PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Aoc&Pres/ClRC/Date0ue.hdd

 

STRUCTURAL AND FUNCTIONAL STUDIES OF PROTEINS INVOLVED IN
MITOCHONDRIAL FUNCTION AND STRUCTURE

By

Yanfeng Zhang

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Genetics

2009

ABSTRACT

STRUCTURAL AND FUNCTIONAL STUDIES OF PROTEINS INVOLVED IN
MITOCHONDRIAL FUNCTION AND STRUCTURE

Yanfeng Zhang

The dynamics of continuous fission and fusion events maintain normal mitochondrial
morphology and reduce the number of functional defects that could lead to a variety of
diseases. DLP-l and MFNs are essential protein components of human mitochondrial
fission and fusion machineries, and functional and structural studies of these proteins
would increase our understanding of the molecular mechanisms of mitochondrial
dynamics, function, and structure.

In this thesis, the biochemical and structural properties of recombinant DLP-l and
selected mutants have been studied. The G350D and R3658 mutants in the middle
domain severely impair the GTPase activity, but have no significant impact on the
protein’s oligomeric state, indicating that these two mutations interrupted the
intramolecular but not intermolecular interactions, and therefore, the middle domain of
DLP-l is important for the protein activity probably by facilitating appropriate
connections between the GTPase domain and the GED. The DLP-l and the isolated PH-
like domain bound free phosphoinositides indicated that DLP-l may interact with
membranes directly by binding acidic phospholipids preferentially phosphoinositides,
and the PH-like domain may be responsible for the interactions. Although GTPase
activity is abolished, the APH bound to liposomes, which suggested that in addition to the
PH—like domain, other regions of DLP-l may function as lipids-interacting enhancer as
well as scaffolds for orienting the PH-like domain into appropriate membrane targeting.

Structural studies of DLP-l and MFNs by way of X-ray crystallography have been

attempted. Molecular protein engineering was designed and performed to improve
protein solubility and to increase the likelihood of protein crystallization.

The recently identified (pro)renin receptor ((P)RR) is an important protein molecule
for the renin-angiotensin system (RAS), a mechanism regulating blood pressure and
cardiovascular function. The (P)RR C-terminus including the cytoplasmic tail is involved
in the assembly of the V0 portion of the vacuolar proton-translocating ATPase. The
cytoplasmic tail is short, but functionally important for the pivotal roles of (P)RR in a
number of signal transduction pathways that activated by binding of (pro)renin.

Finally, the last 19 amino acids of the (P)RR corresponding to the cytoplasmic tail
were fused into the C-terminus of E. coli maltose binding protein (MBP), and the chimera
was expressed in E. coli and purified to homogeneity. Protein crystals, in the presence
and absence of the MBP ligand maltose, were obtained, and X-ray diffraction data to 2.0
A resolution were collected. Despite significantly different unit-cell dimensions and
molecular packing, two monomers of the MBP fusion protein were found in the
asymmetric unit for both structures. Although the (P)RR cytoplasmic tail appeared as a
relatively flexible loop without obvious secondary structural elements, it seemed
responsible for the dimerization of MBP fusion protein in the asymmetric unit. The
residues in the cytoplasmic tail, particularly the two tyrosines, dominate the interdimer

interactions, suggesting a role of the cytoplasmic tail in protein oligomerization.

COPyright by
Yanfeng Zhang
2009

Dedicated to my beloved: my wife and my parents

ACKNOWLEDGEMENTS

Fist and foremost, I would like to express my sincere gratitude to my advisor, Dr. R.
Michael Garavito, for his superb guidance throughout my doctoral program. I would like
to thank him for his support, encouragement, patience, and for giving me freedom to
pursue all my ideas and interests, and educating me tremendously on protein chemistry.

I would like to thank my graduate committee members, Dr. Shelagh Ferguson-Miller,
Dr. James Geiger, and Dr. Katherine W. Osteryoung, for their guidance on my research. I
am grateful for their invaluable advice and suggestions.

I would also like to acknowledge present and former Garavito lab members, Dexin Sui,
Yi Zheng, Amy Scharmen, Young-Moon Cho, Michael Dumond, Dr. Nicole Webb, Dr.
Christine Harman, and Dr. Rachel Powers, for their help and discussions on DNA
cloning, protein expression and purification, and crystallography.

I thank Dr. Ling Qin and Jian Liu in the Ferguson-Miller lab for their help on freezing
crystals, collecting data, and building molecular models. I would like to thank Dr. David
DeWitt for reading my research proposal.

I would like to specially thank Dr. Barbara Sears, Director of Genetics Graduate
Program. I am grateful to her for her advice and guidance which helped me tremendously
for my personal and academic growth. I thank Jeannine Lee for her help.

I thank staff scientists at LS—CAT at Advanced Photon Source, Argonne National
Laboratory, for their training and help with data collections. Dr. Zdzislaw Wawrzak
performed the data collection of the apo-form MBP-(P)RR19 crystals.

Finally, I thank my wife, Xiaoli Gao, my parents, and parents-in-law. Without their

love and support, none of these would have been accomplished.

vi

TABLE OF CONTENTS

 

 

 

 

 

 

LIST OF TABLES -- - - -- ix
LIST OF FIGURES - - -- _ - - x
ABBREBIATIONS - xiii
CHAPTER 1

BACKGROUND AND INTRODUCTION - 1
1.1 Mitochondrial fission and fusion, dynamin-like protein 1, and mitofusins ............... 2
1.1.1 Mitochondrial evolution and fimction ....................................................................... 2
1.1.2 Mitochondrial fission and fusion ............................................................................... 3
1.1.3 Molecular mechanisms of Mitochondrial fission and fusion ..................................... 7
1.2 Renin-angiotensin system, (pro)renin receptor, and the membrane connection ...... 15
1.2.1 Renin-angiotensin systems (RAS) ........................................................................... 15
1.2.2 Prorenin and (pro)renin receptor .............................................................................. 18
1.2.3 Roles of C-terminus of (P)RR in functions of intracellular compartments ............. 21
1.2.4 Pivotal role of the cytoplasmic domain of (P)RR in signal transduction ................ 22
1.2.5 Interactions of the (P)RR with Promyelocytic Zinc Finger Protein ........................ 23
References. . ...................................................................................................................... 26
CHAPTER 2

FUNCTIONAL EXPRESSION OF DLP-l 44
2.1 Introduction .............................................................................................................. 45
2.2 Materials and methods ............................................................................................. 46
2.3 Results and discussion ............................................................................................. 50
2.3.1 Protein exprssion and purification ........................................................................... 50
2.3.2 GTPase activities ..................................................................................................... 57
2.3.3 Oligomeric states ..................................................................................................... 61
2.3.4 Interactions with lipids and membranes .................................................................. 72
2.4 Conclusion ............................................................................................................... 79
References. ...................................................................................................................... 81
CHAPTER 3

STRUCTURAL STUDIES OF DLP-l AND MFN-l - -- - ................ 85
3.1 Introduction .............................................................................................................. 86
3.2 Crystallization trials of DLP-l ................................................................................ 86
3.3 Protein engineering methods for crystallization enhancement ................................ 88
3.3.1 General introduction and rationale ........................................................................... 89
3.4 Experimental procedures, results, and discussion on DLP-l ................................... 92
3.4.1 Fragments construction ............................................................................................ 92
3.4.2 Limited proteolysis ................................................................................................. 98
3.4.3 Surface-entropy reduction ...................................................................................... 104

vii

3.4.4 Fusion proteins for crystallization aids .................................................................. 110

 

 

3.4.5 Homology model of the GTPase domain of DLP-l .............................................. 115
3.5 MFNs experimental procedures and results .......................................................... 116
3.5.1 Protein expression and purification ....................................................................... 116
3.5.2 MFN fragments construction ................................................................................. 118
3.5.3 Protein renaturation ................................................................................................ 125
3.6 Summary ................................................................................................................ 126
References. . .................................................................................................................... 128
CHAPTER 4

STUDIES ON THE (PRO)RENIN RECEPTOR AND THE PROMYELOCYTIC
ZINC FINGER PROTEIN - 131
4.1 Introduction ............................................................................................................ 132
4.2 (P)RR wild type protein ......................................................................................... 133
4.3 Structural determination of the cytoplasmic tail of the (P)RR .............................. 136
4.3.1 Experimental procedures ...................................................................................... 136
4.3.2 Results and discussion ........................................................................................... 141
4.4 Cloning and expression of the promyelocytic zinc finger protein (PLZF) ............. 172
4.4.1 Experimental procedures ....................................................................................... 172
4.4.2 Results and discussion ........................................................................................... 178
4.5 Conclusion ............................................................................................................ 178
References. .................................................................................................................... 180
CHAPTER 5

FUTURE DIRECTIONS _ - - 189
5.1 Alternative protein expression systems and crystallization screen methods ......... 190
5.2 Further structural studies on (P)RR and roles of tyrosines on the cytoplasmic tail in
protein oligomerization and signal transduction ............................................................ 192
References. .................................................................................................................... 195
APPENDIX

PRELIMINARY MAPPING OF PLZF FRAGMENTS BINDING TO THE (P)RR
CYTOPLASMIC TAIL ..... 197

 

viii

LIST OF TABLES

Table 2.1 Specific GTPase activity of DLP-l WT and mutants ....................................... 53
Table 2.2 Calculated molecular weight of DLP-l WT and mutants by gel filtration ....... 71

Table 2.3 Illustration of the binding strength of DLP-l WT and the PH-like domain to
each phospholipids by FAT-Westem blot analysis .......................................................... 74

Table 4.1 Data-collection and processing statistics ........................................................ 163

Table 4.2 Potential hydrogen bonds formed between two MBP-(P)RR19 monomers... 169

ix

LIST OF FIGURES

Figure 1.1 Dynamin superfarnily members in animals and plants ............................. 6

Figure 1.2 Schematic nonproteolytic activation of prorenin bound to the (pro)renin

receptor ............................................................................................... 20
Figure 2.1 Sequence alignments of partial GTPase domain (upper) and the middle
domain (lower) of dynamin related proteins ....... . ............................................................. 52
Figure 2.2 Purification of DLP-l WT and mutants .......................................................... 54
Figure 2.3 GTPase activity of purified DLP-l WT and mutants by malachite green
colorimetric assay ............................................................................................................. 55
Figure 2.4 Schematic model of DLP-l intramolecular interactions ................................. 60
Figure 2.5 Size exclusion chromatography of DLP-l WT and mutants ........................... 66
Figure 2.6 FAT-Westem blot analysis of phospholipids binding by DLP-l WT and the
PH-like domain ................................................................................................................. 73
Figure 2.7 Liposome—binding of DLP-l WT and mutants ................................................ 77
Figure 3.1 Comparison of secondary structure predictions of DLP-l from PSI-PRED and
SABLE severs by visualizing the predictions .................................................... 94
Figure 3.2 Multiple-sequence alignments of DLP-l ......................................................... 95
Figure 3.3 Schematic illustration of DLP-l constructs ..................................................... 99
Figure 3.4 SDS-PAGE analysis of Limited proteolysis of DLP-l by different proteases101
Figure 3.5 Mass determination of the DLP-l proteolytic fragment by mass spectrometriy)2
Figure 3.6 Surface—entropy prediction of DLP-l indicating that the PH-like domain
contains many residue clusters with high conformational entropy ................................. 103
Figure 3.7 Illustration of designed RMG series expression vectors. .............................. 106
Figure 3.8 Homology model of the GTPase domain of DLP-l ...................................... 116

Figure 3.9 Expression and purification of MFN-1 in E.coli. .......................................... 119

Figure 3.10 Size exclusion chromatography of MFN-l by a Superdex 200 10/30 column

......................................................................................................................................... 121
Figure 3.11 Multiple-sequence alignments of MFN-1 by T-Coffee and ESPript ........... 122
Figure 3.12 Comparison of secondary structure predictions of MFN-1 from PSI-PRED
and SABLE severs by visualizing the predictions .......................................................... 124
Figure 3.13 Schematic illustration of MFN-1 constructs. ............................................... 125
Figure 4.1 Putative crystals of the (pro)renin receptor in hanging drops ....................... 136
Figure 4.2 SDS-PAGE of MBP-(P)RR19 purification by amylose column ................... 143
Figure 4.3 Purification of the MBP-(P)RR19 by ion exchange chromatography ........... 144
Figure 4.4 Accurate molecular weight of purified MBP-(P)RR19 measured by mass
spectrometry .................................................................................................................... 147
Figure 4.5 Crystals of apo-MBP-(P)RR19 from hanging drop vapor diffusion ............. 149
Figure 4.6 Crystals of ligand-bound MBP-(P)RR19 ...................................................... 151
Figure 4.7 Structure of MBP-(P)RR19 with maltose bound ........................................... 152
Figure 4.8 Structure of MBP-(P)RR19 without maltose ................................................ 153

