NB used for this study the total Akt levels in the cytosol were not reduced (Fig 1C),
suggesting that Akt degradation may not be occurring at these times and at this dose of
NB treatments.
GA is a benzoquinoid ansamycin antibiotic that affects the N-terminal ATPbinding region of HSP90 (Whitesell, 1994; Panaretou 1998). Interestingly, GA does not
enter the mitochondrion and accumulate in this organelle (Kang, 2007). GA (5 μM)
treatment also resulted in marked reductions in Akt levels in the mitochondria (Fig 1D).
After 15 min of GA treatment there was a 40% decrease in Akt levels in the mitochondria
and the levels of Akt in the mitochondria were reduced further after 90 min of GA
treatment, resulting in a 94% reduction of Akt in the mitochondria. The level of Akt in
the cytosol was unaffected by GA treatments. Treatment of cells with higher
concentrations and longer times of NB and GA treatments did result in degradation of
Akt (data not shown). The effect of NB and GA treatments on Akt levels in the
mitochondria was also confirmed in another cell line. Human HEK293 cells were treated
for 45 min with 625 μM NB or 5 μM GA, and Akt levels in the mitochondria were
assessed. There were significant 50% and 80% reductions in Akt levels in the
mitochondria following NB and GA treatments, respectively (Fig 1E). Thus, two
chemically distinct inhibitors of HSP90 used under these specific conditions were found
to markedly reduce the level of Akt in the mitochondria but not the cytosol, which was
confirmed in two different cell lines.
To test if Hsp90 regulates Akt levels in brain mitochondria, 625 μM NB was
stereotaxically injected into both lateral ventricles of adult C57/Bl6 mice. There was a
significant 40% decrease in mitochondrial Akt levels in NB injected mice compared to
40
saline controls (Figure 1E). Levels of Akt in the cytosol were unchanged by this
treatment (1E). These data indicate that, in in vivo as well as in cell culture models,
inhibition of Hsp90 leads to a decrease in the level of Akt in the mitochondria.
The concern with using inhibitors of HSP90 such as NB and GA is that these
agents are known to elicit the proteolysis of the HSP90 client protein such as Akt. To
allay this concern, HEK293 stable cell lines were created in which HSP90 protein
expression was reduced using HSP90 siRNA. HSP90 expression in the HSP90 siRNA
cells was reduced by 70% of control (Fig 2A), but the expression of HSP70 and heat
shock cognate-70 (HSC70) was unaffected. Interestingly, the level of HSP40 was
noticeably increased. The total level of Akt was unaffected (Fig 2A), but in unstimulated
HSP90 siRNA cells the Akt level in the mitochondria was significantly reduced by 50%
compared to control cells (Fig 2B). Taken together, these results show that HSP90
activity markedly influences the amount of Akt accumulation in the mitochondria.
However, a total blockade of Akt mitochondrial accumulation could never be achieved
using any of these methods of HSP90 inhibition, suggesting that there may also be an
HSP90-independent mode by which Akt accumulates in the mitochondria.
HSP90 mediates import of Akt into isolated mitochondria
To directly assess Akt mitochondrial import by HSP90, an in-vitro mitochondrial
import assay was employed. We also examined if the activation state of Akt affects its
mitochondrial import. Two forms of Akt were tested; an unmodified wtAkt, and a
constitutively-active mutant Akt in which the activity-associated phosphorylation sites at
Thr308 and Ser473 sites were mutated to aspartic acid (DDAkt). These two Akt proteins
were in-vitro translated using rabbit reticulocyte lysates. Initially, the relative activities of
41
the wtAkt and DDAkt within the mitochondria were determined. Isolated intact
mitochondria were incubated with the translated Akt proteins. The external mitochondrial
proteins were digested with proteinase K and the mitochondria were lysed. The
mitochondrial lysates were immunoblotted with a phospho-Akt substrate (PAS) antibody
which detects phosphorylated substrates of Akt. Incubation of intact mitochondria with
wtAkt resulted in increased PAS antibody immunoreactivity against internal
mitochondrial proteins compared to the control (Fig 3A) indicating that the translated
wtAkt can enter the mitochondria and has some basal kinase activity. PAS
immunoreactivity of internal mitochondrial proteins was increased further with the
incubation of mitochondria with DDAkt, indicating that DDAkt has greater activity than
the wtAkt, and also can enter the mitochondria. The activities of the translated proteins
was also measured by immunoprecipitating wtAkt and DDAkt and reacting the
immobilized proteins with an Akt substrate (GSK3α/β crosstide). The Akt activity
mirrored the PAS immunoreactivity (Fig 3A); the wtAkt had some basal kinase activity,
while the DDAkt had increased kinase activity.
For the import assay, wtAkt and the constitutively active DDAkt were in-vitro
translated in the presence of
35
S-methionine which is incorporated into the synthesized
proteins. The protein products were incubated with intact mitochondria for the import
assay as described in Materials and Methods. The mitochondrial import of the
35
S-
methionine-labeled Akt proteins was examined in mitochondria isolated from the control
pcDNA cells and the siRNA HSP90 knockdown cells. The level of HSP90 bound to the
mitochondria in the HSP90 siRNA cells was first determined. Figure 3B shows the
reduced level of HSP90 bound to the mitochondria in the HSP90 siRNA cells compared
42
to the pcDNA control cells. Figure 3C shows that both the wtAkt and DDAkt were found
to be attached to the surface of the mitochondria at about equivalent levels in the
undigested mitochondrial samples from the pcDNA cells and the HSP90 siRNA cells. In
the digested mitochondria, however, substantially less wtAkt and DDAkt were detected
within the mitochondria from the HSP90 siRNA cells compared to mitochondria from the
control pcDNA cells, indicating that HSP90 is mostly required for the mitochondrial
import of both forms of Akt. To further test if HSP90 inhibition would block
mitochondrial import of wtAkt and DDAkt, isolated mitochondria from wild-type
HEK293 cells were incubated with the in-vitro translated proteins in the presence of NB.
NB did not affect Akt binding to the mitochondria surface (Fig 3D), but in agreement
with the previous result, NB-mediated HSP90 inhibition mostly blocked the
mitochondrial import of both forms of Akt. However, in both of these cases a total
blockade of Akt import into the mitochondria was not seen, further suggesting that there
exists a second mechanism by which Akt is imported into the mitochondria. These results
further verify that HSP90 partially facilitates Akt transport into the mitochondria.
Growth factor-stimulated mitochondrial Akt accumulation is not hampered by HSP90
inhibition.
In a previous study it was shown that stimulation of HEK293 cells and SH-SY5Y
cells with IGF-1, a potent activator of the PI3K signaling pathway, induced the
mitochondrial translocation of Akt (Bijur, 2003). Because the mitochondrial import of
constitutively active DDAkt was blocked by HSP90 inhibition, it stood to reason that
mitochondrial accumulation of Akt following IGF-1 stimulation could also be blocked by
HSP90 inhibition. HEK293 cells were treated with IGF-1 alone or with NB and GA prior
43
to IGF-1 treatment. Mitochondrial lysates were immunoblotted for Akt levels. Neither
NB nor GA treatments caused any diminution of Akt import into the mitochondria (Fig
4A and B). Similarly, in SH-SY5Y cells the HSP90 inhibitors had no effect on IGF-1mediated Akt mitochondrial import (data not shown). Furthermore, in the siRNA HSP90
knockdown cells IGF-1-mediated Akt mitochondrial translocation was also not
diminished compared to control cells (Fig 4C). Taken together, our data suggests that the
normal flux of Akt into the mitochondria is HSP90-dependent, and appears to be distinct
from growth factor-stimulated Akt mitochondrial transport, which does not appear to be
dependent upon HSP90 activity.
Akt alters mitochondrial matrix configuration.
The next goal was to examine how the accumulation of Akt within the
mitochondria would affect mitochondrial function. Mitochondria are very dynamic and
pliant organelles whose shapes change with their function; therefore, a visible feature of
the mitochondrion is its morphology. For example, changes in metabolic activity,
respiration state, and apoptotic signaling can markedly affect mitochondrial shape
(Hackenbrock, 1966; Bossy-Wetzel, 2003; Gottlieb, 2003). In this experiment, HEK293
cells were grown in serum-free media for 24 h. Mitochondria were isolated and then
incubated with the constitutively active DDAkt in the presence or absence of 625 μM
novobiocin. Control reactions were incubated with rabbit reticulocyte lysates. The
isolated mitochondria were viewed by transmission electron microscopy. Mitochondria
from control cells in serum-free media have highly condensed cristae (indicated by an
arrow), and decreased matrix density. This configuration was likely due to serum
withdrawal as mitochondria from cells grown in serum typically had highly
44
compartmentalized cristae (Fig 5A), although the size of the mitochondria was much
smaller compared to mitochondria from serum-free cultured cells. Mitochondria
incubated with the DDAkt had highly compartmentalized cristae and appeared to have
more of an orthodox configuration. The cristae were also well defined (Fig 5B, indicated
by an arrow in the middle right panel) within the mitochondria. Densitometric
quantitation in a blind study revealed that DDAkt treatment caused a significant increase
in the matrix density of about 115% (Fig 5C) compared to the control mitochondria.
Cotreatment of the mitochondria with novobiocin and DDAkt significantly blocked the
compartmentalization of the cristae (Fig 5B and C), which appears as decreased matrix
density. However, some closed cristal regions could still be seen, probably indicating that
some DDAkt was still entering the mitochondria and causing cristal remodeling.
Discussion
Previously it was reported that Akt exists in the mitochondria (Bijur 2003). The
import of some proteins into the mitochondria has been shown to be dependent on HSP90
(Young, 2003; Young, 2004; Fan, 2006). Given that Akt is a known client protein of
HSP90 we theorized that HSP90 activity may influence Akt levels in the mitochondria.
This turned out to be the case as the amount of Akt in the mitochondria was significantly
reduced by HSP90 inhibition.
NB and GA are two well known HSP90 inhibitors which have been shown to
block the import of proteins into the mitochondria (Young, 2003; Fan, 2006). Treatment
of cells with both of these inhibitors resulted in the rapid reduction of Akt levels in the
mitochondria. Since GA does not enter the mitochondria (Kang, 2007), it should not
45
affect HSP90 within the mitochondria. Therefore, it can be reasoned that GA prevented
endogenous cytoplasmic Akt from entering the mitochondria by only inhibiting
extramitochondrial HSP90. The rapidity of the decline in mitochondrial Akt levels,
occurring within 15 min of treatment with the inhibitors, raised the possibility that the
reduction of Akt levels in the mitochondria may be due to Akt proteolysis. NB and GA
treatments have previously been shown to cause the destabilization and degradation of
Akt (Marcu, 2000b; Basso, 2002; Kim, 2003; Yun 2004). In background work for these
experiments, high concentrations of NB and GA, greater than those shown in this study,
did cause pronounced decreases in Akt levels in the cytosol and the mitochondria (data
not shown). However, 625 μM NB and 5 μM GA for up to 60 and 90 min, respectively,
did not cause reductions in the cytosolic Akt levels. The reason for the rapid decline in
mitochondrial Akt level following NB and GA treatments is unclear. It is possible that
Akt is normally unstable in the mitochondria and is continuously replenished by import
of Akt from the cytosol. Thus, the import of Akt from the cytosol may maintain a steadystate level of Akt inside the mitochondria. The rapid decrease in Akt levels in the
mitochondria following treatment with the HSP90 inhibitors may be due to an imbalance
in these steady state levels, as replenishment of Akt from the cytosol is prevented. We
also tested the effect of HSP90 inhibition on brain mitochondrial Akt levels by injecting
NB into brains of adult mice. As seen in cell culture models, a two-hour treatment with
625 μM NB caused a significant reduction in Akt levels in mitochondria isolated from
cortex of treated animals, confirming that HSP90 regulates basal mitochondrial Akt
levels in-vivo.
46
To further test if decreased HSP90 activity affects Akt levels in the mitochondria,
HSP90 expression was blocked by siRNA-mediated knockdown of HSP90. Interestingly,
stable HSP90 siRNA HEK293 cell lines were able to thrive despite a significant
reduction in HSP90 protein level, possibly due to a compensatory increase in another
surrogate heat shock protein. For example, HSP40 was substantially increased in the
HSP90 knockdown cells. Nevertheless, cytosolic Akt levels in HSP90 siRNA cell lines
were unaffected, while the level of Akt in the mitochondria was found to be significantly
reduced. Together these results show that decreased HSP90 activity affects the level of
Akt in the mitochondria.
We could not discount the possibility that total cellular Akt levels may be
decreased with HSP90 inhibition, which may be occurring imperceptibly in the cytosolic
fraction, but is amplified in the mitochondrial fraction. As an alternative method to test if
HSP90 affects Akt mitochondrial import, a mitochondrial import assay was employed.
An unmodified wtAkt and constitutively active DDAkt were both tested for their ability
to be imported into the mitochondria. Surprisingly, both forms of Akt accumulated in the
mitochondria. In addition, the import of both forms was inhibited by HSP90 inhibition,
and by decreased HSP90 expression. This indicates that the normal flux of Akt into the
mitochondria is independent of the state of Akt activation. Indeed, Akt is constitutively
bound to HSP90 regardless of its activation state (Meares, 2004), thus it is reasonable to
assume that the levels of Akt in the mitochondria under unstimulated conditions are not
entirely dependent on Akt activity.
Previously it was shown that IGF-1 stimulation, which induces Thr308 and
Ser473 phosphorylation and activation of Akt, causes a rapid translocation of Akt into the
47
mitochondria (Bijur, 2003). Since the DDAkt somewhat mimics Akt phosphorylation at
Thr308 and Ser473, and this mutant Akt was mostly blocked from entering the
mitochondria by decreased HSP90 activity, we surmised that HSP90 inhibition would
also block Akt translocation into the mitochondria after IGF-1 stimulation. This was not
the case, as none of the modes of HSP90 inhibition blocked IGF-1-induced import of Akt
into the mitochondria. This finding clearly points to a second, HSP90-independent mode,
by which Akt is imported into the mitochondria, and may partially explain why a total
inhibition of mitochondrial Akt accumulation was never achievable by any of the HSP90
inhibitors. In addition to HSP90, HSP70, HSC70, and HSP40 have also been shown to
mediate the import of proteins into the mitochondria (Young, 2003: Bhangoo 2007). In
the HSP90 siRNA cells, the levels of HSP70 proteins were unchanged and the level of
HSP40 was markedly increased. Akt has been shown to be a client protein of HSP70, and
the HSP40 co-chaperones are known to interact with HSP70 (Gao, 2002; Cajo, 2006).
Thus, it is conceivable that in addition to HSP90, Akt mitochondrial translocation may
also be facilitated by HSP70, possibly via the HSP40 co-chaperone system described by
Bhangoo et al (2007).
To determine how Akt flux affects the mitochondrion, we began by observing
mitochondrial morphology. One of the most salient features of the mitochondrion is its
appearance. Mitochondria are highly plastic organelles which rapidly change their
configuration in response to their immediate surroundings and respiration states. In
addition, substrate availability and apoptosis signaling also affect mitochondrial
morphology (Gottlieb, 2003; Rasola, 2007). Mitochondria can change their size, shape,
and their cristae. A normally respiring mitochondrion has a classic “orthodox”
48
appearance (Hackenbrock, 1966; 1968), with an intact outer and inner membrane
separated by an intermembrane space, and a densely-packed matrix. Mitochondria
isolated from cells that have been subjected to a toxic stressor typically have highly
condensed cristae with an enlarged matrix (Gottlieb, 2003). For example, serum
deprivation is stressful to cultured cells and can enhance the effects of many toxic stimuli.
We found that mitochondria from HEK293 cells cultured in serum-free media for 24h
had enlarged mitochondria with highly condensed cristae. The addition of exogenous
constitutively active Akt caused the mitochondria to acquire a more orthodox appearance,
with densely-packed cristae. The Akt-induced alteration of the mitochondria was partially
blocked by the addition of NB, paralleling the finding that Akt import was also partially
blocked by HSP90 inhibition.
The initial step in the execution phase of apoptosis is the release of cytochrome c
from the mitochondrial intermembrane space (Haupt, 2003). It has been reported that
matrix remodeling to the condensed state, as was seen in the mitochondria from cells
cultured in serum-free media, exposes cytochrome c to the intermembrane space priming
its release from the mitochondria during apoptosis, whereas mitochondria resistant to
apoptosis maintain an orthodox conformation and this keeps cytochrome c tightly
sequestered within the cristae (Gottlieb, 2003). Akt is a well known cell survivalassociated protein, and overactive Akt is known to be tumorigenic and block anti-cancer
treatments that activate apoptosis (Haupt, 2003). In addition, activated Akt is known to
bolster neuronal survival in the face of toxic or ischemic insult (Mangi, 2003; Ohba,
2004). There are multiple mechanisms by which Akt promotes cell survival. The present
study suggests that an additional mode by which Akt is protective is by maintaining
49
mitochondria in their orthodox state, possibly facilitating the tight sequestration of
cytochrome c in the cristae, and preventing its release into the cytosol to activate
apoptosis.
In conclusion, this study shows that the HSP90 activity markedly influences the
level of Akt in the mitochondria. Furthermore, we have shown for the first time that
HSP90, via Akt, can dramatically alter mitochondrial configuration. Taken together,
these results provide new insight regarding Akt signaling in the mitochondria and its
effects on this organelle.
Acknowledgments
This work was supported by NIH grant NS044853. We wish to thank Dr. Richard
S. Jope for the kind gift of the Akt constructs. We also wish to thank Dr. Buffie
Clodfelder-Miller and Mr. Robert Mans for technical assistance in blinded studies.
50
References
Alessi DR, Andjelkovic M, Caudwell B, Cron P, Morrice N, Cohen P, Hemmings BA.
(1996) Mechanism of activation of protein kinase B by insulin and IGF-1. EMBO
J 15, 6541-51.
Alessi DR, James SR, Downes CP, Holmes AB, Gaffney PR, Reese CB, Cohen P. (1997)
Characterization of a 3-phosphoinositide-dependent protein kinase which
phosphorylates and activates protein kinase Bα. Curr Biol. 7, 261-9.
Andjelković M, Alessi DR, Meier R, Fernandez A, Lamb NJ, Frech M, Cron P, Cohen P,
Lucocq JM, Hemmings BA. (1997) Role of translocation in the activation and
function of protein kinase B. J Biol Chem 272, 31515-24.
Basso AD, Solit DB, Chiosis G, Giri B, Tsichlis P, Rosen N. (2002) Akt forms an
intracellular complex with heat shock protein 90 (HSP90) and Cdc37and is
destabilized by inhibitors of HSP90 function. J. Biol Chem. 277, 39858-66.
Bhangoo MK, Tzankov S, Fan AC, Dejgaard K, Thomas DY, Young JC. (2007) Multiple
40-kDa heat-shock protein chaperones function in TOM70-dependent
mitochondrial import. Mol Biol Cell. 18, 3414-28.
Bijur GN, Jope RS. (2003) Rapid accumulation of Akt in mitochondria following
phosphatidylinositol 3-kinase activation. J Neurochem 87, 1427-35.
Borgatti P, Martelli AM, Bellacosa A, Casto R, Massari L, Capitani S, Neri LM. (2000)
Translocation of Akt/PKB to the nucleus of osteoblast-like MC3T3-E1 cells
exposed to proliferative growth factors. FEBS Lett 477, 27-32.
Bossy-Wetzel E, Barsoum MJ, Godzik A, Schwarzenbacher R, Lipton SA. (2003)
Mitochondrial fission in apoptosis, neurodegeneration and aging. Curr Opin Cell
Biol 15, 706-16.
Cajo GC, Horne BE, Kelley WL, Schwager F, Georgopoulos C, Genevaux P. (2006) The
role of the DIF motif of the DnaJ (HSP40) co-chaperone in the regulation of the
DnaK (HSP70) chaperone cycle. J Biol Chem. 281, 12436-44.
Chacinska A, Pfannschmidt S, Wiedemann N, Kozjak V, Sanjuán Szklarz LK, SchulzeSpecking A, Truscott KN, Guiard B, Meisinger C, Pfanner N. (2004) Essential
role of Mia40 in import and assembly of mitochondrial intermembrane space
proteins. EMBO J. 23, 3735-46.
