One key to two doors
Dual targeting peptides and membrane mimetics
Weihua Ye
©Weihua Ye, Stockholm University 2015
Cover picture: Magnetman 2015 by Weihua Ye
ISBN 978-91-7649-159-1, pp. 1-69
Printed in Sweden by Universitetsservice AB, Stockholm 2015
Distributor: Department of Biochemistry and Biophysics, Stockholm University
To my family
Abstract
A targeting peptide at the N-terminus of a precursor protein usually directs
the protein synthesized in the cytosol to a specific organelle in the cell. Interestingly, some targeting peptides, so-called dual targeting peptides (dTPs)
can target their protein to both mitochondria and chloroplasts. In order to
understand the mechanism of dual targeting, a dTP from threonyl tRNA
synthetase (ThrRS-dTP) was investigated as a model dTP in this thesis work.
The results suggest that ThrRS-dTP is intrinsically disordered in solution but
has an α-helical propensity at the N-terminal part. Tom20 and Toc34 are the
two primary receptors on the outer membranes of mitochondria and chloroplasts, respectively. We found that the N-terminal half of the ThrRS-dTP
sequence, including an amphiphilic helix, is important for the interaction
with Tom20. This part also contains a  motif, where  represents a
hydrophobic/aromatic residue and  represents any amino acid residue. In
contrast, neither the amphiphilic helix nor the  motif in ThrRS-dTP
has any special role for its interaction with Toc34. Instead, the entire sequence of ThrRS-dTP is important for Toc34 interaction, including the Cterminal part which is barely affected by Tom20 interaction.
In addition, the role of lipids in the organelle membrane for the recognition
of dual targeting peptides during protein import is also the focus of this thesis. The tendency to form α-helix in ThrRS-dTP, which is not observable in
solution by CD, becomes obvious in the presence of lipids and DPC micelles. To be able to study such interactions, DMPC/DHPC isotropic bicelles
under different conditions have also been characterized. These results
demonstrate that bicelles with a long-chained/short-chained lipid ratio q =
0.5 and a concentration larger than 75 mM should be used to ensure that the
classical bicelle morphology persists. Moreover, we developed a novel
membrane mimetic system containing the galactolipids, MGDG or DGDG,
which have been proposed to be important for protein import into chloroplasts. Up to 30% MGDG or DGDG lipids were able to be integrated into
bicelles. The local dynamics of the galactolipids in bicelles displays two
types of behavior: the sugar head-group and the glycerol part are rigid, and
the acyl chains are flexible.
i
List of papers
The following papers, referred to in the text by their Roman numerals, are
included in this thesis. Reprints were made with permission from the publishers.
I.
Ye W, Spånning E, Unnerståle S, Gotthold D, Glaser E, and Mäler L.
NMR investigations of the dual targeting peptide of Thr-tRNA
synthetase and its interaction with the mitochondrial Tom20 receptor
in Arabidopsis thaliana. FEBS J. 279: 3738-3748 (2012).
II.
Ye W*, Spåning E*, Glaser E, and Mäler L. Interaction of the dual
targeting peptide of Thr-tRNA synthetase with the chloroplastic
receptor Toc34 in Arabidopsis thaliana. In press, FEBS Open Bio.
(2015).
III.
Ye W, Lind J, Eriksson J, and Mäler L. Characterization of the
morphology of fast-tumbling bicelles with varying composition.
Langmuir 30: 5488-5496 (2014).
IV.
Ye W, Liebau J, and Mäler L. New membrane mimetics with
galactolipids: Lipid properties in fast-tumbling bicelles. J. Phys.
Chem. B 117: 1044-1050 (2013).
Additional papers
Ariöz C, Ye W, Bakali A, Ge C, Liebau J, Götzke H, Barth A, Wieslander
Å, and Mäler L. Anionic lipid binding to the foreign protein MGS provides a
tight coupling between phospholipid synthesis and protein overexpression in
Escherichia coli. Biochemistry 52: 5533-5544 (2013).
Brown C*, Szpryngiel S*, Kuang G, Srivastava V, Ye W, McKee LS, Tu Y,
Mäler L, and Bulone V. Structural and functional characterization of the
“Microtubule Interacting and Trafficking” domain of two oomycete chitin
synthases. Manuscript (2015).
*These authors contributed equally to this work.
ii
Table of contents
Abstract
i
List of papers
ii
Table of contents
iii
Abbreviations
v
1. Introduction ............................................................................................... 1
2. Background ............................................................................................... 3
2.1 Protein import into mitochondria and chloroplasts .............................. 3
2.1.1 Import into mitochondria .............................................................. 4
2.1.2 Import into chloroplasts .............................................................. 10
2.2 Dual targeting and the dual targeting peptide..................................... 14
2.2.1 Dual targeting ............................................................................. 14
2.2.2 The dual targeting peptide of Thr-tRNA synthetase ................... 16
2.3 Intrinsically disordered proteins/peptides........................................... 18
2.4 Biological membranes ........................................................................ 19
2.4.1 Membrane lipids ......................................................................... 19
2.4.2 Membrane proteins ..................................................................... 22
2.4.3 Protein lipid interactions ............................................................. 22
2.5 Membrane mimetic systems ............................................................... 23
2.5.1 Micelles ...................................................................................... 24
2.5.2 Bicelles ....................................................................................... 25
2.5.3 Vesicles ....................................................................................... 29
2.5.4 Other systems ............................................................................. 29
3. Methods.................................................................................................... 30
3.1 Nuclear magnetic resonance (NMR) spectroscopy ............................ 30
3.1.1 Chemical shifts and assignment.................................................. 30
3.1.2 NMR relaxation .......................................................................... 31
3.1.3 Line broadening .......................................................................... 33
3.1.4 Diffusion NMR ........................................................................... 34
3.2 Dynamic light scattering (DLS) ......................................................... 35
3.3 Circular dichroism (CD) spectroscopy ............................................... 36
4. Results and discussion ............................................................................ 38
iii
4.1 ThrRS-dTP(2-60) (Paper I and Paper II) ............................................ 38
4.1.1 Structural properties of ThrRS-dTP(2-60) .................................. 38
4.1.2 The interaction between ThrRS-dTP(2-60) and AtTom20 ......... 40
4.1.3 The interaction between ThrRS-dTP(2-60) and AtToc34........... 42
4.2 Isotropic bicelles (Paper III and Paper IV) ......................................... 43
4.2.1 DMPC/DHPC bicelles under different conditions...................... 43
4.2.2 Bicelles with galactolipids introduced ........................................ 45
4.2.3 Lipid dynamics in bicelles with galactolipids............................. 46
5. Conclusions and outlook......................................................................... 49
Sammanfattning på svenska ...................................................................... 51
Acknowledgements ..................................................................................... 53
References .................................................................................................... 55
iv
Abbreviations
aaRS
CD
CMC
CSA
cTP
DHPC
DLS
DMPC
DMPG
DPC
dTP
HSQC
Hsp
IDP
LUV
MPP
mTP
NMR
NOE
NOESY
PFG
POPC
PreP
SAM
SDS
SPP
SUV
TEM
amino acyl tRNA-synthetase
circular dichroism
critical micelle concentration
chemical shift anisotropy
chloroplastic targeting peptide
L-a-1,2-dihexanoylphosphatidylcholine
dynamic light scattering
1,2-dimyristoyl-sn-glycero-3-phosphocholine
1,2-dimyristoyl-sn-glycero-3-phospho-1-glycerol
dodecylphosphocholine
dual targeting peptide
heteronuclear single quantum coherence
heat shock protein
intrinsically disordered protein
large unilamellar vesicles
mitochondrial processing peptidase
mitochondrial targeting peptide
nuclear magnetic resonance
nuclear Overhauser effect
nuclear Overhauser enhancement spectroscopy
pulsed field gradient
1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine
presequence protease
Sorting and assembly complex
sodium dodecyl sulfate
stromal processing peptidase
small unilamellar vesicle
transmission electron microscopy
v
ThrRS
TIC
TIM
TOC
TOCSY
TOM
threonyl-tRNA synthetase
translocon at inner envelope membrane of chloroplasts
translocase of inner mitochondrial membrane
translocase of outer envelope membrane of chloroplasts
total correlation spectroscopy
translocase of outer mitochondrial membrane
*More abbreviations of lipids can be found in Table 1 in Section 2.4.1.
vi
1. Introduction
The term “cell” was coined 350 years ago when Robert C. Hooke reported
what he had observed in a thin slice of cork through a self-made microscope
(Hooke. 1665, Mazzarello. 1999). Cells are now considered to be the smallest building blocks of all life forms. As a boundary, the semi-permeable
membrane separates the outside from the inside of a cell. In most cells, compartments with specialized functions have been further enclosed by their
membranes and these subcellular structures are called organelles. The biological membrane is of great importance not just because it is the boundary
but even more because it carries out the communication between the inside
and the outside. With the communication and the selective permeability,
biological membranes control the exchange of material between the cells and
organelles and therefore sustain the cell viability.
For a plant cell, two of the most important organelles are mitochondria and
chloroplasts, both of which are involved in energy conversion. The mitochondrion, known as “cellular power plant”, is where cellular respiration
takes place, while the chloroplast is where photosynthesis occurs, converting
light energy into biochemical energy. Most mitochondrial and chloroplastic
proteins are synthesized in the cytosol and imported into the respective organelle post-translationally. The majority of them harbor a signal peptide at the
N-terminus that can be specifically recognized by certain proteins located on
the mitochondrial and chloroplastic membrane, acting as receptors. Figuratively speaking, these membrane proteins or other components that regulate
the entering are doors of the organelles while the N-terminal signal peptides
are the keys. In this way, only proteins destined for mitochondria can be
imported into mitochondria and the same applies to chloroplastic proteins.
This door-key relation in the cell is usually very specific. Interestingly, there
is a group of proteins that can be delivered to both mitochondria and chloroplasts, namely dual-targeted proteins. How does a protein use the same key
to unlock two doors? In order to shed some light on this question, a dual
targeting peptide from threonyl tRNA synthetase (ThrRS-dTP) has been
studied in this thesis. The interactions of this peptide and two gatekeeper
proteins from mitochondria and chloroplasts, respectively, have been investigated.
1
In addition, membrane lipids of organelles play significant roles for protein
import into mitochondria and chloroplasts. However, it is difficult to work
with the mitochondrial and chloroplastic membranes due to the complexity.
Therefore, systems that mimic different membranes are used in different
research areas. In this thesis, membrane mimetic systems, bicelles, with different compositions have been characterized, providing tools for investigating the interaction between lipids and proteins, which include the prospect of
studying the role of lipids in the protein import into mitochondria and chloroplasts. Moreover, studies of membrane mimetic media is an important
topic on its own, since we need to know more about the role of lipids in
membrane protein function.
Nowadays, more and more advanced methods are developed to solve biological questions. During my thesis work, mainly biophysical methods have
been applied, including different optical spectroscopic methods and nuclear
magnetic resonance (NMR) spectroscopy.
2
2. Background
2.1 Protein import into mitochondria and chloroplasts
Both mitochondria and chloroplasts contain their own genetic material
(DNA), transcription and translation systems. Among other evidence, this
indicates that mitochondria and chloroplasts evolved from bacteria by endosymbiotic events, which in brief mean that two organisms lived together
with one inside the other. The endosymbiotic theory can be traced back to
late 19th century when it was observed that the division of chloroplasts in
plant cells closely resembles that of free-living cyanobacteria (Schimper.
1883). Later in the 1920s, this idea was extended to the origin of mitochondria (Wallin. 1923, Wallin. 1927). Though initially dismissed, the endosymbiotic theory is now widely accepted after more evidence is found (Ris et al.
1961, Sagan. 1967, Stocking et al. 1959).
Over evolutionary time, most of the genes from the preexisting invading
prokaryotes were transferred to the nucleus (Kleine et al. 2009, Martin.
2003). However, DNA encoding some proteins and RNAs remained in the
organelles in a circular form. Due to evolutionary pressure, the mitochondrial genome is highly divergent. The smallest mitochondrial DNA genome
(mtDNA) contains 6000 base pairs (bp) encoding only 3 proteins, which is
from Plasmodium falciparum, a protozoan parasite causing malaria in humans (Gray. 2012, Lang et al. 1999). Human mtDNA contains 17 kbp encoding 13 proteins, 2 rRNAs and 22 tRNAs (Lodish et al. 2002). While
mammalian mtDNA lacks introns and other non-coding sequences, plant
mtDNAs contain these together with pseudogenes and pieces of foreign
DNA, which gives rise to much larger genomes, e.g. mtDNA in Arabidopsis
thaliana contains 367 kbp encoding 31 proteins among other things. The size
of the chloroplastic genome also varies from 120 kbp (liverwort) to 160 kbp
(tobacco) depending on the species. In A. thaliana, the chloroplastic genome
(ctDNA) is composed of 154 kbp including 87 protein-coding genes, 4
rRNA genes and 37 tRNA genes (Sato et al. 1999). The retained genes are
essential for the organelle functions, such as e.g. respiratory chain reactions
and carbon assimilation.
3
Despite the fact that a small number of proteins are directly synthesized inside mitochondria and chloroplasts, the proteomes of both organelles are
relatively large. There are 1000-1500 mitochondrial proteins and more than
2000 chloroplastic proteins, most of which are synthesized in the cytosol and
imported into the respective organelle (Millar et al. 2006, Sickmann et al.
2003). Proteins usually exert their functions at specific locations in a cell, e.g.
the endoplasmic reticulum (ER), but most proteins are nuclear encoded.
How does a predestined protein reach the correct location? Importantly,
many of them have to be transported into or across one or two membranes.
This process, namely intracellular protein sorting or protein trafficking, is
very complex. Taking proteins destined for the ER membrane as an example,
some are directed to the ER at the same time as being synthesized while
others after the full length protein is synthesized, i.e. co-/post-translational
targeting. Targeting to mitochondria and chloroplasts is a post-translational
process in most cases. Albeit not common, co-translational import exists
indeed, e.g. import of CoxIVp and fumarase into mitochondria (Yogev et al.
2011). In short, mitochondrial/chloroplastic import is implemented via the
assistance of organelle-specific targeting signals within the precursor protein,
cytosolic factors, receptors on the organelle membrane and organellespecific protein import machineries. More details will be described in the
following part of this section.
Glossary (Kmiec et al. 2014)
Targeting signal: a general term for a signal in a sequence that directs protein
to its destination in a cell. The sequence may or may not be cleaved off. The
destination may be an organelle or an intra organellar compartment.
Targeting peptide: a cleavable peptide that directs protein to its destination
in a cell. The peptide is in most cases at the N-terminus of a precursor protein.
Targeting peptide includes both presequence and transit peptides.
Presequence: a cleavable peptide that directs protein to mitochondria.
Transit peptide: a cleavable peptide that directs protein to chloroplasts.
2.1.1 Import into mitochondria
The mitochondrion is typically round or oval in shape and 0.5-10 µm in diameter. Though it can be found in almost all cells, the number of mitochondria per cell varies greatly, ranging from 100 to 10,000 in different plant
cells (Mauseth. 2014). In extreme cases, human red blood cells do not have
any mitochondria while 100,000 were found in the cells of giant amoeba
Pelomyxa (Wolfe. 1972). However, regardless of their size and number in
different species or tissues, they have very similar structures. They are composed of the outer membrane (OM), the inner membrane (IM), the intermembrane space (IMS), and the viscous matrix, which is bounded by the IM.
The OM is freely permeable to small molecules and contains many special
4
channels, called porins (Alberts et al. 1994). The IM has a large surface area
due to the convolution structure of cristae. In contrast to the OM, the IM is
only permeable to oxygen, carbon dioxide and water. The large impermeability of ions across the IM maintains the membrane potential that is formed
by the action of the enzymes of the electron transport chain and drives the
production of adenosine triphosphate (ATP). Therefore, sophisticated translocases exist in the OM and IM to assist specific molecules to cross the two
membranes.
Figure 1. Import pathways of mitochondrial precursor proteins in plants. Blue
line indicates import to matrix: a. general import pathway. Purple lines indicate import to IM: b. stop transfer pathway, c. conservative sorting pathway
and d. carrier pathway. OXA is the cytochrome oxidase assembly. Orange
line indicates import to IMS: e. MIA pathway. Green lines indicate import to
OM: f. SAM pathway and g. TOM-only pathway. The TOM complex is engaged in the insertion of proteins into or across the OM. The TIM23 complex
is engaged in translocation of precursor proteins with presequence, mainly to
the matrix but also to the IM. The TIM22 complex is engaged in translocation
of precursor proteins without presequences to the IM. The SAM (sorting and
assembly machinery) complex mediates the insertion of β-barrel protein to the
OM. MIA (mitochondrial IMS assembly) facilitates locating cysteine-rich
proteins to the IMS. Cleavable N-terminal targeting peptides are indicated in
red. Hydrophobic sorting domains at the N-terminal are indicated as box in
grey. Adapted from (Murcha et al. 2014, Pon et al. 2007).
As mentioned at the beginning of this section, the vast majority of mitochondrial proteins are nuclear encoded and imported into mitochondria after
being synthesized on the cytosolic ribosomes. Sorting them to the correct
5
mitochondrial sub-compartment is essential for their individual function as
well as mitochondrial biogenesis and cell viability. An overview of protein
import into different mitochondrial compartments is shown in Fig. 1.
Import into matrix
Proteins destined for the mitochondrial matrix are imported via a pathway
that involves mainly two complexes in the OM and IM, respectively: translocase of the outer mitochondrial membrane (TOM complex/machinary) and
translocase of the inner mitochondrial membrane (TIM complex/machinary).
These precursor proteins are synthesized with a signal targeting peptide at
the N-terminus. The mitochondrial targeting peptide (mTP), also known as
presequence, is first recognized by receptors of the TOM complex. After this
the precursor protein is translocated through the TOM machinery. With the
C-terminal part still anchored to the TOM machinery, the protein is further
transferred to the TIM23 machinery via certain contacts and finally translocated through TIM23 into the matrix in an ATP and membrane potential
dependent way. In addition, the presequence translocase associated motor
(PAM) is required for TIM23 association, which drives further inward pulling of preprotein together with Heat shock protein 70 (Hsp70) in an ATPdependent manner. After translocation, the targeting peptide is cleaved off
by the mitochondrial processing peptidase (MPP) and degraded by presequence protease (PreP) while the mature protein is folded with the assistance
of molecular chaperones (Bolender et al. 2008, Lister et al. 2005, Murcha et
al. 2014, Neupert et al. 2007, Stahl et al. 2002).
The TOM complex has a molecular mass of about 500-600 kDa and it is
composed of the initial receptors, the import pore Tom40 and some small
Tom proteins facilitating the dynamics and stability of the machinary. In
yeast and plants, the cytosolic facing receptors include the initial receptor
Tom20, which can bind to both cleavable and non-cleavable signals via hydrophobic interactions, Tom22 (Tom9 in plants), which recognizes cleavable
presequences via inonic interactions and Tom70 (not present in plant), which
is responsible for importing polytopic membrane proteins with internal targeting signals. The Tom40s are β-barrel proteins in the OM and form cationselective translocation channels with other TOM components. The TIM23
complex includes a membrane sector and a motor sector. The membrane
sector consists of Tim50, Tim23, Tim17 and Tim21, which form the protein
translocation pore. The motor sector maintains the driving force for translocation with many proteins involved, e.g. Tim44, Tim14 (Pam18), Tim16
(Pam16), mtHsp70 and Mge1 (Neupert et al. 2007). During the translocation, a TOM-TIM supercomplex together with the precursor probably forms
(Chacinska et al. 2010, Murcha et al. 2014).
Import into other submitochondrial compartments
6
The IMS proteins are mainly imported through the MIA pathway. They usually contain internal cysteine-rich signals, which can be recognized by
Mia40. Then intramolecular disulfides in the precursor protein can be
formed. Depending on the topology and the location of the import signal in
the sequence, membrane proteins destined for the OM and IM follow different import and insertion pathways. This will be mentioned only briefly as it
is beyond the scope of this thesis. The OM proteins are imported mainly
through two routes, β-barrel proteins follow the SAM (sorting and assembly
machinery) pathway despite the lack of an N-terminal targeting peptide,
while proteins with either an N-terminal or a C-terminal transmembrane
segment are imported via the TOM complex with Tom40 involved but not
Tom20 or Tom70. The IM proteins can be divided into three groups following the “carrier pathway”, the “stop transfer pathway” and the “conservative
sorting pathway”. The hydrophobic metabolite carriers, like ADP/ATP carriers, usually contain multiple internal targeting signals instead of an Nterminal cleavable mTP and they follow a TOM/TIM22 pathway. Proteins
following the stop transfer pathway are arrested in the TIM23 complex and
then laterally released into the IM. Proteins following the conservative sorting pathway are inserted into the IM after translocation into the matrix. The
anchor of a few IM proteins can be cleaved, which gives rise to an IMS protein (Schulz et al. 2014).
Import signals
There are different types of mitochondrial import signals in the precursor
proteins: cleavable N-terminal targeting peptides are usually applied for
matrix proteins; N-terminal or C-terminal hydrophobic segments flanked
with positive charge residues is used by the OM proteins with single transmembrane segment; a presequence followed by a hydrophobic sorting domain at the N-terminus is present in some proteins destined for the IM or
IMS and many polytopic proteins that reside in the OM and IM contain multiple internal targeting signals spread over the preprotein. However, matrixdestined proteins devoid of a cleavable N-teminal presequence also exist,
some with a non-cleavable targeting signal (Hammen et al. 1996, Jarvis et
al. 1995, Waltner et al. 1995) and some without any conventional matrixtargeting signal (Sanyal et al. 1995).