Figure 4.9 The (P)RR19 in a 2Fo-Fc electron-density map contoured at one standard
deviation above the mean density ................................................................................... 154

Figure 4.10 Dimeric interface of apo-MBP-(P)RR19 is mediated by hydrogen bonds

formed by residues from (P)RR19 peptide ..................................................................... 155
Figure 4.11 Dimeric interface of 1i gand-bound-MBP-(P)RR1 9 is predominated by

hydrogen bonds formed by residues from (P)RR19 peptide .......................................... 158
Figure 4.12 Size exclusion chromatography of MBP-(P)RR19 ..................................... 161
Figure 4.13 Structure of the (P)RR cytoplasmic tail ....................................................... 170
Figure 4.14 Purification of PLZF fiagments by Ni-NTA column .................................. 175
Figure A1 Binding of the PLZF RD2 domain to the MBP-(P)RR19 ............................. 199

xi

Figure A2 Binding of the PLZF BTB domain to the MBP-(P)RR19 ............................. 200

Figure A3 Binding of the PLZF zinc finger domain to the MBP-(P)RR19 .................... 201
Figure A4 Purification of E. coli maltose-binding protein (MBP) by ion exchange

chromatography .............................................................................................................. 202
Figure A5 Binding of the PLZF RD2 domain to the MBP ............................................. 202

(Images in this dissertation are presented in color)

xii

ACE:

ADL2:

Ang:
AT:

BAR:

BDLP:

BTB:

DLP-l :

Dnml:

DsRed:

DTT:

Dyn A:

ER:
on:
GAP:
GED:
GEF:

GST:

hGBPl :

HR:

ABBREVIATIONS
Angiotensin-converting enzyme
Arabidopsis dynamin-like protein 2
Angiotensin
Angiotensin receptor
Bin/amphiphysin/Rvs domain
Cyanobacterial dynamin-like protein
Bric-a-brac domain
Dynamin-like protein 1
Dynamin-like protein in yeast
A red fluorescent protein fi'om Discosoma sp. reef coral
Dithiothreitol
Dictyostelium discoideum dynamin A
Endoplasmic reticulum
Fuzzy onion protein
GTPase-activating protein
GTPase effector domain
Guanine nucleotide-exchange factors
Glutathione-S-transferase
Human Guanylate-binding protein 1

Heptad repeat region

xiii

IPTG:

LB:

M6P/IGF2R:

MBP:

Isopropyl B-D-l-thiogalactopyranoside
Luria-Bertani media
Mannose 6-phosphate/insulin-like growth factor II receptor

Malto'se—binding protein

MBP-(P)RR19: (P)RR19 fused in the C-terminus of MBP

MFN:
MtDN A:
PA:

Pfu:

PH:
PI3K:
PI(4)P:
PIs:
PLZF:

PRD:

(Pro)renin:

(P)RR:
(P)RR19:
PS:
APH:

RAS:

Mitofusin
Mitochondrial DNA
Phosphatidic acid
Pyrococcus fim’osus
Pleckstrin-homology domain
Phosphatidylinositol-3 kinase
Phosphatidylinositol 4-phosphate
Phosphoinositides
Promyelocytic Zinc Finger Protein
Proline-rich domain
Prorenin and renin
Prorenin and renin receptor
The last 19 amino acid of (P)RR encoding the cytoplasmic tail
Phosphatidylserine

DLPl mutant that the PH-like domain is deleted

Renin-angiotensin system

xiv

RD2:
Sarcosyl:
SH:

TM:

Second repressor domain
N—lauryl-sarcosine
Src homology domain

Transmembrane domain

XV

CHAPTER 1

Background and introduction

1.1 Mitochondrial fission and fusion, dynamin-like protein 1 and mitofusins
1.1.1 Mitochondrial evolution and function

Mitochondria are essential organelles of eukaryotes. It is generally believed that
mitochondria were originally derived fiom prokaryotes by endosymbiosis (Osteryoung
and Nunnari 2003; Dyall, Brown et al. 2004; Gray, Lang et a1. 2004; Dolezal, Likic et a1.
2006; Embley and Martin 2006). The evolutionary scenario is that an aerobic prokaryote
(probably an alpha-proteobacterium) was first engulfed by an ancestor of eukaryotes
about 1.5 billion years ago. Then, the two organisms developed a symbiotic relationship
in which the host provides nutrients for the endosymbiont and takes advantage of the
energy generated by the endosymbiont through aerobic respiration. After years of
adaptation and evolution, genome reduction eventually occurred by which most genes of
the endosymbiont were lost, while some were transferred to nucleus of the host. One of
the key events of endosyrnbiotic organelle biogenesis is the development of
mitochondrial division machinery for reproduction. Another key step is evolution of a
protein translocation mechanism that allows movement of nuclear-encoded proteins into
the mitochondria.

The main function of mitochondria is to produce energy for cellular activities by the
process of oxidative phosphorylation (OXPHOS). Maintenance of normal mitochondrial
function is essential for cellular energy metabolism. Dysfunction of nuclear-encoded or
mitochondrial DNA (m'tDNA)—encoded mitochodrial genes results in various

mitochondrial diseases such as autosomal dominant optical atrophy, Charcot-Marie—Tooth

(CMT) type 2A, Friedreich’s ataxia (FRDA), and Kearns-Sayre syndrome (KSS)
(Chinnery and Schon 2003; Newmeyer and Ferguson-Miller 2003; Rotig, Lebon et al.
2004; Zeviani and Di Donato 2004; Sato, Nakada et al. 2006; Debray, Lambert et al.
2008). Mitochondrial diseases damage a wide range of organs, including brain, heart, and
muscles, and impact the progression of disease states such as diabetes, heart disease,

kidney failure, mental and developmental defects.

1.1.2 Mitochondrial fission and fusion

Mitochondria exist as highly dynamic tubular networks, which are thought to be the
normal morphological state. The dynamic morphology is maintained by tightly regulated
fission and fusion processes (Osteryoung 2000; Osteryoung 2001; Chen, Chomyn et al.
2005; Okamoto and Shaw 2005; Heath-Engel and Shore 2006; Santel 2006; Cerveny,
Tamura et a1. 2007; Hoppins, Lackner et al. 2007; Berman, Pineda et al. 2008; Santel and
Frank 2008; Benard and Karbowski 2009; Hoppins and Nunnari 2009). The balance
between the fusion and fission events regulates the morphology of mitochondria
throughout the cell cycle stages. During cell division, the mitochondria divide and are
distributed to daughter cells. They also undergo continuous fusion to process genetic
recombination with one another to prevent dysfunction arising from mutated genes.
Disruption of fission machinery causes formation of clusters that contain elongated,
interconnected mitochondria (Smimova, Griparic et a]. 2001; Yoon, Krueger et al. 2003;

Stojanovski, Koutsopoulos et al. 2004). Disruption of fusion results in fragmentation of

normal mitochondria (Chen, Detmer et a1. 2003; Chen, Chomyn et al. 2005).

Although still controversial, mitochondrial fission and firsion have been suggested to
be involved in apoptosis, 3 form of programmed cell death that is essential for embryonic
development (Perfettini, Roumier et al. 2005; Youle and Karbowski 2005; Parone and
Martinou 2006; Cheng, Leach et al. 2008; J eong and Sec] 2008; Suen, Norris et al. 2008).
The remarkable morphological characteristic of mitochondria during apoptosis is that
they are fragmented, which indicates that the mitochondrial fission is related to apoptosis.
Disruption of protein components of the fission machinery before induction of apoptosis
not only inhibits mitochondrial fission, but also affects apoptosis (Frank, Gaume et al.
2001; Lee, Jeong et al. 2004). Overexpression of the fission proteins induces apoptosis
(James, Parone et al. 2003). On the other hand, upregulation of the mitochondrial fusion
machinery inhibits apoptosis (Sugioka, Shimizu et al. 2004).

The human mtDNA is 16.6 kb in size and it is circular and double stranded. It encodes
13 respiratory chain subunits, 22 transfer RNAs and 2 ribosomal RNAs (Anderson,
Bankier et a1. 1981). Other protein components of the respiratory chain are encoded by
the nuclear DNA. Theoretically, normal mitochondria are thought to contain wild type
mtDNAs. In reality, the mtDNA is a mixture of wild type and mutated mtDNA
(heteroplasmy) (Sato, Nakada et al. 2006). Because mtDNA has a much higher mutation
rate than nuclear DNA, the extensive and continuous fusion among mitochondria may be
a specific defense mechanism to complement mutated mtDNA and prevent mitochondria

diseases. However, cells only tolerate the mtDNA mutation within a specific level. Once

a particular threshold of the content of mutated mtDNA is passed, normal functions of the
respiratory chain are disrupted and mitochondria-related diseases would occur (Zeviani

2004; Sato, Nakada et al. 2006).

.313 H Amvm BE :00 62 5.x 32 .shoEooEE qofimu Ea couflsna 28580:.
iambic: "meatomzm 5:893 unis. “8:362 H .m can ._. d .oxomofim 88m :83. .90388 comma can commmm 35503838
E 3302: bozmfioxo 8a A Ewaoxfi was 2438 7589a ofi~-§§v 05 .53 con. 93 Ba. 58qu mo smash 582%
08 t3 :0 8x3 conga mo bo§> 05 950% Pawn £5. .35... can 33:5 5 Eon—:2: bun—aroma £5.33 :4 Ban:—

 

520 552% EEO Sgsmeozz 0
European 3838:qu U .EMSS $5. 5895 oxzéfiugn—O ESE? 32330 O

 

 

 

EouoaooEszonE 85 :95 . 1 .

     

ES a
Beta _ tvcoaoofiz

03285

33960 >00 oEomowS—m

:8 283 =3 EEE<

1.1.3 Molecular mechanisms of mitochondrial fission and firsion
1.1.3.1 Dynamin superfamily

Correct mitochondrial mechanisms depend on the functions of dynamin-related
proteins, which are large GTPases in the dynamin superfarnily (Fig. 1.1). Proteins in the
dynamin superfarnily can be divided into two major groups: classical dynamins and
dynamin-related proteins (Praefcke and McMahon 2004). Classical dynamins are proteins
involved in scission of clathrin-coated vesicles during endocytosis (Grigliatti, Hall et al.
1973; Obar, Collins et al. 1990). They have five identifiable domains: GTPase domain,
Middle domain, Pleckstrin-homology domain (PH domain), GTPase effector domain
(GED) and Proline-rich domain (PRD). The dynamin-related proteins are involved in
various membrane tubulation and remodeling events mainly involving fission and fusion
of organelles (Staeheli, Horisberger et al. 1984; Staeheli, Haller et al. 1986; Rothman,
Raymond et al. 1990; Gu and Venna 1996; Hales and Fuller 1997; Hermann, Thatcher et
al. 1998; Kang, J in et al. 1998; Labrousse, Zappaterra et al. 1999; Santel and Fuller 2001;
Olichon, Emorine et al. 2002; Gao, Kadirjan—Kalbach et al. 2003; Santel, Frank et al.
2003; Gao, Sage et al. 2006; Glynn, Froehlich et al. 2008; Glynn, Yang et al. 2009). They
lack one or more domains (such as the PH domain or the PRD) or have additional
domains (such as insertions or organelle-localization signals) compared to the classical
dynamins.

Oligomerization plays an essential role in the functions of proteins in the dynamin

family. Most of them form ring-like or helical structures that bind to the target membrane

and stimulate GTPase activity. Two models have been proposed for molecular functions
of proteins in the dynamin family. One is that the stimulated GTP hydrolysis upon
membrane binding results in a conformational change to generate mechanical force,
which facilitate the membrane constriction and scission (Sweitzer and Hinshaw 1998).
The opposing model is that GTPase activity regulates the function of other molecules that
are actually involved in the membrane fission and remodeling (Scheffzek, Ahmadian et al.
1997; Sever, Muhlberg et al. 1999; Sever, Damke et al. 2000). However, the exact
mechanism remains controversial.

Dynamin family GTPases are unique, and are clearly different from canonical small
GTPases such as Ras-like and heterotrimeric GTP-binding proteins. They are much larger
(70-100 kDa) containing a large GTPase domain (30-40 kDa), and have relatively lower
affinity for guanine nucleotides. They are stable in the absence of guanine nucleotides but
have high turnover rates of GTP (Schweins, Geyer et al. 1995; Binns, Helms et al. 2000;
McEwen, Gee et al. 2001). The GTPase activity of proteins in the dynamin family is
regulated by self-oligomerization, while canonical small GTPases require guanine
nucleotide-exchange factors (GEF) and GTPase-activating protein (GAP) to catalyze

GTPase activity. All of these properties distinguish them fiom canonical small GTPases.