51
Fan AC, Bhangoo MK, Young JC. (2006) HSP90 functions in the targeting and outer
membrane translocation steps of TOM70-mediated mitochondrial import. J Biol
Chem. 281, 33313-24.
Fayard E, Tintignac LA, Baudry A, Hemmings BA. (2005) Protein kinase B/Akt at a
glance. J Cell Sci. 118, 5675-8.
Fujita N, Sato S, Ishida A, Tsuruo T. (2002) Involvement of HSP90 in signaling and
stability of 3-phosphoinositide-dependent kinase-1. J Biol Chem. 277, 10346-53.
Gabriel K, Egan B, Lithgow T. (2003) Tom40, the import channel of the mitochondrial
outer membrane, plays an active role in sorting imported proteins. EMBO J. 22,
2380-6.
Gao T, Newton AC. (2002) The turn motif is a phosphorylation switch that regulates the
binding of HSP70 to protein kinase C. J Biol Chem. 277, 31585-92.
Geissler A, Krimmer T, Bömer U, Guiard B, Rassow J, Pfanner N. (2000) Membrane
potential-driven protein import into mitochondria. The sorting sequence of
cytochrome b(2) modulates the ΔΨ-dependence of translocation of the matrixtargeting sequence. Mol Biol Cell 11, 3977-91.
Gottlieb E, Armour SM, Harris MH, Thompson CB. (2003) Mitochondrial membrane
potential regulates matrix configuration and cytochrome c release during
apoptosis. Cell Death Differ 10, 709-17.
Hackenbrock CR. (1966) Ultrastructural bases for metabolically linked mechanical
activity in mitochondria. I. Reversible ultrastructural changes with change in
metabolic steady state in isolated liver mitochondria. J Cell Biol 30, 269-97.
Hackenbrock CR. (1968) Ultrastructural bases for metabolically linked mechanical
activity in mitochondria. II. Electron transport-linked ultrastructural
transformations in mitochondria. J Cell Biol. 37, 345-69.
Haupt S, Berger M, Goldberg Z, Haupt Y. (2003) Apoptosis - the p53 network. J Cell
Sci. 116, 4077-85.
Kang BH, Plescia J, Dohi T, Rosa J, Doxsey SJ, Altieri DC. (2007) Regulation of tumor
cell mitochondrial homeostasis by an organelle-specific HSP90 chaperone
network.
Cell 131, 257-70.
Kerscher O, Holder J, Srinivasan M, Leung RS, Jensen RE. (1997) The Tim54p-Tim22p
complex mediates insertion of proteins into the mitochondrial inner membrane. J
Cell Biol. 139, 1663-75.
52
Kim S, Kang J, Hu W, Evers BM, Chung DH. (2003) Geldanamycin decreases Raf-1 and
Akt levels and induces apoptosis in neuroblastomas. Int J Cancer 103, 352-9.
Komiya T, Rospert S, Schatz G, Mihara K. (1997) Binding of mitochondrial precursor
proteins to the cytoplasmic domains of the import receptors TOM70 and TOM20
is determined by cytoplasmic chaperones. EMBO J. 16, 4267-75.
Kutik S, Guiard B, Meyer HE, Wiedemann N, Pfanner N. (2007) Cooperation of
translocase complexes in mitochondrial protein import. J Cell Biol. 179, 585-91.
Mangi AA, Noiseux N, Kong D, He H, Rezvani M, Ingwall JS, Dzau VJ. (2003)
Mesenchymal stem cells modified with Akt prevent remodeling and restore
performance of infarcted hearts. Nat Med. 9, 1195-201.
Marcu MG, Chadli A, Bouhouche I, Catelli M, Neckers LM. (2000) The heat shock
protein 90 antagonist novobiocin interacts with a previously unrecognized ATPbinding domain in the carboxyl terminus of the chaperone. J Biol Chem. 275,
37181-6.
Marcu MG, Schulte TW, Neckers L. (2000) Novobiocin and related coumarins and
depletion of heat shock protein 90-dependent signaling proteins. J Natl Cancer
Inst. 92, 242-8.
Martin J, Mahlke K, Pfanner N. (1991) Role of an energized inner membrane in
mitochondrial protein import. ΔΨ drives the movement of presequences. J Biol
Chem 266, 18051-7.
Meares GP, Zmijewska AA, Jope RS. (2004) Heat shock protein-90 dampens and directs
signaling stimulated by insulin-like growth factor-1 and insulin. FEBS Lett. 574,
181-6.
Ohba N, Kiryu-Seo S, Maeda M, Muraoka M, Ishii M, Kiyama H. (2004) Transgenic
mouse overexpressing the Akt reduced the volume of infarct area after middle
cerebral artery occlusion. Neurosci Lett. 359, 159-62.
Panaretou B, Prodromou C, Roe SM, O'Brien R, Ladbury JE, Piper PW, Pearl LH. (1998)
ATP binding and hydrolysis are essential to the function of the HSP90 molecular
chaperone in vivo. EMBO J. 17, 4829-36.
Pfanner N, Rassow J, Guiard B, Söllner T, Hartl FU, Neupert W. (1990) Energy
requirements for unfolding and membrane translocation of precursor proteins
during import into mitochondria. J Biol Chem 265, 16324-9.
Pfanner N, Geissler A. (2001) Versatility of the mitochondrial protein import machinery.
Nat Rev Mol Cell Biol. 2, 339-49.
53
Rasola A, Bernardi P. (2000) The mitochondrial permeability transition pore and its
involvement in cell death and in disease pathogenesis. Apoptosis. 12, 815-33
(2007)
Sato S, Fujita N, Tsuruo T. (2000) Modulation of Akt kinase activity by binding to
HSP90. Proc Natl Acad Sci U S A. 97, 10832-7.
Whitesell L, Mimnaugh EG, De Costa B, Myers CE, Neckers LM. (1994) Inhibition of
heat shock protein HSP90-pp60v-src heteroprotein complex formation by
benzoquinone ansamycins: essential role for stress proteins in oncogenic
transformation. Proc Natl Acad Sci U S A 91, 8324-8.
Xu W, Yuan X, Jung YJ, Yang Y, Basso A, Rosen N, Chung EJ, Trepel J, Neckers L.
(2003) The heat shock protein 90 inhibitor geldanamycin and the ErbB inhibitor
ZD1839 promote rapid PP1 phosphatase-dependent inactivation of Akt in ErbB2
overexpressing breast cancer cells. Cancer Res. 63, 7777-84.
Yano M, Terada K, Mori M. (2004) Mitochondrial import receptors TOM20 and TOM22
have chaperone-like activity. J Biol Chem. 279, 10808-13.
Young JC, Hoogenraad NJ, Hartl FU. (2003) Molecular chaperones HSP90 and HSP70
deliver preproteins to the mitochondrial import receptor TOM70. Cell 112, 4150.
Young JC, Agashe VR, Siegers K, Hartl FU. (2004) Pathways of chaperone-mediated
protein folding in the cytosol. Nat Rev Mol Cell Biol. 5, 781-91.
Yun BG, Huang W, Leach N, Hartson SD, Matts RL. (2004) Novobiocin induces a
distinct conformation of HSP90 and alters HSP90-cochaperone-client interactions.
Biochemistry. 43, 8217-29.
54
55
Fig. 1. HSP90 inhibitors decrease mitochondrial Akt levels in SH-SY5Y and HEK293
cells. (A) Mitochondrial and cytosolic fractions from SH-SY5Y cells were
immunoblotted for β-tubulin (a cytosolic protein) and pyruvate dehydrogenase (PDH), a
mitochondrial protein, to assess mitochondrial purity. (B) Proteinase K digestion of
mitochondrial fractions, followed by immunoblotting, reveals HSP90 existence on the
mitochondrial surface. (C) Serum-starved SH-SY5Y cells were treated time-dependently
with 625 μM novobiocin to inhibit HSP90. Cytosolic (cyto) and mitochondrial (mito)
fractions were immunoblotted for Akt. The Akt bands in the mitochondria were
quantitated by scanning densitometry and are shown in the corresponding graphs as
percent of control, n=3, *p<0.05 compared to values from controls, ANOVA. (D) Serumstarved SH-SY5Y cells were treated with 5 μM geldanamycin to inhibit HSP90.
Fractions were immunoblotted for Akt, n=3, *p<0.05 compared to values from controls,
ANOVA. (E) HEK293 cells were treated for 45 min with 625 μM novobiocin or 5 μM
geldanamycin to inhibit HSP90. For in-vivo studies, “cortex”, novobiocin (625 μM) was
stereotaxically injected into ventricles of mice as described in “Materials and methods”.
The cytosolic and mitochondrial fractions from the cells and cortical tissue were
immunoblotted for Akt and protein bands were quantitated by scanning densitometry.
Percent of control, n=3, *p< 0.05 compared to values from controls, Student’s t-test.
56
Fig. 2. siRNA-mediated knockdown of HSP90 expression results in decreased levels of
mitochondrial Akt. (A) Cells stably transfected with HSP90 siRNA were lysed and
immunoblotted for expression of HSP90, Akt, HSP70, HSC70, and HSP40. *p<0.05
compared with values from control cell line, Student’s t-test. (B) Cells transfected with
HSP90 siRNA or control pcDNA plasmid were fractionated into cytosolic and
mitochondrial fractions and the two fractions were immunoblotted for Akt. *p<0.05
compared with values from control cell line, Student’s t-test.
57
58
Fig. 3. HSP90 mediates the mitochondrial translocation of Akt. (A) Top panel: Isolated
mitochondria were incubated with either wild type (wt) or DDAkt, digested with
proteinase K, and immunoblotted with a phospho-Akt substrate antibody. Bottom panel:
In-vitro translated Akt proteins were measured for enzymatic activity as described in
Material and methods. (B) Mitochondria were isolated from pcDNA control cells and
HSP90 siRNA cells. Mitochondrial lysates were immunoblotted for HSP90. (C)
Representative autoradiograph showing mitochondrial translocation of 35S-Methionine
labeled wtAkt and DDAkt in control cells and HSP90 siRNA cells. Isolated mitochondria
from each cell line were incubated with radiolabled exogenous wtAkt or DD-Akt protein.
Half of each sample was digested with proteinase K, and the other half was left
undigested. The Akt bands were viewed with a phosphorimager. (D) Representative
autoradiograph showing mitochondrial translocation of wtAkt (left panel) and DDAkt
(right panel) in the presence and absence of 625 μM novobiocin.
59
Fig. 4. HSP90 inhibitors do not block IGF-1-induced mitochondrial Akt accumulation.
(A) Serum-starved HEK293 cells were pretreated for 15 minutes with 625 μM NB prior
to treatment with IGF-1, or cells were treated with IGF-1 (50 ng/mL for 15 minutes)
alone. Mitochondrial and cytosolic fractions were immunoblotted for Akt. (B) HEK293
cells were treated with 50 ng/mL IGF-1 for 15 minutes, or pretreated 5μM NB for 15 min
prior to treatment with IGF-1. Mitochondrial and cytosolic fractions were immunoblotted
for Akt. (C) pcDNA control cells and HSP90 knockdown cells were treated for 15
minutes with 50ng/mL IGF-1. Mitochondrial and cytosolic fractions were immunoblotted
for Akt.
60
61
Fig. 5. Akt affects mitochondrial morphology. (A) Mitochondrial pellets were fixed and
prepared for transmission electron microscopy as described in the Materials and Methods
section. Representative transmission electron micrographs of mitochondria isolated from
HEK293 cells in serum-containing versus serum-containing media (size bar = 100 nm).
(B) Mitochondria were isolated from serum-starved HEK293 cells and incubated with
empty reticulocyte (control), DD-Akt, or 625 μM NB and DD-Akt for 45 minutes (size
bar = 100 nm). Representative images are shown. (C) Optical density was quantified in a
blind study from transmission electron micrographs of isolated mitochondria incubated
with DD-Akt or NB + DD-Akt versus control. N=54, *p<0.05, compared to control
mitochondria incubated with reticulocyte lysate. **p<0.05, compared to control
mitochondria incubated with DD-Akt. ANOVA.
62
LOCALIZATION AND FUNCTION OF AKT IN BRAIN MITOCHONDRIA
by
Keri A. Barksdale, Emma Perez-Costas, Rosalinda C. Roberts,
Miguel Melendez-Ferro, and Gautam N. Bijur
In preparation for submission to
Journal of Neurochemistry
Format adapted for dissertation
63
Abstract
The ubiquitously expressed serine/threonine kinase Akt is well recognized as a
major player in cell survival and a member of multiple cell signaling pathways, notably
the phosphatidylinositol 3-kinase (PI3K) signaling pathway. Akt signaling is known to
have substantial roles in brain development and function; however, no ultrastructural
localization studies of Akt in the brain under basal conditions have been done. Here we
report the localization of Akt within neurons and in subcellular organelles of mouse brain,
including the observation of mitochondrial Akt localization. Akt preferentially exists in
mitochondria of neuronal processes (pMito) compared to mitochondria in the soma
(cMito). Functional studies of these two mitochondrial populations reveal that pMito
have enhanced ATP production rates compared to cMito. Furthermore, ICV injection of
insulin like growth factor-1 into adult mice, which activates PI3K signaling and causes
dramatic increases in mitochondrial Akt levels in pMito, caused concurrent increases in
ATP production rates in pMito. ICV injection of LY294002, an inhibitor of PI3K, caused
significant decreases in pSer473 Akt levels in pMito as well as decreases in ATP
production rates. Taken together, these data indicate that the level of Akt in mitochondria
affects mitochondrial function in the form of ATP production.
64
Introduction
The role of the ubiquitously expressed serine/threonine kinase Akt on signaling,
particularly in the brain, has been the subject of numerous studies. In fact, Akt is known
to play roles in significant processes in the brain such as establishment of neuronal
polarity, adult neurogenesis, and long-term potentiation (Yan 2006; Peineau 2007; BruelJungerman 2009). Furthermore, because of its involvement in psychiatric and
neurological disease pathology, Akt has become particularly interesting to researchers
exploring this field. Akt is widely known to phosphorylate and inactivate GSK3β (Ali
2001), which is notable because aberrant GSK3β activity in the brain has been associated
with the pathology of neurodegenerative and psychiatric disorders such as Alzheimer’s
disease, Parkinson’s disease, bipolar disorder, and schizophrenia (Hanger 1992; King
2001; Klein 1996; Kozlovsky 2000).
It is likely that the activity and distribution of Akt in the brain during development
and adulthood can contribute to its physiological roles within the brain. While it is known
that Akt exists in multiple cell compartments, very few studies have examined the
distribution of Akt at the light and electron microscopy level (Eto 2008; Znamensky 2003)
under basal conditions. No studies have shown basal distribution of Akt and paired the
ultrastructural localization data with a physiological function of Akt distribution within a
specific organelle, in this case the mitochondria.
Mitochondria have a specific distribution in neurons, and nascent mitochondria
are generated in the neuronal cell body (Miller 2004), near the cellular transcription and
translation machinery. Mitochondria with a high membrane potential produce more ATP
and are rapidly transported toward axonal and dendritic terminals where the energy
65
demand is high (Miller 2004). Suppression of this transport by disease processes or
mitochondrial toxins is known to abolish synaptic potentiation (Tong 2007).
Mitochondria that are more depolarized, indicating aging or pharmacological inhibition,
are transported back to the cell body (Miller 2004). Thus, mitochondria exist in neurons
in two major populations which we will refer to as mitochondria from cell bodies (cMito)
and neuronal processes (pMito).
Many investigators have examined the effect of Akt signaling on mitochondrial
function as it relates to cell death and metabolism of the cell. It is known, for example,
that increased Akt leads to more efficient sequestration of cytochrome c, a protein
essential for apoptosis (Kennedy 1999; Arciuch 2009; Stiles 2009), which, in turn,
increases cellular survival. Experiments performed in cultured adult sensory neurons as
well as human muscle tissue showed that insulin infusion, which causes Akt activation,
increases levels of ATP as well as rates of ATP production and enhances the
mitochondrial membrane potential (Stump 2003; Huang 2004). Serum withdrawal is
known to cause marked decreases in mitochondrial respiration in multiple cell culture
models using a variety of tissues, and treatment with small doses IGF-1 rescues deficits
in respiration due to growth factor/serum withdrawal (Unterluggauer 2008). Furthermore,
treatment of aging rats with low doses of IGF-1 has been shown to prevent age-related
oxidative damage, loss of mitochondrial membrane potential and deficits in ATP
production (Puche 2008). Lastly, nerve growth factor signaling in the brain is also known
to enhance the mitochondrial membrane potential when applied locally to axons of
cultured dorsal root ganglia cells (Verburg 2008). These experiments show that Akt
66
activity has a profound effect on mitochondrial function in general, but despite these
findings, the exact role of the mitochondrial pool of Akt has remained elusive.
In the present study, immunohistochemistry for detection of Akt as well as
phospho-Ser473-Akt (pAkt) was used in combination with light microscopy and
transmission electron microscopy. Mitochondria were isolated from neuronal processes
(pMito) and cell bodies (cMito) and tested for function under basal conditions. Further
studies were conducted in which mice were intracerebroventrically (ICV) injected with
insulin-like growth factor-1 (IGF-1) or LY294002 (LY) to activate or inhibit the PI3Kinase pathway. Results revealed that mitochondria containing more Akt (pMito) have
enhanced energy production capacity, and this effect can be enhanced or inhibited by the
addition of IGF-1 or LY. This study not only illustrates the intracellular distribution of
Akt within subcellular compartments, but also points to a distinct Akt function within
neuronal mitochondria.
Materials and Methods
Tissue preparation and fixation for immunohistochemistry
All animal procedures were in accordance with the University of Alabama at
Birmingham Institutional Animal Care and Use Committee (IACUC) guidelines. Mice
were euthanized according to an approved IACUC protocol. Adult male C57BL6 mice
(10 to 12 weeks old) were used in this study. Mice were euthanized by decapitation and
the brains were immediately removed, quickly rinsed in cold 0.1M phosphate buffer (PB)
and fixed by immersion in a 4% paraformaldehyde and 0.1% glutaraldehyde in PB, pH
7.4 solution, at 4oC overnight. The brains were then sectioned in the coronal plane on a
67
vibratome and 100 μm free-floating sections were obtained. The sections were stored in
PB at 4°C until processed for immunohistochemistry.
For perfusion fixation, mice were anesthetized using 50 mg/kg sodium
pentobarbital and manually perfused with 0.1 M PB followed by 4% paraformaldehyde
and 0.1% glutaraldehyde in PB. Brains were removed and sectioned as described above.
Immunohistochemistry
For the immunohistochemical localization of Akt and phospho-Ser473 Akt, freefloating sections were rinsed in phosphate buffered saline (PBS), quenched in 1% sodium
borohydride in PBS for 15 minutes, rinsed multiples times in PBS, and the endogenous
peroxide was blocked in a solution of 1.5% hydrogen peroxide in PBS for five minutes.
After rinsing in PBS, non-specific binding sites in the sections were blocked with 2%
normal goat serum in PBS for 30 minutes. The sections were then incubated for 72 hours
at 4°C in a 1:500 dilution of a polyclonal rabbit Akt antibody (Cell Signaling, Danvers,
MA). In addition, two types of controls were performed: some sections were incubated in
the absence of primary antibody, while others were incubated with the Akt or phosphoSer473 Akt antibody pre-adsorbed with two micrograms of Akt blocking peptide or
phospho-Ser473 blocking peptide (Cell Signaling). After rinsing in PBS (four rinses, five
minutes each), all sections were incubated with a biotinylated goat anti-rabbit secondary
antibody (Vector laboratories, Burlingame, CA) diluted 1:200 for 1h at room temperature.
Then, sections were rinsed in PBS and incubated for 30 minutes with an avidinbiotinylated horseradish peroxidase complex (Vectastain ABC system, Vector
Laboratories). After rinsing in PBS, the sections were developed in a diaminobenzidine
solution (10 mg/15 ml PBS; Sigma, St Louis, MO) containing 0.03% hydrogen peroxide
68
for 2-5 min to visualize the reaction product. The sections were then rinsed in PB and
processed for light or electron microscopy.