Among the different types of mitochondrial import signals, the cleavable Nterminal signal has been best characterized. Matrix-targeting sequences indeed have some common properties although they do not share sequence
homology. Presequences, which are necessary and sufficient to direct proteins into the matrix, vary in length from 6 to 122 amino acids, usually about
20-70. In general, plant presequences are longer, e.g. the lengths of most A.
thaliana presequences are 42-50 amino acid residues (Huang et al. 2009,
Zhang et al. 2002). Presequences have a high content of hydrophobic, hy7
droxylated and positively charged amino acids while negatively charged
residues are rarely present. Amino acid distribution analysis of a large number of mitochondrial preproteins from A. thaliana shows that serine is the
most abundant amino acid at almost every position but less abundant towards the C-terminus, and that leucine, arginine and alanine are also particularly enriched (Bhushan et al. 2006, von Heijne. 1986, Zhang et al. 2002). In
addition, arginine has been shown to be significant for import by mutagenesis studies. Another property that most mitochondrial presequences possess
in their sequences is the potential to form an amphiphilic α-helix in which
positively charged residues and hydrophobic residues are located on opposite
faces of the helix. The importance of amphiphilicity has been verified by the
import ability of the synthetic amphiphilic α-helix made up of only leucine,
serine and arginine (Roise et al. 1988, Roise et al. 1986, von Heijne. 1986).
Interaction on the surface of the OM
Interactions of presequence with lipid bilayers have been observed in several
studies (Hammen et al. 1996, Leenhouts et al. 1993, Leenhouts et al. 1994,
Toeroek et al. 1994). Several presequences are found to interact with lipids
in a charge-dependent manner in membrane models. It was proposed that the
amphiphilic helix facilitates the insertion into the membrane and positively
charged residues like arginine are crucial for recognition of the presequence
by the initial receptor Tom20 (Pon et al. 2007). The NMR structure of rat
Tom20 in complex with a presequence, however, revealed that it was the
hydrophobic interactions rather than ionic ones that dominated the interaction (Abe et al. 2000) (Fig. 2A). In addition, the amphiphilic helix was present without a lipid environment. Therefore, the lipid membrane is not considered essential for the recognition of a presequence although the role of
lipids is not eliminated.
Some of the Tom and Tim proteins, e.g. Tom7, Tom20 and Tom40 have
counterparts in different organisms (Endo et al. 2002, Perry et al. 2006).
Interestingly, the functional/structural counterparts from different organisms
may be homologous while in some cases they are not (Perry et al. 2006).
The structure of rat Tom20 in complex with a presequence was first determined by solution NMR in 2000 and later crystal structures with different
modifications were made (Abe et al. 2000, Saitoh et al. 2007, Saitoh et al.
2011). As shown in Fig. 2A, Tom20 from rat has five helices, four of which
are well defined while the fifth is loose and separated from the four-helix
domain. The presequence binds to a groove in Tom20, which consists of
hydrophobic residues and surrounded by hydrophilic residues. The peptide is
in fact buried in a shallow location with positive residues exposed. The
structures shown here represent the cytosolic domain of Tom20, in which the
membrane anchoring parts have been removed. Interestingly, although
Tom20 from animals and plants share the same function and some structural
8
properties, they are anchored to the membrane in reversed order. The animal
Tom20 is N-terminally anchored to the outer membrane while the plant
Tom20 is C-terminally anchored. It is believed to be the results of convergent evolution (Perry et al. 2006). AtTom20 is larger than RnTom20, having
seven helices. Along with two more helices, there is also one more tetratricopeptide repeat (TPR) motif in AtTom20, compared to RnTom20. The TPR
motif is a potential region where protein-protein interaction takes place
(Blatch et al. 1999, Das et al. 1998). For both RnTom20 and AtTom20, TPR
motifs are involved in the interaction with the targeting peptide, but do not
fully overlap with the position where the largest chemical shift perturbation
happens upon addition of targeting peptide. As shown in Fig. 2B, the hydrophobic binding groove of AtTom20 is larger but is still very shallow and
there appears to be two possible binding sites.
A
B
N
C
N
C
Figure 2. Structures of the cytosolic domain of Tom20. A showes one of the
20 NMR-derived structures of RnTom20 (PDB: 1OM2, (Abe et al. 2000) ).
The bound presequence, pALDH(12-22), from aldehyde dehydrogenase is
displayed in transparent cyan. B showes one of the 20 NMR-derived structures of AtTom20 (PDB: 1ZU2, (Perry et al. 2006)). Helices 1 and 3 in
RnTom20 and Helices 1 to 4 together with the small helix at the C-terminus
in AtTom20 are the regions that are involved in the interaction with presequences based on NMR chemical shift perturbation experiements (in colors).
The hydrophobic residues in the interaction patch are indicated in yellow
(Ala, Val, Ile, Leu and Phe). Asp, Gln, Asp and Glu in the peripheral regions
are indicated in red. Positively charged residues in these regions are indicated
in green.
Amphiphilic α-helices have been shown to play an important role in mitochondrial import (Moberg et al. 2004). Based on the structure of rat Tom20
with a bound presequence peptide, a hydrophobic patch in the Tom20
groove was found to provide the interaction surface between the receptor and
the presequence peptide. Mutagenesis studies showed that the ionic interaction between the positive residues from presequence and the negative residues from the periphery of the groove of RnTom20 are dispensable, while
the hydrophobic interactions are essential (Abe et al. 2000). However, the
9
positively charged residues of the presequence have been shown to be important for import efficiency (Tanudji et al. 1999). Therefore, the positively
charged residues are potentially recognition elements for other protein components in the mitochondrial import pathway. A common pattern corresponding to  was identified by NMR chemical shift perturbation experiments on a large number of presequences, where  represents a hydrophobic/aromatic residue and  represents any amino acid residue, but in
most cases  has a long aliphatic side-chain (Muto et al. 2001). This fiveresidue motif is not always localized to the N-terminal part of the presequence although the presequences have the same orientation when bound to
Tom20. Furthermore, two bound states are proposed as different hydrophobic residues in the preseqence  motif are found in the hydrophobic
patch of Tom20 in the snapshots of the crystal structures (Saitoh et al. 2007).
Notably, these two different states are not due to the slightly different tether
between the peptide and Tom20. NMR relaxation analysis further indicates
the existence of an equilibrium between different states, as well as extensive
motion at the interface between the presequence and Tom20 receptor (Saitoh
et al. 2007, Saitoh et al. 2011). In A.thaliana there are four different
isoforms of Tom20 (Lister et al. 2004). The structure derived from NMR is
for isoform AtTom20-3. In Paper I of this thesis, the interaction position of a
dual targeting peptide was mapped by titration of AtTom20-4 isoform to the
peptide.
2.1.2 Import into chloroplasts
In the cells of plants and algae, chloroplasts are a type of plastids where important chemical compounds are produced and stored. It is specialized in
photosynthesis. In addition, it also carries out syntheses of amino acids, fatty
acids and lipid components of its own membranes. Chloroplasts in land
plants have an oval shape about 5-10 µm long and 3-4 µm thick. In analogy
with mitochondria, the number of chloroplasts per cell varies largely. In
algae, there is usually only one chloroplast with a shape that varies significantly. In plants like wheat, there are as many as 20 to 100 chloroplasts per
leaf cell (Milo et al. 2014). In spinach, the range of 100 to 1000 per cell was
reported (Possingham et al. 1969).
Chloroplasts are composed of three membrane systems and three fluid spaces in between these membranes. The outer envelope membrane (OEM) and
the inner envelope membrane (IEM) form the boundaries of the organelle
and give rise to the aqueous inter membrane space (IMS). Inside the doublemembrane envelope is the semi-gel-like stroma, where the thylakoid system
floats. The thylakoid system is the third membrane system, which consisits
of stacks of flattened lamellar vesicles where the light reactions of photosynthesis occur. The third fluid space is located inside the thylakoids, namely
10
the lumen. The OEM is permeable to small molecules like most ions and
metabolites while the IEM is well sealed but allows passage under the control of specialized transporter proteins. Nonetheless, chloroplastic proteins
synthesized in the cytosol cannot freely pass any of the two membranes.
Protein import into chloroplasts is similar to mitochondrial protein import.
However, it is even more complicated. Analogous to the TOM/TIM pathway
for mitochondrial proteins, the TOC/TIC pathway is the “general import
pathway” for chloroplastic proteins. TOC stands for translocase of outer
envelope membrane of chloroplast, and TIC for translocase of inner envelope membrane of chloroplast. However, an alternative TOC/TICindependent pathway has also been found, e.g. some proteins enter chloroplasts via vesicle trafficking from ER (Aronsson et al. 2009). Representative
import pathways to chloroplast can be found in Fig. 3. In recent reviews,
new progresses have been incorporated to improve the conventional models
(Aronsson et al. 2009, Kovács-Bogdán et al. 2010, Paila et al. 2015, Richardson et al. 2014, Soll et al. 2004).
Figure 3. Outlines of protein import into the chloroplast. A general import
pathway to the stroma is indicated with solid lines. Other possible pathways
are indicated with dashed lines. Adapted from (Bruce. 2001).
Import into stroma
Precursor proteins destined for chloroplast stroma, contain an N-terminal
extension, namely a transit peptide or chloroplastic targeting peptide (cTP).
It can be recognized by the receptors of the TOC complex and translocated
11
through the TOC and TIC complexes in an energy-dependent manner. The
cTP binds to the import apparatus reversibly and cannot be delivered further
in the absence of an energy source (Perry et al. 1994). The binding becomes
irreversible when GTP and low concentrations of ATP are provided and the
cTP can reach the IMS as a component in the intermediate complex (Young
et al. 1999). To complete the translocation process, ATP at higher concentration as well as certain chaperones are required in stroma (Pain et al. 1987,
Theg et al. 1989). After import, the cTP is cleaved off by the stromal processing peptidase (SPP) and then further degraded by the presequence protease (PreP) (Benz et al. 2009, Jarvis. 2008, Stahl et al. 2002).
The TOC complex is about 550 kDa, mainly consists of Toc34, Toc159 and
Toc75. The Toc34 and Toc159 components are membrane-integrated receptors, but which one is the primary receptor is debated. Toc75, belonging to
the same family as Tom40, is a cation selective ion channel and forms a βbarrel in the OEM. An integral outer membrane protein, Toc64, is also involved in protein import via Hsp90 in the cytosol (Aronsson et al. 2007).
Although the composition of the TIC complex is still a question under debate, it is known that it involves the channel-forming proteins, a motor complex and regulatory factors. The candidates for the channel components include Tic110, Tic20, Tic21 and a protein import related anion channel (PIRAC) (Kovács-Bogdán et al. 2010). The motor complex driving transport
into the stroma upon large consumption of ATP putatively includes Tic110,
Tic40 and Hsp93. Tic55, Tic62 and Tic32 are the three putative regulator
factors that control import by sensing the chloroplast redox status (KovácsBogdán et al. 2010). Note that the components here are for the “general import pathway”. Mutagenesis studies in Arabidopsis indicated the existence of
additional components and different combinations of these components in
the complexes/supercomplex (Aronsson et al. 2009, Paila et al. 2015).
Import into other chloroplast compartments
The majority of proteins destined for OEM are inserted into the membrane
with an intrinsic non-cleavable targeting signal via the TOC complex in an
unknown way. However, Toc75 is imported with a targeting signal which
has two parts, so-called bipartite TP. It is cleaved off in the stroma and the
IMS in steps (Inoue et al. 2003). Like the situation in mitochondria, most
proteins targeted to the IEM in chloroplasts follow two routes: the stop transfer pathway and conservative sorting pathway. Proteins using the former
route diffuse laterally into the membrane after reaching the TIC complex
(Firlej-Kwoka et al. 2008). Proteins using the latter route insert into the
membrane after being imported into the stroma. Only a limited number of
proteins destined for IMS have been characterized. Additionally, a group of
proteins is further imported into thylakoids after the first part of their bipartite TP is cleaved off in the stroma. Then according to the sorting signal at
12
the second part of the TP, they are transported into the thylakoid membrane
or lumen (Jarvis et al. 2004, Schünemann. 2007).
Import signals
Chloroplast targeting peptides (cTPs), again, are rich in hydrophobic, hydroxylated and positively charged amino acids, but lack of negatively
charged amino acids. However, there are also some differences including
length, amino acid distribution and structural properties. Compared to mTPs,
cTPs contain more serines and prolines but fewer arginines within the 16 Nterminal amino acids (Ge et al. 2014). Secondly, cTPs are generally longer
than mTPs. They have an average length of 60 amino acids although the
length varies within a very wide range, from 13 to 146 residues (Bhushan et
al. 2006, Zhang et al. 2002). Structurally, cTPs have been found to be unstructured in solution in most studies (Krimm et al. 1999, Lancelin et al.
1996, Roise et al. 1988, von Heijne et al. 1991) but some cTPs have been
reported to form helical structures in a membrane mimetic environment
(Berglund et al. 2009a, Bruce. 2000, Krimm et al. 1999).
Interaction on the surface of OEM
Several mechanisms have been suggested for the initial contact of a preprotein with the chloroplast OEM. Firstly, cytosolic factors may be involved in
this process. Protein 14-3-3 and Hsp70 form a so-called guidance complex
that targets the preprotein to the TOC complex in the OEM. Binding to the
guidance complex requires a certain serine or threonine in the cTP to be
phosphorylated. Albeit not a prerequisite, the guidance complex provides a
“fast track” where the transport is about 3-4 times faster (May et al. 2000). A
variant route for a subset of proteins is to use Hsp90 from the cytosol and
docking at the surface of the OEM via the putative receptor Toc64 (Aronsson et al. 2007). Secondly, specific lipids from the chloroplastic membrane have been reported to facilitate interactions between cTPs and the
TOC complex (Bruce. 1998). However, the role of lipids in the import process remains controversial (See Section 1.4.3). Finally, direct interactions
between preproteins and the receptors of the TOC complex have been suggested to be important. The two integral membrane GTPases, Toc34 and
Toc159 have been proposed to be primary receptors in the so-called “motor
model” and “targeting model”, respectively (Aronsson et al. 2009). Both
proteins bind preproteins specifically based on crosslinking studies (Jelic et
al. 2003, Kouranov et al. 1997, Ma et al. 1996, Smith et al. 2004, Sveshnikova et al. 2000). Toc34 consists of a cytosolic GTPase domain and a Cterminal transmembrane segment, while Toc159 consists of an N-terminal
acidic domain, a Toc34 related GTPase domain and a membrane anchor
domain about 54 kDa (Gutensohn et al. 2000, Hirsch et al. 1994, Li et al.
1997, Schnell et al. 1994). The homodimerization and heterodimerization of
13
Toc34 and Toc159 led to a hypothesis that Toc34 switches between the two
forms of dimer in the translocation cycle (Paila et al. 2015).
Toc34 was first identified in pea (Pisum sativum) and the structures of its
cytosolic part with GDP and GMP-PNP have been determined by crystallography (Sun et al. 2002). PsToc34 forms a dimer in endogenous membranes.
There are two isoforms of Toc34 in A. thaliana: AtToc33 and AtToc34,
which have different preprotein specificity. AtToc33 is responsible for targeting proteins related to photosynthesis and AtToc34 is responsible for proteins related to house-keeping functions to chloroplasts (Jarvis. 2008, Jarvis
et al. 1998, Kubis et al. 2003). Knockouts of AtToc33 and AtToc34 show a
phenotype in photosynthetic tissues and root respectively, while a double
null mutant is embryo-lethal. Crystal structures of AtToc33 with GDP and
GMP-PNP are also available in the PDB. Different from PsToc34, AtToc33
is in a monomer form and the dissociation factor for AtToc33 is much higher
than PsToc34 (Koenig et al. 2008, Paila et al. 2015, Voigt et al. 2005). Both
isoforms from A. thaliana are very similar to PsToc34 in sequence (Jarvis et
al. 1998, Voigt et al. 2005). The overall structures of PsToc34 and AtToc33
are also very similar. Moreover, both of them have a hydrophobic cavity
close to Helix-1, which is considered to be a putative preprotein binding site.
In addition, a shallow pocket acting as an extention of the cavity in
Toc33/Toc34 is formed where a polyethylene glycol is bound in the
Toc33/Toc34 crystal structures. In Paper II, the interaction of AtToc34 and
the dual targeting peptide ThrRS-dTP has been investigated by NMR spectroscopy.
2.2 Dual targeting and the dual targeting peptide
2.2.1 Dual targeting
Protein import is generally very specific, however, a growing number of
proteins that are products of a single gene, are found to be distributed to
more than one location in a cell (Chabregas et al. 2003, Karniely et al. 2005,
Peeters et al. 2001). This is understandable as the organelles share some
functions that require the same proteins, e.g. three NAD(P)H dehydrogenases are reported to be dual-targeted to peroxisomes and mitochondria in A.
thaliana (Carrie et al. 2009b). The most common cases in plant are proteins
dual-targeted to mitochondria and chloroplasts, as they share many functional properties including DNA replication, transcription, translation, and protection against oxidative stress (Peeters et al. 2001, Silva-Filho. 2003). To
date more than 100 proteins have been proposed to be dual-targeted to mitochondria and chloroplasts based on different experimental data (Carrie et al.
14
2013, Heazlewood et al. 2007). Most of them are from A. thaliana, although
dual-targeted proteins are also found in other plants like pea, maize and tobacco (Carrie et al. 2009a, Creissen et al. 1995, Rokov-Plavec et al. 2002).
As already mentioned in Section 1.1, most mitochondrial and chloroplastic
proteins are synthesized as precursor proteins equipped with import targeting
peptides at the N-terminus. For most dual-targeted proteins, it is also the
targeting peptide (TP) that functions as a recognition signal, but in this case
for both organelle import pathways, i.e. dual targeting peptide (dTP). The
basic properties of dTPs, such as amino acid contents and structural propensities, are similar to the mTP and the cTP. Again, hydroxylated, hydrophobic
and positively charged amino acid residues are abundant in dTPs and they
have few negatively charged amino acid residues (Berglund et al. 2009b, Ge
et al. 2014). Compared to both mTPs and cTPs, dTPs contain more phenylalanines, more leucines, more serines, but even fewer negatively charged
residues. On the other hand, dTPs contain more arginines than cTPs, but
fewer than mTPs. As for proline, it is the opposite. Accordingly, the physico-chemical property of dTPs is intermediate between mTPs and cTPs,
which is likely to allow the dual targeting capacity.
How can proteins be dual-targeted? One hypothesis is that there is a different
receptor specific for dTPs shared by both mitochondria and chloroplasts.
However, this is unlikely considering a lack of obvious properties of dTPs to
distinguish them from mTPs and cTPs (Berglund et al. 2009b, Carrie et al.
2013). Moreover, it has been observed that dual-targeted proteins inhibit the
import of mitochondrial proteins as well as choloroplastic proteins, indicating that they use the same receptors (Berglund et al. 2009a, Berglund et al.
2009b).
Two (or three) mechanisms have been proposed for different dual targeting
peptides (Fig. 4). Firstly, two twin targeting peptides, coming from alternative translation of the same gene, direct the protein in two different precursor
forms to mitochondria and chloroplasts, respectively. The matured protein
after the cleavage is the same in mitochondria and chloroplasts. It is usually
the protein with a longer TP being targeted to chloroplasts and the shorter
one to mitochondria (Chabregas et al. 2001, Peeters et al. 2001, Watanabe et
al. 2001). Secondly, the dual-targeted protein utilizes an identical peptide to
be recognized by both organelles, which can be further divided into two
categories based on the organization of the peptides. Import experiments of
dTPs with deletion of different segments show some dTP sequences organized in two separate domains, each of which is responsible for one organelle import (Chabregas et al. 2003) (Berglund et al. 2009a, Chabregas et al.
2003). For example, the N-terminal truncation of the targeting peptide for
glutathione reductase (Creissen et al. 1995) abolished the import into mito15
chondria but not into chloroplasts and the N-terminal truncation of PresequenceProtease (Bhushan et al. 2003) and RNA polymerase RpoT:2 (Hedtke
et al. 2000) can inhibit import to mitochondria or chloroplasts respectively.
The other possibility, probably more common, is the use of a truly ambiguous targeting sequence, which uses the same sequence with its properties
weaker than in either of the two TPs exclusively for one organelle, i.e. intermediate between the two TPs (Berglund et al. 2009a, Karniely et al. 2005,
Peeters et al. 2001, Silva-Filho. 2003). In addition, the mature portion of a
dual-targeted protein can also influence the targeting ability (Chew et al.
2003, Rudhe et al. 2002).
A. Twin targeting peptides
B. Ambiguous dTPs
1. Domain organized
2. Overlapping organelle determinants
Figure 4. Dual-targeted precursor proteins from one single gene. (A).Two
twin targeting peptides can be alternatively produced by one single gene, one
for mitochondrial import and the other for chloroplastic import. One single
dual targeting peptide can also be produced as dTPs with separate domains
(B.1) or with overlapping organell determinants (B.2) (Peeters et al. 2001).
2.2.2 The dual targeting peptide of Thr-tRNA synthetase
The ability of being imported into both mitochondria and chloroplasts turns
out to be a common feature for most aminoacyl-tRNA synthetases (aaRSs)
in A. thaliana. Among 24 organellar aaRSs with a targeting peptide, 18 are
identified to be dual-targeted (Duchêne et al. 2005). Interestingly, these dual-targeted aaRSs appear to have different origins, some from a plastid ancestor, some from mitochondria in animals and yeast, and some of eubacteri16
al origin (Duchêne et al. 2005). More interestingly, the dual targeting signals
are organized in different ways. N-terminal truncations of the dual targeting
peptides from 7 aaRSSs gave all four possiblities: no targeting to any organelle (TyrRS∆2-27, ValRS∆2-22 and ThrRS∆2-23), only targeting to mitochondria (ProRS∆2-20), only targeting to chloroplasts (AspRS∆2-16), and
still targeting to both organelles (IleRS∆2-22 and LysRS∆2-16) (Berglund et
al. 2009a). Therefore, it appears that TyrRS, ValRS and ThrRS are using
ambiguous targeting peptides with overlapping organelle determinan, and
that it is also the most common way of achieving dual targeting. However,
the molecular mechanisms of how the ambiguous dTPs mediate dual targeting to both organelles have not yet been elucidated.