1.1.3.2 Mitochondrial fission and DLP-I
Human mitochondrial fission is regulated by dynamin-like protein 1 (DLP-l), which

represents a class of highly conserved GTPases (Dnml in yeast / Drpl in C. elegans)

(Hales and Fuller 1997; Hermann, Thatcher et al. 1998; Smirnova, Shurland et al. 1998;
Bleazard, McCaffery et al. 1999; Labrousse, Zappaterra et al. 1999; Smirnova, Griparic et
al. 2001). The molecular mechanism of mtichondrial fission is mostly studied in yeast. In
yeast, in addition to Dnml, Mdvl and Fisl are required to form fission complexes that
catalyze mitochondrial fission (Fekkes, Shepard et a1. 2000; Mozdy, McCafi‘ery et al.
2000; Tieu and Nunnari 2000; Cerveny, McCaffery et al. 2001; Lackner, Homer et al.
2009). F isl is an outer mitochondrial membrane protein with the C-terminus inserted into
the membrane and the N-terminus forming a tetratricopeptide repeat domain (TPR) that
faces the cytoplasm (Suzuki, Neutzner et al. 2005). Mdvl is a WD-40 repeat-containing
protein which probably functions as a molecular adaptor to mediate the formation of
fission complex (Tieu, Okreglak et al. 2002; Lackner, Homer et al. 2009). It is believed
that the membrane-anchored Fisl recruits Mdvl first, and then this F isl-Mdvl complex
consequently recruits Dnml (Mozdy, McCaffery et al. 2000; Cerveny, McCaffery et a1.
2001; Tieu, Okreglak et a1. 2002; Karren, Coonrod et al. 2005). In humans, DLP-l
mediated mitochondrial fission has been thought to have a mechanism similar to that of
yeast. The structure of human Fisl (hFisl) has been shown to be similar to that of Fisl in
yeast (Suzuki, Jeong et al. 2003; Dohm, Lee et al. 2004; Suzuki, Neutzner et al. 2005).
However, the homologue of Mdvl has not been identified in humans.

DLP-l also undergoes various posttranslational modifications for regulation of
mitochondrial fission. CAMP—dependant protein kinase-dependent phosphorylation on

residues in the C-terminus affects the DLP-l GTPase activity, promotes mitochondrial

fission in mitotic cells, and alters mitochondria morphology (Chang and Blackstone 2007;
Cribbs and Strack 2007; Taguchi, Ishihara et al. 2007). Nitric oxide can trigger
mitochondrial fission, synaptic loss, and neuronal damage, possibly due to the
S-nitrosylation of DLP-l (Cho, Nakamura et al. 2009). The mitochondrial E3 ubiquitin
ligase MARCH V regulates mitochondrial fission by facilitating DLP-l binding to actual
mitochondrial division sites (N akamura, Kimura et al. 2006; Yonashiro, Ishido et al. 2006;
Karbowski, Neutzner et al. 2007). In addition, small ubiquitin-like modifier (SUMO)
proteins are reported to be involved in the DLP-l mediated mitochondrial fission (Harder,
Zunino et al. 2004; Di Bacco and Gi112006; Wasiak, Zunino et al. 2007; Zunino, Schauss
et al. 2007).

DLP-I is comprised of four domains: an N-terminal GTPase domain (1-340 aa) with
conserved GTP-binding motifs; a middle domain (341-500 a) with a potential role in
self-assembly; an insertion (501-607 aa) of unknown firnction (a “putative” PH-like
domain); and a GED (608-710 a) with potential roles in not only self-assembly, but also
cooperative stimulation of GTPase activity.

The GTPase domain contains conserved G1-G4 GTP-binding motifs, which are spread
over the whole domain. Based on the crystal structure of other GTP-binding proteins (Pai,
Krengel et al. 1990; Vetter and Wittinghofer 2001), the G1 motif, or P-loop
(G32xxxxG37K3gs39 in DLP-l) is involved in the binding of phosphates, while the G2
motif (T59 in DLP-l) coordinates magnesium ion and water for catalysis. The G3 motif

(D156xxG159 in DLP-l) is hydrogen-bonded with gamma-phosphates of GTP. The G4

10

motif (T/N215K/R216xD213 in DLP-l) coordinates the base and ribose. Conformational
change caused by hydrolysis of GTP occurs in the switch 1 and switch 2 regions, which
overlap with the G2 and G3 motif, respectively. Mutation of DLP-l-K38A abolishes the
GTPase activity of DLP-l in vitro, and overexpression of DLP-l-K38A in COS-7 cells
markedly reduces the mitochondrial fission (Yoon, Pitts et al. 2001; Zhu, Patterson et a1.
2004). Experiments using labeled GTP have demonstrated that DLP—l-K38A binds but
does not hydrolyze or release GTP (Yoon, Pitts et al. 2001). The crystal structures of the
GTPase domain of Dictyostelium discoideum dynamin A (Dyn A) and rat dynamin 1 have
shown that the overall fold is similar with but larger than those of the canonical GTPase
(Niemann, Knetsch et al. 2001; Reubold, Eschenburg et al. 2005). The structures display
an eight-stranded beta-sheet with six parallel and two antiparallel strands surrounded by
nine helices.

The PH domain of classical dynamins is involved in binding to negatively charged
lipid membranes. The single PH domain of dynamin binds to the lipids with a relatively
low affinity compared to other PH domains. The oligomerized PH domains lead to strong
binding of dynamins to the membranes (Klein, Lee et al. 1998; Lemon and Ferguson
2000). The crystal structure of the PH domain of classical dynamin shows a
seven-stranded beta sheet followed by an alpha helix in the C-terminus. Three variable
loops form a positive surface that may be sites for interaction with lipids (Ferguson,
Lemrnon et al. 1994). For DLP-l, sequence analysis does not indicate any region that

shows a high degree of homology to the PH domains of classical dynamins or other PH

11

domains. However, DLP-l has been shown to tubulate membrane in vitro (Yoon, Pitts et
al. 2001). Although this behavior is reminiscent of the PH domain in classical dynamins,
whether DLP-l binds to specific lipids and which DLP-l domain is critical for membrane
binding remain unknown.

The middle domain and the GED of classical dynamins are thought to be important for
oligomerization and subsequent multimerization (Zhang and Hinshaw 2001). In yeast,
Mutant Dnm1-G335D, which contains a point mutation in the middle domain of Dnml,
fails to self-assemble and forms stable dimers (Ingerman, Perkins et al. 2005). This
mutation inhibits mitochondrial fission but still can interact with the fission complex
containing Mdvl and Fisl (Bhar, Karren et al. 2006). Point mutaions in the middle
domain of the human dynamin forms a dimer instead of a tetramer and fails to integrate
into higher order structures under conditions which stimulate assembly. The dimeric form
also markedly reduces the GTPase activity of the dynamin (Rarnachandran, Surka et al.
2007)

The GED of DLP-l has a potential role in not only self-assembly, but also cooperative
stimulation of GTPase activity. The crystal structure of a member of the dynamin family,
human Guanylate-binding protein 1 (hGBPl) in both nucleotide-free and GTP analogue
GppNHp-bound forms has been solved (Prakash, Praefcke et al. 2000; Prakash, Renault
et a1. 2000). hGBPl is a protein induced by gamma-interferon to mediate antiviral
pathway (Anderson, Carton et al. 1999). Unlike DLP-l, hGBPl lacks the PH-like domain.

The structure is composed of two parts: the large global domain that is the GTPase

12

domain, and the long, purely alpha-helical domain that contains the Middle domain and
the GED. The structure also reveals that the (GED forms a long helix that folds back to
interact with the helical bundle in the middle domain, and contacts the global GTPase
domain. The three-dimensional map of classical dynamin determined from cryo-electron
micrographs at a resolution of 20 I suggested a similar structural property (Zhang and
Hinshaw 2001). Yeast two-hybrid experiments showed that the GED of DLP-l strongly
interacts with the middle and GTPase domains. Mutant DLP-l-K679A, which contains a
mutation in the GED of DLP-l , impairs the GTPase activity and affects the intra- and
intermolecular interactions (Zhu, Patterson et al. 2004). The crystal structure of the
GTPase domain of DynA reveals a hydrophobic groove, suggesting a GED interacting
site (Niemann, Knetsch et al. 2001). The crystal structure of the GTPase domain of rat
dynamin 1 also supports the model that the C-terminus, probably GED, folds back to
stimulate GTPase activity (Reubold, Eschenburg et al. 2005). However, elucidation of
how the GED is involved in the assembly and the stimulation of GTPase activity will

require the detailed structure of the firll-length protein.

1.1.3.3 Mitochondrial fusion and MFNS

Mitochondrial outer membrane fusion is mediated by mitofusins (MFNs), belonging to
a group of highly conserved mitochodrial transmembrane GTPase homologues (szol in
yeast / F 20 in Drosophila) (Hales and Fuller 1997; Hermann, Thatcher et al. 1998; Santel

and Fuller 2001). The two mammalian MFNs, MFN—1 and MFN-2 share the same

13

structural motifs with 77% sequence similarity (Santel, Frank et al. 2003). They contain
four domains: an N-terminal GTPase domain (1-370 a in MFN-1) with conserved
GTP-binding motifs; two heptad repeat regions, I-IRl (371-580 aa in MFN-1) and HR2
(630-741 a in MFN-1), with the HR2 forming a dimeric, antiparallel coiled coil that
mediates tethering of adjacent mitochondria (Koshiba, Detrner et al. 2004); and a
bipartite transmembrane domain, or TM (581-629 a in MFN-l). Structural and
biochemical studies have established that the MFNs are anchored in the mitochondrial
outer membrane with both N- and C-terminus exposed to the cytosol (Rojo, Legros et al.
2002). MFN-l and MFN-2 may play both redundant and distinct roles in mitochondrial
fusion in a GTPase activity-dependent manner (Ishihara, Eura et al. 2004; Chen, Chomyn
et al. 2005).

Similar to those of DLP-l, the conserved G1-G4 GTP-binding motifs of the GTPase
domain of MFNs are spread over the entire domain. GTP hydrolysis has been shown to
be important for MFN-1 mediated tethering of mitochondria (Ishihara, Eura et al. 2004).
Mutant MFN-I-KggT, which contains a mutation in the G1 motif of the GTPase domain,
blocks the ability of overexpressed MFN-1 to induce formation of elongated networks of
mitochondria (Santel, Frank et al. 2003). Overexpression of mutant MFN-l-TlogA,
containing a mutation in the G2 motif, results in fragmentation of mitochondria (Santel,
Frank et al. 2003).

The HRl, HR2, and TM of MFN-2 are found to be important for mitochondrial

targeting. Deletion of any one of these domains caused partial localization to

14

mitochondria and significant amounts of protein remained in the cytosol (Rojo, Legros et
a1. 2002). The crystal structure of a part of HR2 (HR2660-735) of MFN-1 revealed that it
forms a dimeric antiparallel coiled coil that is 95 ' long (Koshiba, Detrner et a1. 2004).
Mutant HR2660-735 L691P and L705P reduce the stability of the HR2 coiled coil and mutants
MFN-l-L691P and MFN-l-L705P cannot restore mitochondrial tubules in MFN-null cells
to the extent that with wide-type MFN-1, indicating that the HR2 coiled coil is important
for the mitochondrial fusion (Koshiba, Detrner et al. 2004). It is believed that HR2
functions as a mitochondrial fusion tether (Rojo, Legros et al. 2002; Koshiba, Detrner et
a1. 2004). The crystal structure of cyanobacterial DLP (BDLP) in both nucleotide-free
and GDP-associated conformation provided structural insights into the functional
mechanisms of dynamins (Low and Lowe 2006). Sequence analysis shows that the BDLP
is closed related to the Arabidopsis chloroplast FZO-like protein (FZL) (Gao, Sage et al.
2006), suggesting a bacterial ancestry of dynamins (Low and Lowe 2006).

Based on the best-studied virus-mediated fusion and vesicle fusion mechanism, and
genetic and structural studies of MFNs, it has been proposed that MFNs form complexes
in trans that mediate homotypic interactions between adjacent mitochondria and are
likely directly involved in outer membrane fusion (Koshiba, Detrner et al. 2004; Griffin,

Detmer et al. 2006).

1.2 Renin-angiotenisn system, (pro)renin receptor, and the membrane connection

1.2.1 Renin-a_ngioten_sin system (RAS)

15

The renin-angiotensin system (RAS) is an incompletely understood mechanism
regulating blood pressure, cardiac and vascular fimction. The aspartyl protease, renin,
which is released by kidney, cleaves the angiotensinogen to generate the decapeptide
angiotensin (Ang) I. The inactive Ang I was firrther processed by the
angiotensin-converting enzyme (ACE) to an active octopeptide, Ang II. Ang II interacts
with cell membrane receptors ATI and AT2, which belong to the G protein—coupled
receptor family, to active downstream signal pathways and regulate blood pressure and
cardiovascular modeling.