Light microscopy
Immunostained sections were mounted on colorfrost/plus slides (Fisher,
Pittsburgh, PA), air-dried overnight, dehydrated in ascending series of ethanol, cleared in
xylene and coverslipped with Eukitt mounting media (O. Kindler, Germany). Sections
were viewed and photographed using a Nikon DS-Fi1 color digital camera coupled to a
Nikon Eclipse 50i. Images were converted to grey scale and adjusted for brightness and
contrast using Corel PhotoPaint 12 (Corel Corporation, Ottawa, Canada). Photomontage
and lettering were done using CorelDraw 12.
Electron Microscopy
Immunostained sections were rinsed in PB and immersed in a solution of 1%
osmium tetroxide in PB for 1 hour, then rinsed in PB (3 times, 10 minutes each), and
gradually dehydrated on series of ethanol from 30% to 70%. Then, sections were stained
with a solution of 1% uranyl acetate in 70% ethanol for 2 hours and further dehydrated in
ethanol. After dehydration was completed, the sections were cleared in propylene oxide
and infiltrated with Epon resin overnight at room temperature. The sections were then
flat-embedded in new Epon resin and allowed to polymerize in an oven at 60°C for 72
hours. Selection of regions of interest for electron microscopy was performed by
visualizing the flat-embedded sections on a Nikon Eclipse 50i light microscope, carefully
identifying anatomical regions and re-dissecting these regions for ultramicrotomy.
Ultrathin sections (90 nm thick) were obtained using a Leica EM UC6 ultramicrotome
(Leica Microsystems, Wetzlar, Germany), mounted on nickel grids and observed and
69
photographed using a FEI-Tecnai T12 Spirit TWIN 20-120kv transmission electron
microscope (FEI, Hillsboro, OR) equipped with an AMT digital camera (Danvers, MA).
Photomontage and lettering was done as for light microscopy.
Isolation of Mitochondrial Populations from Neuronal Processes and Cell Bodies
Tissue from mouse cortices was placed in dounce homogenation tubes containing
ice-cold isolation buffer (215 mM mannitol [Acros Organics, NJ] 75 mM sucrose [Roche
Diagnostics, Indianapolis, IN], 0.1% fatty acids-free bovine serum albumin (FAF-BSA)
[Sigma Aldrich, St. Louis, MO], 20 mM HEPES [Sigma Aldrich], 1 mM ethylene glycol
tetraacetic acid (EGTA), adjusted to pH 7.2 with 5M KOH. The same amount of ice-cold
30% Percoll (GE Healthcare Life Sciences, Uppsala, Sweden) in isolation buffer was
added (about 2X volume) for
a final 15% Percoll solution in the sample. Centrifuge tubes were layered with a 40%
Percoll solution in isolation buffer into the centrifuge tube, followed by a 24% Percoll
solution, and the mitochondrial sample in a 15% Percoll solution. Samples were
centrifuged in a Sorvall RC-5C Plus centrifuge using a fixed angle SS34 rotor at 34,500 x
g for 30 minutes at 4ºC. After centrifugation three hazy bands were formed in the Percoll.
The second and third bands from the top [synaptosomes and cell bodies, respectively]
were removed from the Percoll media and placed in separate tubes, and 10 ml of isolation
buffer was added to wash the Percoll from the sample. Samples were centrifuged at
16,700 x g for 15 minutes. The loose pellets obtained were then resuspended in isolation
buffer and placed in a nitrogen cavitation bomb (Parr Instrument Company, Moline, IL).
The samples were subjected to nitrogen cavitation (800 p.s.i. for 10 min.) for further
homogenization. To further clarify the mitochondria from mitochondrial processes, the
70
pMito homogenates were mixed with an equal volume of 10% Percoll in isolation buffer.
The samples were layered over a 20% Percoll and a 12% Percoll cushion and centrifuged
at 34,500 x g for 30 minutes at 4oC. Following centrifugation, the bottom band containing
pMito was collected. Isolation buffer (10 ml) was added and the sample was centrifuged
at 16,700 x g for 15 minutes. The loose pellet was collected, washed once with isolation
buffer and centrifuged for 10 minutes at 17,600 x g in an Eppendorf 5417R tabletop
centrifuge at 4ºC To collect the cMito following nitrogen cavitation, samples were
centrifuged for 10 minutes at 750 x g in an Eppendorf 5417R tabletop centrifuge at 4ºC.
The supernatant was retained and centrifuged for 15 minutes at 17,600 x g, then the pellet
was collected and washed once with isolation buffer and centrifuged for 10 minutes at
17,600 x g. Mitochondria of both types (pMito and cMito) were kept pelleted on ice until
ready for further use.
Mitochondrial Cryopreservation
All mitochondrial samples were cryopreserved for one week following isolation
and prior to testing. Mitochondria from cell bodies and neuronal processes were
resuspended in isolation buffer containing 20% DMSO (Research Organics, Cleveland,
OH) and 1% FAF-BSA in Nalgene Cryoware cryogenic vials (Nalge Nunc International,
Rochester, NY). Vials were then placed in a foam-insulated thermofreezing container
with an approximate cooling rate of 1ºC per minute (Aladdin Industries Incorporated,
Nashville, TN), and the container was placed at -80oC for 6-8 days. Mitochondrial
suspensions were then thawed at 4oC, washed with isolation buffer, and prepared for
functional studies (mitochondrial membrane potential and ATP production).
71
Introcerebroventricular (ICV) Injections
ICV injections were performed exactly as described previously (Barksdale 2009).
For LY ICV injections, LY294002 was solubilized in DMSO and diluted to a stock
solution in 10% DMSO. 1ul of LY stock was injected in each lateral ventricle for a final
concentration of 20uM. For IGF-1 injections, 1ul of IGF-1 stock solution in 0.9% saline
was injected in each lateral ventricle for a final concentration of 50ng/mL. Lastly, 1 ul of
novobiocin in 0.9% saline stock was injected into each lateral ventricle for a final
concentration of 625 uM. For vehicle control injections, either 10% DMSO in 0.9%
saline or 0.9% saline was used. Following injection, animals were allowed to recover for
2 hours and sacrificed using decapitation. Brains were removed, cortices were dissected
out, and mitochondria were isolated for testing as described above.
ATP Production Assay
After washing and determining protein concentration, 250µg aliquots of pMito
were pelleted and placed on ice until ready for use. For the ATP production assay, pellets
were resuspended in 185 µl respiration buffer [20 mM HEPES-KOH pH 7.4, 80 mM
potassium acetate, 5 mM magnesium acetate, 250 mM sucrose, 1 mM glutamate, and 1
mM malate]. Ten µl of Buffer B [0.5 M Tris-Acetate pH 7.75, 0.8 mM D-Luciferin, 20
μg/ml luciferase] were immediately added to the reaction, and sample luminescence was
read in standard mode on a Turner 20/20 luminometer for 4 seconds with no delay. ADP
(0.1mM) was then added to the sample, and the luminescence was read every 4 seconds
for 180 seconds. Buffer background measurements were taken using buffers containing
all substrates, and these measurements were subtracted from the mitochondrial readings.
72
ATP standards were read using the same conditions and used to extrapolate ATP
generation rates for each sample.
Results
Akt and pSer-473 Akt localization in the brain
It was initially necessary to confirm the specificity of the Akt and pSer473 Akt
(pAkt) antibodies used for this study. These antibodies were first used to immunoblot
mouse cortical homogenates in the presence or absence of blocking peptides specific for
each antibody. Our results show that the chemiluminescence signal disappears when the
Akt or pAkt antibody is pre-incubated with the blocking peptide (Figure 1A).
To characterize localization of Akt in the brain, brains were either perfusion fixed
or immediately immersion-fixed in paraformaldehyde. Coronal slices were stained using
antibodies for Akt and pAkt. First, it was necessary to further confirm specificity of the
antibody, so immunostaining was performed in parallel sections of the mouse brain to
compare the staining obtained with the Akt or pAkt antibodies alone, preadsorption of
each antibody with the blocking peptide, and omission of the primary antibody. Our
results showed that Akt and pAkt strongly labels neurons in the cortex (Figure 1B)
whereas in an adjacent section that was processed with the peptide-preadsorbed antibody
no immunostaining was observed (data not shown). Additionally, incubation with a
nonspecific rabbit IgG, or omission of the primary antibody produced no staining (data
not shown). Together these results confirmed the specificity of Akt and pAkt
immunolabeling and validated the use of this antibody for our studies at the light and
electron microscope.
73
To elucidate Akt localization, light microscopy was first performed on stained
coronal slices. Both Akt and pAkt antibodies labeled neurons in the mouse cortex (Figure
1B). Staining was mostly confined to neurons in the perinuclear regions, and indicates
that under basal conditions Akt is present in abundant amounts in neurons. Furthermore,
by the presence of pAkt, it can be surmised that there is some basal level of Akt
activation in the unstimulated brain.
To characterize ultrastructural localization of Akt within neurons, electron
microscopy of DAB-stained coronal sections was performed. In mouse cortex, Akt
staining was widespread in neurons (Figure 1C). Akt has been previously shown to exist
in mitochondria (Bijur 2003; Barksdale 2009), and indeed electron microscopy indicated
Akt staining on mitochondria in neurons (Figure 1C). Interestingly, the predominant Akt
labeling of mitochondria was found in cross sections of dendrites (Figure 1C, open
arrows) compared to other areas of the cell. In fact, mitochondria that were not localized
to neuronal processes contained no observable Akt staining (Figure 1C, closed arrows).
These data indicate that while Akt exists throughout the neuron, the mitochondrial pool
of Akt is preferentially localized to mitochondria in neuronal processes. To further
confirm this result, pMito and cMito were isolated from mouse cortex and
immunoblotted for Akt and pSer473Akt. pMito were found to contain approximately
70% less Akt compared to cMito, and pAkt was not detected in cMito (Figure 2A, B).
We have previously reported that heat shock protein-90 is responsible for Akt
accumulation in the mitochondria (Barksdale 2009), and interestingly, cMito were found
to contain approximately 40% less Hsp90 (Figure 2A, B). Furthermore, pMito and cMito
74
fractions from postmortem human cortex were immunoblotted for Akt levels. As in the
mouse model, Akt was not detectable in cMito (Figure 2C).
ATP Production rates in pMito and cMito
Since less Akt exits in cMito compared to pMito, and increased Akt signaling in
neurons has been linked to increased energetic capacity, rates of ATP production were
examined in pMito and cMito using a luciferase-based ATP assay. The data revealed that
pMito produce 130.92 nmoles ATP per minute per μg of mitochondrial protein,
compared with only 92.173 nmoles ATP per minute produced by cMito (Table 1). As a
control, the mitochondrial toxin CCCP was added to the pMito preparation and ATP
production was measured. The presence of CCCP caused a significant drop in ATP
production rate (data not shown).
Modulation of the PI3K signaling pathway alters mitochondrial Akt levels and rates of
ATP production
It has been shown previously that treatment of cultured cells with IGF-1 causes a
significant, robust increase in Akt levels in the mitochondria (Bijur 2003). In order to test
if pMito and cMito function is affected by modulation of the PI3K pathway, IGF-1
(50ng/mL) was ICV injected into adult male mice and pMito and cMito were isolated.
Mitochondrial lysates were immunoblotted to confirm activation of the PI3K signaling
pathway. pSer473-Akt was significantly increased in pMito isolated from IGF-1 injected
mice compared to vehicle injected mice (Figure 3A, B). No pSer473Akt was detectable
in cMito from vehicle or IGF-1 injected animals, as evidenced by a highly overexposed
immunoblot, suggesting no increase in ATP production should occur in these samples
(Figure 2C). ICV injection of IGF-1 caused a significant 29% increase in the rate of ATP
75
production in pMito (208.39 nmol/min/μg protein) compared with 161.88 nmol/min/μg
protein in pMito isolated from vehicle-injected mice (Table 1). Conversely, IGF-1
injection caused no significant difference in rates of ATP production in mitochondria
isolated from neuronal cell bodies (Table 1), and actually indicated a nonsignificant 25%
decrease in ATP production in cMito isolated from IGF-1 injected mice. Again, the
addition of CCCP to pMito isolated from both vehicle and IGF-1 mice caused a
significant decrease in ATP production to 50.31 and 47.31 nmol/min/μg
protein,
respectively (Table 1).
To inhibit Akt signaling, LY294002 (20μM) was ICV injected into adult male
mice, and following a two-hour recovery, mice were sacrificed and pMito were isolated.
For this experiment, cMito were not tested, because it is not hypothesized that there
would be any further decrease in Akt or pSer473Akt levels or in rates of ATP production.
Western blotting analysis of these samples revealed no change in total Akt levels, but a
significant decrease in levels of pSer473 Akt in pMito from LY injected mice compared
to vehicle injected mice (Figure 3D, E). Analysis of ATP production rates showed that
pMito from LY injected mice produce 137.17 nmoles ATP/min/μg protein, which is a
significant 33% decrease compared to the 203.74 nmoles ATP/min produced by pMito
isolated from vehicle injected mice. In addition, ICV injection of the Hsp90 inhibitor
novobiocin, which we have shown causes a significant decrease in levels of Akt in
cortical mitochondria, caused a significant 35% decrease in ATP production from 111.9
+/- 32.98 nmoles/min/ug protein to 72.54 +/- 28.69 nm/min/ug protein. Together, these
data indicate that mitochondrial Akt plays some role in the rate of ATP production in the
brain, and that Akt signaling is heterogeneous among brain mitochondria.
76
Discussion
Here we report the novel discovery that the level of Akt in brain mitochondria
plays a significant role in mitochondrial function. Previous studies have indicated that
Akt signaling can impact brain functions, but in this study we focus on Akt signaling
within the mitochondria and show for the first time that Akt signaling is divergent within
different mitochondrial populations. Furthermore, the variation in Akt signaling within
mitochondria reflects clear differences in mitochondrial metabolic output.
In this study, light microscopy and electron microscopy of coronal brain sections
were employed to determine Akt staining in the brain and its subcellular localization
within neurons. The results were further confirmed by immunoblotting for Akt. Under
basal conditions, Akt was shown to exist in many areas of neurons throughout the brain,
including mitochondria. Furthermore, Akt preferentially labeled mitochondria in neuronal
processes (axons and dendrites) compared to neuronal cell bodies, results which were
confirmed using western blotting of mouse and human cortical mitochondria. It is
currently unclear why Akt would localize to mitochondria of neuronal processes. Since it
is known that certain mitochondrial signals such as membrane potential can initiate
retrograde and anterograde mitochondrial trafficking and that growth factor signaling is
involved in mitochondrial positioning and movement through cells (Chada 2003, Miller
2004; Hollenbeck 2005), it is possible that degradation of Akt is a signal for
mitochondria to be transported back to the soma. Alternately, it may be that mitochondria
containing large amounts of Akt are automatically transported to neuronal processes.
Since neuronal processes are known to require large amounts of energy, and we show
77
that the presence of Akt bolsters mitochondrial ATP production, the presence of Akt may
be a requirement for delivery of highly functioning mitochondria to neuronal processes.
Since isolated mitochondria from neuronal processes contained more Akt than
mitochondria from neuronal cell bodies, and cellular Akt signaling has been shown to
increase mitochondrial ATP production rates (Stump 2003; Huang 2004), we measured
rates of mitochondrial ATP production in isolated pMito and cMito. It is important to
reiterate that these measurements were performed in isolated mitochondria; therefore a
view of the function of the mitochondrial pool of Akt could be obtained. The finding that
pMito have enhanced rates of ATP production compared to cMito led to further
experimentation in which IGF-1 and LY, were injected into ventricles of adult male mice.
ICV injection of IGF-1 caused significant increases in levels of Akt and pSer473-Akt in
isolated pMito but not in cMito. Concurrently, ATP production rates were significantly
increased in pMito compared to vehicle controls. Interestingly, there was a nonsignificant
25% decrease in ATP production in pMito from IGF-1 treated mice. The reason for this is
not currently clear. It is possible that IGF-1 stimulation initiates signaling of multiple
signaling pathways in neurons and that in the neuronal cell body this is a detriment to
mitochondrial function, in this case ATP production.
To examine if inhibition of Akt activity could decrease Akt levels and ATP
production rates in isolated mitochondria, LY was ICV injected into adult male mice.
cMito were not tested, because pSer473 Akt could not be detected in cMito under basal
conditions so it stands to reason that there would be no further decrease in either
pSer473-Akt or ATP production rates. In pMito isolated from LY-treated mice, however,
there was a significant decrease in pSer473Akt levels as well as ATP production,
78
indicating that loss of Akt in mitochondria causes a detriment to function. Overall, these
data conform to previous reports that Akt signaling at the plasma membrane plays a role
in mitochondrial function. However, this study specifically isolated the function of the
mitochondrial pool of Akt, which was previously largely unknown.
To further confirm the effects of the mitochondrial pool of Akt on energy
production, other parameters of mitochondrial fitness could be assayed, such as
respiration mitochondrial membrane potential. These examinations could be expanded to
include modulation of the PI3K pathway in the brain, as well as incubation of isolated
mitochondria with constitutively active Akt. Since mitochondrial membrane potential has
been shown to regulate mitochondrial trafficking, modulation of the PI3K pathway may
alter mitochondrial distribution in neurons. Furthermore, morphology of cMito and pMito
should be examined under basal conditions and following treatment of IGF-1 and/or LY
to confirm results previously obtained using cell culture models (Barksdale 2009).
Overall, these data confirm previous reports that PI3K activity at the plasma
membrane affects levels of Akt within the mitochondria (Bijur 2003). This study was
taken a step further to examine the role of Akt within the mitochondria, and the results
suggest that the mitochondrial pool of Akt plays a role in mitochondrial energy output.
The results obtained from this study open doors for further examinations of mitochondrial
function under various conditions of Akt signaling.
79
References
Ali A, Hoeflich KP, Woodgett JR. (2001) Glycogen synthase kinase-3: properties,
functions, and regulation. Chem Rev 101(8):2527-40.
Arciuch VG, Alippe Y, Carreras MC, Poderoso JJ. (2009) Mitochondrial kinases in cell
signaling: Facts and perspectives. Adv Drug Deliv Rev 61(14):1234-49.
Barksdale KA, Bijur GN. (2009) The basal flux of Akt in the mitochondria is mediated
by heat shock protein 90. J Neurochem 108(5):1289-99.
Bijur GN, Jope RS. (2003) Rapid accumulation of Akt in mitochondria following
phosphatidylinositol 3-kinase activation. J Neurochem 87, 1427-35.
Bruel-Jungerman E, Veyrac A, Dufour F, Horwood J, Laroche S, Davis S. (2009)
Inhibition of PI3K-Akt signaling blocks exercise-mediated enhancement of adult
neurogenesis and synaptic plasticity in the dentate gyrus. PLoS One 4(11):e7901.
Chada SR, Hollenbeck PJ. (2003) Mitochondrial movement and positioning in axons: the
role of growth factor signaling. J Exp Biol 206(Pt 12):1985-92.
Eto R, Abe M, Hayakawa N, Kato H, Araki T. (2008) Age-related changes of calcineurin
and Akt1/protein kinase Balpha (Akt1/PKBalpha) immunoreactivity in the mouse
hippocampal CA1 sector: an immunohistochemical study. Metab Brain Dis
23(4):399-409.
Hanger DP, Hughes K, Woodgett JR, Brion JP, Anderton BH (1992) Glycogen synthase
kinase-3 induces Alzheimer's disease-like phosphorylation of τ: generation of
paired helical filament epitopes and neuronal localisation of the kinase. Neurosci
Lett 147: 58-62.
Hollenbeck PJ. (2005) Mitochondria and neurotransmission: evacuating the synapse.
Neuron 47(3):331-3.
Huang TJ, Verkhratsky A, Fernyhough P. (2005) Insulin enhances mitochondrial inner
membrane potential and increases ATP levels through phosphoinositide 3-kinase
in adult sensory neurons. Mol Cell Neurosci 28(1):42-54.
Kennedy SG, Kandel ES, Cross TK, Hay N. (1999) Akt/Protein kinase B inhibits cell
death by preventing the release of cytochrome c from mitochondria. Mol Cell Biol
19(8):5800-10.
King TD, Bijur GN, Jope RS (2001) Caspase-3 activation induced by inhibition of
mitochondrial complex I is facilitated by glycogen synthase kinase-3β and
attenuated by lithium. Brain Res 919: 106-114.
80
Klein PS, Melton DA (1996) A molecular mechanism for the effect of lithium on
development. Proc Natl Acad Sci U S A 93: 8455-8459.