Further studies of the dual targeting peptide of ThrRS (ThrRS-dTP) using Cterminal truncations of different lengths (50, 60, and 65 amino acids) revealed that the shortest sequence with dual targeting capacity was 60 amino
acid residues, although the predicted processing site suggested by ChloroP
for chloroplasts was after 38 residues and by MitoProt for mitochondria after
35 residues (Berglund et al. 2009b). In addition, import of known mitochondrial and chloroplastic precursors was inhibited in the presence of ThrRSdTP(2-60), which indicates that it uses the same import pathways as organelle specific proteins. Therefore, ThrRS-dTP(2-60) should be imported into
mitochondria via the TOM/TIM pathway and into chloroplasts via the
TOC/TIC pathway. Thus, the primary receptors in the two pathways have
been chosen and the interactions between ThrRS-dTP(2-60) and these receptors have been investigated in Paper I and Paper II, respectively.
2
10
20
A S S H S L L F S S S F L S K P S S F
21
30
40
T S S L R R F V Y L P T R Q F W P R Q R
41
50
H G F S T V F A V A T E P A I
60
S S S G P
Figure 5. Primary structure of ThrRS-dTP(2-60). The key motif L24RRFV28
for Tom20 binding is indicated in the red frame.
ThrRS-dTP(2-60) has a molecular weight of 6.6 kD and a pI value as high as
11.8. As seen from the sequence (Fig. 5), ThrRS-dTP(2-60) is very abundant
in Ser residues, 15 out of 59 residues, including 3 Ser residues at the Cterminus. This was the first time that a dual targeting peptide had been purified in chemical quantities (Berglund et al. 2009b). A pGEX construct with
the DNA sequence corresponding to amino acids residue 1-60 fused Cterminally to glutathione S-transferase (GST) was used. The GST-fused
ThrRS-dTP was overexpressed in Escherichia coli and purified from inclu17
sion bodies and then the GST was cleaved off, using CNBr, giving rise to the
peptide ThrRS-dTP(2-60), which could be further purified by ion exchange
chromatography. Previous CD spectroscopic results revealed that ThrRSdTP(2-60) is mainly unstructured in aqueous environment (Berglund et al.
2009b), and that 25-30% helix can be induced by different membrane mimetic media. In Paper I, additional characterization of this dual targeting
peptide has been done.
2.3 Intrinsically disordered proteins/peptides
More than a century ago, a “lock and key” model was proposed to explain
enzyme specificity for a substrate, in which the enzyme active site and the
substrate have complementary geometric shapes (Fischer. 1894). About half
a century ago, protein structures started to be determined with increasing
speed, giving information about their respective functions. As more and
more proteins structures became known, a structure-function paradigm has
gradually evolved. However, a growing number of proteins or regions of
proteins were found to be unstructured, namely intrinsically disordered protein (IDP). Although structural disorder was observed earlier in many studies, the generality of IDPs came to notice in the late 90s last century (Dunker
et al. 1998, Garner et al. 1998). Thereafter, the “lock and key” model and the
structure-function paradigm were reassessed or extended (Tompa. 2009,
Wright et al. 1999).
It is difficult to give an accurate definition of intrinsically disordered proteins. Nonetheless, IDPs are usually defined as proteins that do not adopt a
well-defined native structure in a solution under near-physiological conditions. IDPs are more and more recognized now as being involved in a variety
of biological processes via different molecular functions. The biological
processes include transcription regulation, and cellular signaling (Dunker et
al. 2015, Wright et al. 2015). As for molecular functions, IDPs are in most
cases involved in interactions with a molecular partner and the molecular
partner can be DNA, cytoskeleton, metal ion or another protein that can be
either folded or not (Tompa. 2009).
IDPs by the common definitions have no defined uniform structure. Instead
an ensemble of individual structures can be defined. However, information
about their structures from different perspectives and in different environment is of great importance to their functions. Firstly, IDPs are enriched in
certain amino acids (Dunker et al. 2001): Amino acids with high flexibility are enriched in IDPs; in contrast, hydrophobic residues providing hydrophobic contacts for protein folding are not common in IDPs. Generally, Ala,
18
Arg, Gly, Gln, Ser, Pro, Glu and Lys are disorder-promoting while Trp, Cys,
Phe, Ile, Tyr, Val, and Leu are order-promoting (Tompa. 2009). Secondly,
IDPs that display no detectable structure in one environment may be detected to contain a large portion of structure in another. As mentioned before,
mitochondrial targeting peptides are frequently reported to be unstructured in
buffer solutions but have α-helical structure in a membrane environment and
so does the dual targeting peptide ThrRS-dTP (Abe et al. 2000, Berglund et
al. 2009b). Another circumstance is that IDPs turn to be structured in the
presence of their functional partners although this most likely happens upon
the interaction. The α-helical structure of the Tom20 bound presequence of
pALDH is a good example. However, folding upon interactions is not always the case for IDP involved interactions (Sigalov. 2011). Last but not
least, many IDPs have transient local structures that are difficult to detect
from other observables than NMR chemical shifts. These structural properties together with the inherent dynamics of IDPs can give them functional
advantages over ordered proteins in protein-protein interactions (Dobson.
1993).
As mentioned previously in this thesis, the targeting peptides of different
sorts are unstructured in solution on their own. In comparison with other
IDPs, which have been extensively studied, e.g. β-amyloid peptide Aβ and
α-synuclein, targeting peptides have not drawn as much attention as IDPs.
How can they facilitate protein import with this disordered property? Do the
mTPs and cTPs utilize the same molecular mechanism to get recognized by
the corresponding receptors? Answers to these questions are essential to
understand the dual capacity of dTPs and elucidate a new type of functional
disorder in cells.
2.4 Biological membranes
Biological membranes constitute not only the boundaries of cells, but also
the boundaries of intracellular organelles. In addition, many important biological reactions and processes take place within/at membranes, e.g. photosynthesis, respiration, and signaling (Garrett et al. 2005). Biomembranes are
composed of a lipid bilayer matrix and a multitude of proteins. In addition,
some membrane lipids also play functional roles.
2.4.1 Membrane lipids
Biomembranes contain a large variety of lipid species, which vary in structure, e.g. in the head-group, and in length and saturation of their hydrophobic
acyl chains. Phospholipids, glycolipids and sterols are the three major clas19
ses of membrane lipids. Furthermore, both phospholipids and glycolipids
may consist of either glycerol or sphingosine. Hence, phospholipids include
phosphoglycerides and sphingomyelin, and glycolipids include glyceroglycolipids and glycosphingolipids. Lipids are amphiphilic, with a hydrophilic
head-group and one or two hydrophobic fatty acyl chains. Phospholipids
commonly found in biological membranes include phosphatidic acid (PA),
phsophatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), cardiolipin (CL), and phosphatidylinositol (PI). Common glycolipids in plants include monogalactosyl-diacylglycerol (MGDG/GalDAG),
digalactosyl-diacylglycerol (DGDG/GalGalDAG), and sulfoquinovosyldiacylglycerol (SQDG). Structures of those common glycerol lipids together
with diacylglycerol are given in Table 1.
Different membrane systems have different compositions of lipids. Prokaryotic E. coli inner membranes consist mainly of PE (~75%, (Cronan. 2003)),
while the cellular membrane of Acholeplasma laidlawii consists mainly of
glycolipids, mono- and diglucosyl-diacylglycerol (GlcDAG and
GlcGlcDAG, ~50%, (Österberg et al. 1995)). In eukaryotic cells, PC is the
most abundant lipid in almost all the organellar and plasma membranes for
mammalian cells although the percentage varies substantially (van Meer et
al. 2008, van Meer et al. 2011). On the other hand, the minor lipid components in these membranes are also key players for membrane functions, e.g.
cholesterol as a structural component and PI as a signaling molecule. In
plants, the main structural components of chloroplastic membranes are
galactolipids, a group of glycolipids with galactose moieties in the headgroups. The predominant ones are MGDG and DGDG. Analysis of the composition of chloroplastic envelope membranes isolated from leaves of spinach, sunflower, and maize showed that about 30% of the dry weight of the
total lipid was MGDG and about 30% was DGDG (Poincelot 1976). In addition to MGDG and DGDG, some unusual lipids are also present in chloroplastic membranes, e.g. oligogalactolipids and sulpholipids (Benning et al.
2005).
MGDG is a non-bilayer-forming lipid and can form hexagonal phase. However, as much as 60% can be incorporated into a bilayer when bilayerforming lipids are also present. DGDG is a bilayer-prone lipid. Both galactolipids are of importance for plant cells. They are involved in many cell activities, e.g. cell growth, membrane lipid homeostasis under stress conditions
and even potentially in photosynthesis (Dörmann et al. 2002, Hölzl et al.
2007, Torres‐Franklin et al. 2007).
20
Table 1. Structures of common lipids. R1, R2, R1’ and R2’ refer to fatty acyl chains.
Lipid class
Abbreviation
Diacylglycerol
Strucuture
DAG/DG
Phosphatidic acid
PA
Phosphatidylcholine
PC
Phosphatidylethanolamine
PE
Phosphatidylglycerol
PG
Phosphatidylserine
PS
Cardiolipin
CL
Phosphatidylinositol
PI
Monogalactosyl-diacylglycerol
MGDG
Digalactosyl-diacylglycerol
DGDG
Sulfoquinovosyldiacylglycerol
SQDG
21
2.4.2 Membrane proteins
Membrane proteins account for 20-30% of proteins in most proteomes
(Krogh et al. 2001). Most of the pharmaceutical drugs are designed to target
membrane proteins. Membrane proteins fall into two main categories, integral and peripheral membrane proteins. Typical integral membrane proteins
are polytopic proteins with more than one transmembrane (TM) segment,
such as TM transporters. There are also proteins with a single TM segment,
namely bitopic membrane proteins such as many enzymes and receptors that
are anchored to the membrane via a transmembrane α-helix (Luckey. 2008,
von Heijne. 1996). Monotopic proteins are those inserted into a membrane
but not spanning the entire membrane. Peripheral membrane proteins associate with the lipid matrix of a membrane in different ways due to structural
and functional multiplicity, but neither permanently bind to nor span it
(Luckey. 2008).
Taking mitochondrial and chloroplastic membranes as examples, the receptors, Tom20 and Toc34 are bitopic membrane proteins at the outer membrane of the respective organelle and they are anchored to the membrane
with either the C-terminal part or N-terminal part. Tom40 and Toc75 are βbarrel proteins, which are only found in the outer membranes of mitochondria, chloroplasts, and gram-negative bacteria and both of them function as
cation selective channels. Structural information on other components of
complexes for protein import is limited. Including protein import, many
events at the membranes usually take place via complexes of membrane
proteins and sometimes different complexes are required to work in concert,
e.g. in respiration and photosynthesis.
2.4.3 Protein lipid interactions
The interaction between proteins and lipids is an interplay. In general, it is
considered to be of two types, i.e. specific interactions and nonspecific interactions. Typical nonspecific interactions involve annular lipids that are lipids
around membrane proteins, like a lipid shell. Specific interactions involve
so-called non-annular lipids, which are considered to be essential for protein
structure and function. Conventionally, interactions between non-bonded
atoms are electrostatic interactions, van der Waals forces and hydrogen
bonds (Fersht. 1999). Moreover, the hydrophobic effect due to displacement
of aqueous molecules is also of utmost importance (Fersht. 1999). However,
the mechanism of protein-lipid interaction in a specific case needs to be
studied individually since the details of the interaction mechanism as well as
the specificities and affinities of binding give rise to significant biological
consequences.
22
Protein lipid interactions play multiple roles in protein import processes to
both mitochondria and chloroplasts. However, little is known about the molecular basis of these interactions. First of all, nonspecific interactions are
presumably required for the assembly of the annular lipids and the different
components of import complexes. Secondly, specific interactions have been
reported for different proteins and different lipids. Mitochondrial outer
membrane proteins usually contain a non-cleavable targeting peptide which
often has transmembrane domains with flanking segments. The special designs suggest hydrophobic interactions between transmembrane domains and
lipid acyl chains and electrostatic interactions between the charged flanking
segments and lipid headgroups. The interactions between targeting peptides
and lipids are of special interest in this thesis work. As is the case for the
dual targeting peptide ThrRS-dTP, many targeting peptides have been shown
to be able to fold in different membrane models where lipids can work as
“molecular chaperones” to facilitate peptide folding (Berglund et al. 2009b,
Moberg et al. 2004). Some also demonstrated that the anionic lipid content
was important for the degree of structure induction (Moberg et al. 2004).
The role of lipids for protein recognition at the surface of the outer membranes has been debated. Nevertheless, specific interactions between certain
lipids and targeting peptides were observed in different membrane mimetic
systems in vitro. CL and PE were shown to be required for lipid mixing induced by mitochondrial targeting peptides, and CL could not be replaced by
other negatively charged lipids (Mandieau et al. 1995, Snel et al. 1995). In
addition, galactolipid MGDG, sulfolipid SQDG and negatively charged PG
were found to be predominantly involved in the interactions with chloroplastic targeting peptides (van't Hof et al. 1991, van't Hof et al. 1993). However, an MGDG-deficient mutant showed no deficiency in protein import to
chloroplasts in vivo compared to wild type (Aronsson et al. 2008). Note that
the MGDG content in the MGDG-deficient mutant was only reduced by
about 40% (Jarvis et al. 2000). On the other hand, the phenotype of the null
mutant of the MGDG synthase MGD1 was so severe that it could not be
used for import studies (Kobayashi et al. 2007). Instead, the DGDGdeficient mutant showed significant influence on protein import (Chen et al.
1998). These results suggest specific interactions between targeting peptides
and lipids, albeit further studies are required for a detailed interpretation.
2.5 Membrane mimetic systems
23
As it is very challenging to work directly with biological membranes, a variety of membrane mimetic systems have been developed for different purposes and still new membrane models are under exploration. The most frequently used membrane models in biological studies include three classes: micelles, bicelles, and vesicles (Fig. 6). Studies on these membrane models not
only help to understand biophysical properties of biomembranes, but also
provide information on how to choose a proper membrane mimetic to study
membrane proteins and their interaction with membranes. With the assistance of membrane mimetics, considerable progress in the understanding of
membrane protein structure, folding, and stability has been made. However,
one must keep in mind that they are membrane-like to different degrees and
from different perspectives, but cannot fully represent a biological membrane (Mäler. 2012, Sanders et al. 2006, Sanders. 2008, Warschawski et al.
2011). For example, a structure of a membrane protein or peptide determined
in a membrane mimetic should not be assumed to be the native structure, but
to reflect a native-like structure.
A. Micelles
B. DMPC/DHPC bicelle (q = 0.5)
C. LUV
Figure 6. Three model membrane systems. A. Micelles made up of singlechained detergents or short-double-chained lipids. B. A typical bicelle made
up of DMPC and DHPC. C. A representative large unilamellar vesicle (LUV)
composed of POPC.
2.5.1 Micelles
Once the concentration of a detergent (or a short-chained lipid) in water
reaches a certain value, the excess detergent forms micelles. It is energetically favorable for zwitterionic or amphiphilic molecules to form micelles
where the hydrophilic head-groups point towards the water solution while all
the hydrophobic tails point towards each other. Typical micelles are spherical
with a diameter of ca 40 Å, although some may be slightly ellipsoidal or
even worm-like (Lin et al. 1987, Lipfert et al. 2007) (Fig. 6A). The concentration that a particular detergent/short-chained lipid has to exceed to form
micelles is called its critical micelle concentration (CMC).
24
Micelles formed by sodium dodecylsulfate (SDS), dodecylphosphocholine
(DPC) and dihexanoylphosphatidylcholine (DHPC) have been extensively
studied and commonly applied in biophysical studies of membrane protein/peptide interactions using circular dichroism (CD) and solution nuclear
magnetic resonance (NMR). Their different biophysical properties provide
advantages for different requirements. Micelles composed of anionic detergents, such as SDS, give relatively better quality NMR spectra, but at the
same time have a higher risk to destabilize a protein (Sanders et al. 2006).
The phospholipid DHPC with two short chains has a lower risk to destabilize
proteins. DPC has a lower CMC (1.5 mM) (Lauterwein et al. 1979, le Maire
et al. 2000), and thus at the same detergent concentration, more micelles
form as the micelle sizes are believed to be concentration independent
(Wennerström et al. 1979).
As small objects with only one component, micelles have dramatically simplified the studies of membrane proteins. More importantly, it gives better
NMR spectra, compared to larger membrane mimetics. Structures of many
important parts of membrane proteins are solved in micelles (Moberg et al.
2004, Saviano et al. 1999, Tessmer et al. 1997). Unfortunately, micelles are
oversimplified in several perspectives. A lack of bilayer means that micelles
do not mimic biologic membranes well enough. The additional curvature of
micelles distorts the peptide/protein binding to micelles. Furthermore, studies of the membrane location of a protein, and in particular of peptides, are
not realistic.
2.5.2 Bicelles
Bicelles, which are also known as bilayered phospholipid micelles, are composed of water-insoluble long-chained lipids and soluble detergents (shortchained lipids). In a bicelle, the lipids are predominantly present at the bilayer surfaces and the detergent molecules within the rim around the bilayer,
giving a disk shape (Fig. 6B and Fig. 7). The molar ratio between the longchained lipid and the detergent, the q-value, and the concentration of the two
components, cL in weight/weight % or C in mM, are two characteristics of
importance for the mixtures. Another important parameter for bicelles is the
aggregation number, i.e. the numbers of lipids and detergents in each bicelle.
Bicelles are increasingly applied in studies of membrane structural biology
(Mäler. 2012, Sanders et al. 2006, Sanders. 2008, Warschawski et al. 2011).
In addition, bicelles are also explored in fields other than membrane associated peptide/protein studies, e.g. in studies of macromolecular delivery (Mäler et al. 2009), effect of bicelles on skin (Barbosa-Barros et al. 2008, Rodríguez et al. 2010), and conformation and location studies on membrane25
bound drugs (Houdai et al. 2008, Matsumori et al. 2006). The problematic
crystallization of G protein-coupled receptors was made possible in bicelles
and in lipidic cubic phase (Rasmussen et al. 2007, Thompson et al. 2011).
Bicelles with q ≤ 1 are small sphere-like disks and therefore they tumble
rapidly and isotropically which is a favorable characteristic for high resolution NMR. On the other hand, bicelles with q > 2 are large disks and typically used in solid-state NMR investigations. It is interesting that they align
spontaneously when a high magnetic field is applied. This property has been
employed for the refinement of NMR structures by measuring residual dipolar couplings (Bax. 2009, Prestegard et al. 2001, Tjandra et al. 1997).
Bicelles were originally made by mixing lecithin, which is a mixture of longchained lipids, phosphatidylcholine (PC), phosphatidylethanolamin (PE) and
phosphatidylinositol (PI), and bile salts (Mazer et al. 1980, Small et al.
1969). The most commonly used bicelles nowadays are made of dimyristoyl
phosphatidylcholine (DMPC) and dihexanoyl phosphatidylcholine (DHPC).
Nevertheless, different lipids and detergents are employed for different purposes. In order to study the role of negatively charged lipids, 10-30% of
DMPC in the DMPC/DHPC bicelle may be replaced with negatively charged
lipids, such as dimyristoyl phosphatidylglycerol (DMPG) (Struppe et al.
2000, Szpryngiel et al. 2011, Unnerståle et al. 2011). To meet the needs of
specific cases other lipids are introduced to the bicelles, and new mixtures
have been explored. For example, to study peptide-bilayer mismatch, lipids
with longer or shorter fatty acyl chains, e.g. DLPC, DPPC and POPC have
been introduced (Chou et al. 2004, Lind et al. 2008, Whiles et al. 2002). In
this thesis work, new membrane mimicking bicelle systems with galactolipids are studied in Paper IV. They can be used to further investigate the
interaction between galactolipids and e.g. targeting peptides and plant glycosyltransferases.
The size of a bicelle can be manipulated by the long-chained/short-chained
lipid ratios (Vold et al. 1996). Other factors also affect bicelle size. For the
same q-value, a decrease of the total lipid concentration causes an increase in
bicelle size if the concentration of short-chained lipid/detergent is still above
its CMC. Addition of certain bivalent ions increases bicelle size while addition of monovalent ions does not affect bicelle size (Arnold et al. 2002). One
must pay attention to how to prepare bicelle samples since different procedures also give different sizes (Lu et al. 2012). To obtain optical clarity, there
are four strategies: vortex-centrifuge cycles, heating-cooling cycles, sonication, and freeze-thaw cycles. Sander’s group reported that the bicelles
prepared by the combination of the first two strategies are generally larger
than the combination of the latter two (Lu et al. 2012). Moreover, mixtures
of the two bicellar components may adopt various morphologies depending
on the amphiphile components, composition and temperature (Triba et al.
26
2005, van Dam et al. 2004). In Paper III, the morphology of bicelles with
varying composition (q, C, and T), were characterized.
A
B
C
Figure 7. Sketches of an ideal bicelle. A. The long-chained lipids are indicated with the head-group in blue and the tail in brown while the short-chained
lipids or detergents are indicated in grey (Lind. 2009). B. Top view. C. Cross
section view. R is the radius of the planar domain and r is the internal radius
for the rim region as indicated in B and C.