Renin is considered to catalyze the rate-limiting step of RAS to generate the precursor
of active end product, Ang I (Ang I; Asp]-Arg2-Val3-Tyr4—Ile5-His6—Pro7-Phe8-Hisg-LeuIO)
(de Gasparo, Catt et al. 2000). The X-ray crystal structure shows that the general fold of
renin is comprised of two homologous domains (Sielecki, Hayakawa et al. 1989). The
active site and ligand-binding motif are located in between the two domains (Rahuel,
Priestle et al. 1991). The two major catalytic residues Asp 32 and Asp 215 are in each part.
Renin cleaves the Leu10 -Valll peptide bond and releases Ang I. The ACEs bind Ang I and
cleave off the two C-terminal residues and create active Ang II. The heptapeptide
Ang-(1-7) and dodecapeptide Ang-(l-12) are among those angiotensins discovered
recently and involved in different signal transduction pathways.

The signals of renin and angiotensins were mediated by two major G protein-coupled
receptors, AT] and AT2. Although Ang II binds to both ATI and AT2, the majority of the

Ang II signal was transducted by ATl (Timmerrnans, Wong et al. 1993). The qu family

16

of G proteins dominates the downstream interactions of AT] (Wang, J ayadev et al. 1995).
The Go, Gum, and G12” 3 are other G protein interaction partners of the ATI (Shirai,
Takahashi et al. 1995; Ushio-Fukai, Griendling et al. 1998; Fujii, Onohara et al. 2005).
Besides G proteins, ATl also interacts with beta arrestins to activate a mitogen activated
protein kinase (MAPK) cascade (McDonald, Chow et al. 2000; Tohgo, Pierce et al. 2002).
The AT] receptor-associated protein (ATRAP), the epidermal growth factor (EGF)
receptor, and the nicotinamide adenine dinucleotide phosphate oxidase
(NADPH)—generated reactive oxygen species (ROS) are other partners with which ATl
interacts (Griendling, Minieri et al. 1994; Sabri, Govindarajan et al. 1998; Daviet,
Lehtonen et a1. 1999; Zuo, Ushio-Fukai et al. 2005; Mehta and Griendling 2007; Tamura,
Tanaka et al. 2007). The blood pressure regulation mechanisms that mediated the AT2
receptor are less understood. AT2 interacts with G, (Kang, Richards et al. 1995). The
vasodilation effect mediated by the cascade of bradykirrin (BK), nitric oxide (NO), and
cGMP is thought to be induced by AT2 (Siragy and Carey 1996; Siragy, J affa et al. 1996;
Siragy and Linden 1996).

Since renin, ACE, and AT] and AT2 are the major protein components of the RAS
system, inactivating renin or ACE or blocking the Ang II-receptor interaction are current
therapeutic strategies in hypertension drug development. According to molecular
modeling and X-ray crystal structure of the active site of renin, a number of renin
inhibitors have been created for direct renin inhibition (Rahuel, Priestle et al. 1991;

Rahuel, Rasetti et al. 2000; Holsworth, Powell et al. 2005; Tice, Xu et a1. 2009). These

17

inhibitors occupy the active site of renin so that its substrate angiotensinogen could not
bind and be processed. One of the representative direct renin inhibitors is aliskiren which
has a very high binding affinity for renin (Rahuel, Rasetti et al. 2000; Wood, Maibaum et
al. 2003). RAS blockers such as ACE inhibitors or Ang II ATl blockers cause
accumulation of Ang I and decrease of Ang H. However, they also stimulate the renin
activity probably because of disruption of the feedback loop which inhibits renin (Vander
and Geelhoed 1965; Bing 1973; Borghi, Boschi et al. 1993; Roig, Perez-Villa et al. 2000;

Azizi and Menard 2004).

1.2.2 horenirrind (pro)renin receptor

Prorenin is the renin inactive precursor and it has a 43-amino acid prosegrnent in the
N-terminus(Fukamizu, Nishi et al. 1988; Inagami 1991; Morris 1992; Morris 1992). The
prosegrnent has been thought to block the interaction between the active site and
angiotensins (Baxter, James et al. 1989; Heinrikson, Hui et al. 1989; Shiratori, Nakagawa
et al. 1990). Prorenin can be activated proteolytically by cleaving off the prosegrnent or
non-proteolytically at low pH, low temperature or by interaction with specific antibodies
(Sealey and Laragh 1975; Derkx, von Gool et al. 1976; Leckie and McGhee 1980; Derkx,
Schalekamp et al. 1987; Pitarresi, Rubattu et al. 1992; Reudelhuber, Brechler et al. 1998;
Suzuki, Hatano et al. 1999). The presence of non-proteolytic activation has led to the
identification of the “gate” and “handle” regions in the prosegrnent which may control

prorenin activation (Suzuki, Hayakawa et al. 2003). In blood, the level of prorenin is

18

about 10 times higher that that of renin (Sealey, Glorioso et al. 1986; Leckie, Bimie et al.
1994). However, the exact function of the circulating prorenin remains unclear.

There are two proteins that are generally accepted to be (pro)renin receptors. One is the
mannose 6-phosphate/insulin-like growth factor II receptor (M6P/IGF2R) (van Kesteren,
Danser et al. 1997; van den Eijnden, Saris et al. 2001). The M6P/IGF2R binds renin and
prorenin but does not stimulate any protein activity. However, on binding to the
M6P/IGF2R, prorenin is processed to renin by removing the prosegrnent. Therefore, the
M6P/IGF2R is considered as a clearance receptor of (pro)renin (van den Eijnden, Saris et
a1. 2001; Saris, van den Eijnden et al. 2002).

The second recently identified receptor is the (pro)renin receptor ((P)RR) (Nguyen,
Delarue et a1. 2002). The (P)RR binds both renin and prorenin and it increases the renin
catalytic activity of converting angiotensinogen to Ang I up to four fold (Nguyen,
Delarue et al. 2002; Nabi, Kageshima et al. 2006). Moreover, binding of (P)RR probably
causes a conformational change of prorenin prosegrnent to activate the prorenin (Nguyen,
Delarue et al. 2002; Batenburg, Krop et al. 2007) (Fig. 1.2). One controversial hypothesis
is that the “gate” region of T7FKR and the “handle” region of I] IFLKR on the prosegrnent
of prorenin may be critical for its binding to the (P)RR (Suzuki, Hayakawa et al. 2003).
The (P)RR gene encodes a 350-amino acid protein with a short signal peptide in
N-terminus, a putative 20-amino acid transmembrane region near the C-terminus, and a

short 19-amino acid cytoplasmic tail.

19

.232 “82 Steam ._o£aoz .58 .56
2.32058 :88 can 330356.80 E 37% Egg—063853 ”SERVE £9583 2E3 U 393% Z 8on =83. dowofichEwa
0283 8 a £52? £033 .Bguoaowsc 02a :8 :onoav :5 .9353 mivam 05 0638s was SE82 Euoxoav 05 8 EB
58 55:83 Ho 5:8 SEQ ”533?. 55.223 2: 3 c.52— Enouoa .«o .8333: ark—88.55: gunman—om ”NA unaut—

 

Ewm o .
a = o R A: ........... . AL AL
ESQSEE \ LOP TEL

320:2

20

558$

fl\ _W BB 268 m=to>oo O
A EoEwomoE %

A: w=<v = Emceofifix m2? EEG ME , 5:5 558$
Q wc<v H Emcflomwfix

Gofiswosmaaewca. g mo<

 

.oEocoE mam 5:2 2:82

 
 

1.2.3 Roles of the C-terminus of (P)RR in functimf intracellular compartments

The N—terminal extracellular domain of the (P)RR receptor displays no sequence
similarity to any known protein, but the C-terminal domain of (P)RR is highly
homologous to ATP6M8-9, a highly conserved accessory protein involved in the
assembly of the V0 portion of the vacuolar proton-translocating ATPase (V-ATPases)
(Ludwig, Kerscher et a1. 1998). As a consequence, (P)RR is also known as ATP6AP2
(adaptor protein type II vacuolar H+-ATPase). The V-ATPases play essential roles in
regulating cytoplasmic pH maintenance and the acidifying intracellular compartments
including endosomes, lysosomes and secretory vesicles (Nishi and Forgac 2002). Some
important cellular activities such as endocytosis, intracellular targeting of lysosomal
enzymes, protein processing and degradation, and small molecule trafficking, are
functioned by the V—ATPases (Stevens and Forgac 1997; Forgac 1999; Bowman and
Bowman 2000; Nishi and Forgac 2002).

The ATP6M8-9 is a 70 amino acid portion of the (P)RR C-terminus, which includes a
small portion of the extracellular domain, the transmembrane region, and the cytoplasmic
tail. How the (P)RR is cleaved and which molecule(s) is involved remain unclear,
although the arginine 277 is the putative cutting site for the protease firrin (Bader 2007).
The (P)RR is also present in lower species that do not have RAS such as C. elegans and
Drosophila. The protein sequence of the extracellular part before the putative cutting site
shows little homology between vertebrates’ and invertebrates’ (P)RR, although high

sequence similarity was observed among mammals’ (P)RR. However, The C-terminus

21

after the putative cutting site exhibits about 40% - 50% sequence identity (Bader 2007).
Moreover, a mutation of (P)RR caused X-linked mental retardation and epilepsy
syndrome in humans (Rarnser, Abidi et al. 2005). Zebrafish with mutant (P)RR died in
early development (Amsterdam, Nissen et a1. 2004) and mouse embryonic stem cells that
are deficient for the (P)RR could not generate chimeras after injection into blastocysts
(Burckle and Bader 2006). These results indicate that besides roles in binding (pro)renin,
the (P)RR may be involved in other important cellular functions and the conserved
C-terminus ATP6M8-9 may be a major component in those functions. Therefore, it has
been hypothesized that the (P)RR was evolved fiom an old version of ATP6M8-9 that
was essential for basic cellular firnctions (Burckle and Bader 2006). The extracelluar
domain adapted to a new environment and evolved to take over new function of binding

(pro)renin.

1.2.4 Pivotal role of the cytoplaamic domain of the (P)RR in sigaal transduction

On binding of (pro)renin to (P)RR, a series of signal transduction pathways are
triggered that are independent of the RAS pathway. The binding of renin to (P)RR
induces phosphorylation of serine and tyrosine residues on the (P)RR and activates the
extracellular signal-regulated kinases (ERK) 1/2 and stimulate growth factor TGF- B 1
and plasminogen activator inhibitor—1 (PAIl) (Nguyen, Delarue et al. 2002; Huang,
Wongamorntham et al. 2006; Huang, Noble et al. 2007; Sakoda, Ichihara et al. 2007;

Feldt, Batenburg et al. 2008). Small interfering RNA targeting the (P)RR abolished those

22

stimulations indicating the (P)RR is involved in these signaling pathways. In
cardiomyocytes, prorenin binding activates the p38 mitogen-activated protein kinase
(MAPK) in a concentration-dependent manner and simultaneously phosphorylates
Hsp-27 (Saris, t Hoen et al. 2006). Although short, the cytoplasmic domain is the only
region that is exposed to a downstream receptor molecule, suggesting that it is a very
important mediator of these signal transduction mechanisms.

The (P)RR was first reported to localize on the cell surface (Nguyen, Delarue et al.
2002). However, later research indicated that the (P)RR is mainly localized in the
intracellular perinuclear region, with a minor portion on the cell membrane (Saris, t Hoen
et al. 2006; Schefe, Menk et al. 2006; Feldt, Maschke et al. 2008). The sequence analysis
of the (P)RR revealed two putative intracellular targeting motifs which are in the
cytoplasmic tail (Burckle and Bader 2006), a tyrosine-based motif Y335DSI and a
C-terminal dibasic motif K3461RMD. The Yxx¢ (where x is a random residue and 4) is a
large hydrophobic residue) is typical for protein sorting to endosomes and lysosomes.
The K(x)Kxx or R(x)Rxx are conventional ER retention/retrieval signals. Although
further experimental evidence is needed, the intracellular location may be another cellular
function of the (P)RR that has not yet been elucidated. The signal sequences on the

cytoplasmic tail may also play an important role in the protein localization.