Kozlovsky N, Belmaker RH, Agam G (2000) Low GSK-3β immunoreactivity in
postmortem frontal cortex of schizophrenic patients. American Journal of
Psychiatry 157: 831-833.
Miller KE, Sheetz MP.(2004) Axonal mitochondrial transport and potential are correlated.
J Cell Sci 117(Pt 13):2791-804.
Peineau S, Taghibiglou C, Bradley C, Wong TP, Liu L, Lu J, Lo E, Wu D, Saule E,
Bouschet T, Matthews P, Isaac JT, Bortolotto ZA, Wang YT, Collingridge GL.
(2007) LTP inhibits LTD in the hippocampus via regulation of GSK3beta. Neuron
53(5):703-17.
Puche JE, García-Fernández M, Muntané J, Rioja J, González-Barón S, Castilla Cortazar
I. (2008) Low doses of insulin-like growth factor-I induce mitochondrial
protection in aging rats. Endocrinology 149(5):2620-7.
Stiles BL. (2009) PI-3-K and AKT: Onto the mitochondria. Adv Drug Deliv Rev
61(14):1276-82.
Stump CS, Short KR, Bigelow ML, Schimke JM, Nair KS. (2003) Effect of insulin on
human skeletal muscle mitochondrial ATP production, protein synthesis, and
mRNA tanscripts. Proc Natl Acad Sci U S A 100(13):7996-8001.
Tong JJ. (2007) Mitochondrial delivery is essential for synaptic potentiation. Biol Bull
212(2):169-75.
Unterluggauer H, Hütter E, Viertler HP, Jansen-Dürr P. (2008) Insulin-like growth
factor-induced signals activate mitochondrial respiration. Biotechnol J 3(6):813-6.
Verburg J, Hollenbeck PJ. (2008) Mitochondrial membrane potential in axons increases
with local nerve growth factor or semaphorin signaling. J Neurosci 28(33):830615.
Yan D, Guo L, Wang Y. (2006) Requirement of dendritic Akt degradation by the
ubiquitin-proteasome system for neuronal polarity. J Cell Biol 174(3):415-24.
Znamensky V, Akama KT, McEwen BS, Milner TA. (2003) Estrogen levels regulate the
subcellular distribution of phosphorylated Akt in hippocampal CA1 dendrites. J
Neurosci. 23(6):2340-7.
81
A
pSer473-Akt
Akt
Blocking peptide
-
+
B
-
+
p-Ser473 Akt
Akt
Cortex
Scale Bar: 100 μm
Scale Bar: 50 μm
Cortex
Akt
C
82
Fig 1. Localization of Akt and pSer473 Akt in mouse cortex. (a) Immunoblot
confirmation of Akt and pAkt antibody specificity was performed in cortical
homogenates from adult male mice. (b) Coronal sections were taken from adult male
mouse brain and stained for Akt and pSer473 Akt followed by DAB incubation. Left
panels: Scale bar =100 μm. Right panels: scale bar = 50 μm. (c) Coronal sections from
mouse cortex were stained for Akt followed by electron microscopy. Closed arrows
indicate mitochondria which are not labeled for Akt. Open arrows indicate mitochondria
in cross sections of neuronal processes which are labeled for Akt.
83
B
pMito
cMito
120
Mitochondrial protein levels
(% of pMito)
A
100
Akt
pSer473-Akt
Hsp90
ATP Synthase
80
*
60
40
*
20
0
pMito
cMito
Akt
C
(Human)
pMito
pMito
cMito
HSP90
cMito
Akt
ATP Synthase
Fig 2. Confirmation of Akt levels in two populations of mitochondria. (a)
Immunoblotting of pMito and cMito populations to detect levels of Akt, pAkt, and
HSP90. ATP Synthase was immunoblotted as a loading control. (b) Quantitation of
immunoblots for Akt and HSP90. Akt and HSP90 bands were quantitated as “% of
pMito” levels, n=4, *p<0.05, Student’s two-tailed T-test. (c) pMito and cMito fractions
isolated from human cortex were immunoblotted for Akt levels to confirm results in
obtained in mouse models. ATP synthase was used as a loading control.
84
A
B
400
pSer473 Akt and total Akt
(% of control)
pMito
pSer473-Akt
Akt
ATP Synthase
Saline
IGF-1
*
350
300
250
200
150
100
50
0
C
(Overexposed)
pMito
Control
IGF-1
pSer473-Akt
Control
IGF-1
Total Akt
pMito
D
cMito
pSer473-Akt
pSer473-Akt
Akt
Akt
Saline
IGF-1
+
+
-
ATP Synthase
+
DMSO
160
pSer473 Akt and total Akt
(% of control)
E
+
-
140
120
100
80
60
*
40
20
0
Control
LY
pSer473-Akt
Control
LY
Total Akt
85
LY
Fig 3. Modulation of mitochondrial Akt levels by activating or inhibiting PI3K signaling
in the brain. (a) Adult male mice were ICV injected with IGF-1, then pMito were isolated
and immunoblotted for pAkt and Akt levels. (b) Quantitiation of pAkt and Akt levels in
pMito following ICV injection of IGF-1. Akt and pAkt bands were quantified using a
densitometer as “% of control”, n=6, *p<0.05, Student’s two tailed T-test. (c)
Comparison of Akt and pSer473Akt levels in pMito and cMito under control and IGF-1
stimulated conditions following ICV injection. The overexposed immunoblot indicates
very little Akt and no pAkt present in cMito, even under stimulated conditions. (d) Adult
male mice were ICV injected with LY294002, then pMito were isolated and
immunoblotted for pAkt and Akt levels. (e) Quantitation of pAkt and Akt levels in pMito
following ICV injection of LY. Akt and pAkt bands were quantified using a densitometer
as “% of control, n=5, *p<0.05, Student’s two tailed T-test.
86
ICV Treatment
ATP Production Rate
(nmol/min/μg protein)
ATP Production Rate
(% of Control)
cMito
no injection
92.173 +/- 11.19
100%
pMito
no injection
130.92 +/- 3.8*
142%
pMito
0.9 % saline
161.88 +/- 25.69
100%
pMito
50 ng/ml IGF-1
208.39 +/- 30.68*
129%
cMito
0.9 % saline
100%
cMito
50 ng/ml IGF-1
219.92 +/- 48.01
165.96 +/- 26.82
pMito
10% DMSO in saline
pMito
20 μM LY204002
203.74 +/- 31.65
137.17 +/- 22.90*
100%
67%
pMito
0.9% Saline (MG/CCCP)
50.31 +/- 7.11#
pMito
50 ng/ml IGF-1 (MG/CCCP)
Mitochondrial Population
47.31 +/-
7.91#
75%, NS
31%
22%
Table 1. ATP production rates in cMito and pMito under basal and stimulated conditions.
pMito and cMito were isolated from mouse cortex and tested for ATP production rates.
In addition, mice were ICV injected with either IGF-1 or LY204002, followed by ATP
production assays in isolated mitochondria. ATP production rates are given as raw values
+/- SEM, as well as % change from control. n=3 for basal (no injection) and n=6 for all
other conditions, *p<0.05, Student’s two tailed T-test. #p<0.05, Student’s two tailed Ttest.
87
MITOCHONDRIAL VIABILITY IN MOUSE AND HUMAN POSTMORTEM
BRAIN
by
Keri A. Barksdale, Emma Perez-Costas, Johanna C. Gandy, Miguel Melendez-Ferro,
Rosalinda C. Roberts, and Gautam N. Bijur
In Revision
FASEB Journal
Format adapted for dissertation
88
ABSTRACT
Neuronal function in the brain requires energy in the form of ATP, and
mitochondria are canonically associated with ATP production in neurons. The
electrochemical gradient which underlies the mitochondrial transmembrane potential
(ΔΨmem) is harnessed for ATP generation. Here we show that ΔΨmem and ATP
production can be engaged in mitochondria isolated from human brains up to 8.5 hours
postmortem. Also, a time course of postmortem intervals from 0 hours to 24 hours using
mitochondria isolated from mouse cortex reveals that ΔΨmem in mitochondria can be
reconstituted beyond 10 hours postmortem. It was found that complex-I of the
mitochondrial electron transport chain was adversely affected with increasing
postmortem interval. Mitochondria isolated from postmortem mouse brains maintain the
ability to produce ATP, but rates of production decreased with longer postmortem
intervals. Furthermore, we show that postmortem brain mitochondria retain their ΔΨmem
and ATP production capacities following cryopreservation. Our findings that ΔΨmem
and ATP generating capacity can be reinitiated in brain mitochondria hours after death
indicates that human postmortem brains can be an abundant source of viable
mitochondria to study metabolic processes in health and disease, and it is also possible to
archive these mitochondria for future studies.
89
INTRODUCTION
Unmitigated ATP supply is important in many tissues, and is especially important for
the brain. The homeostatic maintenance of the ionic gradient by neuronal membranes,
propagation of the neuronal action potential, and synaptic transmission require a readily
available stock of ATP, in addition to energy required for the other internal functions of
the neurons. Mitochondria are classically known for ATP production, and they are the
major source of ATP for neurons. It is widely accepted that detrimental alterations in
mitochondrial function can negatively affect neuronal function, and it has been shown
that mitochondria play a key role in neuronal plasticity and death (1). Numerous studies
have reported the occurrence of altered or decreased mitochondrial function in the aging
brain, in neurodegenerative diseases such as Alzheimer’s, Parkinson’s, and Huntington’s
disease, in psychiatric disorders such as schizophrenia and bipolar disorder (2, 3, 4, 5, 6),
and in neurometabolic disorders such as Leigh syndrome (7). Thus, it is clear that
mitochondria play a fundamental role in normal brain function and in brain disease
processes.
To understand the role of mitochondria in the central nervous system, direct
assessment of mitochondrial activity is typically necessary. However, one of the vexing
problems of assessing mitochondrial function is that many functional assays require
abundant amounts of intact mitochondria. Animal models are commonly used for the
study of mitochondrial function and dysfunction in the central nervous system because
they can be manipulated genetically and pharmacologically, and also because they are an
abundant source of mitochondrial preparations. However, given that there are obvious
species-specific variations in metabolic activity, mitochondria from laboratory animals
90
cannot accurately reflect the functional or structural status of mitochondria in the human
central nervous system. To assess mitochondrial activity in human brain, investigators
have previously utilized frozen postmortem brain tissues to analyze mitochondrial
enzymatic activities (8), mitochondrial protein levels (9), and mitochondrial DNA (10, 3).
However, there are several other vital indices that could provide insight into brain
mitochondrial activity which cannot be conducted in frozen tissue samples. These include
measurements of the mitochondrial membrane potential (ΔΨmem), ATP production,
calcium buffering capacity, and respiration, which together give an overall assessment of
mitochondrial health and activity. For example, the ΔΨmem, which is the
electrochemical gradient across the inner mitochondrial membrane, serves as an
important overall indicator of mitochondrial activity. It is also a fundamental component
of respiring mitochondria. The ΔΨmem is linked to many crucial mitochondrial functions
including ATP synthesis, calcium homeostasis, mitochondrial protein import, and
mitochondrial metabolite transport (11) all of which are typically analyzed in real-time
measurements. However, these analyses can only be conducted in structurally intact and
functional mitochondria. Furthermore, if these measurements are undertaken in isolated
mitochondria, large quantities of relatively pure mitochondrial preparations are required.
The need for mitochondrial purity, structural integrity, and abundance for functional
studies are common barriers which can encumber brain mitochondrial research. An
additional hindrance for research on human brain mitochondria is the extremely limited
source of human brain tissues in general.
Postmortem human brains, when available, are potentially an abundant source of
mitochondrial material. However, it is generally thought that direct measurements of
91
mitochondrial activity, such as the ΔΨmem and ATP production are not possible in
postmortem brain samples. This is because of the common belief that most organelle
functions cease after death, and in fact, with increasing postmortem interval (PMI), there
is a rapid deterioration of the brain tissue. In this study, we report that postmortem mouse
and human brains with relatively short PMIs are an abundant source of structurally intact
and functional mitochondria, and we report that mitochondria can be stored frozen for
functional measurements at a later time.
MATERIALS AND METHODS
C57/Bl6 mice were obtained commercially. All animal experiments were conducted
in accordance with the University of Alabama at Birmingham Institutional Animal Care
and Use Committee (IACUC) approved protocols. Postmortem human tissue was
obtained from the Alabama Brain Collection (ABC, Department of Psychiatry and
Behavioral Neurobiology, University of Alabama at Birmingham), and human brain
tissues were obtained in accordance with the University of Alabama at Birmingham
Institutional Review Board (IRB). All tissue donations were obtained through the
Alabama Organ Center in compliance with the Alabama uniform anatomical gift act.
Mouse postmortem time course
To test the effect of postmortem interval on mitochondrial viability, mice were
euthanized by cervical dislocation and the bodies were stored at 4oC for 0 (control
animals), 10, 18, and 24 hours after death. Brains were dissected and the mouse brain
cortices were used for mitochondrial isolations.
92
Postmortem human tissue
Postmortem human brain samples were obtained from the Alabama Brain Collection
in collaboration with the Alabama Organ Center. Consent was obtained from the next of
kin. Serology was performed for hepatitis A, B or C, HIV, cytomegalovirus and syphilis.
One case was positive for cytomegalovirus (case S2). Brain tissue from the prefrontal
cortex (Brodmann Area 10, BA10) was obtained from 3 different cases with postmortem
intervals below 10 hours (Table 1). For each case, approximately 5 cubic centimeters of
fresh tissue was placed in isolation buffer (215 mM mannitol [Acros Organics, NJ] 75
mM sucrose [Roche Diagnostics, Indianapolis, IN], 0.1% fatty acids-free bovine serum
albumin (FAF-BSA) [Sigma Aldrich, St. Louis, MO], 20 mM HEPES [Sigma Aldrich], 1
mM ethylene glycol tetraacetic acid (EGTA), adjusted to pH 7.2 with 5M KOH)
immediately after dissection for further processing.
Isolation of Mitochondrial Populations from Neuronal Processes and Cell Bodies
Methods for this isolation were adapted from Brown et al., 2006 (4). Tissues collected
from either mouse or human were dissociated by Dounce homogenization in ice-cold
isolation buffer. The same amount of ice-cold 30% Percoll (GE Healthcare Life Sciences,
Uppsala, Sweden) in isolation buffer was added (about 2X volume) for a final 15%
Percoll solution in the sample. The Percoll solutions were prepared in isolation buffer and
filtered to remove contaminants. Centrifuge tubes were layered from bottom to top, first
with 40% Percoll, followed by 24% Percoll, and finally the 15% Percoll mitochondrial
sample. Samples were centrifuged in a Sorvall RC-5C Plus centrifuge using a fixed angle
SS34 rotor at 34,500 x g for 30 min at 4ºC. After centrifugation three hazy bands were
formed in the Percoll. The second and third bands from the top [synaptosomes and cell
93
bodies, respectively] were removed from the Percoll media and placed in separate tubes,
and 10 ml of isolation buffer was added to wash the Percoll from the samples. Samples
were centrifuged at 16,700 x g for 15 min. The loose pellets obtained were then
resuspended in isolation buffer and placed in a nitrogen cavitation bomb (Parr Instrument
Company, Moline, IL). The samples were subjected to nitrogen cavitation for further
homogenization (800 p.s.i. for 10 min for mouse samples and 1000 p.s.i. for 10 min for
human samples). The mitochondria were then separated into two general populations:
Mitochondria from cellular processes, designated pMito, are mitochondria that exist
predominantly in the axons, neurites, and the synapses. Mitochondria in the cell soma,
designated cMito, are mitochondria which are mostly resident in the perinuclear area and
in the cell bodies.
To purify the pMito fraction, the homogenates were mixed with an equal volume of
10% Percoll in isolation buffer. The samples were layered over a 20% and a 12% Percoll
cushions and centrifuged at 34,500 x g for 30 min at 4oC. Following centrifugation, the
bottom band containing pMito was collected. Isolation buffer (10 ml) was added and the
sample was centrifuged at 16,700 x g for 15 min. The loose pellet was collected, washed
once with isolation buffer and centrifuged for 10 min at 17,600 x g in an Eppendorf
5417R tabletop centrifuge at 4ºC. To purify the cMito following nitrogen cavitation,
samples were centrifuged for 10 min at 750 x g in an Eppendorf 5417R tabletop
centrifuge at 4ºC. The supernatant was retained and centrifuged for 15 min at 17,600 x g,
then the pellet was collected and washed once with isolation buffer and centrifuged for 10
min at 17,600 x g. Mitochondria of both types (pMito and cMito) were kept pelleted on
ice for a maximum of 40 min until used for the assays.
94
Mitochondrial Cryopreservation
Mitochondria isolated from mouse cortices at 0 and 10 hours postmortem were used.
Previous reports have shown that mitochondria isolated from brain tissue can maintain
structural and functional integrity after being frozen in a buffer containing 20% dimethyl
sulfoxide (DMSO) and 1% FAF-BSA (1). pMito and cMito were resuspended in isolation
buffer containing sterile 20% DMSO (Research Organics, Cleveland, OH) and 1% FAFBSA in Nalgene Cryoware cryogenic vials (Nalge Nunc International, Rochester, NY).
Vials were then placed in a foam-insulated thermofreezing container with an approximate
cooling rate of 1ºC per min (Aladdin Industries Incorporated, Nashville, TN), and the
container was placed at -80oC for seven days. Mitochondrial suspensions were then
thawed at 4oC, washed with isolation buffer, and prepared for functional studies (ΔΨmem
and ATP production) and assessment of mitochondrial integrity and morphology by
transmission electron microscopy.
Mitochondria functional assays
Rhodamine 123 Assay for Analysis of ΔΨmem: Mitochondria were washed with
isolation buffer containing no BSA, and then assayed for protein concentration using the
bicinchoninic acid assay (Pierce, Rockford, IL). During the protein assay incubation,
mitochondria were pelleted (2 min centrifugation at 17,600 x g) in isolation buffer to
avoid exposure to air. Both cMito and pMito were tested for ΔΨmem. Mitochondria (60
μg) were resuspended in 200 μl isolation buffer containing 0.1% BSA and added to a 96well clear-bottom plate. Rhodamine 123 (Rh123, 300 nM) was added to each sample just
before reading. Samples were read in a Biotek Synergy 2 spectrofluorometer at 25oC.
Readings were taken every 45 seconds at 485 ± 20 nm excitation and 528 ± 20 nm
95
emission wavelengths. Baseline readings were measured for 5 min before the addition of
any substrates or treatments. Malate (7 mM), glutamate (7 mM), and 100 µM ADP were
added and fluorescence was read for 10 min. The mitochondrial complex I inhibitor
rotenone (5 µM) was added, and samples were monitored for 10 minutes followed by the
addition of succinate (7 mM) and ADP (100 µM) and another 10 min of fluorescence
readings. Finally, carbonyl cyanide m-chlorophenylhydrazone (CCCP) (50 µM) was
added and fluorescence was monitored for the last 10 minutes. Water or DMSO were
used as vehicle controls.
ATP Production Assay: pMito were assessed for ATP production. After determining
protein concentration, 250 µg aliquots of pMito were pelleted and placed on ice until
ready for use. For the ATP production assay, pellets were resuspended in 185 µl
respiration buffer [20 mM HEPES-KOH pH 7.4, 80 mM potassium acetate, 5 mM
magnesium acetate, 250 mM sucrose, 1 mM glutamate, and 1 mM malate]. Buffer B [0.5
M Tris-Acetate pH 7.75, 0.8 mM D-Luciferin, 20 μg/ml luciferase] (10 µl) was
immediately added to the reaction, and sample luminescence was read in standard mode
on a Turner 20/20 luminometer for 4 sec with no delay. ADP (0.1 mM) was then added to
the sample, and the luminescence was read every 4 sec for a total of 60 sec. Buffer
background measurements were taken using buffers containing all substrates, and these
measurements were subtracted from the mitochondrial readings. ATP standards were
read using the same conditions and used to extrapolate ATP generation rates for each
sample.
96
Transmission Electron Microscopy
Isolated mitochondria were pelleted and resuspended in cold fixative [4%
paraformaldehyde and 0.1% glutaraldehyde in 0.1M phosphate buffer pH 7.4 (PB)].