To understand bicelles and their physical properties better, an ideal bicelle
model was proposed for DMPC/DHPC bicelles by Vold and Prosser (Fig. 6)
(Vold et al. 1996). Later this model was modified by considering the CMC
of DHPC and the different head-group areas between DMPC and DHPC to
explain dynamic light scattering results (Glover et al. 2001). In the ideal
model, the two lipid species purely segregate into two regions. As in Fig.
6A, DMPC molecules are exclusively present in the planar domain and
DHPC molecules are exclusively present in the rim region. The surface areas
for the two regions Ap and Ar can be calculated using the radius of the planar
domain R and the internal radius of the rim domain r. In the ideal model, the
q-value is the ratio of the concentration of DMPC and that of DHPC in the
sample and all the lipids are present in the bicelles. Therefore, q also equals
to the surface ratio:
𝑞=
𝐴𝑝
𝑅2
=
𝐴𝑟 𝑟(𝜋𝑅 + 2𝑟)
(1.1)
By reorganizing the equation, an expression for the radius R of the central
planar region can be given by
1
8
𝑅 = 𝑟𝑞 (𝜋 + √𝜋 2 + )
2
𝑞
(1.2a)
27
In the refined model, the effective q-value, qeff, is used instead of q due to the
existence of free DHPC molecules in the solution. Another factor considered
in this model is the ratio between the head-group area of the long-chained
lipids and that of the detergent. Taking this ratio as k leads to the modified
relation between R and qeff:
1
8
𝑅 = 𝑘𝑟𝑞𝑒𝑓𝑓 (𝜋 + √𝜋 2 +
)
2
𝑘𝑞𝑒𝑓𝑓
(1.2b)
In agreement with the experimental data mentioned before, one can see that
as the q-value increases, the planar domain becomes larger and the size of
the bicelle increases for both models. R = 2r, when q is about 0.5. This is
also in agreement with the experimental data on isotropic bicelles with q =
0.5, where r = 20 Å and R = 40 Å (Eum 1989 Vold and Prosser1996). With
these equations and values as well as estimated values for the head-group
area for DMPC (60 Å2) and that for DHPC (66 Å2 and/or 102 Å2) (van Dam
et al. 2004), the aggregate number can be calculated, which provides important information for studies on membrane proteins reconstituted in bicelles, i.e. to maintain a reasonable lipid/protein ratio. For the ideal
DMPC/DHPC bicelle with q = 0.5, the estimated number of DMPC molecules per bicelle is around 150 and the number of DHPC molecules is
somewhere between 200 and 350 depending on different values for the
DHPC head-group area (van Dam et al. 2004).
As a simplified membrane model which still shows typical lipid motions,
bicelles provide a tool to study membrane associated dynamics as well, e.g.
the influence of a peptide on lipid dynamics and vice versa. Generally, a
variety of motions on different time-scales can take place for a molecule in a
bilayer: lateral diffusion within the bilayer, rotation around its principal axis
and internal fast motions (Ellena et al. 1993, Mayer et al. 1990). Extensive
motions must take place to allow the detergent to exchange between the ones
in the bicelle and in the solution. Carbon-13 relaxation measurements as well
as model-free analysis and MD simulations have been employed to investigate lipid dynamics in different membrane mimetic systems including bicelles. In aligned bicelles, quadrupolar splittings of deuterated lipids are
often measured to obtain information about the segmental order parameter.
In Paper IV, carbon-13 relaxation parameters were measured and analyzed to
understand the dynamics of galactolipids in bicelles and their influence on
the overall dynamics of bicelles.
28
2.5.3 Vesicles
Vesicles are closed spherical bilayers made up of lipids. Vesicles with only
one bilayer are called unilamellar vesicles. Depending on the size, they can
be divided into three categories: small unilamellar vesicles (SUVs), which
have diameters of 25-50 nm, large unilamellar vesicles (LUVs) with diameters of 50-200 nm and giant unilamellar vescicles (GUVs), which are in the
size range of cells. Vesicles with many bilayers, i.e. multilamellar vescicles
first form after dissolving lipids in water and vortexing the samples. Sonication or freeze-thaw cycles can decrease the lamellarity. To ensure a uniform size in the sample, ultracentrifugation or extrusion is required. Compared to bicelles, vesicles are even more native-like as membrane mimicking
systems due to the small curvature and large bilayer surfaces. However, the
large size causes slow tumbling leading to poor quality of NMR spectra, and
the large surfaces may cause light scattering leading to problems with optical
spectroscopic measurements.
2.5.4 Other systems
In addition, there are other membrane mimetic media like nanodisks, amphipols, reversed micelles and supported lipid membranes (Mäler. 2012,
Warschawski et al. 2011). Within the same size range, nanodisks are similar
to bicelles. Instead of detergents, nanodisks have a lipid bilayer stabilized by
apolipoprotein, the length of which is used to control the size of the nanodisk
(Mäler. 2012, Warschawski et al. 2011). Recently, a novel nanodisk system,
so-called native nanodisk was introduced. It uses a styrene maleic acid copolymer (SMA) instead of detergent or scaffold protein and the native lipids
isolated from original cellular membranes instead of other lipids or detergents (Dorr et al. 2014, Knowles et al. 2009, Orwick-Rydmark et al. 2012).
Amphipols are made of designed short amphiphilic polymers carrying many
hydrophobic chains (Warschawski et al. 2011). They can keep membrane
proteins soluble in aqueous solutions. In an organic solvent environment, the
lipids arrange as reversed micelles with the head-group towards the soluble
part of membrane protein leaving the hydrophobic tails outside. Reversed
micelles are developed with the aim of reducing the membrane protein complex size for NMR (Luckey. 2008, Warschawski et al. 2011). Supported
lipid membranes are based on the ability of vesicles to adsorb to certain
types of solid supports or surfaces and form planar bilayers (Richter et al.
2006, Tamm et al. 1985).
29
3. Methods
3.1 Nuclear magnetic resonance (NMR) spectroscopy
NMR spectroscopy is a versatile technique applied in many different disciplines of modern science. In biophysics NMR can be used for protein structure determination and mapping biomolecular interactions as well as for
dynamic studies. It is the main method used in this thesis work. In Papers I
and II, 15N HSQC of ThrRS-dTP was used as a fingerprint for studies on the
interactions between the targeting peptide and two receptors. In Paper IV,
diffusion experiments verified that bicelles persist when galactolipids were
introduced to a certain degree. In addition, relaxation parameters were collected for the galactolipids and the local dynamics were analyzed for the
galactolipids in bicelles.
3.1.1 Chemical shifts and assignment
NMR Chemical shifts origin from local chemical environments sensed by
the nuclei. Therefore, they carry a lot of information of the surroundings of
the nuclei, including conformational information of the molecules where the
nuclei sit. The typical chemical shift ranges for specific nuclei in specific
molecules are usually known. For example, 13C chemical shifts for carbons
in a methyl group are in the range of 13-16 ppm and 13C chemical shifts for
carbons from a double bond in a fatty acid chain are expected to be in the
range of 115-140 ppm. Another textbook example is that 1H chemical shifts
for the backbone amide proton in the common amino acid residues in a random coil are in the range of 8.1-8.8 ppm (Wuthrich. 1986).
For a protein, backbone chemical shifts have been used to calculate backbone dihedral angles (Cheung et al. 2010, Shen et al. 2009). More frequently, they have been used to identify secondary structural elements of a protein.
The chemical shift deviations from the corresponding random coil values are
named secondary chemical shifts. Based on the secondary chemical shifts,
the chemical shift index (CSI) analysis was introduced by Wishart et al.
(Wishart et al. 1992, Wishart et al. 1994, Wishart et al. 1995a, Wishart et al.
1995b), to give information about the secondary structure for each residue.
For example, a positive deviation in the chemical shift for Cα and CO atoms
30
indicates α-helix propensity and a negative deviation indicates β-strand propensity. For proteins in partially structured states, the secondary structure
propensity (SSP) method has been developed by Forman-Kay and coworkers (Marsh et al. 2009). In addition, the scores using the SSP method
give quantitative characterization of secondary structure elements: the larger
the absolute value, the larger the corresponding propensity.
15
N HSQC, as a fingerprint of a protein, can be used for further characteristics of the protein after the 1H and 15N resonances are assigned. Two procedures have been developed for sequence-specific backbone assignments
(Sattler et al. 1999, Wuthrich. 1986). The first one makes use of heteronuclear scalar couplings for connectivities between neighboring spin systems
(residues). There are three main pairs of experiments designed for this:
HNCACB (Grzesiek et al. 1992b) and CBCA(CO)NH (Grzesiek et al.
1992c); HNCA (Grzesiek et al. 1992a, Kay et al. 1990) and HN(CO)CA
(Bax et al. 1991, Grzesiek et al. 1992a); HNCO (Grzesiek et al. 1992a, Kay
et al. 1990) and HN(CA)CO (Clubb et al. 1992). The second sequential assignment procedure makes use of 1H-1H NOEs to identify adjacent residues.
The corresponding experiments are 15N NOESY-HSQC (Marion et al.
1989a, Marion et al. 1989b, Zuiderweg et al. 1989) and 15N TOCSY-HSQC
(Marion et al. 1989b). In general, assignment for proteins with only 15Nlabelling is harder than that for proteins with 15N, 13C-labelling. In some
cases, one still has to choose some of the triple resonance experiments as a
supplement.
3.1.2 NMR relaxation
The process by which a perturbed system returns to its equilibrium is known
as relaxation. NMR relaxation is typically due to molecular reorientation,
internal motions or chemical exchange involving the nuclei under investigation. These motions modulate anisotropic interactions, such as dipole-dipole
interactions, chemical shift anisotropy, paramagnetic interactions, quadrupolar interactions etc. Among them, dipole-dipole (DD) interactions and chemical shift anisotropy (CSA) are the most common spin relaxation mechanisms for spin-½ nuclei in the absence of paramagnetic species and there is
also interference between the two mechanisms, i.e. cross-correlated relaxation. Details about fundamental NMR relaxation theory can be found in textbooks (Keeler. 2005, Levitt. 2001).
The longitudinal relaxation time T1, transverse relaxation time T2 and steady
state NOE are frequently measured NMR relaxation parameters. Dynamic
information on the spin systems and the molecules can be extracted from
these parameters. In nuclear spin relaxation theory, the mathematical description is too complicated to be described here, and can be found in several
31
books (Abragam. 1961, Kowalewski et al. 2006). Basically, the experimental
parameters are described in terms of spectral densities of molecular motions.
The spectral density function, J(ω), comes from the Fourier transform of the
correlation function, G(τ), which describes the correlation between the orientation of a bond vector at two time points with an interval of τ. For 13C nuclei
in a lipid in solution, the relaxation is dominated by 1H-13C dipole-dipole
interactions and also by chemical shift anisotropy in some cases. According
to the Redfield theory, the relaxation parameters T1-1, T2-1 and NOE can be
expressed as linear combinations of spectral densities at different frequencies:
𝑇1 −1 =
𝑑2
[𝐽(𝜔𝐻 − 𝜔𝐶 ) + 3𝐽(𝜔𝐶 ) + 6𝐽(𝜔𝐻 + 𝜔𝐶 )] + 𝑐 2 𝐽(𝜔𝐶 )
4
𝑇2 −1 =
𝑑2
[4𝐽(0) + 𝐽(𝜔𝐻 − 𝜔𝐶 ) + 3𝐽(𝜔𝐶 ) + 6𝐽(𝜔𝐻 )
8
𝑐2
+ 6𝐽(𝜔𝐻 + 𝜔𝐶 )] + [3𝐽(𝜔𝐶 ) + 4𝐽(0)]
6
𝑁𝑂𝐸 = 1 +
𝑑2 𝛾𝐻 𝑇1
[6𝐽(𝜔𝐻 + 𝜔𝐶 ) − 𝐽(𝜔𝐻 − 𝜔𝐶 )]
4𝛾𝐶
(2.1)
(2.2)
(2.3)
where ωH and ωC are the Larmor frequencies of the 1H and 13C nuclei. ωC is
replaced by ωN when the nuclei of interest is 15N. The dipole-dipole interaction constant is d = μ0hγHγC/8π2r3 and the chemical shift anisotropy interaction constant is c = ωC∆σ/√3, where μ0 is the permeability of vacuum, h
Planck’s constant, γ the gyromagnetic ratio, r the distance between the 1H
and 13C nuclei and ∆σ is a measure of the chemical shift anisotropy.
To interpret the relaxation parameters in terms of molecular dynamics, different models have been developed. The “Model-Free” approach is a classical one, widely used to obtain information about site-specific internal motions of molecules (Clore et al. 1990a, Clore et al. 1990b, Halle et al. 1981,
Lipari et al. 1982a, Lipari et al. 1982b, Wennerström et al. 1974). Rather
than fitting the experimental data to any specific physical models with the
assumption of how a bond vector moves, the spectral density function of the
“Model-Free” approach is derived from the mathematical analysis based on
the separation of the global tumbling from internal motion. Therefore, the
amplitude and rate of internal dynamics for individual bond vectors can be
characterized by the generalized order parameter and the internal correlation
time for motions of the bond vector in a molecular frame. In order to include
motions both on a fast timescale and a slow timescale, the spectral density
function was extended to the extended model-free formalism (Clore et al.
1990a, Clore et al. 1990b, Farrow et al. 1994).
32
In bicelles, the main contributions to dynamics are from overall bicelle reorientation, overall lipid reorientation within the bicelle and local motion of
individual bond vectors in the lipid. Assuming they are on distinct timescales, one can divide the overall correlation function of lipid motion into
three components: Coverall(t) = Cbicelle(t)∙Clipid(t)∙Clocal,j(t). Considering the
large rotational correlation time of the bicelle due to the large size (Halle.
1991), the overall motion of the bicelle is assumed to be too slow to affect
the spectra density function at all frequencies except ω=0. Therefore the
correlation function for a 1H-13C bond vector j becomes: Clipid(t)∙Clocal,j(t) =
[(1-Slipid2)∙exp(-t/τlipid)+Slipid2]∙[(1-Slocal,j2)∙exp(-t/τlocal,j)+Slocal,j2], and the spectral density function is:
2
2
2
2
− 𝑆𝑙𝑜𝑐𝑎𝑙,𝑗
𝑆𝑙𝑖𝑝𝑖𝑑
)𝜏𝑙𝑖𝑝𝑖𝑑 (1 − 𝑆𝑙𝑜𝑐𝑎𝑙,𝑗
)𝜏 ′
2 (𝑆𝑙𝑜𝑐𝑎𝑙,𝑗
𝐽𝑗 (𝜔) = [
+
]
2
5
1 + 𝜔 2 𝜏 ′2
1 + 𝜔 2 𝜏𝑙𝑖𝑝𝑖𝑑
(2.4)
where 1/τ’ =1/τlipid +1/τlocal,j, Slipid and τlipid are the generalized order parameter
and the correlation time for the overall lipid motion. Slocal,j and τlocal,j are the
generalized order parameter and the correlation time for the individual bond
vectors at position j (Andersson et al. 2005, Ellena et al. 1993). Note that the
transverse relaxation parameter, T2-1, is strongly affected by the slow motion,
since they contain the spectral density function at zero frequency, so it cannot be included in the analysis based on this model.
3.1.3 Line broadening
Line broadening is frequently seen in NMR spectra due to a variety of
sources. Fourier transform of the NMR time domain signals give a Lorentzian lineshape with a linewidth ∆ν = R2/π. Therefore, the larger the transverse
relaxtion rate (R2 = T2-1), the wider the line. Thus, the relaxation mechanisms
mentioned in the previous section influence the linewidth as well. In order to
be distinguished from other sources, the corresponding transverse relaxation
rate and linewidth are often denoted as intrinsic transverse relaxation rate R20
and natural linewidth ∆ν0.
Chemical exchange between states that give different chemical shifts is also
one important source of line broadening. Taking two-state chemical exchange as an example, the linewidth is influenced by the exchange rates
together with the chemical shift difference between the two states, δν (Hz),
and the populations of the two states. If the exchange rate is much smaller
than the chemical shift difference, kex << δν, there are two separate peaks in
the spectrum and the extra linewidths due to exchange are proportional to the
exchange rates to the two states resepctively. In the other limit, if kex >> δν,
33
the two peaks merge to a single resonance peak at the population weighted
average frequency. In contrast to slow exchange, the extra linewidth in fast
exchange is inversely proportional to the exchange rate. An increase in the
exchange rate in the fast exchange limit causes therefore line narrowing.
When kex is comparable to δν, the lines are severely broadened and often
beyond detection. However, whether the exchange is slow, intermediate or
fast may be altered by changing temperature or magnetic field strength. Accordingly, the line broadening may be reduced. In addition, special experiments have been developed to capture information on intermediate exchange
events, e.g. relaxation disperison experiments (Palmer et al. 2001).
3.1.4 Diffusion NMR
Translational diffusion coefficients can be measured by NMR for various
objects, from a single molecule to a large lipid aggregates. It has been well
employed for different purposes, e.g. to identify ligand-protein interactions
(Hajduk et al. 1997), to extract structural information of proteins (Danielsson et al. 2002, Mittag et al. 2008) and to probe peptide-bicelle or micelle interactions (Szpryngiel et al. 2011, Unnerståle et al. 2011, Unnerståle
et al. 2009). The pulsed field gradient spin echo (PFG-SE) experiment is a
typical diffusion experiment (Stejskal et al. 1965). The coherences of molecules are spatially labeled by the first magnetic field gradient. As translational diffusion occurs during the following evolution time, the refocusing of
the second gradient is not complete, giving an attenuated signal for the experiment. The intensity decays exponentially according to Stejskal-Tanner
equation, I = I0exp[-Dγ2g2δ2(∆-δ/3)]. Here, I and I0 are the intensity with and
without gradient respectively, D the diffusion coefficient, γ the gyromagnetic
ratio of the nucleus, g the strength of the gradients, δ the length of gradient
pulse and Δ is the diffusion time. By incrementing the gradient strength in a
series, the corresponding attenuated intensity can be measured. Thus, diffusion coefficients can be deduced by fitting the signal attenuation to the
Stejskal-Tanner equation.
The translational diffusion coefficient is determined by many physical parameters, e.g. size and shape. For a rigid spherical molecule in infinite dilution, a relation between D and the hydrodynamic radius Rh of a molecule or
object is given by the Stokes-Einstein equation:
𝐷=
𝑘𝑇
6𝜋𝜂𝑅ℎ 𝐹
(2.5)
where k is the Boltzmann constant, T the absolute temperature, η the viscosity of the solution and Rh the hydrodynamic radius of the molecule. For better
34
estimations, a shape factor, F, is considered to modify Rh when it is not
spherical. For an oblate object like bicelle, the Perrin factor Fp is used:
𝐹𝑃 =
√(𝑎/𝑏)2 − 1
(𝑎/𝑏)2/3 ∙ 𝑎𝑟𝑐𝑡𝑎𝑛√(𝑎/𝑏)2 − 1
(2.6)
where a is the length of the long axes and b is that of the short axis (Cantor
et al. 1980, Perrin. 1934). For an ideal bicelle as described before, a is 50 Å
and b is 10 Å, leading to a shape factor value of 1.22.
3.2 Dynamic light scattering (DLS)
DLS is mainly used to determine the size distribution profile of particles and
polymers in solution. It is also referred to as photon correlation spectroscopy
or quasi-elastic light scattering. In a DLS experiment, light passes through
the sample and hits small particles in it. The scattering will be isotropic if the
particles are small compared to the laser wavelength (typically below 250
nm). Imagine if particles in solution were stationary, a screen of the sample
would give a stationary speckle pattern. However, the particles undergo
Brownian motion. Thus the intensity of each speckle fluctuates because the
phase addition of the scattered light from the surrounding particles may be
either constructive or destructive. The dynamic information for the particles
can be derived from the autocorrelation of the intensity over the experiments. As in the Stokes-Einstein equation (Eq. 2.5), larger particles diffuse
slower, and thus the intensity fluctuates slower (Fig. 8).
For samples with monodisperse particles, the correlation function is an exponentially decaying function with the decay rate, Γ = D[(4πn/λ)sin(θ/2)] 2,
where n is the refractive index of the medium, λ, the wavelength of the laser,
θ, the scattering angle and D, the diffusion rate as in the Stokes-Einstein
equation. For samples with polydisperse particles, the correlation function is
a sum of all the exponential decays determined by particles with different
sizes. Various algorithms have been applied to extract the size distribution
from the correlation function. For example, the Cumulants analysis fits a
single exponential to the correlation function and estimates the width of the
distribution from the fitting while Non-negative least squares (NNLS) and
CONTIN methods fit a multiple exponential to obtain the size distribution.
The size distribution obtained directly from these fittings has the intensity of
scattered light in y-axis and the diameter of the particle in x-axis. As the
intensity is proportional to the sixth power of the radius while the volume is
proportional to the third power of the radius, additional process is needed to
convert the intensity distribution to a volume distribution or number distribu35
tion. An ideal monodisperse sample would give a single smooth symmetric
peak, in which case the center of the peak would be the same in all the three
types of distribution. However, in practice, the center of the peak is sensitive
to different weighting strategies.
A
B
C
D
Figure 8. Examples of DLS data for large particles and small particles. A and
C show intensity fluctuations for a vesicle sample and a micelle sample, respectively. B and D are the corresponding correlation function curves. B gives
a typical behavior for a sample containing large particles in which the correlation of the signal decays slower than in D.
3.3 Circular dichroism (CD) spectroscopy
CD spectroscopy is based on the fact that optically active chiral molecules
have different absorbance of left circularly polarized light and right circularly polarized light. Previously, it was measured by changing the polarization
of linearly polarized light (compensator technique), giving a result as ellipticity () (Nordén et al. 1997). Nowadays differential absorbance is directly
measured, but ellipticity is kept as the two parameters can be easily converted. To estimate conformation information independent of polymer lengths,
ellipticity () is usually normalized to mean residue molar ellipticity, typically, [θ]MRE in deg cm2 dmol-1.