1.2.5 Interactioas of the (P)RR with Promyelocytic Zinc Finger Protein

The transcriptional factor Promyelocytic Zinc Finger Protein (PLZF) was identified as

23

a downstream partner that interacts with the cytoplasmic domain of the (P)RR by yeast
two-hybrid screening and coimmunoprecipitation (Schefe, Menk et al. 2006; Danser,
Batenburg et al. 2008; Schefe, Unger et al. 2008). The PLZF belongs to the family of
Kruppel-type zinc finger proteins (Chowdhury, Deutsch et al. 1987). It is a transcriptional
repressor involved in regulating cell cycle and growth suppression, and has been
suggested in limb development, differentiation of myeloid cells, and sperrnatogenesis
(Melnick and Licht 1999; Yeyati, Shaknovich et al. 1999; Bama, Hawe et al. 2000;
Takahashi and Licht 2002). The human PLZF has a Broad-Complex, Tramtrack, and
Bric-a-brac (BTB), or Poxvirus and Zinc Finger (POZ) domain in the N-terminal 120
amino acids, a central second repressor domain (RD2) of about 250 amino acids, and nine
C2H2 Kruppel-type zinc fingers in the C-terminus. The BTB domains in transcription
factors are usually involved in regulation of gene expression by controlling the chromatin
conformation (Albagli, Dhordain et al. 1995). The RD2 contains a proline—rich region,
which may be responsible for protein-protein interactions (Li, English et al. 1997;
Melnick, Westendorf et al. 2000). The zinc fingers are responsible for DNA binding, or
interactions with RNA or other protein partners. The X-ray crystal structure of the BTB
domain of human PLZF revealed a tightly intertwined dimer with about 25% of the
monomer surface involved in the hydrophobic interface (Ahmad, Engel et al. 1998).

On binding of (pro)renin by (P)RR, the PLZF is activated and translocated to the
nucleus to be recruited to the cis element of the (P)RR promoter. The transcription of the

(P)RR is repressed by the activated PLZF and therefore, a short negative feedback loop

24

was created (Schefe, Menk et al. 2006). These results support a novel signal transduction
pathway in which (pro)renin, (P)RR, and PLZF are involved. This negative feedback
pathway, which downregulates the (P)RR expression, explains why accumulation of
(pro)renin during renin inhibitor exposure does not increase the (P)RR activity to a
significant level. Activation of (P)RR will cause a six-fold increase in recruitment of
PLZF to (P)RR promoter region and therefore prevent further (P)RR activation through
suppression of (P)RR expression (Schefe, Menk et al. 2006). Direct interaction of PLZF
with the cytoplasmic domain of the (P)RR was demonstrated by the yeast two-hybrid
study and coimmunoprecipitation with truncated (P)RR fragments (Schefe, Menk et al.
2006). However, which region(s) of the PLZF is responsible for the interaction remains
unclear.

Together with PLZF, the promoter activity of the p85a subunit of the
phosphatidylinositol-3 kinase (PI3K-p85a) was stimulated by 45% (Schefe, Menk et al.
2006). Nuclear PLZF binds to the PI3K-p85a consensus sequence and positively regulate
the gene. The PI3K-p85a is a protein involved in activation of protein synthesis and
cardiac hypertrophy (Senbonmatsu, Saito et al. 2003). What the exact roles of the

(P)RR-PLZF-PI3K-p85a pathway are in cellular function remain unclear.

25

References

Ahmad, K. F., C. K. Engel and G G Prive (1998). "Crystal structure of the BTB domain
from PLZF." Proc Natl Acad Sci U S A 95(21): 12123-8.

Albagli, O., P. Dhordain, C. Deweindt, G Lecocq and D. Leprince (1995). "The
BTB/POZ domain: a new protein-protein interaction motif common to DNA- and
actin-binding proteins." Cell Growth Differ 6(9): 1193-8.

Amsterdam, A., R. M. Nissen, Z. Sun, E. C. Swindell, S. Farrington and N. Hopkins
(2004). "Identification of 315 genes essential for early zebrafish development."
Proc Natl Agad Sci U S A 101(35): 12792-7.

Anderson, S., A. T. Bankier, B. G Barrell, M. H. de Bruijn, A. R. Coulson, J. Drouin, I. C.
Eperon, D. P. Nierlich, B. A. Roe, F. Sanger, P. H. Schreier, A. J. Smith, R. Staden
and I. G Young (1981). "Sequence and organization of the human mitochondrial
genome." Nature 290(5806): 457-65.

 

Anderson, S. L., J. M. Carton, J. Lou, L. Xing and B. Y. Rubin (1999).
"Interferon-induced guanylate binding protein-1 (GBP-l) mediates an antiviral
effect against vesicular stomatitis virus and encephalomyocarditis virus."
Virology 256(1): 8-14.

Azizi, M. and J. Menard (2004). "Combined blockade of the renin-angiotensin system
with angiotensin-converting enzyme inhibitors and angiotensin II type 1 receptor
antagonists." Circulation 109(21): 2492-9.

Bader, M. (2007). "The second life of the (pro)renin receptor." J Renin Angiotenaia
Aldosterone Syst 8(4): 205-8.

Barna, M., N. Hawe, L. Niswander and P. P. Pandolfi (2000). "Plzf regulates limb and
axial skeletal patterning." Nat Genet 25(2): 166-72.

Batenburg, W. W., M. Krop, I. M. Garrelds, R. de Vries, R. J. de Bruin, C. A. Burckle, D.
N. Muller, M. Bader, G Nguyen and A. H. Danser (2007). "Prorenin is the
endogenous agonist of the (pro)renin receptor. Binding kinetics of renin and
prorenin in rat vascular smooth muscle cells overexpressing the human (pro)renin
receptor." J Hypertens 25(12): 2441-53.

Baxter, J. D., M. N. James, W. N. Chu, K. Duncan, M. A. Haidar, C. T. Carilli and T. L.
Reudelhuber (1989). "The molecular biology of human renin and its gene." Yale J

26

Biol Med 62(5): 493-501.

Benard, G and M. Karbowski (2009). "Mitochondrial fusion and division: Regulation
and role in cell viability." Semin Cell Dev Biol 20(3): 365-74.

Berman, S. B., F. J. Pineda and J. M. Hardwick (2008). "Mitochondrial fission and fusion
dynamics: the long and short of it." Cell Death Differ 15(7): 1147-52.

Bhar, D., M. A. Karren, M. Babst and J. M. Shaw (2006). "Dimeric Dnml-G385D
interacts with Mdvl on mitochondria and can be stimulated to assemble into
fission complexes containing Mdvl and Fisl ." J Biol Chem 281(25): 17312-20.

Bing, J. (1973). "Rapid marked increase in plasma renin in rats treated with inhibitors of
the renin system. Effects of 1-sar-8-ala-angiotensin H and of a synthetic
converting enzyme inhibitor (nonapeptide, SQ 20.881) on normal and
adrenalectomized rats" Acta Pathol Microbiol Scand A 81(3): 376—8.

Binns, D. D., M. K. Helms, B. Barylko, C. T. Davis, D. M. Jameson, J. P. Albanesi and J.
F. Eccleston (2000). "The mechanism of GTP hydrolysis by dynamin II: a
transient kinetic study." Biochemisfl 39(24): 7188-96.

Bleazard, W., J. M. McCaffery, E. J. King, S. Bale, A. Mozdy, Q. Tieu, J. Nunnari and J.
M. Shaw (1999). "The dynamin-related GTPase Dnml regulates mitochondrial
fission in yeast." Nat Cell Biol 1(5): 298-304.

Borghi, C., S. Boschi, E. Ambrosioni, G Melandri, A. Branzi and B. Magnani (1993).
"Evidence of a partial escape of renin-angiotensin-aldosterone blockade in
patients with acute myocardial infarction treated with ACE inhibitors." LCM
Pharrnacol 33(1): 40-5.

Bowman, E. J. and B. J. Bowman (2000). "Cellular role of the V-ATPase in Neurospora
crassa: analysis of mutants resistant to concanamycin or lacking the catalytic

subunit A." J Exp Biol 203(Pt 1): 97-106.

Burckle, C. and M. Bader (2006). "Prorenin and its ancient receptor." Hypertension 48(4):
549-51.

Cerveny, K. L., J. M. McCaffery and R. E. Jensen (2001). "Division of mitochondria
requires a novel DMNl-interacting protein, Net2p." Mol Biol Cell 12(2): 309-21.

Cerveny, K. L., Y. Tamura, Z. Zhang, R. E. Jensen and H. Sesaki (2007). "Regulation of

27

mitochondrial fusion and division." Trends Cell Biol 17(11): 563-9.

Chang, C. R. and C. Blackstone (2007). "Cyclic AMP-dependent protein kinase
phosphorylation of Drpl regulates its GTPase activity and mitochondrial
morphology." J Biol Chem 282(30): 21583-7.

Chen, H., A. Chomyn and D. C. Chan (2005). "Disruption of fusion results in
mitochondrial heterogeneity and dysfirnction." J Biol Chem 280(28): 26185-92.

Chen, H., S. A. Detrner, A. J. Ewald, E. E. Griffin, S. E. Fraser and D. C. Chan (2003).
"Mitofusins Mfrrl and Mfi12 coordinately regulate mitochondrial fusion and are
essential for embryonic development." J Cell Biol 160(2): 189-200.

Cheng, W. C., K. M. Leach and J. M. Hardwick (2008). "Mitochondrial death pathways
in yeast and mammalian cells." Biochim Biophys Acta 1783(7): 1272-9.

Chinnery, P. F. and E. A. Schon (2003). "Mitochondria." J Neurol Neurosurg Psychim
74(9): 1188-99.

Cho, D. H., T. Nakamura, J. Fang, P. Cieplak, A. Godzik, Z. Gu and S. A. Lipton (2009).
"S-nitrosylation of Drpl mediates beta-amyloid—related mitochondrial fission and
neuronal injury." Science 324(5923): 102-5.

 

Chowdhury, K., U. Deutsch and P. Gruss (1987). "A multigene family encoding several
"finger" structures is present and differentially active in mammalian genomes."
M 48(5): 771-8.

Cribbs, J. T. and S. Strack (2007). "Reversible phosphorylation of Drpl by cyclic
AMP-dependent protein kinase and calcineurin regulates mitochondrial fission
and cell death." EMBO Rap 8(10): 939-44.

Danser, A. H., W. W. Batenburg, J. H. van Esch and M. Krop (2008). "Prorenin anno
2008." J Mol Med 86(6): 655-8.

Daviet, L., J. Y. Lehtonen, K. Tamura, D. P. Griese, M. Horiuchi and V. J. Dzau (1999).
"Cloning and characterization of ATRAP, a novel protein that interacts with the
angiotensin 11 type 1 receptor." J Biol Chem 274(24): 17058-62.

de Gasparo, M., K. J. Catt, T. Inagami, J. W. Wright and T. Unger (2000). "International

union of pharmacology. XXIII. The angiotensin 11 receptors." Pharrnacol Rev
52(3): 415-72.

28

Debray, F. G, M. Lambert and G A. Mitchell (2008). "Disorders of mitochondrial
function." Curr Onin Pedia_tr 20(4): 471-82.

Derkx, F. H., M. P. Schalekamp and M. A. Schalekamp (1987). "Two-step prorenin-renin
conversion. Isolation of an intermediary form of activated prorenin." J Biol Chem
262(6): 2472-7.

Derkx, F. H., J. M. von Gool, G J. Wenting, R. P. Verhoeven, A. J. Man in 't Veld and M.
A. Schalekamp (1976). "Inactive renin in human plasma." Lancet 1(7984): 496-9.

 

Di Bacco, A. and G Gill (2006). "SUMO-specific proteases and the cell cycle. An
essential role for SENPS in cell proliferation." Cell Cycle 5(20): 2310-3.

Dohm, J. A., S. J. Lee, J. M. Hardwick, R. B. Hill and A. G Gittis (2004). "Cytosolic
domain of the human mitochondrial fission protein Fisl adopts a TPR fold."

Proteins-Structure Function and Genetics 54(1): 153-156.

Dolezal, P., V. Likic, J. Tachezy and T. Lithgow (2006). "Evolution of the molecular
machines for protein import into mitochondria." Science 313(5785): 314-8.

Dyall, S. D., M. T. Brown and P. J. Johnson (2004). "Ancient invasions: from
endosymbionts to organelles." Science 304(5668): 253-7.

 

Embley, T. M. and W. Martin (2006). "Eukaryotic evolution, changes and challenges."
Nature 440(7084): 623-30.

Fekkes, P., K. A. Shepard and M. P. Yaffe (2000). "Gag3p, an outer membrane protein
required for fission of mitochondrial tubules." J Cell Biol 151(2): 333-40.

Feldt, S., W. W. Batenburg, I. Mazak, U. Maschke, M. Wellner, H. Kvakan, R. Dechend,
A. Fiebeler, C. Burckle, A. Contrepas, A. H. Jan Danser, M. Bader, G Nguyen, F.
C. Luft and D. N. Muller (2008). "Prorenin and renin-induced extracellular
signal-regulated kinase 1/2 activation in monocytes is not blocked by aliskiren or
the handle-region peptide." Hypertension 51(3): 682-8.

Feldt, S., U. Maschke, R. Dechend, F. C. Luft and D. N. Muller (2008). "The putative
(pro)renin receptor blocker HRP fails to prevent (pro)renin signaling." J Am Soc
Naphrol 19(4): 743-8.

Ferguson, K. M., M. A. Lemmon, J. Schlessinger and P. B. Sigler (1994). "Crystal
structure at 2.2 A resolution of the pleckstrin homology domain from human

29

dynamin." @ 79(2): 199-209.