Mitochondria were pelleted using centrifugation (17,600 x g for 5 minutes), and fresh
fixative was added to the pellet. The mitochondrial pellets were maintained in fixative for
12 hours at 4oC. Pellets were then washed thoroughly with filtered PB, postfixed for 1
hour with 1% osmium tetroxide in PB, rinsed with PB, and dehydrated in 50% and 70%
ethanol baths. A 2% uranyl acetate in 70% ethanol solution was applied for contrast (1
hour at 25oC in the dark) followed by 2 rinses in 70% ethanol (5 min each). Dehydration
in ethanol was completed with 2 steps of 5 min in 95% and 100% ethanol baths. Then,
100% propylene oxide (Electron Microscope Sciences, Fort Washington, PA) was
applied to the samples (2 baths of 10 min each). Samples were then placed gradually in a
1:1 mixture of propylene oxide and Epon resin (25.9% Embed 812 + 15.5% Araldite 502
+ 55.9% DDSA + 2.7% BDMA, all resin components from Electron Microscopy
Sciences) for 1h at 25oC, a 1:2 mixture of propylene oxide-Epon resin (1h at 25oC), and
finally transferred to pure Epon resin and kept overnight at 25oC. The following day the
samples were placed in freshly prepared resin for an additional hour at 25oC and finally
allowed to polymerize in an oven at 70ºC for a minimum of 4 days. Ultrathin sections (90
nm thick) were obtained using a Leica EM UC6 ultramicrotome (Leica Microsystems,
Wetzlar, Germany), mounted on formvar-coated copper grids and observed and
photographed using a Hitachi transmission electron microscope, model H7650 (Hitachi,
Japan), equipped with a Hamamatsu ORCA-HR digital camera (Bridgewater, NJ, USA).
97
Figure plates were prepared using Corel PhotoPaint 12 and CorelDraw 12 (Corel, Ottawa,
Canada).
RESULTS
The presence of a robust ΔΨmem in postmortem mouse cortex
The goal of this study was to examine if mitochondria maintain their integrity and
activity subsequent to the death of an individual. To test if functional mitochondria could
be isolated from postmortem brain, adult male C57/Bl6 mice were euthanized by cervical
dislocation and the intact bodies were placed at 4oC for increasing lengths of time (0, 10,
18, and 24 hours) to test the possible time-dependent degradation of brain mitochondria.
These time points were used because the typical PMI for human postmortem tissues is
beyond 8 hours, although on very rare occasions human brain tissues with shorter PMIs
are available. Furthermore, this mode of death and body storage was used to mimic a
generalized scenario of human death and subsequent body handling. Mitochondria from
the postmortem mice were compared to mitochondria from “0 hours” animals in which
the mice were euthanized and cortical mitochondria were isolated immediately after
death. In the initial time-course study, cell body mitochondria (cMito) were isolated from
mouse cortex. The potentiometric dye Rh123 was used to measure ΔΨmem. Rh123 is
electrophoretically transported into intact mitochondria and sequestered within the
organelle due to the charge difference across the mitochondrial inner membrane (12, 13).
Non-respiring mitochondria and mitochondria lacking structural integrity cannot
accumulate Rh123 (13). Thus, when the mitochondrial inner membrane is intact and
polarized, Rh123 is sequestered within mitochondria and Rh123 fluorescence decreases
98
(13). Upon depolarization, the sequestered Rh123 is released from the mitochondria into
the surrounding buffer, resulting in increased fluorescence (13). A PMI time course was
established to examine how increasing PMI would affect ΔΨmem. Treatment with the
respiratory substrates malate, glutamate, and ADP caused efficient sequestration of
Rh123, resulting in decreased fluorescence in all the time points tested (0 to 24 hours
PMI), and all samples of mitochondria responded equivalently to the addition of
respiratory substrates (Fig. 1A-D). These results indicate that isolated mouse brain
mitochondria are intact and can be induced to hyperpolarize, with respect to baseline
readings, for at least 24 hours after death.
Manipulation of the ΔΨmem with respiratory toxins provides an additional
assessment of mitochondrial integrity and viability. The addition of the complex I
inhibitor rotenone resulted in the rapid depolarization of mitochondria, indicated by the
upward shift in Rh123 fluorescence across all time points (Fig. 1A-D), but depolarization
of the 10, 18 and 24 hours PMI mitochondria was clearly more robust than the 0 hour
mitochondria. Bypass of complex I by the addition of succinate (which activates complex
II) resulted in complete restoration of the ΔΨmem at all the PMI time points (Fig. 1A-D).
The addition of the protonophore CCCP equivalently dissipated the ΔΨmem for all the
PMI time points (Fig. 1A-D). At PMIs greater than 10 hours, the brain architecture and
the tissues overall were substantially degraded, and there was an approximate 70%
reduction in mitochondrial yield. Therefore, for the remainder of the studies, the 0 and 10
hours PMI were tested.
The next goal was to test if there was any statistical variation between 0 and 10 hours
PMI. The only variation between the two time points occurred in response to rotenone
99
treatments, whereby there was a significantly increased rotenone-induced membrane
depolarization at the 10 hours PMI (Fig. 1E-F) Thus, complex I may have acquired more
sensitivity to rotenone with increasing PMIs, but the adjacent complex II was unaffected.
Dimethyl sulfoxide (DMSO) was used as a vehicle control for rotenone and CCCP, but
no effect of DMSO on ΔΨmem was evident (Fig. 1E-F). The fluorescence of the
hyperpolarized mitochondria from the 10 hour PMI remained stable and constant
throughout the 45-minute time frame (Fig. 1F, open squares). Thus, there was no leakage
of Rh123 from the mitochondria, indicating that the mitochondrial membranes remained
fully intact. This result indicates that postmortem mouse brain cortical mitochondria can
be induced to engage their ΔΨmem 24 hours postmortem.
In order to get a visual confirmation of mitochondrial structural integrity,
mitochondrial pellets from the 0 and 10 hours PMI were visualized by electron
microscopy. In Figures 1G and 1I, it is evident that the mitochondrial fraction is mostly
free from non-mitochondrial debris, which could occlude accurate measurements of
mitochondrial function. Higher magnification images (Fig. 1H, J) illustrate that most of
the mitochondria in the 0 and 10 hours PMI samples appear to be intact and maintain
their orthodox configuration, which has been shown to be indicative of mitochondria with
normal respiration, energy production, and cell survival signaling (14, 15, 16). These data
together indicate that mitochondria isolated from mouse postmortem brains are
structurally intact, as well as functionally viable.
Mitochondrial ΔΨmem and ATP production in postmortem human cortex
The next objective was to extend these observations to the human brain, and to
examine if ΔΨmem in mitochondria isolated from postmortem human cortex also
100
responded to respiratory substrates and toxins. Human cortical tissue was extracted from
BA10 from three deceased individuals with no documented history of neurological or
psychiatric disorders (Table I). In all three cases the PMI was below 10 hours and the pH
was between 6.3 and 6.4. Mitochondria from cell bodies (cMito) and neuronal processes
(pMito) were isolated as described in Materials and Methods and tested for ΔΨmem. In
Subject 1 (S1) the pMito and cMito hyperpolarized in response to the respiratory
substrates (Fig. 2A-B), and the pMito robustly depolarized in response to treatments with
rotenone and CCCP (Fig. 2A). By comparison, upon the addition of rotenone, the cMito
from S1 maximally depolarized, and were then refractory to further succinate and CCCP
treatments (Fig. 2B), suggesting that complex II in this cMito sample was impaired. The
pMito from Subject 2 (S2) also displayed the characteristic responses to substrates and
toxins (Fig. 2C), but the cMito from this same individual showed a noticeably mitigated
response to the stimuli (Fig. 2D) compared to the matched pMito. In contrast, pMito and
cMito isolated from Subject 3 (S3), were neither responsive, nor were there correlative
responses in their ΔΨmem to malate/glutamate, rotenone, and succinate (Fig. 2E-F).
However, ΔΨmem was still amenable to robust depolarization with CCCP. Since the
mitochondrial sample from S3 displayed an aberrant ΔΨmem profile, we wanted to test
further if these mitochondria were metabolically active. ATP production was measured in
comparison to S2 mitochondria, which displayed the more characteristic ΔΨmem profile
in our studies and were isolated from an equivalent age (77 years for S2 and 70 years for
S3) and PMI brain sample (8.25 hours for S2 and 8.5 hours for S3). Since the
mitochondria from S3 did not hyperpolarize in response to glutamate/malate, it was
interesting to find that there was ample ATP production capacity in response to these
101
substrates, and there was only 20% less ATP production than those from S2 (Fig. 2G).
Furthermore, ATP production in S3 mitochondria was inhibited by 87.5% by the addition
of CCCP (Fig. 2G). This suggests that the mitochondria from S3 retain their ATPgenerating capability despite the lack of substrate-induced changes in the ΔΨmem.
Mitochondrial integrity from all the human samples was also verified visually. The
electron micrographs (Fig. 3A-H) show that the isolated human postmortem brain
mitochondria are predominantly intact and retain much of their orthodox internal
morphology. Overall, these data show for the first time that mitochondria in postmortem
human brain remain intact and metabolically active for at least 8.5 hours following the
death of the individual.
Cryogenic preservation of postmortem mitochondria minimally affects ΔΨmem and
bolsters ATP production
Cryogenic storage of freshly isolated mitochondria has been reported previously (1,
17, 18). The purpose here was to test if mitochondria from postmortem brains could also
withstand cryogenic storage conditions. This would be an important assessment since
there are occasions where comparative studies of mitochondrial functional indices must
be measured simultaneously under standardized conditions. Given the sporadic nature of
human tissue procurement, these types of studies in fresh samples would be impossible.
For these studies, cMito and pMito isolated from mouse cortex 0 and 10 hours after
death were stored as described in the Materials and Methods. After 7 days of storage,
mitochondrial preparations were thawed on ice and tested for ΔΨmem and ATP
production, and the stored mitochondria were examined for structural changes. Cryogenic
storage did not appear to cause any alterations in the ΔΨmem profile of the cMito and
102
pMito in response to respiratory substrates and toxins (Fig. 4A-D). Furthermore, the
ΔΨmem profile of the frozen cMito was not significantly altered when compared to
freshly isolated cMito (Fig. 4E), although there was a slight, but consistent, potentiation
in the responses of the frozen mitochondria to the respiratory substrates and to rotenone.
For extended storage, future studies will be necessary to determine the maximum length
of time cryogenic storage that is feasible without loss of mitochondrial integrity.
However, the present findings show that cryogenic storage of postmortem brain
mitochondria for at least 7 days does not cause overt damage to mitochondrial membrane
integrity (Fig. 4F-G).
To assess if the postmortem interval combined with cryogenic storage can affect
metabolic output, ATP production rate was also measured in pMito. The rate of ATP
production was significantly decreased, by 56%, in the freshly-isolated 10 hours PMI
mitochondria, compared to the 0 hours freshly-isolated mitochondria (Table 2), indicating
that the postmortem interval does cause a decrement in mitochondrial ATP output.
Subsequent to cryogenic storage it was interesting to find that the mitochondria had an
augmented rate of ATP production, with and without CCCP treatment in both 0 and 10
hours PMI mitochondria. However, the stored 10 hours PMI mitochondria had a
proportional, 57.9% reduction in the rate of ATP production compared to the 0 hours
stored mitochondria (Table 2), as was found in the freshly-isolated samples. This
indicates that cryogenic conditions did not alter the percent reduction in ATP output from
the 10 hours PMI mitochondria. These results highlight the fact that isolated
mitochondria can be cryogenically archived and then reanimated at a later time. However,
103
studies of comparative mitochondrial ATP measurements between different postmortem
mitochondrial samples must be matched for PMI and storage conditions.
DISCUSSION
Moments after death, the brain is placed in a condition of asphyxia, and tissues are
deprived of oxygen which is required as the final electron acceptor at the terminus of the
electron transport chain. As a consequence, mitochondrial ATP generation ceases.
Among the early ramifications due to the loss of ATP production is the destruction of
cellular membranes, because phospholipids in the membranes become rapidly oxidized
and are not replaced by the ATP-dependent processes of membrane repair and renewal.
The loss of cell membrane integrity disgorges normally sequestered hydrolytic enzymes
and the process of cell and organelle degradation commences. These are some of the
events that are known to occur after death (19). During the process of postmortem
cellular breakdown, there appears to be a window of time in which ΔΨmem and ATP
production can be re-engaged in brain mitochondria isolated from deceased individuals.
This is a valuable finding, given that there is limited availability of human mitochondrial
samples from brain tissue for research purposes. Direct, real-time examination of human
brain mitochondrial function is very difficult due to the limited availability of human
brain tissues. We now report that it is possible to isolate copious amounts of intact and
functional mitochondria from human postmortem brain samples for direct investigation
into human brain mitochondrial activity and for archival purposes.
Previously it was unknown how differences in PMI would affect mitochondrial
integrity and function. An initial time course of PMIs in mouse revealed that the ΔΨmem
104
in brain mitochondria can be reconstituted for at least 24 hours after death. However, an
increased PMI also resulted in the rapid degradation of brain tissues, and the overall yield
of intact mitochondria was substantially reduced at 18 and 24 hours PMI. Nonetheless, it
was interesting to find that mitochondrial activity could be reconstituted from
postmortem mouse brains stored in the intact mouse bodies under controlled conditions.
It was also noteworthy that the two different populations of mitochondria in this study,
cMito and pMito, both displayed similar ΔΨmem profiles at 0 and 10 hours PMI, and
equivalently withstood cryopreservation. Mitochondria are not a homogenous population
in the brain, and mitochondria in neuronal processes (pMito) and mitochondria in the cell
soma (cMito), have been shown to have functionally distinct characteristics (20, 4, 21).
However, the findings here suggest that PMI and cryopreservation have little or no
differential effects on ΔΨmem between cMito and pMito.
In contrast to the standardized mouse husbandry and controlled storage conditions of
mouse tissues, which yields mitochondria of consistent quality, the condition of
mitochondria from human postmortem brain tissues is far more variable. This is because
the premortem state of the individual, including age, drug use, and possible CNS
infections, as well as cause of death, storage condition of the body, tissue pH, and PMI,
altogether impact the overall yield and quality of human brain mitochondria. The human
brain tissues used in this study were all below the 10 hours PMI, however, there were
marked inter-individual and inter-mitochondrial (cMito versus pMito) variations, as each
of the different mitochondrial samples displayed uniquely distinct ΔΨmem profiles,
which was not evident in the mouse brain mitochondrial samples. One of the interindividual differences between samples was age. Samples from S1, which was 32 years
105
old, responded most robustly to the different treatments, while the other two samples
(from S2 and S3, which were around 70 years old) produced more variable results. This
likely reflects the inherent complexity of working with heterogeneous human samples; a
problem which can be overcome by the inclusion of greater sample numbers. However,
the finding presented here with the three human samples clearly shows the feasibility of
using human postmortem brains as an abundant source of metabolically active
mitochondria for research purposes.
The remarkable resiliency of the mitochondrion is likely due to the combination of its
enclosed asymmetric double membrane barrier, the dissimilar composition of the
mitochondrial membranes, and the intricate crisscrossing cristae which together scaffold
the mitochondrial structure and protect it from the normal tissue degradation processes.
The resistance of mitochondria to proteolytic degradation is further illustrated by findings
from mitochondria isolated from cultured cells (22, 23). When these mitochondria were
incubated in a 30oC water bath in the presence of the proteolytic enzymes trypsin or
proteinase K, the proteins that were exposed on the cytosolic face of the outer
mitochondrial membrane were rapidly digested, but the proteins encapsulated within the
mitochondrial membranes were unaffected (22, 23). This suggests that the double
mitochondrial membrane is mostly impervious to proteolytic enzymes, thereby insulating
the internal contents of the mitochondria from the digestive milieu. Therefore, in the
dying brain tissues, mitochondria may be able to withstand the early degradation
processes that typically destroy other cellular structures.
One of the beneficial consequences of utilizing human postmortem brains is the
abundant yield of mitochondria. The storage of mitochondria for archival purposes and
106
subsequent analysis liberates the researcher from the long and tedious process of
analyzing mitochondrial functions one sample at a time. Instead, with appropriate storage,
it may be possible to assess mitochondrial activity, for example ΔΨmem and ATP
production, in a rapid and standardized process. In the present study, cryogenic storage of
mouse brain mitochondria for at least one week clearly did not affect their ΔΨmem
profile. The concept of cryogenic storage of mitochondria dates back to 1961 (24, 25).
More recent studies have confirmed that mitochondria within rat brain synaptosomes and
isolated rat brain mitochondria can be reanimated following cryopreservation (26, 1). The
most important consideration of mitochondrial cryopreservation is the addition of a
cryoprotective agent, because freezing in its absence destroys mitochondrial structure and
function (1). Numerous cryoprotectants are available, but the most commonly used are
DMSO and glycerol. A comparative study using these two cryoprotectants revealed that
DMSO is far superior in preserving mitochondrial integrity and function upon freezethawing than glycerol (17). In our study it was interesting to observe that ΔΨmem not
only was unaffected in the DMSO cryopreserved mitochondrial samples, but curiously
there was a substantial enhancement of ATP production for pMito cryogenically stored,
both after 0 and 10 hours PMI. To our knowledge no one has reported ATP production
measurements
from
cryopreserved
brain
mitochondria;
however,
respiration
measurements have revealed decreased respiratory control ratios in cryopreserved rat
brain mitochondria (1). The reason for the enhancement of mitochondrial ATP
production following cryopreservation is unclear, although there is some evidence to
suggest that DMSO can affect the affinity of the F1 subunit of ATP synthase for Pi and
influence ATP formation (27). Although cryopreservation accentuated mitochondrial
107
ATP output in the 0 and 10 hours PMI mitochondria, it did not seem to affect the relative
difference in the rate of ATP production, as in both the freshly-isolated samples and the
cryopreserved samples the 10 hours PMI had 56.2% and 57.9% reduction, respectively,
in ATP production compared to the 0 hours PMI mitochondria. Thus, although ΔΨmem
measurements may be unaffected by cryopreservation, ATP production from
cryopreserved mitochondrial samples must be matched to other cryopreserved samples.
In conclusion, the common perception is that once an individual is deceased all
metabolic activity is irreversibly arrested. Our findings show for the first time that, in
actuality, brain mitochondrial ΔΨmem can be resurrected, and hence mitochondrial
metabolic activity can be effectively reinitiated within at least 10 hours after death. The
ability to access a source of fully functional human brain mitochondria, allows for the
first time the assessment of indices of mitochondrial function that could not be
accomplished previously in human brain tissues. Although the therapeutic implications
are yet to be discovered, these results provide mitochondrial researchers a novel and
additional resource to directly investigate mitochondrial structure and function in the
normal and diseased human brain.
ACKNOWLEDGMENTS
The authors wish to thank Joy K. Roche for assistance in processing of electron
microscopy samples. This work was supported by NIH grant NS044853 to GNB and
MH60744 to RCR. JCG is supported by the Cellular and Molecular Biology Program
training grant GM008111.