CD spectroscopy over different wavelength ranges has many different applications. For biological systems, especially proteins/peptides, CD is quite
powerful to probe changes in macromolecular conformation. Far-UV CD is
36
applied to probe the conformational change of peptides secondary structure
in different environments and addition of small molecules. There are two
transitions in the far-UV region for peptide bond: typically around 190 nm
and 220 nm and they represent π → π* transition and n → π* transition,
respectively. The characteristic of the CD spectra for an α–helical protein is a
positive peak at 192 nm and two negative peaks at 208 nm and 222 nm and
the characteristic for the disordered protein is a negative peak at around 200
nm. The CD signal for β-sheet is generally much smaller and varied, generally a positive band near 195 nm and a negative band of comparable magnitude at 216 nm are seen. CD is also applied to probe the intra-amide π → π*
charge transfer transition between 170 and 180 nm (Gilbert et al. 2004). In
addition, there is a characteristic π → π* absorption between 250 and 300
nm for aromatic residues (Wallace et al. 2009).
37
4. Results and discussion
4.1 ThrRS-dTP(2-60) (Paper I and Paper II)
In order to understand the mechanism of dual targeting to mitochondria and
chloroplasts, the dual targeting peptide ThrRS-dTP(2-60) was used as a
model for dTPs in general. To date ThrRS-dTP(2-60) is still the only dual
targeting peptide that has been identified, produced and purified although a
growing number of dual-targeted proteins have been identified and notable
bioinformatic analyses are available (Carrie et al. 2009b, Carrie et al. 2013,
Ge et al. 2014, Heazlewood et al. 2007). In this thesis work, we characterized this peptide and studied its interactions with the primary receptors of
mitochondria and chloroplasts. For this purpose we have assigned the 15N
HSQC spectrum for ThrRS-dTP(2-60) using a combination of 15N NOESYHSQC and 15N TOCSY-HSQC together with HNCA and HN(CO)CA experiments. Then, the interactions between the targeting peptide and receptors
were mapped onto the peptide sequence by recording the changes in the
spectrum upon addition of AtTom20 or AtToc34. To simplify the studies,
receptors without their transmembrane segments were used.
4.1.1 Structural properties of ThrRS-dTP(2-60)
The primary structure of ThrRS-dTP(2-60), as given in Fig. 5, shows that it
is a typical dual targeting peptide but not a typical IDP. As a dTP, it has a
higher prevalence of Ser (25%), Phe (12%) and Leu (9%) compared to mTPs
and cTPs. As in mTPs, it is enriched in Arg, with 5 Arg out of 59 residues.
At the same time, it is also enriched in Pro, as observed for cTPs (5 out of
59). The high content of hydrophobic residues (Phe and Leu) in ThrRS-dTP
makes it different from a typical IDP, although it has an extremely high content of disorder-promoting residues, e.g. Ser, Arg and Pro. Several sequence
based predictors have been applied to predict the potential secondary structure or disordered domains of ThrRS-dTP (Fig. 8). Two helical segments,
T21-V28 and S44-V49 have been predicted by Jpred although the reliability
is low, especially for the second segment. Regions close to the two ends
have been predicted to be disordered by several predictors. It seems that this
sequence is ambiguous to those predictors that are trained by either proteins
with clear ordered secondary structure or disordered domains.
38
Previously, CD results showed that ThrRS-dTP is mainly random coil in
aqueous solution (Berglund et al. 2009b). In accordance with this, the narrow chemical shift dispersion in the proton dimension (7.9-8.5 ppm) of the
15
N HSQC spectrum again suggests that ThrRS-dTP is largely unstructured
in buffer (Fig. 1, Paper I). Moreover, water-amide hydrogen exchange monitored by CLEANEX experiments (Hwang et al. 1998), in which water is
selectively excitated, indicated very fast exchange for all amide 1H resonances in the peptide. All the peaks appear already at a mixing time of 80 ms.
The results suggest again that ThrRS-dTP is an intrinsically disordered peptide. This is not surprising since most targeting peptides appear to be disordered in buffer, although α-helix has been observed for e.g. alternative oxidase (AOX) (Rimmer et al. 2011) and F1 (Moberg et al. 2004) in the presence of a receptor or lipids. In addition, disorder has several functional advantages in protein interactions, e.g. high specificity with low affinity (Zhou
et al. 2001) and larger binding surfaces (Gunasekaran et al. 2003). Therefore,
it is reasonable for targeting peptides to be disordered in order to interact
with receptors efficiently.
A
B
Figure 9. Secondary structure information of ThrRS-dTP(2-60). A. Histogram of SSP scores based on Hα and Cα chemical shifts. B. Helical regions
predicted by Jpred and disordered regions predicted by GlobPlot(Linding et
al. 2003b), FoldIndex (Prilusky et al. 2005), DisEMBL (Linding et al. 2003a)
and PONDER (Li et al. 1999).
39
Nevertheless, we found that ThrRS-dTP has a tendency to form helical structure at the N-terminal half. The CD spectrum from a previous study suggests
that about 25-30% helical structure can be induced by different membrane
mimetics. In our work, we have measured CD spectra of different fragments
in buffer and DPC micelles, ThrRS-dTP(1-29), ThrRS-dTP(1-19), ThrRSdTP(20-29) and ThrRS-dTP(30-60). They all gave a random coil appearance
in the spectrum in buffer but a large amount of α-helical structure was induced in the N-terminal half ThrRS-dTP(1-29) by DPC. By using secondary
structural propensity scores (SSP) based on Cα and Hα chemical shifts, we
found again the tendency to form helical structure in the N-terminal part, and
a very weak propensity for alternating  structure and α-helix in the middle
part of the sequence (Fig. 8). The N-terminal part with evident α-helical
propensity includes S6-F27 (P17-S19 remained unassigned), and the middle
part with weak secondary structural propensity includes Q34-R38 and Q39F47. Furthermore, according to HELI-QUEST analysis of the peptide, there
is a hydrophobic face with 8 residues in the L11-P37 segment and the S11V28 segment shows an amphiphilic character. These results correspond well
to the characteristics of mTPs that interact with Tom20.
4.1.2 The interaction between ThrRS-dTP(2-60) and AtTom20
In order to map the interaction sites of AtTom20 on ThrRS-dTP, AtTom20
was titrated onto a sample of ThrRS-dTP and changes in the 15N HSQC
werer monitored (Paper I). Although no large chemical shift differences
were observed upon titration of AtTom20, minor but significant chemical
shift changes were found in more than one region. Moreover, line broadening and significant reduction in peak intensity were observed for peaks corresponding to a large part of the peptide sequence. These were the dominant
changes upon addition of AtTom20. It is also worth noting that the Cterminal part seemed not to be affected by AtTom20.
Considering both the chemical shift changes and intensity reductions, the F9Q39 region was the most affected, which most likely involves two segments,
F9-V28 and L30-Q39 as Y29 was not affected as much as other residues in
this region. This corresponds well with the predicted secondary structure
regions from the Jpred prediction and SSP scores as mentioned before. Interestingly, the most affected segment, F9-V28, includes L24RRFV28, which
had both significant chemical shift changes and a large reduction in intensities. L24RRFV28 is the only part of the ThrRS-dTP sequence with the 
pattern, where  represents a hydrophobic/aromatic residue and  represents
any amino acid. The  pattern has previously been identified in different mTPs to be an important feature for Tom20 recognition (Muto et al.
2001). Note that the F9-V28 segment has the potential to form an amphiphilic helix as predicted by HELI-QUEST, and all residues except F20 in
40
the hydrophobic face showed changes in either chemical shift or intensity or
both. F20 appeared to be less affected in this segment, but still more than
average. Moreover, import capacity experiments in vitro showed that ThrRSdTP(1-29) strongly inhibited import into mitochondria (73%) while the truncated forms, ThrRS-dTP(1-19) and ThrRS-dTP(20-29) only exhibited a
weak inhibition (≤ 22%). The weak inhibition of ThrRS-dTP(20-29), which
contains the L24RRFV28 motif, suggests that the  pattern is not the
single requirement for import. Instead, the F9-V28 amphiphilic helix is
probably also a prerequisite.
In addition to the F9-V28 and L30-Q39 segments, F47 and A48 also showed
obvious intensity loss and H41, V46 and F47 also show significant chemical
shift changes. In contrast, crosspeaks from the C-terminal residues were not
affected by addition of Tom20, indicating that they probably do no contribute much to the interaction with Tom20. In accordance with these results, the
ThrRS-dTP(30-60) fragment was observed to only weakly inhibit mitochondrial import (25%). In summary, ThrRS-dTP(2-60) share the main features
with known mTP-Tom20 interaction (Abe et al. 2000, Muto et al. 2001,
Perry et al. 2006, Rimmer et al. 2011, Roise et al. 1988). It is mainly the
residues in ThrRS-dTP that possess transient structure that contribute to the
interaction to Tom20, but also additional residues take part in the interaction,
whereas a long segment (~ 10 residues) in the C-terminus appears not to take
part in Tom20 recognition.
An additional observation worth noticing is that the main effect on the spectra upon addition of Tom20 to the peptide sample is selective line broadening instead of chemical shift changes. This could be due to two reasons. First,
binding of ThrRS-dTP to Tom20 results in the formation of a complex,
which gives rise a slower overall tumbling and further an increase in transverse relaxation rate R2. Second, the line broadening could also come from
chemical exchange between the bound state and the free state of ThrRS-dTP.
In agreement with this, multiple states in a dynamic equilibrium including
more than one peptide-bound states have previously been reported for one
mTP with Tom20 (Saitoh et al. 2007, Saitoh et al. 2011). ThrRS-dTP(2-60)
does not appear to bind rigidly to Tom20 but is in a dynamic equilibrium,
which is also reasonable in the sense that the targeting peptide is also required to lead the protein to the next component in the mitochondrial import
pathway. It is not likely that the dTP (or mTP) should bind tightly to the
Tom20 receptor, as this would inhibit further import of the protein into the
organelle.
41
4.1.3 The interaction between ThrRS-dTP(2-60) and AtToc34
In the same way, we also investigated the interaction between ThrRS-dTP
and AtToc34 by tracing the changes in the 15N HSQC spectrum upon addition of the receptor and testing the import capacity in vitro of different peptide fragments derived from ThrRS-dTP.
First, the spectrum of ThrRS-dTP was quite sensitive to addition of AtToc34.
Most of the crosspeaks had a 50% intensity reduction already at a molar ratio
of 0.1:1 (AtToc34:ThrRS-dTP). To reach a corresponding intensity reduction
when adding AtTom20, a molar ratio of 0.2:1 was needed. At all molar ratios,
the relative signal intensity with and without receptor was always lower for
Toc34, compared to Tom20. This is reasonable as Toc34 is about 1.5 times
larger than Tom20. Since the rotational correlation time τc for a monomeric
protein is empirically proportional to molecular mass (M in kDa times 0.6 ns,
(Cavanagh. 2007)) and the spectral density function is inversely proportional
to the τc in the slow motion regime, the relaxation rate R2 would roughly be
1.5 times as large for Toc34 as it for Tom20. Consequently, the line broadening for the peptide spectrum is more obvious upon addition of Toc34 than
Tom20.
Second, the whole sequence of ThrRS-dTP appeared to play an important
role in the interaction with Toc34, without any specific region being more
affected by the addition of AtToc34 than others. Peaks corresponding to residues along the entire sequence showed evident and uniform line broadening
behavior upon addition of Toc34. On the other hand, no large chemical shift
changes were observed for specific residues. Minor but significant large
chemical shift changes were only observed for two residues, F35 (∆δHN = 0.001 ppm, ∆δNH = -0.017 ppm) and Q39 (∆δHN = -0.072 ppm, ∆δNH = 0.013 ppm). The  pattern, L24RRFV28, which is considered to be very
important for Tom20 interaction showed no specificity for Toc34 interaction.
Moreover, in the import competition experiments, both ThrRS-dTP(1-29)
and ThrRS-dTP(30-60) caused severe inhibition of chloroplast import, 88%
and 90% respectively. This suggests again that the N-terminal and the Cterminal halves of the sequence are equally important for Toc34 interactions,
which is different from the observations made with Tom20.
Interestingly, a subset of new peaks appeared in the spectrum as most of the
original peaks disappeared when the molar ratio of ThrRS-dTP and Toc34
reached 1:1. The new set of peaks had a higher dispersion in proton dimention, indicating that some structure in the targeting peptide might be induced
upon interaction with Toc34. However, these new peaks may also be (in part)
due to degradation of the peptide. Unfortunately, no peaks could be assigned
42
due to the dramatic change. Note that this was never observed when Tom20
was added to ThrRS-dTP up to a molar ratio of 1:1.
In conclusion, our studies have for the first time revealed the interactions of
a dual targeting peptide with the primary import receptors in the TOM complex of mitochondria and the TOC complex of the chloroplasts. Difference
in the interactions with the two receptors was observed. For mitochondrial
import, it appears that an N-terminal sequence with a propensity for forming
an amphiphilic helix is important, whereas for chloroplastic import, much
larger portions of the sequence appears to be involved in the receptor recognition.
4.2 Isotropic bicelles (Paper III and Paper IV)
Isotropic bicelles have been extensively applied for various purposes, especially for investigating membrane proteins. For studies of plant membranes,
e.g. chloroplastic membranes, mimetics containing glycolipids are needed.
However, the physical properties of bicelles under different conditions are
not yet fully understood. In order to understand bicelle morphology, we have
characterized bicelles with different DMPC/DHPC ratios and different concentrations at different temperatures by using dynamic light scattering (DLS)
and cryogenic transmission electron microscopy (cryo-TEM) in Paper III. In
Paper IV, galactolipids were introduced to replace up to 30 % DMPC in the
bicelle, giving a novel bicelle system that mimics the chloroplastic membrane. The dynamics of the galactolipids in the novel bicelles was analyzed
by measuring the NMR relaxation parameters R1 and NOE.
4.2.1 DMPC/DHPC bicelles under different conditions
To be able to use bicelles in a reliable way for investigating effects of protein and peptides on lipids, it is necessary to have good knowledge about the
bicelles. Temperature is an important factor when designing such experiments. Moreover, the commonly used lipid DMPC has a gel to liquid crystal
transition temperature around room temperature. So what temperature should
we choose? What is the temperature dependence of the morphology of isotropic bicelles? To study the effect of temperature, concentration and q ratio
on bicelle properties we designed a combined DLS and cryo-TEM approach
to measure the size of the bicelles as a function of these three parameters.
Our DLS results showed a temperature dependence of the bicelle size for
almost all the q ratios (0.5, 1 and 1.5). With higher concentrations and higher
q-values, the dependence was more evident. More than a two-fold increase
43
in bicelle size was observed over a temperature range from 17 to 40 C for
bicelles with q ratios of 1 and 1.5 at high concentrations. However, this temperature dependence was not observed for the samples with a low total lipid
concentration (75 mM) when q was 0.5 and 1. In comparison with bicelles,
the hydrodynamic radii of DHPC micelles stayed the same for all the experimental conditions. As the temperature affects the packing of the hydrocarbon chains in the DMPC lipids, the temperature dependence behavior in size
is most likely an indicator for the existence of a bilayered structure of the
DMPC lipids. Hence, the bicelle shape probably does not persist for “bicelle”
samples at low concentration and low q ratios.
The ratio between DMPC and DHPC, q, is an important factor for bicelles.
In theory, larger q indicates larger bilayer surfaces, with overall larger size
of the bicelle. The q ratio seems also to be important for bicelle stability. At
a high temperature, especially close to 40 C, the measurements for q = 1.5
bicelles had very large errors, indicating lack of stability of the bicelle morphology. In addition, q > 0.5 bicelles appeared to be multidisperse, containing aggregates of distinct sizes and shapes. Two populations were found for
150 mM q = 1 bicelles in our DLS experiments. One with a smaller population had a hydrodynamic radius, RH, around 2.5 nm, and the other with a
larger population had a radius of 6-8 nm. The size of the smaller species is
close to what was observed for micelles, indicating they are micelle-like.
The cryo-TEM image for q = 1.5 bicelles revealed that the morphology was
even more complex. At a high concentration (150 mM) of DMPC/DHPC,
small disks with a radius around 7 nm dominated, although some flat bandlike structures were also observed. At a low concentration (15 mM), clusters
of aggregated disks were observed, and the size of the dominant disks was
increased (13 nm). After 24 hour incubation, several additional different
shapes were observed. On the contrary, the particles in the q = 0.5 bicelle
sample appeared to be evenly distributed as small spherical particles with a
radius of around 3.7 nm, even when the concentration was low (15 mM), and
after 24 hour incubation.
Concentration is another critical factor for bicelles. As mentioned above, no
evident concentration dependence for either the size or the morphology was
observed for q = 0.5 bicelles. On the othter hand, the bicelles at low concentration and at q = 1 or 1.5 were observed to be larger. This is probably due to
the fact that the effective q-value for these is higher as the critical micellar
concentration of DHPC is considered to be constant at a given temperature.
As mentioned before, at 75 mM, the size of the particles in the sample was
observed to be temperature independent at q = 0.5 and 1, but not at q = 1.5.
Together with the observations of mixed species, we proposed that a critical
bicelle concentration (CBC) might be required to sustain a true “bicelle”
shape with a bilayer structure. The CBC seems to be dependent on the q ratio.
44
Note that a similar study also proposed a “CBC” where it was found that the
free DHPC concentration in bicelles depended on the q ratio and it was defined as CBC (Beaugrand et al. 2014).
In summary, all bicelle samples were homogenous at room temperature but
the bilayered structure most likely does not persist when the total lipid concentration is too low. At high temperatures, close to 40 C, more than one
species appeared for samples with high q ratios.
4.2.2 Bicelles with galactolipids introduced
In Paper IV, the galactolipids MGDG and DGDG were incorporated into
bicelles with the aim to provide novel membrane models for studies of proteins in e.g. plants and chloroplastic membranes in particular. These would
be useful for exploring the functions of both the galactolipids and related
proteins. DMPC/DHPC bicelles with a q value of 0.5 and a total lipid concentration of 300 mM were chosen as a control. Then different mole percents
of MGDG or DGDG were used to replace DMPC in the bicelle preparation.
Pulse field gradient diffusion NMR experiments were applied to check
whether there is a phase separation due to addition of galactolipids and if not,
to determine the size of the bicelles.
Based on the fact that the galactolipids MGDG and DGDG always had the
same diffusion rates as DMPC in the bicelle samples, we concluded that the
galactolipids were well mixed with DMPC without a phase separation. We
found that both MGDG and DGDG can be incorporated into DMPC/DHPC
bicelles up to 30%. In general, MGDG is a non-bilayer forming lipid and
DGDG is a bilayer-forming lipid. High amounts ( 60%) of MGDG in bilayers have been found to induce a hexagonal liquid crystalline phase or
other phases (Castro et al. 2007, Chung et al. 1993). However, the bilayer
can be retained as long as there are enough other bilayer-forming lipids present.
Moreover, the introduction of galactolipids caused a systematic size increase.
To facilitate a comparison, the diffusion rates of bicelles were converted to
hydrodynamic radii using the Stokes-Enstein equation together with an estimated Perrin shape factor of 1.22. Compared to the DMPC/DHPC bicelles
without galactolipids, the size of the bicelles containing 30% MGDG increased significantly (from 3.9 nm to 5.4 nm) while the size of bicelles containing the same amount of DGDG increased less (from 3.9 nm to 4.7 nm).
The increase in the bicelle size is due to several factors. First, the polyunsaturation in the acyl chains of both MGDG and DGDG leads to a loose
packing of the lipid tails, which should lead to a larger size. Second, both
galactolipids have longer chains (18 carbons) than DMPC (14 carbons). The
45
non-bilayer forming property of MGDG is most likely the reason for the
additional increase in size of MGDG-containing bicelles. One argument can
also be whether the shape of the bicelles is still the “ideal bicelle model”. If
the bicelles still have a disk shape with a planar area of DMPC and galactolipids surrounded by a rim area of DHPC, the larger size is due to a larger
planar area. This can be due to both reasons stated above but as well to an
additional number of lipids in the planar area. In this case, there would be an
increase in the amount of free DHPC as the planar area is proportional to R2
while the rim area is approximately proportional to R, and for a bicelle with
an increased size one would expect that there are more lipids added to the
bilayer than DHPC added to the rim. With the same amount of long-chained
lipids and DHPC in the sample, this would just give a smaller number of
bicelles with larger number of lipids involved per bicelle. On the other hand,
it is also possible that there is a higher degree of mixing between DHPC and
long-chained lipids after addition of non-bilayered lipid MGDG. It has previously been argued that different degrees of DHPC mixing was the reason
for different sizes of bicelles (Chupin et al. 1994, Triba et al. 2005).
4.2.3 Lipid dynamics in bicelles with galactolipids
In order to investigate the local dynamics of the lipids in the bicelles, two
relaxation parameters were collected for assigned carbon-13 peaks for
MGDG, DGDG and DMPC: longitudinal relaxation rates, R1, and steady
state heteronuclear NOE values. The experiments were performed with samples containing lipids at natural abundance. In addition, the two parameters
were fitted using a model-free analysis, giving two local motional parameters: the generalized local order parameter squared Slocal,j2 and the correlation
time τlocal,j for each assigned 13C-1H bond vector.