Forgac, M. (1999). "Structure and properties of the vacuolar (H+)-ATPases" Journal of
Biological Chemisgy 274(19): 12951-12954.

Frank, S., B. Gaume, E. S. Bergrnann-Leitrrer, W. W. Leitner, E. G Robert, F. Catez, C. L.
Smith and R. J. Youle (2001). "The role of dynamin-related protein 1, a mediator
of mitochondrial fission, in apoptosis." Dev Cell 1(4): 515-25.

 

Fujii, T., N. Onohara, Y. Maruyarna, S. Tanabe, H. Kobayashi, M. Fukutomi, Y.
Nagarnatsu, N. Nishihara, R. Inoue, H. Sumimoto, F. Shibasaki, T. Nagao, M.
Nishida and H. Kurose (2005). "Galpha12/13-mediated production of reactive
oxygen species is critical for angiotensin receptor-induced NFAT activation in
cardiac fibroblasts." J Biol Chem 280(24): 23041-7.

Fukarnizu, A., K. Nishi, T. Cho, M. Saitoh, K. Nakayama, H. Ohkubo, S. Nakanishi and
K. Murakami (1988). "Structure of the rat renin gene." J Mol Biol 201(2): 443-50.

Gao, H., D. Kadirjan-Kalbach, J. E. Froehlich and K. W. Osteryoung (2003). "ARCS, a
cytosolic dynamin-like protein fi'om plants, is part of the chloroplast division
machinery." Proc Natl Acad Sci U S A 100(7): 4328-33.

Gao, H., T. L. Sage and K. W. Osteryoung (2006). "FZL, an FZO-like protein in plants, is
a determinant of thylakoid and chloroplast morphology." Proc Natl Acad Sci U S
_A_ 103(17): 6759-64.

Glynn, J. M., J. E. Froehlich and K. W. Osteryoung (2008). "Arabidopsis ARC6
coordinates the division machineries of the inner and outer chloroplast
membranes through interaction with PDV2 in the interrnembrane space." flag
_@ 20(9): 2460-70.

Glynn, J. M., Y. Yang, S. Vitha, A. J. Schmitz, M. Hemmes, S. Y. Miyagishima and K. W.
Osteryoung (2009). "PARC6, a novel chloroplast division factor, influences FtsZ
assembly and is required for recruitment of PDVl during chloroplast division in
Arabidopsis." Plant J.

Gray, M. W., B. F. Lang and G Burger (2004). "Mitochondria of protists." Annu Rev
Genet 38: 477-524.

Griendling, K. K., C. A. Minieri, J. D. Ollerenshaw and R. W. Alexander (1994).
"Angiotensin II stimulates NADH and NADPH oxidase activity in cultured

3O

vascular smooth muscle cells." Circ Res 74(6): 1141 -8.

 

Griffin, E. E., S. A. Detrner and D. C. Chan (2006). "Molecular mechanism of
mitochondrial membrane fusion." Biochim Biophys Acta 1763(5-6): 482-9.

Grigliatti, T. A., L. Hall, R. Rosenbluth and D. T. Suzuki (1973). "Temperature-sensitive
mutations in Drosophila melanogaster. XIV. A selection of immobile adults." Moi
Gen Genet 120(2): 107-14.

Gu, X. and D. P. Verma (1996). "Phragrnoplastin, a dynamin-like protein associated with
cell plate formation in plants." Embo J 15(4): 695-704.

Hales, K. G and M. T. Fuller (1997). "Developmentally regulated mitochondrial fusion
mediated by a conserved, novel, predicted GTPase." Call 90(1): 121-9.

Harder, Z., R. Zunino and H. McBride (2004). "Sumol conjugates mitochondrial
substrates and participates in mitochondrial fission." Curr Biol 14(4): 340-5.

Heath-Engel, H. M. and G C. Shore (2006). "Mitochondrial membrane dynamics, cristae
remodelling and apoptosis." Biochim Biophys Acta 1763(5-6): 549-60.

Heinrikson, R. L., J. Hui, H. Zurcher-Neely and R. A. Poonnan (1989). "A structural
model to explain the partial catalytic activity of human prorenin." Am J Hypertens
2(5 Pt 1): 367-80.

Hermann, G J ., J. W. Thatcher, J. P. Mills, K. G Hales, M. T. Fuller, J. Nunnari and J. M.
Shaw (1998). "Mitochondrial fusion in yeast requires the transmembrane GTPase
onlp." J Cell Biol 143(2): 359-73.

Holsworth, D. D., N. A. Powell, D. M. Downing, C. Cai, W. L. Cody, J. M. Ryan, R.
Ostroski, M. Jalaie, J. W. Bryant and J. J. Edrnunds (2005). "Discovery of novel
non-peptidic ketopiperazine-based renin inhibitors." Bioorganic & Medicinfl
Chemisgy 13(7): 2657-2664.

Hoppins, S., L. Lackner and J. Nunnari (2007). "The machines that divide and fuse
mitochondria." Annu Rev Biochem 76: 751-80.

Hoppins, S. and J. Nunnari (2009). "The molecular mechanism of mitochondrial fusion."
Biochim Biophfi Acta 1793(1): 20-6.

Huang, Y., N. A. Noble, J. Zhang, C. Xu and W. A. Border (2007). "Renin-stimulated

31

TGF-betal expression is regulated by a mitogen-activated protein kinase in
mesangial cells." Kidney Int 72(1): 45-52.

Huang, Y., S. Wongamorntham, J. Kasting, D. McQuillan, R. T. Owens, L. Yu, N. A.
Noble and W. Border (2006). "Renin increases mesangial cell transforming
growth factor-beta] and matrix proteins through receptor-mediated, angiotensin
II-independent mechanisms." Kidney Int 69(1): 105-13.

Inagami, T. (1991). "Purification of renin and prorenin." Hypertension 18(3): 241 -51.

Ingerman, E., E. M. Perkins, M. Marino, J. A. Mears, J. M. McCaffery, J. E. Hinshaw and
J. Nunnari (2005). "Dnml forms spirals that are structurally tailored to fit
mitochondria." J Cell Biol 170(7): 1021-7.

Ishihara, N., Y. Eura and K. Mihara (2004). "Mitofusin 1 and 2 play distinct roles in
mitochondrial fusion reactions via GTPase activity." J Cell Sci 117(Pt 26):
6535-46.

James, D. I., P. A. Parone, Y. Mattenberger and J. C. Martinou (2003). "hFisl, a novel
component of the mammalian mitochondrial fission machinery." J Biol Chem
278(38): 36373-9.

Jeong, S. Y. and D. W. Sec] (2008). "The role of mitochondria in apoptosis." BMB Rep
41(1): 11-22.

Kang, J ., E. M. Richards, P. Posner and C. Sumners (1995). "Modulation of the delayed
rectifier K+ current in neurons by an angiotensin 11 type 2 receptor fragment." Air;
J Ph3§i01268(1 Pt 1): C278-82.

Kang, S. G, J. B. Jin, H. L. Piao, K. T. Pih, H. J. Jang, J. H. Lim and I. Hwang (1998).
"Molecular cloning of an Arabidopsis cDNA encoding a dynamin-like protein that
is localized to plastids." Plant Mol Biol 38(3): 437-47.

Karbowski, M., A. Neutzner and R. J. Youle (2007). "The mitochondrial E3 ubiquitin
ligase MARCHS is required for Drpl dependent mitochondrial division." J Cell
Biol 178(1): 71-84.

Karren, M. A., E. M. Coonrod, T. K. Anderson and J. M. Shaw (2005). "The role of

F islp-Mdvlp interactions in mitochondrial fission complex assembly." J Cell Biol
171(2): 291-301.

32

Klein, D. E., A. Lee, D. W. Frank, M. S. Marks and M. A. Lemmon (1998). "The
pleckstrin homology domains of dynamin isoforms require oligomerization for
high afiinity phosphoinositide binding." J Biol Chem 273(42): 27725-33.

Koshiba, T., S. A. Detrner, J. T. Kaiser, H. Chen, J. M. McCaffery and D. C. Chan (2004).
"Structural basis of mitochondrial tethering by mitofusin complexes." Science
305(5685): 858-62.

Labrousse, A. M., M. D. Zappaterra, D. A. Rube and A. M. van der Bliek (1999). "C.
elegans dynamin-related protein DRP-l controls severing of the mitochondrial
outer membrane." Mol Cell 4(5): 815-26.

Lackner, L. L., J. S. Homer and J. Nunnari (2009). "Mechanistic analysis of a dynamin
effector." Science 325(5942): 874-7.

 

Leckie, B. J ., G Birnie and R. Carachi (1994). "Renin in Wilms' tumor: prorenin as an
indicator." J Clin Endocrinol Metab 79(6): 1742-6.

Leckie, B. J. and N. K. McGhee (1980). "Reversible activation-inactivation of renin in
human plasma." Nature 288(5792): 702-5.

 

Lee, Y. J ., S. Y. Jeong, M. Karbowski, C. L. Smith and R. J. Youle (2004). "Roles of the
mammalian mitochondrial fission and firsion mediators Fisl, Drpl, and Opal in
apoptosis." Mol Biol Cell 15(11): 5001-11.

Lemmon, M. A. and K. M. Ferguson (2000). "Signal-dependent membrane targeting by
pleckstrin homology (PH) domains." Biochem J 350 Pt 1: 1-18.

Li, J. Y., M. A. English, H. J. Ball, P. L. Yeyati, S. Waxman and J. D. Licht (1997).
"Sequence-specific DNA binding and transcriptional regulation by the
promyelocytic leukemia zinc finger protein." J Biol Chem 272(36): 22447-55.

Low, H. H. and J. Lowe (2006). "A bacterial dynamin-like protein." Nature 444(7120):
766-9.

 

Ludwig, J ., S. Kerscher, U. Brandt, K. Pfeifl‘er, F. Getlawi, D. K. Apps and H. Schagger
(1998). "Identification and characterization of a novel 9.2-kDa membrane
sector-associated protein of vacuolar proton-ATPase from chromaffin granules." 1
Biol Chem 273(18): 10939-47.

McDonald, P. H., C. W. Chow, W. E. Miller, S. A. Laporte, M. E. Field, F. T. Lin, R. J.

33

Davis and R. J. Lefl<owitz (2000). "Beta-arrestin 2: a receptor-regulated MAPK
scaffold for the activation of JNK3." Science 290(5496): 1574-7.

 

McEwen, D. P., K. R. Gee, H. C. Kang and R. R. Neubig (2001). "Fluorescent
BODIPY—GTP analogs: real-time measurement of nucleotide binding to G
proteins." Anal Biochem 291(1): 109-17.

Mehta, P. K. and K. K. Griendling (2007). "Angiotensin H cell signaling: physiological
and pathological effects in the cardiovascular system." Am J Physiol Cell thsiol
292(1): C82-97.

Mehrick, A. and J. D. Licht (1999). "Deconstructing a disease: RARalpha, its fusion
partners, and their roles in the pathogenesis of acute promyelocytic leukemia."
Blood 93(10): 3167-215.

Melnick, A. M., J. J. Westendorf, A. Polinger, G W. Carlile, S. Arai, H. J. Ball, B.
Lutterbach, S. W. Hiebert and J. D. Licht (2000). "The ETO protein disrupted in
t(8;21)-associated acute myeloid leukemia is a corepressor for the promyelocytic
leukemia zinc finger protein." Mol Cell Biol 20(6): 2075-86.

Morris, B. J. (1992). "Molecular biology of renin. I: Gene and protein structure, synthesis
and processing." J Hypertens 10(3): 209-14.

Morris, B. J. (1992). "Molecular biology of renin. I: Gene and protein structure, synthesis
and processing." J oumil of Hypertension 10(3): 209-14.

Mozdy, A. D., J. M. McCaffery and J. M. Shaw (2000). "Dnmlp GTPase-mediated
mitochondrial fission is a multi-step process requiring the novel integral
membrane component Fislp." J Cell Biol 151(2): 367-80.

Nabi, A. H., A. Kageshima, M. N. Uddin, T. Nakagawa, E. Y. Park and F. Suzuki (2006).
"Binding properties of rat prorenin and renin to the recombinant rat renin/prorenin

receptor prepared by a baculovirus expression system." Int J Mol Med 18(3):
483-8.

Nakamura, N., Y. Kimura, M. Tokuda, S. Honda and S. Hirose (2006). "MARCH-V is a
novel mitofusin 2- and Drpl-binding protein able to change mitochondrial

morphology." EMBO Rep 7(10): 1019-22.

Newmeyer, D. D. and S. F erguson-Miller (2003). "Mitochondria: releasing power for life
and unleashing the machineries of death." §e_ll 112(4): 481-90.

34

Nguyen, G, F. Delarue, C. Burckle, L. Bouzhir, T. Giller and J. D. Sraer (2002). "Pivotal
role of the renin/prorenin receptor in angiotensin II production and cellular
responses to renin." J Clin Invest 109(11): 1417-27.