108
REFERENCES
1. Nukala VN, Singh IN, Davis LM, Sullivan PG. (2005) Cryopreservation of brain
mitochondria: A novel methodology for functional studies. J Neurosci Methods
152(1-2), 48-54
2. Wang X, et al. (2009) The role of abnormal mitochondrial dynamics in the
pathogenesis of Alzheimer's disease. J Neurochem 109, 153-159
3. Vila M, Ramonet D, Perier C. (2008) Mitochondrial alterations in Parkinson's
disease: new clues. J Neurochem 107(2), 317-28
4. Brown MR, Sullivan PG, Geddes JW. (2006) Synaptic mitochondria are more
susceptible to Ca2+overload than nonsynaptic mitochondria. J Biol Chem 281(17),
11658-68
5. Verkhratsky A, Fernyhough P. (2008) Mitochondrial malfunction and Ca2+
dyshomeostasis drive neuronal pathology in diabetes. Cell Calcium 44(1), 112-22
6. Rollins B et al. (2009) Mitochondrial variants in schizophrenia, bipolar disorder,
and major depressive disorder. PLoS One 4(3), 4913
7. Gerards M et al. (2009) Defective complex I assembly due to C20orf7 mutations
as a new cause of Leigh syndrome. J Med Genet PMID 19542079
8. Devi L, Raghavendran V, Prabhu BM, Avadhani NG, Anandatheerthavarada HK.
(2008) Mitochondrial import and accumulation of -synuclein impair complex I
in human dopaminergic neuronal cultures and Parkinson disease brain. J Biol
Chem 283(14), 9089-100
9. Park LC, et al. (2001) Mitochondrial impairment in the cerebellum of the patients
with progressive supranuclear palsy. J Neurosci Res. 66(5), 1028-34
10. Alam ZI, et al. (1997) Oxidative DNA damage in the parkinsonian brain: an
apparent selective increase in 8-hydroxyguanine levels in substantia nigra. J
Neurochem 69(3), 1196-203
11. Hüttemann M, et al. (2008) Regulation of oxidative phosphorylation, the
mitochondrial membrane potential, and their role in human disease. J Bioenerg
Biomembr 40(5), 445-56
12. Foster KA, Galeffi F, Gerich FJ, Turner DA, Müller M. (2006) Optical and
pharmacological tools to investigate the role of mitochondria during oxidative
stress and neurodegeneration. Prog Neurobiol 79(3):136-71
109
13. Chen LB. Mitochondrial membrane potential in living cells. (1988) Annu Rev Cell
Biol. 4, 155-81
14. Hackenbrock CR. (1966) Ultrastructural bases for metabolically linked
mechanical activity in mitochondria. Reversible ultrastructural changes with
change in metabolic steady state in isolated liver mitochondria. J Cell Biol 30(2),
269-97
15. Rossignol R, et al. (2004) Energy substrate modulates mitochondrial structure and
oxidative capacity in cancer cells. Cancer Res 64(3), 985-93
16. Gottlieb E, Armour SM, Harris MH, Thompson CB. (2003) Mitochondrial
membrane potential regulates matrix configuration and cytochrome c release
during apoptosis. Cell Death Differ 10(6), 709-17
17. Kuznetsov AV, et al. (2003) Cryopreservation of mitochondria and mitochondrial
function in cardiac and skeletal muscle fibers. Anal Biochem 319(2), 296-303
18. Sheleg S, Hixon H, Cohen B, Lowry D, Nedzved M. (2008) Cardiac
Mitochondrial Membrane Stability after Deep Hypothermia using a Xenon
Clathrate Cryostasis Protocol - an Electron Microscopy Study. Int J Clin Exp
Pathol 1(5), 440-7
19. Gill-King, H. (1997) in Forensic Taphonomy: The postmortem fate of human
remains, ed WD Haglund (CRC Press LLC)
20. Dennis SC, Lai JC, Clark JB. (1977) Comparative studies on glutamate
metabolism in synaptic and non-synaptic rat brain mitochondria. Biochem J
164(3), 727-36
21. Naga KK, Sullivan PG, Geddes JW. (2007) High cyclophilin D content of
synaptic mitochondria results in increased vulnerability to permeability transition.
J Neurosci 27(28), 7469-75
22. Bijur GN, Jope RS. (2003) Rapid accumulation of Akt in mitochondria following
phosphatidylinositol 3-kinase activation. J Neurochem 87(6), 1427-35
23. Barksdale KA, Bijur GN. (2009) The basal flux of Akt in the mitochondria is
mediated by heat shock protein 90. J Neurochem 108(5), 1289-99
24. Greiff D, Myers M. Effect of dimethyl sulphoxide on the cryo-tolerance of
mitochondria. (1961) Nature 190, 1202-4
25. Greiff D, Myers M, Privetera CA. (1961) The effects of glycerol, freezing and
storage at low temperatures, and drying by vacuum sublimation on oxidative
phosphorylation by mitochondrial suspensions. Biochim Biophys Acta 50, 233-42
110
26. Begley JG, et al. (1998) Cryopreservation of rat cortical synaptosomes and
analysis of glucose and glutamate transporter activities, and mitochondrial
function. Brain Res Brain Res Protoc. 3(1), 76-82
27. Sakamoto J. (1984) Effect of dimethylsulfoxide on ATP synthesis by
mitochondrial soluble F1-ATPase. J Biochem 96(2), 483-7
111
112
Fig 1. Effect of increasing PMIs on mitochondrial ΔΨmem. Mitochondria from cell
bodies (cMito) were isolated from mouse brain cortex. ΔΨmem results are shown as “%
Initial Reading”, where the baseline Rh123 fluorescence reading is set at 100%. A-D)
ΔΨmem was examined in response to the sequential addition of malate/glutamate/ADP
(MGA), rotenone (ROT), succinate/ADP (S/A), and carbonyl cyanide mchlorophenylhydrazone (CCCP) to the mitochondrial suspensions. Cortical mitochondria
were isolated from tissues at incremental postmortem times. E-F) Summary graphs show
the mean values (n=3) of ΔΨmem in mitochondria isolated from 0 (Figure E) and 10
(Figure F) hours PMI [*p<0.05 compared to values from rotenone-treated samples at 0
hours PMI cMito samples, Student’s t-test. Error bars represent S.E.M]. G-J)
Representative electron micrographs show that the isolated mitochondria are intact and
display characteristic orthodox morphology, at 0 hours (G-H) and 10 hours (I-J) PMI.
Scale bars: 2 microns in G, I; 0.5 microns in H, J.
113
114
Fig 2. Mitochondrial ΔΨmem in human postmortem brain. Mitochondria isolated from
human postmortem brains are intact and responsive to substrates and toxins.
Mitochondria from neuronal processes (pMito) and cell bodies (cMito) were isolated as
described in Materials and Methods. ΔΨmem results are shown as “% Initial Reading”.
A-F) pMito and cMito were isolated from BA10 of three different individuals with
different PMIs ranging from 7.5 to 8.5 hours (A, B: Subject 1; C, D: Subject 2; E, F:
Subject 3). pMito and cMito suspensions were treated sequentially with
malate/glutamate/ADP (MGA), rotenone (ROT), succinate/ADP (S/A), and carbonyl
cyanide m-chlorophenylhydrazone (CCCP). G) Comparison of ATP production by
isolated mitochondria from Subject 2 and Subject 3 after the addition of mitochondrial
substrates malate, glutamate, and ADP. As a control for the assay, the addition of CCCP
to the sample was used to verify that ATP production could be inhibited.
115
Fig 3. Structurally intact mitochondria can be obtained from postmortem human brain. AD) Low and high magnification images of isolated mitochondria from case S1. E-H) Low
and high magnification images of isolated mitochondria from case S2. Note that
differences in age do not affect obtaining viable mitochondria from human brain tissue
(see table 1 for demographic information). Scale bars: 2 microns in A, E, G; 1 micron in
C; 0.5 microns in B, D, F, H.
116
117
Fig 4. Cryopreservation for 1 week does not affect ΔΨmem. Mitochondria isolated from
0 and 10h PMI mouse cortex were cryopreserved (20% DMSO, -80oC) for one week. The
responsiveness of the ΔΨmem to substrates and toxins was tested in cryopreserved cMito
and pMito isolated from 0 (A-B), and 10 (C-D) hours PMI mouse cortex. Error bars
represent S.E.M. E) Comparison of ΔΨmem responses to mitochondrial substrates and
toxins between freshly-isolated (Fr) and cryopreserved (CP) mitochondria from 10 hours
PMI mouse cortex. There was no statistical difference in the responsiveness of ΔΨmem
to substrates and toxins between the freshly-isolated and the cryopreserved mitochondria.
(Student’s t-test, p>0.05, n=3). Electron micrographs of cMito cryopreserved in 20%
DMSO for one week after 0 and 10 hours PMI are shown in F and G respectively. Scale
bars: 2 microns in F-G.
118
Table 1. Demographic data of the human cases used in the study. Note that all cases used
are below 10 hours postmortem. pH is provided as a quality control for the tissue. PMI:
postmortem interval.
119
Table 2. Rates of ATP production are expressed as nmol/min/μg mitochondrial protein.
SEM: Standard Error of the Mean; CP: Cryopreserved; *p<0.05, Student’s t-test, in
comparisons made between the following samples: *0 hr PMI vs. 10 hr PMI; **0 hr PMI
vs. 0 hr PMI CP; ***10 hr PMI vs. 10 hr PMI CP; †0 hr PMI CP vs. 10 hr PMI CP
120
CONCLUSIONS
From basic signaling at the plasma membrane to activate the PI3K signaling
pathway to roles in cell survival and neuronal protection, it is clear that Akt has multiple
functions in the brain. New roles for Akt in neurons are constantly being elucidated and it
is likely that many have not yet been discovered. The role that Akt plays in cellular
metabolism is in the forefront of this project. As reviewed above, Akt signaling in many
cell types is linked to increased energy production and neuronal survival in the face of
ischemia and other toxic insults. In addition, Akt signaling in the mitochondria is
postulated to result in increased cellular survival and ATP production. Since Akt levels in
the mitochondria are known to be highly dynamic, it stands to reason that Akt would play
some fundamental role in brain mitochondrial function. For these reasons, the present
study focused on mechanisms of Akt translocation into mitochondria as well as roles of
mitochondrial Akt in cultured cell models and in the rodent brain. The study was further
expanded to examine the possibilities that Akt signaling can be examined in the human
brain.
Akt activation occurs at the cytoplasmic face of the plasma membrane, however
following its activation, Akt rapidly translocates into different subcellular compartments
including the mitochondria (Andjelkovic 1997; Bijur 2003) where it can phosphorylate a
great number of substrates to affect a multitude of cellular functions (Coffer 1998).
Although it is known that mitochondrial import of Akt is blocked by a dissipation of the
121
mitochondrial transmembrane potential (Bijur 2003), the molecular mechanism
underlying Akt mitochondrial transport previous to this study was unknown.
Akt is a well-known client protein of HSP90 and Akt activity is affected by its
binding to HSP90 (Sato, 2000). HSP90 inhibitor treatment has been associated with
destabilization, deactivation, and proteolysis of HSP90 client proteins (Marcu 2000b;
Yun, 2004; Fujita, 2002; Xu, 2003; Kim 2003). Thus, Akt activity, stability, and function
are dependent on its interaction with HSP90.
It has been reported that inhibition of HSP90 with NB and GA can block the
import of proteins into mitochondria (Fan, 2006), indicating that HSP90 plays a major
role in mitochondrial protein translocation. In the initial process of mitochondrial import,
HSP90 (with its preprotein cargo) binds to one of the mitochondrial surface receptors on
the TOM complex, the TOM70 or TOM20 receptor (Yano, 2004; Fan, 2006).
Subsequently, the preprotein is imported through the other components of the
mitochondrial membrane to its final destination (Gabriel, 2003; Chacinska, 2004;
(Kerscher, 1997; Kutik, 2007). Thus, HSP90, in conjunction with the mitochondrial
transport proteins, functions in the efficient translocation of proteins into the
mitochondria and its subcompartments.
To test the hypothesis that HSP90 mediates mitochondrial accumulation of Akt,
the two different HSP90 inhibitors, novobiocin and geldanamycin, were used to treat
cultured SH-SY5Y neuroblastoma and human embryonic kidney-293 (HEK293) cells.
Both inhibitors caused a rapid time-dependent decrease in Akt levels in mitochondria
isolated from these cells without affecting total cellular levels of Akt. Furthermore, ICV
122
injection of novobiocin into ventricles of adult male mice led to a significant decrease in
Akt levels in mitochondria isolated from cortex.
The reason for the extremely rapid decline in mitochondrial Akt levels is currently
not clear. However, it is possible that the cytoplasm serves as a constant source of Akt,
and once inside the mitochondrion, Akt is rapidly turned over. Since geldanamycin does
not enter mitochondria (Kang 2007), it should only affect the cytoplasmic pool of Hsp90.
Therefore, it can be concluded that geldanamycin was able to specifically inhibit the
extra-mitochondrial pool of HSP90 to prevent mitochondrial translocation of cytoplasmic
Akt. To further test the effects of HSP90 on Akt levels in mitochondria, cell lines were
created in which HSP90 expression was blocked by stable expression of HSP90 siRNA.
Surprisingly, these cells were able to thrive despite a significant decrease in cellular
HSP90 levels, possibly because there was a significant increase in compensatory
chaperones. Indeed, we observed a substantial increase in HSP40 levels in these cells.
Nevertheless, we observed no change in overall Akt levels, but a significant decrease in
the mitochondrial pool of Akt. Together, these data indicate that HSP90 activity plays a
significant role in affecting the level of Akt in the mitochondria.
In order to isolate the mitochondrial translocation system of Akt, an in vitro
mitochondrial import assay was used. An unmodified wtAkt and constitutively active
DDAkt were incubated with isolated functional mitochondria from wild type cells and
HSP90 siRNA cells. Both forms of Akt accumulated in the mitochondria, and
translocation of both proteins was inhibited by siRNA knockdown or novobiocin
treatment. These data indicate that the normal flux of Akt in the mitochondria is mediated
by HSP90 and is independent of the activation state of Akt. Since Akt is known to be
123
constitutively bound to HSP90 regardless of activation state (Meares 2004), it is
reasonable to hypothesize that Akt translocation into mitochondria is not entirely
dependent upon its activity.
Notably, in cell culture models, neither inhibition nor siRNA knockdown of
HSP90 inhibited IGF-1 stimulated Akt translocation into mitochondria which could
indicate that under stimulated conditions, Akt may enter mitochondria through another
translocation mechanism. In fact, HSP70 and the HSP40 co-chaperone system have been
shown to mediate the import of proteins into mitochondria (Young 2003; Bhangoo 2007).
Akt is a known client protein of HSP70, and the HSP40 co-chaperones are known to
interact with HSP70 to aid in protein folding and mitochondrial translocation (Gao 2002;
Cajo 2006). Therefore, HSP70 may act as a compensatory mechanism for mitochondrial
translocation of Akt into mitochondria under stimulated conditions, or in the absence of
HSP90.
Since HSP90 could possibly aid in mitochondrial translocation of some of the
TOM and TIM components to mitochondria, additional follow up studies would be
necessary to ensure that the mitochondrial translocation machinery is not compromised in
our pharmacological and cell culture models. Use of the pharmacological inhibitors at
low doses and short time points is not hypothesized to negatively affect the TOM and
TIM machinery expression, since the half life of mitochondria is on the order of weeks
and these machinery are thought to be stable in the membrane. Importantly, some
subunits of ATP synthase interact with HSP90 and HSP70, indicating it can be targeted
to mitochondria via cytosolic chaperones, but there were no observed changes in
124
mitochondrial levels of these proteins, suggesting that the mitochondrial translocation
machinery is intact in our HSP90 siRNA cell line.
In order to determine how Akt affects the mitochondrion, mitochondrial
morphology was first examined. Serum deprivation and loss of substrate availability is
stressful to cultured cells and can enhance the effects of many toxic stimuli (Rasola 2000;
Gottlieb 2003), and in fact, we found that mitochondria from HEK293 cells cultured in
serum-free media for 24h contained less Akt and had enlarged mitochondria with highly
condensed cristae. Incubation of functional, isolated mitochondria from serum-free
conditions with exogenous, constitutively active DDAkt caused a change in
mitochondrial morphology from the condensed to the orthodox configuration, with a
significant increase in mitochondrial optical density. Furthermore, the Akt-induced
change in mitochondrial conformation was blocked by the addition of novobiocin, further
confirming that HSP90 mediates mitochondrial accumulation of Akt and suggesting that
Akt plays a role in maintaining mitochondria in the orthodox configuration. It has been
reported that matrix remodeling to the condensed state, as was seen in the mitochondria
from cells cultured in serum-free media, exposes cytochrome c to the intermembrane
space, priming its release from the mitochondria during apoptosis. Thus, Akt’s effect on
mitochondrial morphology presents an additional mode by which Akt serves as a
protective protein, providing new insight regarding Akt effects on mitochondrial function.
Akt signaling in the brain has been linked to adult neurogenesis, neuroprotection,
and establishment of neuronal polarity. While it is known that Akt exists in the cytoplasm,
the nucleus, and the mitochondria, very few studies have examined the distribution of
Akt at the light and electron microscopy level (Eto 2008; Znamensky 2003) under basal
125
conditions. Therefore, to obtain an idea of Akt cellular and ultrastructural localization,
immunohistochemistry for detection of Akt as well as phospho-Ser473-Akt (pAkt) was
used in combination with light microscopy and transmission electron microscopy. Results
showed that Akt is abundant in neurons compared to glia, and that Akt exists in the
cytoplasm, nucleus, and mitochondria, as previously reported. Surprisingly, Akt
preferentially labeled mitochondria in neuronal processes (axons and dendrites)
compared to neuronal cell bodies and this result was confirmed using immunoblotting in
both mouse and human mitochondrial lysates. Interestingly, immunoblotting studies
revealed that in cMito, both Akt and HSP90 levels were similarly decreased, further
indicating that HSP90 mediates Akt entry into mitochondria. It is currently unclear why
Akt should be so robustly decreased in cMito. Since it is known that certain
mitochondrial signals, such as mitochondrial membrane potential, can initiate retrograde
and anterograde mitochondrial trafficking and that growth factor signaling is involved in
mitochondrial positioning and movement through cells (Chada 2003; Miller 2004;
Hollenbeck 2005), it could be that degradation of Akt is a signal for older mitochondria
to be transported back to the soma. Alternately, it may be that mitochondria containing
large amounts of Akt are automatically transported to neuronal processes. Since neuronal
processes are known to require large amounts of energy, and we show that increased Akt
bolsters mitochondrial ATP production, the presence of Akt may be a requirement for
delivery of highly functioning mitochondria to neuronal processes.
Because isolated mitochondria from neuronal processes contained more Akt than
mitochondria from neuronal cell bodies, and cellular Akt signaling has been shown to
increase mitochondrial ATP production rates (Stump 2003; Huang 2004), rates of ATP
126
production in cMito and pMito were measured so that the function of the mitochondrial
pool of Akt could be elucidated. pMito had significantly enhanced rates of ATP
production compared to cMito, and ATP production could be significantly decreased with
the addition of the protonophore and complex IV inhibitor CCCP, which was used as a
control for the luciferase assay.
These results led to further manipulations of the PI3K signaling pathway in mouse
brain. IGF-1 injected into ventricles of adult male mice caused significant increases in
levels of pSer473-Akt in isolated pMito but not in cMito. Concurrently, ATP production
rates were significantly increased in pMito compared to vehicle controls, and again
CCCP significantly decreased ATP production. Interestingly, there was no change in
ATP production in cMito, and in fact there was a nonsignificant 25% decrease in ATP
production in pMito from IGF-1 treated mice. The reason for this is not currently clear. It
is possible that IGF-1 stimulation initiates signaling of multiple signaling pathways in
neurons and that in the neuronal cell body this is a detriment to mitochondrial function, in
this case ATP production.
To decrease PI3K signaling activity and examine the effects on mitochondrial
function, LY294002 was ICV injected into adult male mice. In pMito isolated from LYtreated mice, there was a significant decrease in pSer473Akt levels as well as ATP
production. Together, these data confirm previous reports that PI3K activity at the plasma
membrane affects levels of Akt within the mitochondria (Bijur 2003) and suggests that
the mitochondrial pool of Akt plays a substantial role in mitochondrial energy output.
To further confirm the effects of the mitochondrial pool of Akt on mitochondrial
function, other parameters of mitochondrial fitness could be assayed, such as respiration,
127
activities of specific complexes in the electron transport chain, and ΔΨmem. These
examinations could be expanded to include modulation of the PI3K pathway in the brain,
as well as incubation of isolated mitochondria with constitutively active Akt. Since
mitochondrial membrane potential has been shown to regulate mitochondrial trafficking,
modulation of the PI3K pathway may alter mitochondrial distribution in neurons.
Furthermore, morphology of cMito and pMito should be examined under basal conditions
and following treatment of IGF-1 and/or LY to confirm results obtained in our cell
culture model, which indicated Akt’s role in mitochondrial morphology.
As numerous studies have reported the occurrence of altered or decreased
mitochondrial function in the aging brain and neurodegenerative diseases it has become
clear that mitochondria play a fundamental role in normal brain function and in brain
disease processes. Because of the ease of genetic and pharmacological manipulation,
animal and cell culture models are often used to study mitochondrial function and
dysfunction in the brain. However, considering the differences between species in
metabolism and gene expression, these models often do not accurately reflect the
mitochondrial function of the human central nervous system. Frozen postmortem brain
tissues from humans have been used in the past to analyze many parameters of
mitochondrial function (Park 2001). However, some indices of mitochondrial function
must be measured in intact, freshly isolated, functional mitochondria, such as ΔΨmem
and ATP production. The need for large amounts of pure, structurally intact mitochondria
is a common barrier for performing brain mitochondrial research, as well as the fact that
human brain tissues are in extremely limited supply.