Not surprisingly, two distinct dynamic regions were observed based on the
local dynamics of the galactolipids in bicelles: the rigid sugar head-group
together with the glycerol part, and the flexible acyl chain region. All R1
values for the carbons in the galactose units or glycerol in both galactolipids
at both fields were above 2 s-1, while the R1 values for the acyl chain carbons
were all below 2 s-1. Turning to the NOEs, carbons in the galactose moieties
took on NOE values below 2 for both galcatolipids, while carbons in the acyl
chains take on values above 2. Moreover, a clear field dependence in the R1
values was observed for the sugar head-groups and the glycerol part in both
the MGDG and DGDG molecules, but was not observed for the acyl chain.
Finally, in both the MGDG and DGDG molecule, the order parameters for
the head-group and glycerol part were much higher than those for the acyl
chain. The former had Slocal,j2 values in the range of 0.6-0.9 while the latter
were in the range of 0.003-0.21. Interestingly, the local correlation times
46
were also much higher for the sugar ring carbon vectors (approaching ns)
than for the 13C-1H bond vectors in the acyl chains.
More details of the local motion were observed by taking a close-look at the
relaxation data. In the acyl chains, more extensive motions were observed
towards the end of the chains. In the head-group, the exocyclic carbon G6 in
MGDG and G6’ in the first galactose moiety of DGDG showed slower R1
relaxation rates compared to the ring carbons in the galactose moiety. This
may indicate that they are more flexible than the ring carbons, but no evident
differences in NOE values were observed for the exocyclic carbon in either
MGDG or DGDG. The model-free analysis showed that the exocyclic carbons had order parameters similar to the other carbons in the sugar moieties
but very short correlation times for the local motion. The equatorial bond
vector for G4 in MGDG showed a lower NOE value and a slightly higher R1
value compared to other carbons in the galactose unit. This is reasonable as
the 13C-1H bond vector at position G4 is equatorial while the other vectors
are axial. The model-free fitting for this bond vector was difficult, but always gave a relatively high order parameter. The high order parameter indicates that G4 in MGDG experiences less motion that is coupled to relaxation.
On the other hand, the unstable fitting suggests that the model is not appropriate and that the anisotropy of the head-group motion probably should be
taken into account in a more comprehensive model. The 13C-1H bond vectors
of G4’ and G4 in the two galactose units in DGDG had different behavior
compared to G4 in MGDG. The dynamics of G4’ was similar to average ring
carbons but G4 in DGDG had an opposite behavior, i.e. a slower relaxation
rate R1 and a larger NOE value. Although the fitting procedure for G4’ was
not successful, the fitting for G4 in the second sugar unit gave a relatively
small Slocal,j2 (0.68) compared to G4 in MGDG. This indicates that the orientation of the head-groups of MGDG and DGDG may be different.
A
B
Figure 10. Structures of MGDG (A) and DGDG (B) with nomenclature.
In addition to the galactolipid dynamics, the relaxation parameters for
DMPC and DHPC were also recorded and evaluated. Generally, three regions of different local lipid dynamics were observed: the glycerol linker
part was the most rigid, the choline carbons were relatively flexible and the
end of the acyl chain was the most flexible. When R1 and NOE values for
DMPC in galactolipid bicelles were compared to previous relaxation data for
47
DMPC in DMPC/DHPC bicelles (Biverståhl et al. 2009), little difference
was seen in the acyl chain. In addition, we observed only small differences
in relaxation data for the bicelles with either MGDG or DGDG. Overall, the
results suggest that the effects of MGDG and DGDG on the dynamics of
DMPC and DHPC are insignificant, again indicating that the bicelles with
galactolipids is a stable membrane mimetic model.
48
5. Conclusions and outlook
The overall aim of this thesis work involved two main objectives: to explore
the mechanism of dual targeting to mitochondria and chloroplasts and to find
appropriate membrane mimetics for NMR studies on membrane related projects including dual targeting. In Paper I and II, the interactions of a dual
targeting peptide with the mitochondrial import receptor Tom20 and chloroplastic import receptor Toc34 were characterized for the first time. In Paper
III, we characterized the most commonly used bicelles, DMPC/DHPC bicelles as a function of temperature, concentration, and q value. In Paper IV,
we investigated a novel bicelle system, bicelles containing galactolipids,
which are the most abundant lipids in chloroplastic membrane.
We found that ThrRS-dTP(2-60) has a helical propensity in the N-terminal
part and weak β/α secondary propensities in the following part. The potential
N-terminal α helix is an amphiphilic helix. ThrRS-dTP(2-60) interacts with
Tom20 with more than one region and probably also in more than one state.
The amphiphilic helix and the  motif play important roles in the dTPTom20 interaction. A relatively long C-terminal sequence of the peptide
seems not to be involved in the interaction with Tom20. In contrast, different
regions of the whole sequence of ThrRS-dTP(2-60) are important for Toc34
interaction. In addition, some structure might be formed upon Toc34 interaction. However, more experiments are required to provide solid evidence for
this.
This is just a beginning of the work aiming to elucidate dual targeting mechanisms at atomic level. First, to obtain a more general picture, several dual
targeting peptides should be studied. Second, more extensive studies would
shed light on the dual targeting mechanism. For example, dissociation constants for the interactions between ThrRS-dTP and the receptors would provide a quantitative measure of the interaction strength. Chemical exchange
rates between the bound state and the free state of ThRS-dTP in the presence
of receptors would help us understand the dynamics during the interactions.
Then there are additional perspectives concerning the dual targeting process
other than the interactions between the dTPs and the receptors, e.g. interaction of dTPs with cytosolic factors, possible phosphorylation of dTPs, the
structures of dTPs or structures of receptors in the presence of dTPs, mutagenesis studies of dTPs and so forth.
49
Moving to the studies on classic DMPC/DHPC bicelles, our studies suggest
that a q value of 0.5 is advisable for NMR studies, since this is the most stable bicelle among the three that we have studied (0.5, 1 and 1.5), with the
least temperature and concentration dependence and with only one species
present. If a high q value is required, additional attention should be paid to at
high temperature. In summary, caution is needed for preparing bicelles with
different concentrations and q values at different temperatures. Similar studies of bicelles containing lipids of different lengths may provide information
about how lipid length affects bicelle morphology. This may improve studies
on membrane protein mismatching.
With regard to the studies on bicelles containing MGDG or DGDG, as much
as 30 % of the DMPC in the q = 0.5, C = 300 mM bicelles can be replaced
by MGDG or DGDG and can be integrated into bicelles. The galactolipids in
the bicelles have two distinct local dynamic regions: the galactose units together with the glycerol linker region are very rigid, and the acyl chains are
quite flexible. In the future, this new membrane model can be applied to
MGDG or DGDG related studies. Specific questions involve targeting to
different organelles, including chloroplastic targeting peptides as well as
dual targeting. Investigations on the interaction between ThrRS-dTP and the
novel MGDG bicelles can help to solve the question on the role of MGDG in
the chloroplastic import pathway.
50
Sammanfattning på svenska
Mitokondrien och kloroplasten är två av de viktigaste subcellulära enheterna
i en växtcell. Eftersom dessa två organeller är separata membranomslutna
strukturer som inte kan tillverka alla sina egna proteiner, så är vissa proteiner
syntetiserade i cytosolen och måste importeras genom membranet.
Mitokondriella importpeptider och kloroplastiska importpeptider fungerar
som nycklar till respektive organell. De har några skillnader i
aminosyrasammansättning och struktur, fast de är väldigt lika. De flesta
proteiner som ska importeras till organeller har motsvarande särskilda
importpeptider men för några proteiner har det visat sig att de anväder
dubbla importpeptider (dTPs: dual targeting peptides) som samtidigt kan
öppna dörrarna till mitokondrier och kloroplaster och leverera proteinerna
dit.
I denna doktorsavhandling har en dubbel importpeptid, ThrRS-dTP,
undersöktes som modell-dTP i syfte att förstå mekanismen för dubbel
lokalisering. Vi har studerat interaktionen mellan denna peptid och
importreceptorerna Tom20 i mitokondrier och Toc34 i kloroplaster med
hjälp av NMR spektroskopi. Våra studier visade att ThrRS-dTP har en
aminosyrasammansättning som är en blandning av egenskapen hos både
mitokondriella såväl som kloroplastisk importpeptider och är ostrukturerad.
Peptiden har dock en benägenhet att bilda α-helix i den N-terminala delen.
Den amfipatiska α-helix-egenskapen tillsammans med ett specifikt motiv
visade sig vara viktigast för interaktionen med den mitokondriella receptorn
Tom20. För interaktionen med den kloroplastiska receptorn Toc34, är alla
delar av ThrRS-dTP lika viktiga, inklusive den C-terminala delen som
knappt påverkas av interaktion med Tom20.
Vi vill också veta om lipider i organellmembranet påverkar proteinimport
och funktionen av dubbla importpeptider. Därför har vi karakteriserat
DMPC\DHPC biceller som kan härma biologiska membran under olika
förhållanden och utvecklat nya biceller som innehåller dom kloroplastiska
galaktolipiderna MGDG eller DGDG. Dessa två lipider har tidigare visat sig
vara viktiga för proteinimport och det finns anledning att tro att
importpeptider växelverkar direkt med denna lipider. Biceller är blandningar
av lipider och surfaktanter (detergent på engelska). Under vissa förhållanden
bildar dessa disk-liknade objekt med ett litet bilager. I den första studien
51
visade vi att lipidkoncentration och förhållandet mellan lipider och
surfaktanter är mycket viktigt och att dessa påverkar bicellernas egenskaper.
Resultaten visar att man behöver ganska koncentrerade lösningar och en
relative hög halt surfaktant för att säkerställa att bicellerna verkligen är diskformade och att inga övriga objekt bildas. I den andra studien visade vi att
det är möjligt att framställa biceller med upp till 30 % galaktolipid. Med
hjälp av spinnrelaxationmätningar fann vi att båda galaktolipiderna
uppvisade relativt rigida sockerenheter medan fettsyrakedjorna var mycket
rörliga. Sammanfattningvis visar dessa resultat att lipiderna betedde sig likt
de som förekommer i mer realistiska membran och därför även att dessa nya
biceller är användbara för att studera t.ex. importpeptider.
52
Acknowledgements
This thesis cannot be finished without the help and support from groups of
people. I am grateful to you all that have surrounded me during my doctoral
journey. I would especially like to thank:
Lena Mäler, my supervisor, for accepting me as a doctoral student, for giving me freedom, giving me fire, and also keeping me on the right track with
your encouraging support and valuable guidance. I really appreciate your
effort in helping me.
Elzbieta Glaser, my second supervisor, for valuable scientific discussions
and for sharing the passion for life as a female scientist.
Erika Spånning, Jesper Lind, Jobst Liebau, Candan Ariöz, Amin Bakali, David Gotthold, Sofia Unnerståle, Scarlett Szpryngiel, Jonny
Eriksson and Katarina Edwards, my collabrators. I’ve learnt a lot from
you and it is a great pleasure to work with you. Erika, that’s a fruitful collaboration and I am always ready for more interesting trip stories. Jesper,
that was definitely the best supervision, and I’ll always appreciate that you
are generous in sharing your enormous knowledge.
Astrid Gräslund, for sharing your visions on various scientific topics. Stefan Nordlund, for taking care of all PhD students through all the checking
points. Åke Wieslander, for your vivid elucidation of the glycosyltransferases. Andreas Barth, for your valuable advice and serious attitude on the
papers. Törbjörn Astlind, for your enormous effort to keep all the machines
running properly, and your patience to explain what is going on especially in
the summers. Haidi Astlind, for your important administrative support.
Britt-Marie Olsson, for helping me the details in the lab. Among others,
you are the great persons that keep this working environment the best.
The present and past members of the biophysic group, for making us like a
big family. Johannes, for daily talks about our lives and this world, and for
showing me how to take the bike apart. Axel, for always being so calm and
cheerful, and for taking me to the activities which are ordinary for you but
exotic for me. Jobst, for making the paper better than I could and staying
after the master project and for keeping the plants as well as the afternoon
break alive. Pontus, for talks about opinions on those TV series and for
53
amazing me with your timing system. I appreciate you all for your input of
physics into my knowledge. Scarlett, for always caring about what we are
doing and for the MIT1 project. Sofia, for guidance during the early conference trips, for organization of excursion and baking events and for the unfinished HsapBK project. Fatemeh, for daily chat about our families here and
abroad and for fighting against work back to back. Anna, for being kind and
keeping things in order. Girls, I miss you. Candan, for the joyful time of
collaboration, for keeping smile and the spirit of “never lose hope” and for
joining us. Chenge, for your bright laughters and for the lovely afternoons in
the woods. Sebastian, for making me not alone at work in some evenings or
on the weekends lately. Ane M., for good discussions on F1β and PreP, and
for coming back from time to time. Nadja, Maurizio, Jüri, Andrea B.,
Henrik B., Jens D., Chang Y., Margareta, Gustav B., Britt-Marie S.,
Göran E., Shu Z. and Jakob for breaks and bonus discussions. Last but not
least, Ann, for saving my cover picture.
Friends since those early days: Changrong, for your kind help since day
one; Erik, for being a nice officemate and for your passion for music; Lumi,
for being a nice officemate and studybuddy in Swedish; Viktor, my Ph.D
twin, for your sense of humor and your effeciency.
People from Mikael Oliveberg’s group, in particular: Mikael for taking me
to the fantastic protein folding course and allowing me around in the lab;
Ellinor, for your guidance in the pre-PhD time. Xin and Fang, for your help
lately in the lab. People from Elzbieta Glaser’s group: Beata and Pedro, for
your suggestions and for the visit to the plant room.
Lab assistance buddies, Martin, Lisa L. and Walter, for the good time and
for answering the difficult questions.
My family, for your great supports to my decisions in work and for your
endless help outside work with love. In particular, Biao, for showing me the
beauty of NMR and getting me interested in it, for your confidence and trust
in me as well as criticism, and for always being there in all the good and
hard times.
54
References
Abe, Y., T. Shodai, T. Muto, K. Mihara, H. Torii, S. Nishikawa, T. Endo, and D.
Kohda. 2000. Structural basis of presequence recognition by the mitochondrial
protein import receptor Tom20. Cell 100:551-560.
Abragam, A. 1961. The principles of nuclear magnetism. Clarendon Press, Oxford.
Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts, J. D. Watson, and A. Grimstone. 1994. Molecular Biology of the Cell. Garland Publishing Inc., New
York.
Andersson, A. and L. Mäler. 2005. Magnetic resonance investigations of lipid
motion in isotropic bicelles. Langmuir 21:7702-7709.
Arnold, A., T. Labrot, R. Oda, and E. J. Dufourc. 2002. Cation modulation of
bicelle size and magnetic alignment as revealed by solid-state NMR and electron microscopy. Biophys. J. 83:2667-2680.
Aronsson, H. and P. Jarvis. 2009. The chloroplast protein import apparatus, its
components, and their roles, p. 89-123. In D. G. Robinson (ed.), The Chloroplast, vol. 13. Springer, Heidelberg.
Aronsson, H., P. Boij, R. Patel, A. Wardle, M. Töpel, and P. Jarvis. 2007.
Toc64/OEP64 is not essential for the efficient import of proteins into chloroplasts in Arabidopsis thaliana. Plant J. 52:53-68.
Aronsson, H., M. A. Schottler, A. A. Kelly, C. Sundqvist, P. Dormann, S. Karim,
and P. Jarvis. 2008. Monogalactosyldiacylglycerol deficiency in Arabidopsis
affects pigment composition in the prolamellar body and impairs thylakoid
membrane energization and photoprotection in leaves. Plant Physiol. 148:580592.
Barbosa-Barros, L., C. Barba, M. Cócera, L. Coderch, C. López-Iglesias, A. de
la Maza, and O. López. 2008. Effect of bicellar systems on skin properties.
Int. J. Pharm. 352:263-272.
Bax, A. and M. Ikura. 1991. An efficient 3D NMR technique for correlating the
proton and 15N backbone amide resonances with the a-carbon of the preceding
residue in uniformly 15N/13C enriched proteins. J. Biomol. NMR 1:99-104.
Bax, A. 2009. Weak alignment offers new NMR opportunities to study protein
structure and dynamics. Protein Sci. 12:1-16.
Beaugrand, M., A. A. Arnold, J. Hénin, D. E. Warschawski, P. T. Williamson,
and I. Marcotte. 2014. Lipid concentration and molar ratio boundaries for the
use of isotropic bicelles. Langmuir 30:6162-6170.
Benning, C. and H. Ohta. 2005. Three enzyme systems for galactoglycerolipid
biosynthesis are coordinately regulated in plants. J. Biol. Chem. 280:23972400.
Benz, J. P., J. Soll, and B. Bölter. 2009. Protein transport in organelles: the composition, function and regulation of the Tic complex in chloroplast protein import.
FEBS J. 276:1166-1176.
55
Berglund, A., C. Pujol, A. M. Duchene, and E. Glaser. 2009a. Defining the determinants for dual targeting of amino acyl-tRNA synthetases to mitochondria
and chloroplasts. J. Mol. Biol. 393:803-814.
Berglund, A., E. Spånning, H. Biverståhl, G. Maddalo, C. Tellgren-Roth, L.
Mäler, and E. Glaser. 2009b. Dual targeting to mitochondria and chloroplasts:
characterization of Thr–tRNA synthetase targeting peptide. Mol. plant 2:12981309.
Bhushan, S., C. Kuhn, A. Berglund, C. Roth, and E. Glaser. 2006. The role of
the N-terminal domain of chloroplast targeting peptides in organellar protein
import and miss-sorting. FEBS Lett. 580:3966-3972.
Bhushan, S., B. Lefebvre, A. Ståhl, S. J. Wright, B. D. Bruce, M. Boutry, and E.
Glaser. 2003. Dual targeting and function of a protease in mitochondria and
chloroplasts. EMBO Rep. 4:1073-1077.
Biverståhl, H., J. Lind, A. Bodor, and L. Mäler. 2009. Biophysical studies of the
membrane location of the voltage-gated sensors in the HsapBK and KvAP K
channels. Biochim. Biophys. Acta 1788:1976-1986.
Blatch, G. L. and M. Lässle. 1999. The tetratricopeptide repeat: a structural motif
mediating protein‐protein interactions. Bioessays 21:932-939.
Bolender, N., A. Sickmann, R. Wagner, C. Meisinger, and N. Pfanner. 2008.
Multiple pathways for sorting mitochondrial precursor proteins. EMBO Rep.
9:42-49.
Bruce, B. D. 2001. The paradox of plastid transit peptides: conservation of function
despite divergence in primary structure. Biochim. Biophys. Acta 1541:2-21.
Bruce, B. D. 1998. The role of lipids in plastid protein transport. Plant Mol. Biol.
38:223-246.
Bruce, B. D. 2000. Chloroplast transit peptides: structure, function and evolution.
Trends Cell Biol. 10:440-447.
Cantor, C. R. and P. R. Schimmel. 1980. Biophysical chemistry: Part II: Techniques for the study of biological structure and function. WH Freeman, New
York.
Carrie, C., E. Giraud, and J. Whelan. 2009a. Protein transport in organelles: dual
targeting of proteins to mitochondria and chloroplasts. FEBS J 276:1187-1195.
Carrie, C., K. Kühn, M. W. Murcha, O. Duncan, I. D. Small, N. O’Toole, and J.
Whelan. 2009b. Approaches to defining dual‐targeted proteins in Arabidopsis.
Plant J. 57:1128-1139.
Carrie, C. and I. Small. 2013. A reevaluation of dual-targeting of proteins to mitochondria and chloroplasts. Biochim. Biophys. Acta 1833:253-259.
Castro, V., S. V. Dvinskikh, G. Widmalm, D. Sandström, and A. Maliniak. 2007.
NMR studies of membranes composed of glycolipids and phospholipids. Biochim. Biophys. Acta 1768:2432-2437.
Cavanagh, J. 2007. Protein NMR spectroscopy: principles and practice. Academic
Press, USA.
Chabregas, S. M., D. D. Luche, L. P. Farias, A. F. Ribeiro, M. Van Sluys, C. F.
Menck, and M. C. Silva-Filho. 2001. Dual targeting properties of the Nterminal signal sequence of Arabidopsis thaliana THI1 protein to mitochondria
and chloroplasts. Plant Mol. Biol. 46:639-650.
Chabregas, S. M., D. D. Luche, M. Van Sluys, C. F. M. Menck, and M. C. SilvaFilho. 2003. Differential usage of two in-frame translational start codons regu-
56
lates subcellular localization of Arabidopsis thaliana THI1. J. Cell. Sci.
116:285-291.
Chacinska, A., M. van der Laan, C. S. Mehnert, B. Guiard, D. U. Mick, D. P.
Hutu, K. N. Truscott, N. Wiedemann, C. Meisinger, N. Pfanner, and P.
Rehling. 2010. Distinct forms of mitochondrial TOM-TIM supercomplexes
define signal-dependent states of preprotein sorting. Mol. Cell. Biol. 30:307318.
Chen, L. and H. Li. 1998. A mutant deficient in the plastid lipid DGD is defective
in protein import into chloroplasts. Plant J. 16:33-39.
Cheung, M. S., M. L. Maguire, T. J. Stevens, and R. W. Broadhurst. 2010.
DANGLE: A Bayesian inferential method for predicting protein backbone dihedral angles and secondary structure. J. Magn. Reson. 202:223-233.
Chew, O. and J. Whelan. 2003. Dual targeting ability of targeting signals is dependent on the nature of the mature protein. Funct. Plant Biol. 30:805-812.
Chou, J. J., J. L. Baber, and A. Bax. 2004. Characterization of phospholipid
mixed micelles by translational diffusion. J. Biomol. NMR 29:299-308.
Chung, J., J. Tolman, K. Howard, and J. Prestegard. 1993. Three-Dimensional
13C-13C-Correlated Experiments on Oligosaccharides. J. Magn. Reson.
102:137-147.