Niemann, H. H., M. L. Knetsch, A. Scherer, D. J. Manstein and F. J. Kull (2001). "Crystal
structure of a dynamin GTPase domain in both nucleotide-free and GDP-bound
forms." Embo J 20(21): 5813-21.

Nishi, T. and M. Forgac (2002). "The vacuolar (H+)-ATPases--nature's most versatile
proton pumps." Mt Rev Mol Cell Biol 3(2): 94-103.

Obar, R. A., C. A. Collins, J. A. Hammarback, H. S. Shpetner and R. B. Vallee (1990).
"Molecular cloning of the microtubule-associated mechanochemical enzyme

dynamin reveals homology with a new family of GTP-binding proteins." Nature
347(6290): 256-61.

Okamoto, K. and J. M. Shaw (2005). "Mitochondrial morphology and dynamics in yeast
and multicellular eukaryotes." Annu Rev Genet 39: 503-36.

Olichon, A., L. J. Emorine, E. Descoins, L. Pelloquin, L. Brichese, N. Gas, E. Guillou, C.
Delettre, A. Valette, C. P. Hamel, B. Ducommun, G Lenaers and P. Belenguer
(2002). "The human dynamin-related protein OPAl is anchored to the
mitochondrial inner membrane facing the inter-membrane space." F EBS Lett
523(1-3): 171-6.

Osteryoung, K. W. (2000). "Organelle fission. Crossing the evolutionary divide." Plant
Physiol 123(4): 1213-6.

Osteryoung, K. W. (2001). "Organelle fission in eukaryotes." Curr Onin Microbiol 4(6):
639-46.

Osteryoung, K. W. and J.‘ Nunnari (2003). "The division of endosymbiotic organelles."
Science 302(5651): 1698-704.

Pai, E. F., U. Krengel, G A. Petsko, R. S. Goody, W. Kabsch and A. Wittinghofer (1990).
"Refined Crystal-Structure of the Triphosphate Conformation of H-Ras P21 at
1.35 a Resolution - Implications for the Mechanism of th Hydrolysis." m
Journal 9(8): 2351-2359.

Parone, P. A. and J. C. Martinou (2006). "Mitochondrial fission and apoptosis: an
ongoing trial." Biochim Biophys Acta 1763(5-6): 522-30.

35

Perfettini, J. L., T. Roumier and G Kroemer (2005). "Mitochondrial fusion and fission in
the control of apoptosis." Trends Cell Biol 15(4): 179-83.

Pitarresi, T. M., S. Rubattu, R. Heinrikson and J. E. Sealey (1992). "Reversible
cryoactivation of recombinant human prorenin." J Biol Chem 267(17): 11753-9.

Praefcke, G J. and H. T. McMahon (2004). "The dynamin superfamily: universal
membrane tubulation and fission molecules?" Mt Rev Mol Cell Biol 5(2):
133-47.

Prakash, B., G J. Praefcke, L. Renault, A. Wittinghofer and C. Hemnann (2000).
"Structure of human guanylate-binding protein 1 representing a unique class of
GTP-binding proteins." Nature 403(6769): 567-71.

 

Prakash, B., L. Renault, G J. Praefcke, C. Hernnann and A. Wittinghofer (2000).
"Triphosphate structure of guanylate-binding protein 1 and implications for
nucleotide binding and GTPase mechanism." Embo J 19(17): 4555-64.

Rahuel, J ., J. P. Priestle and M. G Grutter (1991). "The crystal structures of recombinant
glycosylated human renin alone and in complex with a transition state analog
inhibitor." J Struct Biol 107(3): 227-36.

Rahuel, J., V. Rasetti, J. Maibaum, H. Rueger, R. Goschke, N. C. Cohen, S. Stutz, F.
Cumin, W. Fuhrer, J. M. Wood and M. G Grutter (2000). "Structure-based drug
design: the discovery of novel nonpeptide orally active inhibitors of human
renin." Chem Biol 7(7): 493-504.

Ramachandran, R., M. Surka, J. S. Chappie, D. M. Fowler, T. R. Foss, B. D. Song and S.
L. Schmid (2007). "The dynamin middle domain is critical for tetramerization and
higher-order self-assembly." Embo J 26(2): 559-66.

Rarnser, J., F. E. Abidi, C. A. Burckle, C. Lenski, H. Toriello, G Wen, H. A. Lubs, S.
Engert, R. E. Stevenson, A. Meindl, C. E. Schwartz and G Nguyen (2005). "A
unique exonic splice enhancer mutation in a family with X-linked mental
retardation and epilepsy points to a novel role of the renin receptor." Hum Mol

Genet 14(8): 1019-27.
Reubold, T. F., S. Eschenburg, A. Becker, M. Leonard, S. L. Schmid, R. B. Vallee, F. J.

Kull and D. J. Manstein (2005). "Crystal structure of the GTPase domain of rat
dynamin 1." Proc Natl Acad Sci U S A 102(37): 13093-8.

36

Reudelhuber, T. L., V. Brechler, I. Jutras, C. Mercure and D. Methot (1998). "Proteolytic
and non-proteolytic activation of prorenin." Adv Exp Med Biol 436: 229-38.

Roig, E., F. Perez-Villa, M. Morales, W. Jimenez, J. Orus, M. Heras and G Sanz (2000).
"Clinical implications of increased plasma angiotensin 11 despite ACE inhibitor
therapy in patients with congestive heart failure." Eur Heart J 21(1): 53-7.

Rojo, M., F. Legros, D. Chateau and A. Lombes (2002). "Membrane topology and
mitochondrial targeting of mitofusins, ubiquitous mammalian homologs of the
transmembrane GTPase on." J Cell Sci 115(Pt 8): 1663-74.

Rothman, J. H., C. K. Raymond, T. Gilbert, P. J. O'Hara and T. H. Stevens (1990). "A
putative GTP binding protein homologous to interferon-inducible Mx proteins
performs an essential function in yeast protein sorting." @ 61(6): 1063-74.

Rotig, A., S. Lebon, E. Zinovieva, J. Mollet, E. Sarzi, J. P. Bonnefont and A. Munnich
(2004). "Molecular diagnostics of mitochondrial disorders." Biochim Bionhys
Acta 1659(2-3): 129-35.

Sabri, A., G Govindarajan, T. M. Griffin, K. L. Byron, A. M. Samarel and P. A. Lucchesi
(1998). "Calcium- and protein kinase C-dependent activation of the tyrosine

kinase PYK2 by angiotensin II in vascular smooth muscle." Circ Res 83(8):
841-51.

Sakoda, M., A. Ichihara, Y. Kaneshiro, T. Takemitsu, Y. Nakazato, A. H. Nabi, T.
Nakagawa, F. Suzuki, T. Inagami and H. Itoh (2007). "(Pro)renin
receptor-mediated activation of mitogen-activated protein kinases in human
vascular smooth muscle cells." Hypertens Res 30(11): 1139-46.

Santel, A. (2006). "Get the balance right: mitofusins roles in health and disease." Biochim
Bionhvs Am 1763(5-6): 490-9.

Santel, A. and S. Frank (2008). "Shaping mitochondria: The complex posttranslational
regulation of the mitochondrial fission protein DRPl." IUBMB Life 60(7):
448-55.

Santel, A., S. Frank, B. Gaume, M. Herrler, R. J. Youle and M. T. Fuller (2003).
"Mitofusin-1 protein is a generally expressed mediator of mitochondrial fusion in

mammalian cells." J Cell Sci 116(Pt 13): 2763-74.

Santel, A. and M. T. Fuller (2001). "Control of mitochondrial morphology by a human

37

mitofirsin." J Cell Sci 114(Pt 5): 867-74.

Santel, A. and M. T. Fuller (2001). "Control of mitochondrial morphology by a human
mitofusin." Journal of Cell Science 114(5): 867-874.

Saris, J. J., P. A. t Hoen, I. M. Garrelds, D. H. Dekkers, J. T. den Dunnen, J. M. Lamers
and A. H. Jan Danser (2006). "Prorenin induces intracellular signaling in
cardiomyocytes independently of angiotensin II." Hypertension 48(4): 564-71.

Saris, J. J ., M. M. van den Eijnden, J. M. Lamers, P. R. Saxena, M. A. Schalekamp and A.
H. Danser (2002). "Prorenin-induced myocyte proliferation: no role for

intracellular angiotensin II." Hypertension 39(2 Pt 2): 573-7.

Sato, A., K. Nakada and J. Hayashi (2006). "Mitochondrial dynamics and aging:
Mitochondrial interaction preventing individuals fiom expression of respiratory
deficiency caused by mutant mtDNA." Biochim Biophfi Acta 1763(5-6): 473-81.

Schefe, J. H., M. Menk, J. Reinemund, K. Efi‘ertz, R. M. Hobbs, P. P. Pandolfi, P. Ruiz, T.
Unger and H. Funke-Kaiser (2006). "A novel signal transduction cascade
involving direct physical interaction of the renin/prorenin receptor with the
transcription factor promyelocytic zinc finger protein." Circ Res 99(12): 1355-66.

 

Schefe, J. H., T. Unger and H. Funke-Kaiser (2008). "PLZF and the (pro)renin receptor."
J Mol Med 86(6): 623-7.

Scheffzek, K., M. R. Ahmadian, W. Kabsch, L. Wiesmuller, A. Lautwein, F. Schmitz and
A. Wittinghofer (1997). "The Ras-RasGAP complex: structural basis for GTPase
activation and its loss in oncogenic Ras mutants." Science 277(5324): 333-8.

 

Schweins, T., M. Geyer, K. Scheffzek, A. Warshel, H. R. Kalbitzer and A. Wittinghofer
(1995). "Substrate-assisted catalysis as a mechanism for GTP hydrolysis of p2lras
and other GTP-binding proteins." N_at Struct Biol 2(1): 36-44.

Sealey, J. E., N. Glorioso, J. Itskovitz and J. H. Laragh (1986). "Prorenin as a
reproductive hormone. New form of the renin system." Am J Med 81(6): 1041-6.

Sealey, J. E. and J. H. Laragh (1975). ""Prorenin" in human plasma?" Circ Res 36(6
Suppl 1): 10-6.

Senbonmatsu, T., T. Saito, E. J. Landon, O. Watanabe, E. Price, Jr., R. L. Roberts, H.
Irnboden, T. G Fitzgerald, F. A. Gaffney and T. Inagami (2003). "A novel

38

angiotensin II type 2 receptor signaling pathway: possible role in cardiac
hypertrophy. " Embo J 22(24): 6471-82.

Sever, S., H. Damke and S. L. Schmid (2000). "Dynamin:GTP controls the formation of
constricted coated pits, the rate limiting step in clathrin-mediated endocytosis." J
Cell Biol 150(5): 1137-48.

Sever, S., A. B. Muhlberg andS. L. Schmid (1999). "Impairment of dynamin's GAP
domain stimulates receptor-mediated endocytosis." Nature 398(6727): 481-6.

 

Shirai, H., K. Takahashi, T. Katada and T. Inagami (1995). "Mapping of G protein
coupling sites of the angiotensin H type 1 receptor." Hypertension 25(4 Pt 2):
726-30.

Shiratori, Y., S. Nakagawa, H. Hori, K. Murakami and H. Umeyama (1990). "Protein
modeling of human prorenin using the molecular dynamics method." J Mol Graph
8(3): 163-7, 150.

Sielecki, A. R., K. Hayakawa, M. Fujinaga, M. E. Murphy, M. Fraser, A. K. Muir, C. T.
Carilli, J. A. Lewicki, J. D. Baxter and M. N. James (1989). "Structure of
recombinant human renin, a target for cardiovascular-active drugs, at 2.5 A
resolution." Science 243(4896): 1346-51.

 

Siragy, H. M. and R. M. Carey (1996). "The subtype-2 (AT2) angiotensin receptor
regulates renal cyclic guanosine 3', 5'-monophosphate and ATI receptor-mediated
prostaglandin E2 production in conscious rats." J Clin Invest 97(8): 1978-82.

Siragy, H. M., A. A. Jaffa, H. S. Margolius and R. M. Carey (1996). "Renin-angiotensin
system modulates renal bradykinin production." Am J Physiol 271(4 Pt 2):
R1090-5.

Siragy, H. M. and J. Linden (1996). "Sodium intake markedly alters renal interstitial fluid
adenosine." Hypertension 27(3 Pt 1): 404-7.

Smirnova, E., L. Griparic, D. L. Shurland and A. M. van der Bliek (2001).
"Dynamin-related protein Drpl is required for mitochondrial division in
mammalian cells." Mol Biol Cell 12(8): 2245-56.

Smirnova, E., D. L. Shurland, S. N. Ryazantsev and A. M. van der Bliek (1998). "A

human dynamin-related protein controls the distribution of mitochondria."
Molecular Biology of the Cell 9: l97a-197a.