128
Before this study could be expanded to include mitochondrial Akt research in the
human brain, we first had to determine if intact, functional mitochondria could be
obtained from postmortem brains. An initial time course of PMI in mice stored under
controlled conditions revealed that the mitochondrial membrane potential remains intact
up to 24 hours post-mortem. However, at the 18 hour PMI, massive structural
degradation occurred and mitochondrial yield decreased significantly. It is important to
note here that at the 0 and 10 hour time points, both cMito and pMito responded to all
substrates and toxins equivalently in the ΔΨmem assay and both populations withstood
cryopreservation at -80oC and in the presence of 20% DMSO as tested in the ΔΨmem
and ATP production assays. Furthermore, electron microscopy of mitochondria from post
mortem mouse cortices revealed that mitochondria are structurally intact and mostly in
the orthodox configuration. These data clearly indicate that ΔΨmem and ATP production
capacity can be reconstituted in mitochondria isolated from postmortem brains by the
presence of mitochondrial substrates. Furthermore, although specific functional
differences between cMito and pMito have been observed, these findings suggest that
PMI and cryopreservation have no differential effects on the two populations of neuronal
mitochondria.
Examinations of postmortem human brain mitochondria followed, and in contrast
to the standardized conditions of mouse care and tissue storage, the condition of human
tissues is extremely variable. The cause of death, medication and illicit drug use, age, and
possible infections are all pre-mortem conditions which can affect the quality of tissue
obtained from postmortem human. Furthermore, tissue pH, PMI, and storage conditions
of the body affect the integrity of postmortem tissue. Postmortem human tissue for this
129
study was obtained from individuals below the 10 hour PMI, though there were
substantial variations between individuals as well as between mitochondrial populations.
Both populations of mitochondria isolated from cortex from subject 1, who was a 32-year
old female, responded robustly in the ΔΨmem assay to all substrates and toxins. However,
mitochondria isolated from subject 2 and subject 3, significantly older subjects, showed
much more variation between mitochondrial populations. Electron microscopy of the
human samples revealed that structurally intact mitochondria are obtained from
postmortem human brain, and look remarkably similar to mitochondria isolated from
mouse brain. ATP production assays indicate that these mitochondria have the ability to
engage the electron transport chain and oxidative phosphorylation processes. Thus, it is
possible to obtain functional, structurally intact mitochondria from post-mortem human
and mouse brain, and that these mitochondria can be cryopreserved for further testing in
the ΔΨmem and ATP production assays. Though the reason for the ATP production
enhancement after cryopreservation is unclear, it has been shown that DMSO can bind
the F1 subunit of ATP synthase and bolster ATP production rates (Sakamoto 1984).
Nevertheless, postmortem human brains, when available, are a potential source of large
amounts of functional mitochondria. With controlled storage and archival conditions, it
could be argued that postmortem human mitochondrial “banking” can feasibly be used
for high-throughput screening of large numbers of samples, which would decrease
variability between assays and provide copious amounts of information about
mitochondrial function in the normal and diseased human brain. Furthermore, these data
show that mitochondria in the brain are remarkably resilient to proteolytic degradation. In
cell culture models, it has been shown that isolated mitochondria resist proteolytic
130
degradation, possibly because of the enclosed double membrane barrier which protects
the delicate matrix from the digestive properties of proteases (Bijur 2003; Barksdale
2009).
Since our findings show for the first time ever that the ΔΨmem in isolated
mitochondria from postmortem brain can be reconstituted, and that these mitochondria
maintain ATP production ability, it allows for assessment of mitochondrial function that
was thought previously to be impossible in human brain tissue. Even though there are
complexities that come with heterogeneous human brain samples, some of these can be
overcome by working with large sample sets. Together, these results provide a novel
approach for the study of mitochondrial function in the healthy and diseased human brain.
Results from this study have provided new insight into the mechanism of Akt
entry into mitochondria as well as what role Akt localization may play in neuronal and
mitochondrial function (Figure 2). In the first goal of this study, it was determined that
HSP90 binds cytoplasmic Akt and targets it to one of the receptors on the outer
mitochondrial membrane. In cell culture models, it was shown that Akt works to maintain
mitochondria in an orthodox, or highly respiring, state. In mouse brain, localization
studies revealed widespread Akt expression in neurons throughout the brain, and
enhanced labeling of mitochondria in neuronal processes. Consequently, mitochondria
isolated from neuronal processes have enhanced rates of ATP production, which can be
modulated by modulating the activity of the PI3K signaling pathway, suggesting that the
mitochondrial pool of Akt plays a role not only in modulating mitochondrial morphology,
but also in the rate of mitochondrial ATP production. This makes sense considering that
Akt maintains the orthodox position of mitochondria which is indicative of increased
131
ΔΨmem and energy output. Therefore, it is possible that HSP90 chaperones Akt to
mitochondria, causing maintenance of mitochondria in the orthodox configuration. The
Akt-containing mitochondria, in turn, are highly polarized and able to generate ATP at a
higher rate. Subsequently, these mitochondria are transported to neuronal processes,
where energy requirements are high.
In the last goals of the study, we attempted to extend work done in cell culture
and mouse models to the human brain. Since no previous characterization of postmortem
brain mitochondria had been done, our initial studies showed that isolated mitochondria
from postmortem mouse and human brains are indeed structurally intact and functional as
determined by electron microscopy and the ATP production and ΔΨmem assays. The
next logical step for these studies would be to examine function of Akt in mitochondria
isolated from normal and diseases human brain. Taken together, these data show novel
roles for Akt function in the mitochondria and provide new methodology for the study of
mitochondria in human brain.
132
Dendrites
Axon
Terminals
Akt
Akt
Akt
Akt
Akt
Nucleus (Soma)
Key:
Transport to processes
Transport to cell soma
ATP
Fig 1. Hypothesized model for the function of mitochondrial Akt in neurons.
Mitochondria of neuronal processes (pMito) have been shown in this project to contain
more Akt than mitochondria in the cell soma (cMito), and that Akt works to maintain
mitochondria in an orthodox configuration. It was also shown that pMito have enhanced
rates of ATP production compared with cMito. We therefore hypothesize that
mitochondria with high levels of Akt with high ΔΨmem (outlined in red) are transported
to neuronal processes, where there is a high energy requirement for neuronal signaling.
Degradation or lack of Akt may signal retrograde transport to the neuron soma where
ATP requirements are not as high.
133
LIST OF GENERAL REFERENCES
Ahmad N, Wang Y, Haider KH, Wang B, Pasha Z, Uzun O, Ashraf M. (2006) Cardiac
protection by mitoKATP channels is dependent on Akt translocation from cytosol
to mitochondria during late preconditioning. Am J Physiol Heart Circ Physiol,
290(6):H2402-8.
Alam ZI, Jenner A, Daniel SE, Lees AJ, Cairns N, Marsden CD, Jenner P, Halliwell B.
(1997) Oxidative DNA damage in the parkinsonian brain: an apparent selective
increase in 8-hydroxyguanine levels in substantia nigra. J Neurochem 69(3):1196203.
Alekseeva MG, Elizarov SM, Bekker OB, Lubimova IK, Danilenko VN (2009) F0F1 ATP
Synthase of Streptomycetes: modulation of Activity and Oligomycin Resistance
by Protein Ser/Thr Kinases. Membrane and Cell Biology 3(1) 16-23.
Alessi DR, Andjelkovic M, Caudwell B, Cron P, Morrice N, Cohen P, Hemmings
BA.(1996) Mechanism of activation of protein kinase B by insulin and IGF-1.
EMBO J 15(23):6541-51.
Alessi DR, Deak M, Casamayor A, Caudwell FB, Morrice N, Norman DG, Gaffney P,
Reese CB, MacDougall CN, Harbison D, Ashworth A, Bownes M. (1997) 3Phosphoinositide-dependent protein kinase-1 (PDK1): structural and functional
homology with the Drosophila DSTPK61 kinase. Curr Biol 7(10):776-89.
Ali A, Hoeflich KP, Woodgett JR. (2001) Glycogen synthase kinase-3: properties,
functions, and regulation. Chem Rev. 101(8):2527-40.
Anderson MF, Sims NR. (1999) Mitochondrial respiratory function and cell death in
focal cerebral ischemia. J Neurochem 73(3):1189-99.
Anderson S, Bankier AT, Barrell BG, de Bruijn MH, Coulson AR, Drouin J, Eperon IC,
Nierlich DP, Roe BA, Sanger F, Schreier PH, Smith AJ, Staden R, Young IG.
(1981) Sequence and organization of the human mitochondrial genome. Nature
290(5806):457-65.
Andjelković M, Alessi DR, Meier R, Fernandez A, Lamb NJ, Frech M, Cron P, Cohen P,
Lucocq JM, Hemmings BA. (1997) Role of translocation in the activation and
function of protein kinase B. J Biol Chem 272(50):31515-24.
134
Antico Arciuch VG, Galli S, Franco MC, Lam PY, Cadenas E, Carreras MC,
Poderoso JJ. (2009) Akt1 intramitochondrial cycling is a crucial step in the redox
modulation of cell cycle progression. PLoS One 4(10):e7523.
Arciuch VG, Alippe Y, Carreras MC, Poderoso JJ. (2009) Mitochondrial kinases in cell
signaling: Facts and perspectives. Adv Drug Deliv Rev 61(14):1234-49. Epub
2009 Sep 4. PubMed PMID: 19733603.
Arimura N, Kaibuchi K. (2005) Key regulators in neuronal polarity. Neuron 48(6):881-4.
Baker MJ, Frazier AE, Gulbis JM, Ryan MT. (2007) Mitochondrial proteinimport machinery: correlating structure with function. Trends Cell Biol
17(9):456-64.
Baker MJ, Frazier AE, Gulbis JM, Ryan MT. (2007) Mitochondrial protein-import
machinery: correlating structure with function. Trends Cell Biol. 17(9):456-64.
Barksdale KA, Bijur GN. (2009) The basal flux of Akt in the mitochondria is mediated
by heat shock protein 90. J Neurochem 108(5):1289-99.
Basso AD, Solit DB, Chiosis G, Giri B, Tsichlis P, Rosen N. (2002) Akt forms an
Intracellular complex with heat shock protein 90 (Hsp90) and Cdc37 and is
destabilized by inhibitors of Hsp90 function. J Biol Chem 277(42):39858-66.
Becker L, Bannwarth M, Meisinger C, Hill K, Model K, Krimmer T, Casadio R, Truscott
KN, Schulz GE, Pfanner N, Wagner R. (2005) Preprotein translocase of the outer
mitochondrial membrane: reconstituted Tom40 forms a characteristic TOM pore.
J Mol Biol 353(5):1011-20.
Bhangoo MK, Tzankov S, Fan AC, Dejgaard K, Thomas DY, Young JC. (2007) Multiple
40-kDa heat-shock protein chaperones function in Tom70-dependent
mitochondrial import. Mol Biol Cell 18(9):3414-28.
Bijur GN, Jope RS. (2003) Rapid accumulation of Akt in mitochondria following
phosphatidylinositol 3-kinase activation. J Neurochem 87(6):1427-35.
Bolender N, Sickmann A, Wagner R, Meisinger C, Pfanner N. (2008) Multiple pathways
for sorting mitochondrial precursor proteins. EMBO Rep 9s (1):42-9.
Bossy-Wetzel E, Barsoum MJ, Godzik A, Schwarzenbacher R, Lipton SA. (2003)
Mitochondrial fission in apoptosis, neurodegeneration, and aging. Curr Opin Cell
Biol 15: 706-716
Brazil DP, Hemmings BA. (2001) Ten years of protein kinase B signaling: a hard Akt to
follow. Trends Biochem Sci 26(11):657-64.
135
Browne SE. (2008) Mitochondria and Huntington's disease pathogenesis: insight from
genetic and chemical models. Ann N Y Acad Sci 1147:358-82.
Bruel-Jungerman E, Veyrac A, Dufour F, Horwood J, Laroche S, Davis S.(2009)
Inhibition of PI3K-Akt signaling blocks exercise-mediated enhancement of adult
neurogenesis and synaptic plasticity in the dentate gyrus. PLoS One 4(11):e7901.
Brywe KG, Mallard C, Gustavsson M, Hedtjärn M, Leverin AL, Wang X, Blomgren K,
Isgaard J, Hagberg H. (2005) IGF-I neuroprotection in the immature brain after
hypoxia-ischemia, involvement of Akt and GSK3beta? Eur J Neurosci
21(6):1489-502.
Cajo GC, Horne BE, Kelley WL, Schwager F, Georgopoulos C, Genevaux P. (2006) The
role of the DIF motif of the DnaJ (Hsp40) co-chaperone in the regulation of the
DnaK (Hsp70) chaperone cycle. J Biol Chem 281(18):12436-44.
Castellani R, Hirai K, Aliev G, Drew KL, Nunomura A, Takeda A, Cash AD,
Obrenovich ME, Perry G, Smith MA. (2002) Role of mitochondrial dysfunction
in Alzheimer's disease. J Neurosci Res 70(3):357-60.
Chada SR, Hollenbeck PJ. (2003) Mitochondrial movement and positioning in axons: the
role of growth factor signaling. J Exp Biol 206(Pt 12):1985-92.
Chada SR, Hollenbeck PJ. (2004) Nerve growth factor signaling regulates motility and
docking of axonal mitochondria. Curr Biol 14(14):1272-6.
Chacinska A, Pfannschmidt S, Wiedemann N, Kozjak V, Sanjuán Szklarz LK, SchulzeSpecking A, Truscott KN, Guiard B, Meisinger C, Pfanner N. (2004) Essential
role of Mia40 in import and assembly of mitochondrial intermembrane space
proteins. EMBO J. 23:3735-46.
Chandana R, Mythri RB, Mahadevan A, Shankar SK, Srinivas Bharath MM. (2009)
Biochemical analysis of protein stability in human brain collected at different
post-mortem intervals. Indian J Med Res 129(2):189-99.
Chang DT, Reynolds IJ. (2006) Mitochondrial trafficking and morphology in healthy and
injured neurons. Prog Neurobiol 80(5):241-68.
Cho H, Thorvaldsen JL, Chu Q, Feng F, Birnbaum MJ. Akt1/PKBalpha is required
for normal growth but dispensable for maintenance of glucose homeostasis in
mice. J Biol Chem. 2001 Oct 19;276(42):38349-52.
Coffer PJ, Jin J, Woodgett JR. (1998) Protein kinase B (c-Akt): a multifunctional
mediator of phosphatidylinositol 3-kinase activation. Biochem J 335(Pt 1), 113.
136
Csermely P, Schnaider T, Soti C, Prohászka Z, Nardai G. (1998) The 90-kDa molecular
chaperone family: structure, function, and clinical applications. A comprehensive
review. J. Pharmacol Ther 79(2):129-68.
Devi L, Raghavendran V, Prabhu BM, Avadhani NG, Anandatheerthavarada HK. (2008)
Mitochondrial import and accumulation of alpha-synuclein impair complex I in
human dopaminergic neuronal cultures and Parkinson disease brain. J Biol Chem
283(14):9089-100.
Dietmeier K, Hönlinger A, Bömer U, Dekker PJ, Eckerskorn C, Lottspeich F, Kübrich M,
Pfanner N. (1997)Tom5 functionally links mitochondrial preprotein receptors to
the general import pore. Nature 388(6638):195-200.
Easton RM, Cho H, Roovers K, Shineman DW, Mizrahi M, Forman MS, Lee VM,
Szabolcs M, de Jong R, Oltersdorf T, Ludwig T, Efstratiadis A, Birnbaum MJ.
(2005) Role for Akt3/protein kinase Bgamma in attainment of normal brain size.
Mol Cell Biol (5):1869-78.
Eto R, Abe M, Hayakawa N, Kato H, Araki T. (2008) Age-related changes of calcineurin
and Akt1/protein kinase Balpha (Akt1/PKBalpha) immunoreactivity in the mouse
hippocampal CA1 sector: an immunohistochemical study. Metab Brain Dis
23(4):399-409.
Fan AC, Bhangoo MK, Young JC. (2006) Hsp90 functions in the targeting and outer
membrane translocation steps of Tom70-mediated mitochondrial import. J Biol
Chem 281(44):33313-24.
Fayard E, Tintignac LA, Baudry A, Hemmings BA. (2005) Protein kinase B/Akt at a
glance. J Cell Sci 118(Pt 24):5675-8.
Frederick RL, Shaw JM. (2007) Moving mitochondria: establishing distribution of an
essential organelle. Traffic 8(12):1668-75.
Frezza C, Cipolat S, Martins de Brito O, Micaroni M, Beznoussenko GV, Rudka T,
Bartoli D, Polishuck RS, Danial NN, De Strooper B, Scorrano L. (2006) OPA1
controls apoptotic cristae remodeling independently from mitochondrial fusion.
Cell 126(1):177-89.
Fujita N, Sato S, Ishida A, Tsuruo T. (2002) Involvement of Hsp90 in signaling and
stability of 3-phosphoinositide-dependent kinase-1. J Biol Chem. 277(12):1034653.
Gabriel K, Egan B, Lithgow T. (2003) Tom40, the import channel of the mitochondrial
outer membrane, plays an active role in sorting imported proteins. EMBO J 22,
2380-6.
137
Gao T, Newton AC. (2002) The turn motif is a phosphorylation switch that regulates
the binding of Hsp70 to protein kinase C. J Biol Chem 277(35):31585-92.
Gill-King, H. (1997) in Forensic Taphonomy: The postmortem fate of human remains,
Ed WD Haglund (CRC Press LLC).
Gottlieb E, Armour SM, Harris MH, Thompson CB. (2003) Mitochondrial membrane
potential regulates matrix configuration and cytochrome c release during
apoptosis. Cell Death Differ 10(6):709-17.
Gottlob K, Majewski N, Kennedy S, Kandel E, Robey RB, Hay N. (2001) Inhibition of
early apoptotic events by Akt/PKB is dependent on the first committed step of
glycolysis and mitochondrial hexokinase. Genes Dev. 15(11):1406-18.
Gray MW. (1993) Origin and evolution of organelle genomes. Curr Opin Genet Dev
3(6):884-90.
Hackenbrock CR. (1966) Ultrastructural bases for metabolically linked mechanical
activity in mitochondria. I. Reversible ultrastructural changes with change in
metabolic steady state in isolated liver mitochondria. J Cell Biol 30(2):269-97.
Hackenbrock CR. (1968) Ultrastructural bases for metabolically linked mechanical
activity in mitochondria. II. Electron transport-linked ultrastructural
transformations in mitochondria. J Cell Biol 37(2):345-69.
Hanger DP, Hughes K, Woodgett JR, Brion JP, Anderton BH. (1992) Glycogen synthase
kinase-3 induces Alzheimer's disease-like phosphorylation of tau: generation of
paired helical filament epitopes and neuronal localisation of the kinase. Neurosci
Lett. 147(1):58-62.
Heales SJ, Bolaños JP, Stewart VC, Brookes PS, Land JM, Clark JB. (1999) Nitric
oxide, mitochondria and neurological disease. Biochim Biophys Acta
1410(2):215-28.
Højlund K, Wrzesinski K, Larsen PM, Fey SJ, Roepstorff P, Handberg A, Dela F,
Vinten J, McCormack JG, Reynet C, Beck-Nielsen H. (2003) Proteome analysis
reveals phosphorylation of ATP synthase beta -subunit in human skeletal muscle
and proteins with potential roles in type 2 diabetes. J Biol Chem 278(12):1043642.
Hollenbeck PJ. (2005) Mitochondria and neurotransmission: evacuating the synapse.
Neuron 47(3):331-3.
138
Hoshi M, Takashima A, Noguchi K, Murayama M, Sato M, Kondo S, Saitoh Y, Ishiguro
K, Hoshino T, Imahori K. (1996) Regulation of mitochondrial pyruvate
dehydrogenase activity by tau protein kinase I/glycogen synthase kinase 3beta in
brain. Proc Natl Acad Sci U S A. 93(7):2719-23.