Chupin, V., R. van't Hof, and B. de Kruijff. 1994. The transit sequence of a chloroplast precursor protein reorients the lipids in monogalactosyl diglyceride
containing bilayers. FEBS Lett. 350:104-108.
Clore, G. M., P. C. Driscoll, P. T. Wingfield, and A. M. Gronenborn. 1990a.
Analysis of the backbone dynamics of interleukin-1. beta. using twodimensional inverse detected heteronuclear nitrogen-15-proton NMR spectroscopy. Biochemistry 29:7387-7401.
Clore, G. M., A. Szabo, A. Bax, L. E. Kay, P. C. Driscoll, and A. M. Gronenborn. 1990b. Deviations from the simple two-parameter model-free approach
to the interpretation of nitrogen-15 nuclear magnetic relaxation of proteins. J.
Am. Chem. Soc. 112:4989-4991.
Clubb, R. T., V. Thanabal, and G. Wagner. 1992. A new 3D HN (CA) HA experiment for obtaining fingerprint H N-H α cross peaks in 15 N-and 13 C-labeled
proteins. J. Biomol. NMR 2:203-210.
Creissen, G., H. Reynolds, Y. Xue, and P. Mullineaux. 1995. Simultaneous targeting of pea glutathione reductase and of a bacterial fusion protein to chloroplasts and mitochondria in transgenic tobacco. Plant J. 8:167-175.
Cronan, J. E. 2003. Bacterial membrane lipids: where do we stand? Annu. Rev.
Microbiol. 57:203-224.
Danielsson, J., J. Jarvet, P. Damberg, and A. Gräslund. 2002. Translational diffusion measured by PFG-NMR on full length and fragments of the Alzheimer
Ab (1–40) peptide. Determination of hydrodynamic radii of random coil peptides of varying length. Magn. Res. Chem. 40:S89-S97.
Das, A. K., P. T. W. Cohen, and D. Barford. 1998. The structure of the tetratricopeptide repeats of protein phosphatase 5: implications for TPR-mediated protein–protein interactions. EMBO J. 17:1192-1199.
Dobson, C. M. 1993. Flexible friends. Curr. Biol. 3:530-532.
Dörmann, P. and C. Benning. 2002. Galactolipids rule in seed plants. Trends Plant
Sci. 7:112-118.
57
Dorr, J. M., M. C. Koorengevel, M. Schafer, A. V. Prokofyev, S. Scheidelaar, E.
A. van der Cruijsen, T. R. Dafforn, M. Baldus, and J. A. Killian. 2014. Detergent-free isolation, characterization, and functional reconstitution of a tetrameric K+ channel: the power of native nanodiscs. Proc. Natl. Acad. Sci. U.
S. A. 111:18607-18612.
Duchêne, A., A. Giritch, B. Hoffmann, V. Cognat, D. Lancelin, N. M. Peeters,
M. Zaepfel, L. Maréchal-Drouard, and I. D. Small. 2005. Dual targeting is
the rule for organellar aminoacyl-tRNA synthetases in Arabidopsis thaliana .
Proc. Natl. Acad. Sci. U. S. A. 102:16484-16489.
Dunker, A. K., S. E. Bondos, F. Huang, and C. J. Oldfield. 2015. Intrinsically
disordered proteins and multicellular organisms. Sem. Cell Dev. Biol. 37:4455.
Dunker, A. K., J. D. Lawson, C. J. Brown, R. M. Williams, P. Romero, J. S. Oh,
C. J. Oldfield, A. M. Campen, C. M. Ratliff, and K. W. Hipps. 2001. Intrinsically disordered protein. J. Mol. Graph. Model. 19:26-59.
Dunker, A., E. Garner, S. Guilliot, P. Romero, K. Albrecht, J. Hart, Z. Obradovic, C. Kissinger, and J. Villafranca. 1998. Protein disorder and the evolution of molecular recognition: theory, predictions and observations. Pac. Symp.
Biocomput. 3:473-484.
Ellena, J. F., L. S. Lepore, and D. S. Cafiso. 1993. Estimating lipid lateral diffusion in phospholipid vesicles from carbon-13 spin-spin relaxation. J. Phys.
Chem. 97:2952-2957.
Endo, T. and D. Kohda. 2002. Functions of outer membrane receptors in mitochondrial protein import. Biochim. Biophys. Acta 1592:3-14.
Farrow, N. A., R. Muhandiram, A. U. Singer, S. M. Pascal, C. M. Kay, G. Gish,
S. E. Shoelson, T. Pawson, J. D. Forman-Kay, and L. E. Kay. 1994. Backbone dynamics of a free and a phosphopeptide-complexed Src homology 2
domain studied by 15N NMR relaxation. Biochemistry 33:5984-6003.
Fersht, A. 1999. Structure and mechanism in protein science: a guide to enzyme
catalysis and protein folding. W.H. Freeman and Company, New York.
Firlej-Kwoka, E., P. Strittmatter, J. Soll, and B. Bölter. 2008. Import of preproteins into the chloroplast inner envelope membrane. Plant Mol. Biol. 68:505519.
Fischer, E. 1894. Einfluss der Configuration auf die Wirkung der Enzyme. Ber.
Dtsch. Chem. Ges. 27:2985-2993.
Garner, E., P. Cannon, P. Romero, Z. Obradovic, and A. K. Dunker. 1998.
Predicting disordered regions from amino acid sequence. Genome Inform.
9:201-213.
Garrett, R. H. and C. M. Grisham. 2005. Biochemistry. Thomson Learning, USA.
Ge, C., E. Spanning, E. Glaser, and A. Wieslander. 2014. Import determinants of
organelle-specific and dual targeting peptides of mitochondria and chloroplasts
in Arabidopsis thaliana. Mol. Plant. 7:121-136.
Gilbert, A. and J. Hirst. 2004. Charge-transfer transitions in protein circular dichroism spectra. J. Mol. Struct. Theochem. 675:53-60.
Glover, K. J., J. A. Whiles, G. Wu, N. Yu, R. Deems, J. O. Struppe, R. E. Stark,
E. A. Komives, and R. R. Vold. 2001. Structural evaluation of phospholipid
bicelles for solution-state studies of membrane-associated biomolecules. Biophys. J. 81:2163-2171.
58
Gray, M. W. 2012. Mitochondrial evolution. Cold Spring Harb Perspect. Biol.
4:a011403.
Grzesiek, S. and A. Bax. 1992a. Improved 3D triple-resonance NMR techniques
applied to a 31 kDa protein. J. Magn. Reson. 96:432-440.
Grzesiek, S. and A. Bax. 1992b. An efficient experiment for sequential backbone
assignment of medium-sized isotopically enriched proteins. J. Biomol. NMR
99:201-207.
Grzesiek, S. and A. Bax. 1992c. Correlating backbone amide and side chain resonances in larger proteins by multiple relayed triple resonance NMR. J. Am.
Chem. Soc. 114:6291-6293.
Gunasekaran, K., C. J. Tsai, S. Kumar, D. Zanuy, and R. Nussinov. 2003. Extended disordered proteins: targeting function with less scaffold. Trends Biochem. Sci. 28:81-85.
Gutensohn, M., B. Schulz, P. Nicolay, and U. Flügge. 2000. Functional analysis of
the two Arabidopsis homologues of Toc34, a component of the chloroplast
protein import apparatus. Plant J. 23:771-783.
Hajduk, P. J., E. T. Olejniczak, and S. W. Fesik. 1997. One-dimensional relaxation-and diffusion-edited NMR methods for screening compounds that bind to
macromolecules. J. Am. Chem. Soc. 119:12257-12261.
Halle, B. 1991. Theory of spin relaxation by diffusion on curved surfaces. J. Chem.
Phys. 94:3150-3168.
Halle, B. and H. Wennerström. 1981. Interpretation of magnetic resonance data
from water nuclei in heterogeneous systems. J. Chem. Phys. 75:1928.
Hammen, P., D. Gorenstein, and H. Weiner. 1996. Amphiphilicity determines
binding properties of three mitochondrial presequences to lipid surfaces. Biochemistry 35:3772-3781.
Heazlewood, J. L., R. E. Verboom, J. Tonti-Filippini, I. Small, and A. H. Millar.
2007. SUBA: the Arabidopsis Subcellular Database. Nucleic Acids Res.
35:D213-D218.
Hedtke, B., T. Börner, and A. Weihe. 2000. One RNA polymerase serving two
genomes. EMBO Rep. 1:435-440.
Hirsch, S., E. Muckel, F. Heemeyer, G. von Heijne, and J. Soll. 1994. A receptor
component of the chloroplast protein translocation machinery. Science
266:1989-1992.
Hölzl, G. and P. Dörmann. 2007. Structure and function of glycoglycerolipids in
plants and bacteria. Prog. Lipid Res. 46:225-243.
Hooke, R. C. 1665. Micrographia: or some physiological descriptions of miniature
bodies made by magnifying glasses. John Martyn, and James Allestry, London,
England.
Houdai, T., N. Matsumori, and M. Murata. 2008. Structure of membrane-bound
amphidinol 3 in isotropic small bicelles. Org. Lett. 10:4191-4194.
Huang, S., N. L. Taylor, J. Whelan, and A. H. Millar. 2009. Refining the definition of plant mitochondrial presequences through analysis of sorting signals,
N-terminal modifications, and cleavage motifs. Plant Physiol. 150:1272-1285.
Hwang, T. L., P. C. M. van Zijl, and S. Mori. 1998. Accurate quantitation of water-amide proton exchange rates using the Phase-Modulated CLEAN chemical
EXchange (CLEANEX-PM) approach with a Fast-HSQC (FHSQC) detection
scheme. J. Biomol. NMR 11:221-226.
59
Inoue, K. and K. Keegstra. 2003. A polyglycine stretch is necessary for proper
targeting of the protein translocation channel precursor to the outer envelope
membrane of chloroplasts. Plant J. 34:661-669.
Jarvis, P. 2008. Targeting of nucleus‐encoded proteins to chloroplasts in plants.
New Phytol. 179:257-285.
Jarvis, P. and C. Robinson. 2004. Mechanisms of protein import and routing in
chloroplasts. Curr. Biol. 14:R1064-R1077.
Jarvis, J. A., M. T. Ryan, N. J. Hoogenraad, D. J. Craik, and P. B. Hoj. 1995.
Solution structure of the acetylated and noncleavable mitochondrial targeting
signal of rat chaperonin 10. J. Biol. Chem. 270:1323-1331.
Jarvis, P., P. Dormann, C. A. Peto, J. Lutes, C. Benning, and J. Chory. 2000.
Galactolipid deficiency and abnormal chloroplast development in the Arabidopsis MGD synthase 1 mutant. Proc. Natl. Acad. Sci. U. S. A. 97:81758179.
Jarvis, P., L. J. Chen, H. Li, C. A. Peto, C. Fankhauser, and J. Chory. 1998. An
Arabidopsis mutant defective in the plastid general protein import apparatus.
Science 282:100-103.
Jelic, M., J. Soll, and E. Schleiff. 2003. Two Toc34 homologues with different
properties. Biochemistry 42:5906-5916.
Karniely, S. and O. Pines. 2005. Single translation—dual destination: mechanisms
of dual protein targeting in eukaryotes. EMBO Rep. 6:420-425.
Kay, L. E., M. Ikura, R. Tschudin, and A. Bax. 1990. Three-dimensional tripleresonance NMR spectroscopy of isotopically enriched proteins. J.Magn.Reson.
89:496-514.
Keeler, J. 2005. Understanding NMR spectroscopy. John Wiley & Sons Ltd, Chichester.
Kleine, T., U. G. Maier, and D. Leister. 2009. DNA transfer from organelles to the
nucleus: the idiosyncratic genetics of endosymbiosis. Annu. Rev. Plant Biol.
60:115-138.
Kmiec, B., P. F. Teixeira, and E. Glaser. 2014. Shredding the signal: targeting
peptide degradation in mitochondria and chloroplasts. Trends Plant Sci.
19:771-778.
Knowles, T. J., R. Finka, C. Smith, Y. Lin, T. Dafforn, and M. Overduin. 2009.
Membrane proteins solubilized intact in lipid containing nanoparticles bounded by styrene maleic acid copolymer. J. Am. Chem. Soc. 131:7484-7485.
Kobayashi, K., M. Kondo, H. Fukuda, M. Nishimura, and H. Ohta. 2007. Galactolipid synthesis in chloroplast inner envelope is essential for proper thylakoid
biogenesis, photosynthesis, and embryogenesis. Proc. Natl. Acad. Sci. U. S. A.
104:17216-17221.
Koenig, P., M. Oreb, A. Höfle, S. Kaltofen, K. Rippe, I. Sinning, E. Schleiff, and
I. Tews. 2008. The GTPase cycle of the chloroplast import receptors
Toc33/Toc34: implications from monomeric and dimeric structures. Structure
16:585-596.
Kouranov, A. and D. J. Schnell. 1997. Analysis of the interactions of preproteins
with the import machinery over the course of protein import into chloroplasts.
J. Cell Biol. 139:1677-1685.
Kovács-Bogdán, E., J. Soll, and B. Bölter. 2010. Protein import into chloroplasts:
the Tic complex and its regulation. Biochim. Biophys. Acta 1803:740-747.
60
Kowalewski, J. and L. Mäler. 2006. Nuclear spin relaxation in liquids: theory,
experiments, and applications. CRC Press, Taylor & Francis Group, New York.
Krimm, I., P. Gans, J. Hernandez, G. J. Arlaud, and J. Lancelin. 1999. A coil–
helix instead of a helix–coil motif can be induced in a chloroplast transit peptide from Chlamydomonas reinhardtii . Eur. J. Biochem. 265:171-180.
Krogh, A., B. Larsson, G. Von Heijne, and E. L. Sonnhammer. 2001. Predicting
transmembrane protein topology with a hidden Markov model: application to
complete genomes. J. Mol. Biol. 305:567-580.
Kubis, S., A. Baldwin, R. Patel, A. Razzaq, P. Dupree, K. Lilley, J. Kurth, D.
Leister, and P. Jarvis. 2003. The Arabidopsis ppi1 mutant is specifically defective in the expression, chloroplast import, and accumulation of photosynthetic proteins. Plant Cell 15:1859-1871.
Lancelin, J., P. Gans, E. Bouchayer, I. Bally, G. J. Arlaud, and J. Jacquot. 1996.
NMR structures of a mitochondrial transit peptide from the green alga Chlamydomonas reinhardtii . FEBS Lett. 391:203-208.
Lang, B. F., M. W. Gray, and G. Burger. 1999. Mitochondrial genome evolution
and the origin of eukaryotes. Annu. Rev. Genet. 33:351-397.
Lauterwein, J., C. Bösch, L. R. Brown, and K. Wüthrich. 1979. Physicochemical
studies of the protein-lipid interactions in melittin-containing micelles. Biochim. Biophys. Acta 556:244-264.
le Maire, M., P. Champeil, and J. V. Møller. 2000. Interaction of membrane proteins and lipids with solubilizing detergents. Biochim. Biophys. Acta 1508:86111.
Leenhouts, J. M., J. de Gier, and B. de Kruijff. 1993. A novel property of a mitochondrial presequence: Its ability to induce cardiolipin-specific interbilayer
contacts which are dissociated by a transmembrane potential. FEBS Lett.
327:172-176.
Leenhouts, J. M., Z. Török, R. A. Demel, J. D. Gier, and B. D. Kruijff. 1994.
The full length of a mitochondrial presequence is required for efficient monolayer insertion and interbilayer contact formation. Mol. Membr. Biol. 11:159164.
Levitt, M. H. 2001. Spin dynamics: basics of nuclear magnetic resonance. John
Wiley & Sons, Ltd, Chichester.
Li, X., P. Romero, M. Rani, A. K. Dunker, and Z. Obradovic. 1999. Predicting
protein disorder for N-, C-and internal regions. Genome Informatics 10:30-40.
Li, H. and L. J. Chen. 1997. A novel chloroplastic outer membrane-targeting signal
that functions at both termini of passenger polypeptides. J. Biol. Chem.
272:10968-10974.
Lin, T. L., S. H. Chen, N. E. Gabriel, and M. F. Roberts. 1987. Small-angle neutron scattering techniques applied to the study of polydisperse rodlike diheptanoylphosphatidylcholine micelles. J. Phys. Chem. 91:406-413.
Lind, J. 2009. Dissertation/Thesis. Department of Biochemistry and Biophysics,
Stockholm University, Stockholm.
Lind, J., J. Nordin, and L. Mäler. 2008. Lipid dynamics in fast-tumbling bicelles
with varying bilayer thickness: effect of model transmembrane peptides. Biochim. Biophys. Acta 1778:2526-2534.
Linding, R., L. J. Jensen, F. Diella, P. Bork, T. J. Gibson, and R. B. Russell.
2003a. Protein disorder prediction: implications for structural proteomics.
Structure 11:1453-1459.
61
Linding, R., R. B. Russell, V. Neduva, and T. J. Gibson. 2003b. GlobPlot: Exploring protein sequences for globularity and disorder. Nucleic Acids Res.
31:3701-3708.
Lipari, G. and A. Szabo. 1982a. Model-free approach to the interpretation of nuclear magnetic resonance relaxation in macromolecules. 1. Theory and range
of validity. J. Am. Chem. Soc. 104:4546-4559.
Lipari, G. and A. Szabo. 1982b. Model-free approach to the interpretation of nuclear magnetic resonance relaxation in macromolecules. 2. Analysis of experimental results. J. Am. Chem. Soc. 104:4559-4570.
Lipfert, J., L. Columbus, V. B. Chu, S. A. Lesley, and S. Doniach. 2007. Size and
shape of detergent micelles determined by small-angle X-ray scattering. J.
Phys. Chem. B 111:12427-12438.
Lister, R., J. M. Hulett, T. Lithgow, and J. Whelan. 2005. Protein import into
mitochondria: origins and functions today. Mol. Membr. Biol. 22:87-100.
Lister, R., O. Chew, M. N. Lee, J. L. Heazlewood, R. Clifton, K. L. Parker, A. H.
Millar, and J. Whelan. 2004. A transcriptomic and proteomic characterization of the Arabidopsis mitochondrial protein import apparatus and its response to mitochondrial dysfunction. Plant Physiol. 134:777-789.
Lodish, H., A. Berk, S. L. Zipursky, P. Matsudaira, D. Baltimore, and J. Darnell. 2002. Molecular Cell Biology. W. H. Freeman, New York.
Lu, Z., W. D. Van Horn, J. Chen, S. Mathew, R. Zent, and C. R. Sanders. 2012.
Bicelles at low concentrations. Mol. Pharm. 9:752-761.
Luckey, M. 2008. Membrane structural biology: with biochemical and biophysical
foundations. Cambridge University Press, New York.
Ma, Y., A. Kouranov, S. E. LaSala, and D. J. Schnell. 1996. Two components of
the chloroplast protein import apparatus, IAP86 and IAP75, interact with the
transit sequence during the recognition and translocation of precursor proteins
at the outer envelope. J. Cell Biol. 134:315-327.
Mäler, L. and A. Gräslund. 2009. Artificial membrane models for the study of
macromolecular delivery. Methods Mol. Biol. 480:129-139.
Mäler, L. 2012. Solution NMR studies of peptide-lipid interactions in model membranes. Mol. Membr. Biol. 29:155-176.
Mandieau, V., I. Martin, and J. Ruysschaert. 1995. Interaction between cardiolipin and the mitochondrial presequence of cytochrome c oxidase subunit IV
favours lipid mixing without destabilizing the bilayer structure. FEBS Lett.
368:15-18.
Marion, D., L. E. Kay, S. W. Sparks, D. A. Torchia, and A. Bax. 1989a. Threedimensional heteronuclear NMR of nitrogen-15 labeled proteins. J. Am. Chem.
Soc. 111:1515-1517.
Marion, D., P. C. Driscoll, L. E. Kay, P. T. Wingfield, A. Bax, A. M. Gronenborn, and G. M. Clore. 1989b. Overcoming the overlap problem in the assignment of proton NMR spectra of larger proteins by use of three-dimensional
heteronuclear proton-nitrogen-15 Hartmann-Hahn-multiple quantum coherence
and nuclear Overhauser-multiple quantum coherence spectroscopy: application
to interleukin 1. beta. Biochemistry 28:6150-6156.
Marsh, J. A., V. K. Singh, Z. Jia, and J. D. Forman‐Kay. 2009. Sensitivity of
secondary structure propensities to sequence differences between α‐and γ‐
synuclein: Implications for fibrillation. Protein Sci. 15:2795-2804.
62
Martin, W. 2003. Gene transfer from organelles to the nucleus: frequent and in big
chunks. Proc. Natl. Acad. Sci. U. S. A. 100:8612-8614.
Matsumori, N., A. Morooka, and M. Murata. 2006. Detailed description of the
conformation and location of membrane-bound erythromycin a using isotropic
bicelles. J. Med. Chem. 49:3501-3508.
Mauseth, J. D. 2014. Botany: an introduction to plant biology. Jones & Bartlett
Publishers, Burlington, MA.
May, T. and J. Soll. 2000. 14-3-3 Proteins form a guidance complex with chloroplast precursor proteins in plants. Plant Cell 12:53-64.
Mayer, C., G. Gröbner, K. Müller, K. Weisz, and G. Kothe. 1990. Orientationdependent deuteron spin-lattice relaxation times in bilayer membranes: characterization of the overall lipid motion. Chem. Phys. Lett. 165:155-161.
Mazer, N. A., G. B. Benedek, and M. C. Carey. 1980. Quasielastic light-scattering
studies of aqueous biliary lipid systems. Mixed micelle formation in bile saltlecithin solutions. Biochemistry 19:601-615.
Mazzarello, P. 1999. A unifying concept: the history of cell theory. Nat. Cell Biol.
1:E13-E15.