Huang TJ, Verkhratsky A, Fernyhough P. (2005) Insulin enhances mitochondrial inner
membrane potential and increases ATP levels through phosphoinositide 3-kinase
in adult sensory neurons. Mol Cell Neurosci 28(1):42-54.
Hüttemann M, Lee I, Pecinova A, Pecina P, Przyklenk K, Doan JW. (2008) Regulation of
oxidative phosphorylation, the mitochondrial membrane potential, and their role
in human disease. J Bioenerg Biomembr 40(5):445-56.
Jacinto E, Facchinetti V, Liu D, Soto N, Wei S, Jung SY, Huang Q, Qin J, Su B. (2006)
SIN1/MIP1 maintains rictor-mTOR complex integrity and regulates Akt
phosphorylation and substrate specificity. Cell 127(1):125-37.
Jackson TC, Rani A, Kumar A, Foster TC. (2009) Regional hippocampal differences in
AKT survival signaling across the lifespan: implications for CA1 vulnerability
with aging. Cell Death Differ 16(3):439-48.
Jope RS, Bijur GN. (2002) Mood stabilizers, glycogen synthase kinase-3beta and cell
survival. Mol Psychiatry 7 Suppl 1:S35-45.
Kalkman HO. (2006) The role of the phosphatidylinositide 3-kinase-protein kinase B
pathway in schizophrenia. Pharmacol Ther. 110(1):117-34.
Kang BH, Plescia J, Dohi T, Rosa J, Doxsey SJ, Altieri DC. (2007)Regulation of tumor
cell mitochondrial homeostasis by an organelle-specific Hsp90 chaperone
network. Cell 131(2):257-70.
Kasahara T, Kubota M, Miyauchi T, Noda Y, Mouri A, Nabeshima T, Kato T. (2006)
Mice with neuron-specific accumulation of mitochondrial DNA mutations
show mood disorder-like phenotypes. Mol Psychiatry 11(6):577-93, 523.
Kennedy SG, Kandel ES, Cross TK, Hay N. (1999) Akt/Protein kinase B inhibits cell
death by preventing the release of cytochrome c from mitochondria. Mol Cell Biol
19(8):5800-10.
Kerscher O, Holder J, Srinivasan M, Leung RS, Jensen RE. (1997) The Tim54p-Tim22p
complex mediates insertion of proteins into the mitochondrial inner membrane. J
Cell Biol 139, 1663-75.
139
Kim S, Kang J, Hu W, Evers BM, Chung DH. (2003) Geldanamycin decreases Raf-1 and
Akt levels and induces apoptosis in neuroblastomas. Int J Cancer 103, 352-9.
Kimura K, Hattori S, Kabuyama Y, Shizawa Y, Takayanagi J, Nakamura S, Toki S,
Matsuda Y, Onodera K, Fukui Y. (1994) Neurite outgrowth of PC12 cells is
suppressed by wortmannin, a specific inhibitor of phosphatidylinositol 3-kinase. J
Biol Chem. 269(29):18961-7.
King TD, Bijur GN, Jope RS. (2001) Caspase-3 activation induced by inhibition of
mitochondrial complex I is facilitated by glycogen synthase kinase-3beta and
attenuated by lithium. Brain Res. 919(1):106-14.
Klein PS, Melton DA. (1996) A molecular mechanism for the effect of lithium on
development. Proc Natl Acad Sci U S A.93(16):8455-9.
Koehler CM. (2004) New developments in mitochondrial assembly. Annu Rev Cell Dev
Biol 20:309-35.
Kozlovsky N, Belmaker RH, Agam G. (2000) Low GSK-3beta immunoreactivity in
postmortem frontal cortex of schizophrenic patients. Am J Psychiatry.
157(5):831-3.
Kung L, Roberts RC. (1999) Mitochondrial pathology in human schizophrenic striatum: a
postmortem ultrastructural study. Synapse 31(1):67-75.
Kutik S, Guiard B, Meyer HE, Wiedemann N, Pfanner N. (2007) Cooperation of
translocase complexes in mitochondrial protein import. J Cell Biol. 179, 585-91.
Lai HC, Liu TJ, Ting CT, Sharma PM, Wang PH.(2003) Insulin-like growth factor-1
prevents loss of electrochemical gradient in cardiac muscle mitochondria via
activation of PI3-kinase/Akt pathway. Mol Cell Endocrinol 205(1-2):99-106.
Lapchak PA, Maher P, Schubert D, Zivin JA. (2007) Baicalein, an antioxidant
12/15-lipoxygenase inhibitor improves clinical rating scores following multiple
infarct embolic strokes. Neuroscience 150(3):585-91.
Lessmann E, Ngo M, Leitges M, Minguet S, Ridgway ND, Huber M. (2007) Oxysterolbinding protein-related protein (ORP) 9 is a PDK-2 substrate and regulates Akt
phosphorylation. Cell Signal 19(2):384-92.
Lewis DA. (2002) The human brain revisited: opportunities and challenges in
postmortem studies of psychiatric disorders. Neuropsychopharmacology
26(2):143-54.
140
Liu C, Wu J, Gu J, Xiong Z, Wang F, Wang J, Wang W, Chen J. (2007) Baicalein
Improves cognitive deficits induced by chronic cerebral hypoperfusion in rats.
Pharmacol Biochem Behav 86(3):423-30.
Liu C, Wu J, Xu K, Cai F, Gu J, Ma L, Chen J. (2009) Neuroprotection by baicalein in
ischemic brain injury involves PTEN/AKT pathway. J Neurochem [Epub
ahead of print] PubMed PMID: 20050973.
Lovestone S, Killick R, Di Forti M, Murray R. (2007) Schizophrenia as a GSK-3
dysregulation disorder. Trends Neurosci 30(4):142-9.
Majewski N, Nogueira V, Robey RB, Hay N. (2004) Akt inhibits apoptosis downstream
of BID cleavage via a glucose-dependent mechanism involving mitochondrial
hexokinases. Mol Cell Biol 24(2):730-40.
Mattson MP, Liu D. (2003) Mitochondrial potassium channels and uncoupling proteins in
synaptic plasticity and neuronal cell death. Biochem Biophys Res Commun
304(3):539-49.
Marcu MG, Schulte TW, Neckers L. (2000) Novobiocin and related coumarins and
depletion of heat shock protein 90-dependent signaling proteins. J Natl Cancer
Inst 92(3):242-8.
Marcu MG, Chadli A, Bouhouche I, Catelli M, Neckers LM. (2000) The heat shock
protein 90 antagonist novobiocin interacts with a previously unrecognized ATPbinding domain in the carboxyl terminus of the chaperone. J Biol Chem
275(47):37181-6.
Matute C, Sánchez-Gómez MV, Martínez-Millán L, Miledi R. (1997) Glutamate
receptor-mediated toxicity in optic nerve oligodendrocytes. Proc Natl Acad Sci U
S A. 94(16):8830-5.
McClellan AJ, Xia Y, Deutschbauer AM, Davis RW, Gerstein M, Frydman J. (2007)
Diverse cellular functions of the Hsp90 molecular chaperone uncovered using
systems approaches. Cell 5;131(1):121-35.
Meares GP, Zmijewska AA, Jope RS. (2004) Heat shock protein-90 dampens and directs
signaling stimulated by insulin-like growth factor-1 and insulin. FEBS Lett 574(13):181-6.
Miller KE, Sheetz MP. (2004) Axonal mitochondrial transport and potential are
correlated. J Cell Sci 117(Pt 13):2791-804.
141
Miyamoto S, Murphy AN, Brown JH. (2008) Akt mediates mitochondrial protection in
cardiomyocytes through phosphorylation of mitochondrial hexokinase-II. Cell
Death Differ 15(3):521-9.
Neupert W. (1997) Protein import into mitochondria. Annu Rev Biochem 66:863-917.
Ness JK, Scaduto RC Jr, Wood TL. (2004) IGF-I prevents glutamate-mediated bax
translocation and cytochrome C release in O4+ oligodendrocyte progenitors. Glia.
46(2):183-94.
Nicholls DG, Budd SL. (2000) Mitochondria and neuronal survival. Physiol Rev
80(1):315-60.
Niizuma K, Endo H, Chan PH. (2009) Oxidative stress and mitochondrial dysfunction as
determinants of ischemic neuronal death and survival. J Neurochem 109 Suppl
1:133-8.
Okamoto K, Shaw JM. (2005) Mitochondrial morphology and dynamics in yeast and
multicellular eukaryotes. Annu Rev Genet 39:503-36.
Olichon A, Baricault L, Gas N, Guillou E, Valette A, Belenguer P, Lenaers G. (2003)
Loss of OPA1 perturbates the mitochondrial inner membrane structure and
integrity, leading to cytochrome c release and apoptosis. J Biol Chem
278(10):7743-6.
Orr AL, Li S, Wang CE, Li H, Wang J, Rong J, Xu X, Mastroberardino PG,
Greenamyre JT, Li XJ. (2008) N-terminal mutant huntingtin associates with
mitochondria and impairs mitochondrial trafficking. J Neurosci 28(11):2783-92.
Palay SL. (1956) Synapses in the central nervous system. J Biophys Biochem Cytol
2(4, Suppl):193-202.
Park LC, et al. (2001) Mitochondrial impairment in the cerebellum of the patients with
progressive supranuclear palsy. J Neurosci Res. 66(5), 1028-34.
Parone PA, Da Cruz S, Tondera D, Mattenberger Y, James DI, Maechler P, Barja
F, Martinou JC. (2008) Preventing mitochondrial fission impairs mitochondrial
function and leads to loss of mitochondrial DNA. PLoS One 3(9):e3257.
Parone PA, Martinou JC. (2006) Mitochondrial fission and apoptosis: an ongoing trial.
Biochim Biophys Acta 1763(5-6):522-30.
Panaretou B, Prodromou C, Roe SM, O'Brien R, Ladbury JE, Piper PW, Pearl LH.(1998)
ATP binding and hydrolysis are essential to the function of the Hsp90 molecular
chaperone in vivo. EMBO J 17(16):4829-36.
142
Park LC, Albers DS, Xu H, Lindsay JG, Beal MF, Gibson GE. (2001) Mitochondrial
impairment in the cerebellum of the patients with progressive supranuclear palsy.
J Neurosci Res 66(5):1028-34.
Pearl LH, Prodromou C. (2006) Structure and mechanism of the Hsp90 molecular
chaperone machinery. Annu Rev Biochem 75:271-94.
Peineau S, Taghibiglou C, Bradley C, Wong TP, Liu L, Lu J, Lo E, Wu D, Saule
E, Bouschet T, Matthews P, Isaac JT, Bortolotto ZA, Wang YT, Collingridge GL.
(2007) LTP inhibits LTD in the hippocampus via regulation of GSK3beta. Neuron
53(5):703-17.
Pfanner N, Craig EA, Hönlinger A. (1997) Mitochondrial preprotein translocase. Annu
Rev Cell Dev Biol 13:25-51.
Pilegaard H, Birk JB, Sacchetti M, Mourtzakis M, Hardie DG, Stewart G, Neufer PD,
Saltin B, van Hall G, Wojtaszewski JF. (2006) PDH-E1alpha dephosphorylation
and activation in human skeletal muscle during exercise: effect of intralipid
infusion. Diabetes 55(11):3020-7.
Pitt D, Werner P, Raine CS. (2000) Glutamate excitotoxicity in a model of multiple
sclerosis. Nat Med 6(1):67-70.
Prickaerts J, Moechars D, Cryns K, Lenaerts I, van Craenendonck H, Goris I,
Daneels G, Bouwknecht JA, Steckler T. (2006) Transgenic mice overexpressing
Glycogen synthase kinase 3beta: a putative model of hyperactivity and mania. J
Neurosci 26(35):9022-9.
Quiroz JA, Gray NA, Kato T, Manji HK. (2008) Mitochondrially mediated plasticity in
the pathophysiology and treatment of bipolar disorder.
Neuropsychopharmacology 33(11):2551-65.
Rasola A, Bernardi P. (2007) The mitochondrial permeability transition pore and its
involvement in cell death and in disease pathogenesis. Apoptosis 12(5):815-33.
Rollins B, Martin MV, Sequeira PA, Moon EA, Morgan LZ, Watson SJ, Schatzberg
A, Akil H, Myers RM, Jones EG, Wallace DC, Bunney WE, Vawter MP. (2009)
Mitochondrial variants in schizophrenia, bipolar disorder, and major depressive
disorder. PLoS One 4(3):e4913.
Sakamoto, J. (1984) Effect of dimethyl sufoxide on ATP synthesis by mitochondrial
soluble F1-ATPase. J Biochem 96(2), 483-7.
Santi SA, Lee H. (2009) The Akt isoforms are present at distinct subcellular locations.
Am J Physiol Cell Physiol. PMID 20018949.
143
Sato S, Fujita N, Tsuruo T. (2000) Modulation of Akt kinase activity by binding to
Hsp90. Proc Natl Acad Sci U S A 97(20):10832-7.
Scheffler, I. E. (2008) Mitochondria, 2nd Edition (Wiley-Liss Inc).
Scorrano L. (2005) Proteins that fuse and fragment mitochondria in apoptosis:
con-fissing a deadly con-fusion? J Bioenerg Biomembr 37(3):165-70.
Shao L, Martin MV, Watson SJ, Schatzberg A, Akil H, Myers RM, Jones EG, Bunney
WE, Vawter MP. (2008) Mitochondrial involvement in psychiatric disorders. Ann
Med 40(4):281-95.
Sims NR, Anderson MF. (2002) Mitochondrial contributions to tissue damage in stroke.
Neurochem Int. 2002 May;40(6):511-26. Review. PubMed PMID: 11850108.
Staal SP. (1987) Molecular cloning of the akt oncogene and its human homologues
AKT1 and AKT2: amplification of AKT1 in a primary human gastric
adenocarcinoma. Proc Natl Acad Sci U S A. 84(14):5034-7.
Staal SP, Hartley JW. (1988) Thymic lymphoma induction by the AKT8 murine
retrovirus. J Exp Med 167(3):1259-64.
Staal SP, Hartley JW, Rowe WP. (1977) Isolation of transforming murine leukemia
viruses from mice with a high incidence of spontaneous lymphoma. Proc Natl
Acad Sci U S A 74(7):3065-7.
Scheid MP, Woodgett JR. (2003) Unravelling the activation mechanisms of protein
kinase B/Akt. FEBS Lett 546(1):108-12.
Stiles BL. (2009) PI-3-K and AKT: Onto the mitochondria. Adv Drug Deliv Rev
61(14):1276-82.
Stump CS, Short KR, Bigelow ML, Schimke JM, Nair KS. (2003) Effect of insulin on
human skeletal muscle mitochondrial ATP production, protein synthesis, and
mRNA transcripts. Proc Natl Acad Sci U S A 100(13):7996-8001.
Sui L, Wang J, Li BM. (2008) Role of the phosphoinositide 3-kinase-Akt-mammalian
target of the rapamycin signaling pathway in long-term potentiation and trace
fear conditioning memory in rat medial prefrontal cortex. Learn Mem 15(10):76276.
Szendroedi J, Roden M. (2008) Mitochondrial fitness and insulin sensitivity in humans.
Diabetologia 51(12):2155-67.
144
Takahashi M, Chesley A, Freyssenet D, Hood DA. (1998) Contractile activity-induced
adaptations in the mitochondrial protein import system. Am J Physiol 274(5 Pt
1):C1380-7.
Tong JJ. (2007) Mitochondrial delivery is essential for synaptic potentiation. Biol Bull
212(2):169-75.
Trencia A, Perfetti A, Cassese A, Vigliotta G, Miele C, Oriente F, Santopietro
S, Giacco F, Condorelli G, Formisano P, Beguinot F. (2003) Protein kinase B/Akt
binds and phosphorylates PED/PEA-15, stabilizing its antiapoptotic action. Mol
Cell Biol 23(13):4511-21.
Unterluggauer H, Hütter E, Viertler HP, Jansen-Dürr P. (2008) Insulin-like growth
factor-induced signals activate mitochondrial respiration. Biotechnol J 3(6):813-6.
van der Heide LP, Ramakers GM, Smidt MP. (2006) Insulin signaling in the central
nervous system: learning to survive. Prog Neurobiol 79(4):205-21.
Vanhaesebroeck B, Alessi DR. (2000) The PI3K-PDK1 connection: more than just a road
to PKB. Biochem J 346 Pt 3:561-76.
Verburg J, Hollenbeck PJ. (2008) Mitochondrial membrane potential in axons increases
with local nerve growth factor or semaphorin signaling. J Neurosci 28(33):830615.
Verwer RW, Hermens WT, Dijkhuizen P, ter Brake O, Baker RE, Salehi A, Sluiter
AA, Kok MJ, Muller LJ, Verhaagen J, Swaab DF. (2002) Cells in human
postmortem brain tissue slices remain alive for several weeks in culture. FASEB J
16(1):54-60.
Verwer RW, Hermens WT, ter Brake O, Verhaagen J, Swaab DF. (2002) Life after
death? Neurology 59(9):1355.
Vila M, Ramonet D, Perier C. (2008) Mitochondrial alterations in Parkinson's disease:
new clues. J Neurochem 107(2):317-28.
Wang X, Su B, Lee HG, Li X, Perry G, Smith MA, Zhu X. (2009) Impaired balance of
mitochondrial fission and fusion in Alzheimer's disease. J Neurosci 29(28):9090103.
Wang X, Su B, Zheng L, Perry G, Smith MA, Zhu X. (2009) The role of abnormal
mitochondrial dynamics in the pathogenesis of Alzheimer's disease. J Neurochem
109 Suppl 1:153-9.
Whiteman EL, Cho H, Birnbaum MJ. (2002) Role of Akt/protein kinase B in metabolism.
Trends Endocrinol Metab 13(10):444-51.
145
Whitesell L, Mimnaugh EG, De Costa B, Myers CE, Neckers LM. (1994) Inhibition of
heat shock protein HSP90-pp60v-src heteroprotein complex formation by
benzoquinone ansamycins: essential role for stress proteins in oncogenic
transformation. Proc Natl Acad Sci U S A 91(18):8324-8.
Xu W, Yuan X, Jung YJ, Yang Y, Basso A, Rosen N, Chung EJ, Trepel J, Neckers L.
(2003) The heat shock protein 90 inhibitor geldanamycin and the ErbB inhibitor
ZD1839 promote rapid PP1 phosphatase-dependent inactivation of AKT in ErbB2
overexpressing breast cancer cells. Cancer Res 3(22):7777-84.
Yamamoto H, Fukui K, Takahashi H, Kitamura S, Shiota T, Terao K, Uchida M,
Esaki M, Nishikawa S, Yoshihisa T, Yamano K, Endo T. (2009) Roles of Tom70
in import of presequence-containing mitochondrial proteins. J Biol Chem
284(46):31635-46.
Yan D, Guo L, Wang Y. (2006) Requirement of dendritic Akt degradation by the
ubiquitin-proteasome system for neuronal polarity. J Cell Biol 174(3):415-24.
Yang ZZ, Tschopp O, Baudry A, Dümmler B, Hynx D, Hemmings BA. (2004)
Physiological functions of protein kinase B/Akt. Biochem Soc Trans 32(Pt 2):3504.
Yano M, Terada K, Mori M. (2004) Mitochondrial import receptors TOM20 and TOM22
have chaperone-like activity. J Biol Chem 279, 10808-13.
Young JC, Hoogenraad NJ, Hartl FU. (2003) Molecular chaperones Hsp90 and Hsp70
Deliver preproteins to the mitochondrial import receptor Tom70. Cell 112(1):4150.
Yun BG, Huang W, Leach N, Hartson SD, Matts RL. (2004) Novobiocin induces a
distinct conformation of Hsp90 and alters Hsp90-cochaperone-client interactions.
Biochemistry 43(25):8217-29.
Zhou H, Li XM, Meinkoth J, Pittman RN. (2000) Akt regulates cell survival and
apoptosis at a postmitochondrial level. J Cell Biol 151(3):483-94.
Znamensky V, Akama KT, McEwen BS, Milner TA. (2003) Estrogen levels regulate the
subcellular distribution of phosphorylated Akt in hippocampal CA1 dendrites. J
Neurosci 23(6):2340-7.
146
APPENDIX A
IACUC APPROVAL
147
APPENDIX B
IRB APPROVAL
148