Millar, A. H., J. Whelan, and I. Small. 2006. Recent surprises in protein targeting
to mitochondria and plastids. Curr. Opin. Plant Biol. 9:610-615.
Milo, R. and R. Philips. 2014. How large are chloroplasts?, p. 64-68. In R. Milo
and R. Philips (ed.), Cell Biology By The Number. Draft, Unknown.
Mittag, T., S. Orlicky, W. Y. Choy, X. Tang, H. Lin, F. Sicheri, L. E. Kay, M.
Tyers, and J. D. Forman-Kay. 2008. Dynamic equilibrium engagement of a
polyvalent ligand with a single-site receptor. Proc. Natl. Acad. Sci. U. S. A.
105:17772-17777.
Moberg, P., S. Nilsson, A. Stahl, A. Eriksson, E. Glaser, and L. Mäler. 2004.
NMR solution structure of the mitochondrial F1beta presequence from Nicotiana plumbaginifolia . J. Mol. Biol. 336:1129-1140.
Murcha, M. W., B. Kmiec, S. Kubiszewski-Jakubiak, P. F. Teixeira, E. Glaser,
and J. Whelan. 2014. Protein import into plant mitochondria: signals, machinery, processing, and regulation. J. Exp. Bot. 65:6301-6335.
Muto, T., T. Obita, Y. Abe, T. Shodai, T. Endo, and D. Kohda. 2001. NMR identification of the Tom20 binding segment in mitochondrial presequences. J. Mol.
Biol. 306:137-143.
Neupert, W. and J. M. Herrmann. 2007. Translocation of proteins into mitochondria. Annu. Rev. Biochem. 76:723-749.
Nordén, B. and A. Rodger. 1997. Circular dichroism and linear dichroism. Oxford
University Press, USA.
Orwick-Rydmark, M., J. E. Lovett, A. Graziadei, L. Lindholm, M. R. Hicks,
and A. Watts. 2012. Detergent-free incorporation of a seven-transmembrane
receptor protein into nanosized bilayer Lipodisq particles for functional and
biophysical studies. Nano letters 12:4687-4692.
Österberg, F., L. Rilfors, Å. Wieslander, G. Lindblom, and S. Gruner. 1995.
Lipid extracts from membranes of Acholeplasma laidlawii A grown with different fatty acids have a nearly constant spontaneous curvature. Biochim. Biophys. Acta 1257:18-24.
Paila, Y. D., L. G. Richardson, and D. J. Schnell. 2015. New insights into the
mechanism of chloroplast protein import and its integration with protein quality control, organelle biogenesis and development. J. Mol. Biol. 427:1038-1060.
63
Pain, D. and G. Blobel. 1987. Protein import into chloroplasts requires a chloroplast ATPase. Proc. Natl. Acad. Sci. U. S. A. 84:3288-3292.
Palmer, A. G.,3rd, C. D. Kroenke, and J. P. Loria. 2001. Nuclear magnetic resonance methods for quantifying microsecond-to-millisecond motions in biological macromolecules. Methods Enzymol. 339:204-238.
Peeters, N. and I. Small. 2001. Dual targeting to mitochondria and chloroplasts.
Biochim. Biophys. Acta 1541:54-63.
Perrin, F. 1934. Mouvement brownien d'un ellipsoide. I. Dispersion diélectrique
pour des molécules ellipsoidales. J.phys.radium 5:497-511.
Perry, A. J., J. M. Hulett, V. A. Likic, T. Lithgow, and P. R. Gooley. 2006. Convergent evolution of receptors for protein import into mitochondria. Curr. Biol.
16:221-229.
Perry, S. E. and K. Keegstra. 1994. Envelope membrane proteins that interact with
chloroplastic precursor proteins. Plant Cell 6:93-105.
Poincelot, R. P. 1976. Lipid and fatty acid composition of chloroplast envelope
membranes from species with differing net photosynthesis. Plant Physiol.
58:595-598.
Pon, L. A. and E. A. Schon. 2007. Mitochondria (Methods in Cell Biology, 80).
Academic Press, California; London.
Possingham, J. and W. Saurer. 1969. Changes in chloroplast number per cell during leaf development in spinach. Planta 86:186-194.
Prestegard, J. H. and A. I. Kishore. 2001. Partial alignment of biomolecules: an
aid to NMR characterization. Curr. Opin. Chem. Biol. 5:584-590.
Prilusky, J., C. E. Felder, T. Zeev-Ben-Mordehai, E. H. Rydberg, O. Man, J. S.
Beckmann, I. Silman, and J. L. Sussman. 2005. FoldIndex: a simple tool to
predict whether a given protein sequence is intrinsically unfolded. Bioinformatics 21:3435-3438.
Rasmussen, S. G. F., H. J. Choi, D. M. Rosenbaum, T. S. Kobilka, F. S. Thian,
P. C. Edwards, M. Burghammer, V. R. P. Ratnala, R. Sanishvili, and R. F.
Fischetti. 2007. Crystal structure of the human β2-adrenergic G-proteincoupled receptor. Nature 450:383-387.
Richardson, L. G., Y. D. Paila, S. R. Siman, Y. Chen, M. D. Smith, and D. J.
Schnell. 2014. Targeting and assembly of components of the TOC protein import complex at the chloroplast outer envelope membrane. Front. Plant Sci.
5:269.
Richter, R. P., R. Bérat, and A. R. Brisson. 2006. Formation of solid-supported
lipid bilayers: an integrated view. Langmuir 22:3497-3505.
Rimmer, K. A., J. H. Foo, A. Ng, E. J. Petrie, P. J. Shilling, A. J. Perry, H. D. T.
Mertens, T. Lithgow, T. D. Mulhern, and P. R. Gooley. 2011. Recognition
of mitochondrial targeting sequences by the import receptors Tom20 and
Tom22. J. Mol. Biol. 405:804-818.
Ris, H. and R. N. Singh. 1961. Electron microscope studies on blue-green algae. J.
Biophys. Biochem. Cytol. 9:63-80.
Rodríguez, G., L. Rubio, M. Cócera, J. Estelrich, R. Pons, A. de la Maza, and
O. López. 2010. Application of bicellar systems on skin: diffusion and molecular organization effects. Langmuir 26:10578-10584.
Roise, D., F. Theiler, S. J. Horvath, J. M. Tomich, J. H. Richards, D. S. Allison,
and G. Schatz. 1988. Amphiphilicity is essential for mitochondrial presequence function. EMBO J. 7:649-653.
64
Roise, D., S. J. Horvath, J. M. Tomich, J. H. Richards, and G. Schatz. 1986. A
chemically synthesized pre-sequence of an imported mitochondrial protein can
form an amphiphilic helix and perturb natural and artificial phospholipid bilayers. EMBO J. 5:1327-1334.
Rokov-Plavec, J., S. Lesjak, I. Landeka, I. Mijakovic, and I. WeygandDurasevic. 2002. Maize seryl-tRNA synthetase: specificity of substrate recognition by the organellar enzyme. Arch. Biochem. Biophys. 397:40-50.
Rudhe, C., O. Chew, J. Whelan, and E. Glaser. 2002. A novel in vitro system for
simultaneous import of precursor proteins into mitochondria and chloroplasts.
Plant J. 30:213-220.
Sagan, L. 1967. On the origin of mitosing cells. J. Theor. Biol. 14:225-276.
Saitoh, T., M. Igura, Y. Miyazaki, T. Ose, N. Maita, and D. Kohda. 2011. Crystallographic snapshots of Tom20-mitochondrial presequence interactions with
disulfide-stabilized peptides. Biochemistry 50:5487-5496.
Saitoh, T., M. Igura, T. Obita, T. Ose, R. Kojima, K. Maenaka, T. Endo, and D.
Kohda. 2007. Tom20 recognizes mitochondrial presequences through dynamic equilibrium among multiple bound states. EMBO J. 26:4777-4787.
Sanders, C. R. 2008. Development and application of bicelles for use in biological
NMR and other biophysical studies, p. 233-239. In Anonymous Modern Magnetic Resonance. Springer, Netherlands.
Sanders, C. R. and F. Sönnichsen. 2006. Solution NMR of membrane proteins:
practice and challenges. Magn. Reson. Chem. 44:S24-S40.
Sanyal, A. and G. S. Getz. 1995. Import of transcription factor MTF1 into the yeast
mitochondria takes place through an unusual pathway. J. Biol. Chem.
270:11970-11976.
Sato, S., Y. Nakamura, T. Kaneko, E. Asamizu, and S. Tabata. 1999. Complete
structure of the chloroplast genome of Arabidopsis thaliana. DNA Res. 6:283290.
Sattler, M., J. Schleucher, and C. Griesinger. 1999. Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution
employing pulsed field gradients. Prog. Nucl. Magn. Reson. Spectrosc. 34:93158.
Saviano, G., O. Crescenzi, D. Picone, P. Temussi, and T. Tancredi. 1999. Solution structure of human β‐endorphin in helicogenic solvents: an NMR study. J.
Pept. Sci. 5:410-422.
Schimper, A. 1883. Ueber die Entwicklung der Chlorophyllkörner und Farbkörper
(Nachtrag). Bot. Zeit 105-14, 121-31, 137-46, 153-62.
Schnell, D. J., F. Kessler, and G. Blobel. 1994. Isolation of components of the
chloroplast protein import machinery. Science 266:1007-1012.
Schulz, C., A. Schendzielorz, and P. Rehling. 2014. Unlocking the presequence
import pathway. Trends Cell Biol. 14:211-221.
Schünemann, D. 2007. Mechanisms of protein import into thylakoids of chloroplasts. Biol. Chem. 388:907-915.
Shen, Y., F. Delaglio, G. Cornilescu, and A. Bax. 2009. TALOS : a hybrid method
for predicting protein backbone torsion angles from NMR chemical shifts. J.
Biomol. NMR 44:213-223.
Sickmann, A., J. Reinders, Y. Wagner, C. Joppich, R. Zahedi, H. E. Meyer, B.
Schonfisch, I. Perschil, A. Chacinska, B. Guiard, P. Rehling, N. Pfanner,
65
and C. Meisinger. 2003. The proteome of Saccharomyces cerevisiae mitochondria. Proc. Natl. Acad. Sci. U. S. A. 100:13207-13212.
Sigalov, A. B. 2011. Uncoupled binding and folding of immune signaling-related
intrinsically disordered proteins. Prog. Biophys. Mol. Biol. 106:525-536.
Silva-Filho, M. C. 2003. One ticket for multiple destinations: dual targeting of proteins to distinct subcellular locations. Curr. Opin. Plant Biol. 6:589-595.
Small, D., S. Penkett, and D. Chapman. 1969. Studies on simple and mixed bile
salt micelles by nuclear magnetic resonance spectroscopy. Biochim. Biophys.
Acta 176:178-189.
Smith, M. D., C. M. Rounds, F. Wang, K. Chen, M. Afitlhile, and D. J. Schnell.
2004. atToc159 is a selective transit peptide receptor for the import of nucleusencoded chloroplast proteins. J. Cell Biol. 165:323-334.
Snel, M. M., A. I. de Kroon, and D. Marsh. 1995. Mitochondrial presequence
inserts differently into membranes containing cardiolipin and phosphatidylglycerol. Biochemistry 34:3605-3613.
Soll, J. and E. Schleiff. 2004. Protein import into chloroplasts. Nat. Rev. Mol. Cell
Biol. 5:198-208.
Stahl, A., P. Moberg, J. Ytterberg, O. Panfilov, H. Brockenhuus Von Lowenhielm, F. Nilsson, and E. Glaser. 2002. Isolation and identification of a
novel mitochondrial metalloprotease (PreP) that degrades targeting presequences in plants. J. Biol. Chem. 277:41931-41939.
Stejskal, E. O. and J. E. Tanner. 1965. Spin diffusion measurements: spin echoes
in the presence of a time-dependent field gradient. J. Chem. Phys. 42:288-292.
Stocking, C. and E. M. Gifford. 1959. Incorporation of thymidine into chloroplasts
of Spirogyra. Biochem. Biophys. Res. Commun. 1:159-164.
Struppe, J., J. A. Whiles, and R. R. Vold. 2000. Acidic phospholipid bicelles: a
versatile model membrane system. Biophys. J. 78:281-289.
Sun, Y., F. Forouhar, H. Li, S. Tu, Y. Yeh, S. Kao, H. Shr, C. Chou, C. Chen,
and C. Hsiao. 2002. Crystal structure of pea Toc34, a novel GTPase of the
chloroplast protein translocon. Nat. Struct. Mol. Biol. 9:95-100.
Sveshnikova, N., J. Soll, and E. Schleiff. 2000. Toc34 is a preprotein receptor
regulated by GTP and phosphorylation. Proc. Natl. Acad. Sci. U. S. A.
97:4973-4978.
Szpryngiel, S., C. Ge, I. Iakovleva, A. Georgiev, J. Lind, Å. Wieslander, and L.
Mäler. 2011. Lipid interacting regions in phosphate stress glycosyltransferase
atDGD2 from Arabidopsis thaliana. Biochemistry 50:4451-4466.
Tamm, L. K. and H. M. McConnell. 1985. Supported phospholipid bilayers. Biophys. J. 47:105-113.
Tanudji, M., S. Sjöling, E. Glaser, and J. Whelan. 1999. Signals required for the
import and processing of the alternative oxidase into mitochondria. J. Biol.
Chem. 274:1286-1293.
Tessmer, M. R. and D. A. Kallick. 1997. NMR and structural model of dynorphin
A (1-17) bound to dodecylphosphocholine micelles. Biochemistry 36:19711981.
Theg, S. M., C. Bauerle, L. J. Olsen, B. R. Selman, and K. Keegstra. 1989. Internal ATP is the only energy requirement for the translocation of precursor proteins across chloroplastic membranes. J. Biol. Chem. 264:6730-6736.
66
Thompson, A. A., J. J. Liu, E. Chun, D. Wacker, H. Wu, V. Cherezov, and R. C.
Stevens. 2011. GPCR stabilization using the bicelle-like architecture of mixed
sterol-detergent micelles. Methods 55:310-317.
Tjandra, N. and A. Bax. 1997. Direct measurement of distances and angles in biomolecules by NMR in a dilute liquid crystalline medium. Science 278:11111114.
Toeroek, Z., R. A. Demel, J. M. Leenhouts, and B. de Kruijff. 1994. Presequence-mediated intermembrane contact formation and lipid flow. A model
membrane study. Biochemistry 33:5589-5594.
Tompa, P. 2009. Structure and function of intrinsically disordered proteins. CRC
Press, New York.
Torres‐Franklin, M. L., A. Gigon, D. F. De Melo, Y. Zuily‐Fodil, and A. T.
Pham‐Thi. 2007. Drought stress and rehydration affect the balance between
MGDG and DGDG synthesis in cowpea leaves. Physiol. Plantarum 131:201210.
Triba, M. N., D. E. Warschawski, and P. F. Devaux. 2005. Reinvestigation by
phosphorus NMR of lipid distribution in bicelles. Biophys. J. 88:1887-1901.
Unnerståle, S., L. Mäler, and R. R. Draheim. 2011. Structural characterization of
AS1-membrane interactions from a subset of HAMP domains. Biochim. Biophys. Acta 1808:2403-2412.
Unnerståle, S., J. Lind, E. Papadopoulos, and L. Maler. 2009. Solution structure
of the HsapBK K+ channel voltage-sensor paddle sequence. Biochemistry
48:5813-5821.
van Dam, L., G. Karlsson, and K. Edwards. 2004. Direct observation and characterization of DMPC/DHPC aggregates under conditions relevant for biological
solution NMR. Biochim. Biophys. Acta 1664:241-256.
van Meer, G., D. R. Voelker, and G. W. Feigenson. 2008. Membrane lipids:
where they are and how they behave. Nat. Rev. Mol. Cell Bio. 9:112-124.
van Meer, G. and A. I. de Kroon. 2011. Lipid map of the mammalian cell. J. Cell.
Sci. 124:5-8.
van't Hof, R., R. A. Demel, K. Keegstra, and B. de Kruijff. 1991. Lipid—peptide
interactions between fragments of the transit peptide of ribulose-1, 5bisphosphate carboxylase/oxygenase and chloroplast membrane lipids. FEBS
Lett. 291:350-354.
van't Hof, R., W. van Klompenburg, M. Pilon, A. Kozubek, G. de Korte-Kool,
R. A. Demel, P. J. Weisbeek, and B. de Kruijff. 1993. The transit sequence
mediates the specific interaction of the precursor of ferredoxin with chloroplast
envelope membrane lipids. J. Biol. Chem. 268:4037-4042.
Voigt, A., M. Jakob, R. B. Klösgen, and M. Gutensohn. 2005. At least two Toc34
protein import receptors with different specificities are also present in spinach
chloroplasts. FEBS Lett. 579:1343-1349.
Vold, R. R. and R. S. Prosser. 1996. Magnetically oriented phospholipid bilayered
micelles for structural studies of polypeptides. Does the ideal bicelle exist? J.
Magn. Reson. 113:267-271.
von Heijne, G. 1996. Principles of membrane protein assembly and structure. Prog.
Biophys. Mol. Biol. 66:113-140.
von Heijne, G. and K. Nishikawa. 1991. Chloroplast transit peptides. The perfect
random coil? FEBS Lett. 278:1-3.
67
von Heijne, G. 1986. Mitochondrial targeting sequences may form amphiphilic
helices. EMBO J. 5:1335-1342.
Wallace, B. A. and R. W. Janes. 2009. Modern techniques for circular dichroism
and synchrotron radiation circular dichroism spectroscopy. Ios Press, Amsterdam.
Wallin, I. E. 1923. The mitochondria problem. Am. Nat. 57:255-261.
Wallin, I. E. 1927. Symbionticism and the Origin of Species. Williams & Wilkins
Company, Baltimore.
Waltner, M. and H. Weiner. 1995. Conversion of a nonprocessed mitochondrial
precursor protein into one that is processed by the mitochondrial processing
peptidase. J. Biol. Chem. 270:26311-26317.
Warschawski, D. E., A. A. Arnold, M. Beaugrand, A. Gravel, É. Chartrand,
and I. Marcotte. 2011. Choosing membrane mimetics for NMR structural
studies of transmembrane proteins. Biochim. Biophys. Acta 1808:1957-1974.
Watanabe, N., F. S. Che, M. Iwano, S. Takayama, S. Yoshida, and A. Isogai.
2001. Dual targeting of spinach protoporphyrinogen oxidase II to mitochondria
and chloroplasts by alternative use of two in-frame initiation codons. J. Biol.
Chem. 276:20474-20481.
Wennerström, H. and B. Lindman. 1979. Micelles. Physical chemistry of surfactant association. Physics Reports 52:1-86.
Wennerström, H., G. Lindblom, and B. Lindman. 1974. Theoretical aspects on
the NMR of quadrupolar ionic nuclei in micellar solutions and amphiphilic
liquid crystals. Chem. Scr. 6:97-103.
Whiles, J. A., K. J. Glover, R. R. Vold, and E. A. Komives. 2002. Methods for
studying transmembrane peptides in bicelles: consequences of hydrophobic
mismatch and peptide sequence. J. Magn. Reson. 158:149-156.
Wishart, D. S., B. D. Sykes, and F. M. Richards. 1992. The chemical shift index:
a fast and simple method for the assignment of protein secondary structure
through NMR spectroscopy. Biochemistry 31:1647-1651.
Wishart, D. S. and B. D. Sykes. 1994. The 13 C chemical-shift index: a simple
method for the identification of protein secondary structure using 13 C chemical-shift data. J. Biomol. NMR 4:171-180.
Wishart, D. S., C. G. Bigam, J. Yao, F. Abildgaard, H. J. Dyson, E. Oldfield, J.
L. Markley, and B. D. Sykes. 1995a. 1 H, 13 C and 15 N chemical shift referencing in biomolecular NMR. J. Biomol. NMR 6:135-140.
Wishart, D., C. G. Bigam, A. Holm, R. S. Hodges, and B. D. Sykes. 1995b. 1 H,
13 C and 15 N random coil NMR chemical shifts of the common amino acids.
I. Investigations of nearest-neighbor effects. J. Biomol. NMR 5:67-81.
Wolfe, S. L. 1972. Biology of the Cell. Wadsworth Pub Co, California.
Wright, P. E. and H. J. Dyson. 2015. Intrinsically disordered proteins in cellular
signalling and regulation. Nat. Struct. Mol. Biol. 16:18-29.
Wright, P. E. and H. J. Dyson. 1999. Intrinsically unstructured proteins: reassessing the protein structure-function paradigm. J. Mol. Biol. 293:321-331.
Wuthrich, K. 1986. NMR of proteins and nucleic acids. Wiley, USA.
Yogev, O., A. Naamati, and O. Pines. 2011. Fumarase: a paradigm of dual targeting and dual localized functions. FEBS J. 278:4230-4242.
Young, M. E., K. Keegstra, and J. E. Froehlich. 1999. GTP promotes the formation of early-import intermediates but is not required during the translocation step of protein import into chloroplasts. Plant Physiol. 121:237-244.
68
Zhang, X. P. and E. Glaser. 2002. Interaction of plant mitochondrial and chloroplast signal peptides with the Hsp70 molecular chaperone. Trends Plant Sci.
7:14-21.
Zhou, P., A. A. Lugovskoy, J. S. McCarty, P. Li, and G. Wagner. 2001. Solution
structure of DFF40 and DFF45 N-terminal domain complex and mutual chaperone activity of DFF40 and DFF45. Proc. Natl. Acad. Sci. U. S. A. 98:60516055.
Zuiderweg, E. R. P. and S. W. Fesik. 1989. Heteronuclear three-dimensional NMR
spectroscopy of the inflammatory protein C5a. Biochemistry 28:2387-2391.
69