39

Staeheli, P., O. Haller, W. Boll, J. Lindenrnann and C. Weissmann (1986). "Mx protein:
constitutive expression in 3T3 cells transformed with cloned Mx cDNA confers
selective resistance to influenza virus." CLII 44(1): 147-58.

Staeheli, P., M. A. Horisberger and O. Haller (1984). "Mx-dependent resistance to
influenza viruses is induced by mouse interferons alpha and beta but not gamma."
Virology 132(2): 456-61.

Stevens, T. H. and M. Forgac (1997). "Structure, function and regulation of the vacuolar
(H+)-ATPase." Annu Rev Cell Dev Biol 13: 779-808.

Stojanovski, D., O. S. Koutsopoulos, K. Okamoto and M. T. Ryan (2004). "Levels of
human F isl at the mitochondrial outer membrane regulate mitochondrial
morphology." J Cell Sci 117(Pt 7): 1201-10.

Suen, D. F., K. L. Norris and R. J. Youle (2008). "Mitochondrial dynamics and
apoptosis." Genes Dev 22(12): 1577-90.

Sugioka, R., S. Shimizu and Y. Tsujimoto (2004). "on1, a protein involved in
mitochondrial firsion, inhibits apoptosis." J Biol Chem 279(50): 52726-34.

Suzuki, F., Y. Hatano, T. Nakagawa, K. Terazawa, A. Gotoh, U. M. Nasir, Y. Ishida and Y.
Nakamura (1999). "Non-proteolytic activation of human prorenin by anti-prorenin
prosegment (pf#1: 1P-15P) antiserum." Biosci Biotechnol Biochem 63(3): 550-4.

Suzuki, F., M. Hayakawa, T. Nakagawa, U. M. Nasir, A. Ebihara, A. Iwasawa, Y. Ishida,
Y. Nakamura and K. Murakami (2003). "Human prorenin has "gate and handle"
regions for its non-proteolytic activation." J Biol Chem 278(25): 22217-22.

Suzuki, M., S. Y. Jeong, M. Karbowski, R. J. Youle and N. Tjandra (2003). "The solution
structure of human mitochondria fission protein F isl reveals a novel TPR-like
helix bundle." Journal of Molecular Biology 334(3): 445-458.

Suzuki, M., A. Neutzner, N. Tjandra and R. J. Youle (2005). "Novel structure of the N
terminus in yeast Fisl correlates with a specialized function in mitochondrial

fission." J Biol Chem 280(22): 21444-52.

Sweitzer, S. M. and J. E. Hinshaw (1998). "Dynamin undergoes a GTP-dependent
conformational change causing vesiculation." (fl 93(6): 1021-9.

Taguchi, N., N. Ishihara, A. Jofuku, T. Oka and K. Mihara (2007). "Mitotic

4o

phosphorylation of dynamin-related GTPase Drpl participates in mitochondrial
fission." J Biol Chem 282(15): 11521-9.

Takahashi, S. and J. D. Licht (2002). "The human promyelocytic leukemia zinc finger
gene is regulated by the Evi-l oncoprotein and a novel guanine-rich site binding
protein." Leukemia 16(9): 1755-62.

Tamura, K., Y. Tanaka, Y. Tsurumi, K. Azuma, A. Shigenaga, H. Wakui, S. Masuda and
M. Matsuda (2007). "The role of angiotensin ATl receptor-associated protein in

renin-angiotensin system regulation and function." Curr Hypertens Rep 9(2):
121-7.

Tice, C. M., Z. Xu, J. Yuan, R. D. Simpson, S. T. Cacatian, P. T. Flaherty, W. Zhao, J. Guo,
A. Ishchenko, S. B. Singh, Z. Wu, B. B. Scott, Y. Bukhtiyarov, J. Berbaum, J.
Mason, R. Panemangalore, M. G Cappiello, D. Muller, R. K. Harrison, G M.
McGeehan, L. W. Dillard, J. J. Baldwin and D. A. Claremon (2009). "Design and
optimization of renin inhibitors: Orally bioavailable alkyl amines." Bioorgided
Chem Lett 19(13): 3541-5.

Tieu, Q. and J. Nunnari (2000). "Mdvlp is a WD repeat protein that interacts with the
dynamin-related GTPase, Dnmlp, to trigger mitochondrial division." J Cell Biol
151(2): 353-66.

Tieu, Q., V. Okreglak, K. Naylor and J. Nunnari (2002). "The WD repeat protein, Mdvlp,
functions as a molecular adaptor by interacting with Dnmlp and Fislp during
mitochondrial fission." J Cell Biol 158(3): 445-52.

Timmermans, P. B. M. W. M., P. C. Wong, A. T. Chiu, W. F. Herblin, P. Benfield, D. J.
Carini, R. J. Lee, R. R. Wexler, J. A. M. Saye and R. D. Smith (1993).
"Angiotensin-Ii Receptors and Angiotensin-Ii Receptor Antagonists." Pharmacol
3g 45(2): 205-251.

Tohgo, A., K. L. Pierce, E. W. Choy, R. J. Leflrowitz and L. M. Luttrell (2002).
"beta-Arrestin scaffolding of the ERK cascade enhances cytosolic ERK activity
but inhibits ERK-mediated transcription following angiotensin ATla receptor
stimulation." J Biol Chem 277(11): 9429-36.

Ushio-Fukai, M., K. K. Griendling, M. Akers, P. R. Lyons and R. W. Alexander (1998).
"Temporal dispersion of activation of phospholipase C-beta 1 and -gamma
isoforms by angiotensin II in vascular smooth muscle cells - Role of alpha(q/l l),
alpha(12), and beta gamma G protein subunits." Journal of Biological Chemistry

41

273(31): 19772-19777.

van den Eijnden, M. M., J. J. Saris, R. J. de Bruin, E. de Wit, W. Sluiter, T. L.
Reudelhuber, M. A. Schalekamp, F. H. Derkx and A. H. Danser (2001). "Prorenin

accumulation and activation in human endothelial cells: importance of mannose
6-phosphate receptors." Arterioscler Thromb Vasc Biol 21(6): 911-6.

van Kesteren, C. A., A. H. Danser, F. H. Derkx, D. H. Dekkers, J. M. Lamers, P. R.
Saxena and M. A. Schalekamp (1997). "Mannose 6-phosphate receptor-mediated

internalization and activation of prorenin by cardiac cells." Hypertension 30(6):
1389-96.

Vander, A. J. and G W. Geelhoed (1965). "Inhibition of renin secretion by angiotensin.
II." Proc Soc Exp Biol Med 120(2): 399-403.

Vetter, I. R. and A. Wittinghofer (2001). "The guanine nucleotide-binding switch in three
dimensions." Science 294(5545): 1299-304.

 

Wang, C., S. Jayadev and J. A. Escobedo (1995). "Identification of a domain in the
angiotensin 11 type 1 receptor determining Gq coupling by the use of receptor
chimeras." J Biol Chem 270(28): 16677-82.

Wasiak, S., R. Zunino and H. M. McBride (2007). "Bax/Bak promote sumoylation of
DRPl and its stable association with mitochondria during apoptotic cell death." 1
Cell Biol 177(3): 439-50.

Wood, J. M., J. Maibaum, J. Rahuel, M. G Grutter, N. C. Cohen, V. Rasetti, H. Ruger, R.
Goschke, S. Stutz, W. Fuhrer, W. Schilling, P. Rigollier, Y. Yamaguchi, F. Cumin,
H. P. Baum, C. R. Schnell, P. Herold, R. Mah, C. Jensen, E. O'Brien, A. Stanton
and M. P. Bedigian (2003). "Structure-based design of aliskiren, a novel orally
effective renin inhibitor." Biochem Biophys Res Commun 308(4): 698-705.

Yeyati, P. L., R. Shaknovich, S. Boterashvili, J. Li, H. J. Ball, S. Waxman, K.
Nason-Burchenal, E. Dmitrovsky, A. Zelent and J. D. Licht (1999). "Leukemia
translocation protein PLZF inhibits cell growth and expression of cyclin A."
Oncogene 18(4): 925-34.

Yonashiro, R., S. Ishido, S. Kyo, T. Fukuda, E. Goto, Y. Matsuki, M. Ohmura-Hoshino, K.
Sada, H. Hotta, H. Yamamura, R. Inatome and S. Yanagi (2006). "A novel
mitochondrial ubiquitin ligase plays a critical role in mitochondrial dynamics."
Embo J 25(15): 3618-26.

42

Yoon, Y., E. W. Krueger, B. J. Oswald and M. A. McNiven (2003). "The mitochondrial
protein hFisl regulates mitochondrial fission in mammalian cells through an
interaction with the dynamin-like protein DLPl ." Mol Cell Biol 23(15): 5409-20.

Yoon, Y., K. R. Pitts and M. A. McNiven (2001). "Mammalian dynamin-like protein
DLPl tubulates membranes." Mol Biol Cell 12(9): 2894-905.

Youle, R. J. and M. Karbowski (2005). "Mitochondrial fission in apoptosis." Nat Rev
Mol Cell Biol 6(8): 657-63.

Zeviani, M. (2004). "Mitochondrial disorders." Suppl Clin Neurophysiol 57: 304-12.

Zeviani, M. and S. Di Donato (2004). "Mitochondrial disorders." Brain 127(Pt 10):
2153-72.

Zhang, P. and J. E. Hinshaw (2001). "Three-dimensional reconstruction of dynamin in the
constricted state." Nat Cell Biol 3(10): 922-6.

Zhu, P. P., A. Patterson, J. Stadler, D. P. Seeburg, M. Sheng and C. Blackstone (2004).
"Intra- and intermolecular domain interactions of the C-terrninal GTPase effector
domain of the multimeric dynamin-like GTPase Drpl." J Biol Chem 279(34):
35967-74.

Zunino, R., A. Schauss, P. Rippstein, M. Andrade-Navarro and H. M. McBride (2007).
"The SUMO protease SENPS is required to maintain mitochondrial morphology
and function." J Cell Sci 120(Pt 7): 1178-88.

Zuo, L., M. Ushio-Fukai, S. lkeda, L. Hilenski, N. Patrushev and R. W. Alexander (2005).
"Caveolin-l is essential for activation of Racl and NAD(P)H oxidase after
angiotensin 11 type 1 receptor stimulation in vascular smooth muscle cells: role in
redox signaling and vascular hypertrophy." Arterioscler Thromb Va_sc Biol 25(9):
1 824-30.

43

CHAPTER 2

Functional expression of DLP-1

2.1 Introduction

Functional and structural studies on DLP-1 would shed light onto the molecular
mechanisms of mitochondrial fission, increasing our understanding of mitochondrial
defects and prevent human diseases. The middle domain and the GED of dynamins and
related proteins are considered to be important for protein intra- and intermolecular
interactions (Shin, Takatsu et al. 1999; Smirnova, Shurland et al. 1999; Zhang and
Hinshaw 2001; Zhu, Patterson et al. 2004). Mutations in the yeast homolog Dnml and
the human dynamin have been showed to impair the protein GTPase activities and disrupt
the protein oligomeric states (Ingerman, Perkins et al. 2005; Ramachandran, Surka et al.
2007). However, the functions of the DLP-1 middle domain on protein GTPase activities
and oligomerization have not been directly studied.

The dynamins also bind to lipids, particularly to negatively charged phospholipids,
primarily through their PH domains (Klein, Lee et al. 1998; Lemon and Ferguson 2000).
Usually the regions between the middle domains and the GEDs of dynamin-related
proteins are topologically analogous to the dynamin PH domain, and these regions
contain high content of positively charged residues that have the potential to interact with
negatively charged membrane lipids. However, for the dynamin-related proteins, there is
no domain or motif that is obviously similar to the canonical PH domain in dynamin.
DLP-1 binds and tubulates membranes (Yoon, Pitts et al. 2001), but the specificities of
the binding and what region(s) is responsible for the binding remain unclear.

Finally, one bottleneck that impacts protein structural studies, especially

45

membrane-interacting proteins of large size, such as DLP-l, is obtaining sufficient
amounts of highly purified protein suitable for functional and structural studies.
Towards solving the issues mentioned above, this chapter describes the recombinant
DLP-1 expression and purification, and biochemical analyses for functional

characterizations and future structural studies.

2.2 Materials and methods

2.2.] DNA Cloning of DLP-1 and Mutants - The DNA encoding for wild type human
DLP-1 isoform 2, the GTPase domain (1-30m), and the PH-like domain (497-607aa)
were amplified from cDNA IMAGE (clone ID 3882922) by PCR. The PCR amplification
was comprised of 35 cycles of denaturing at 94 °C for 30 sec, annealing at 55 °C for 45
sec, and elongation at 72 °C for 2 min and 30 sec for wild type DLP-l; 1 min for the
GTPase domain; and 40 sec for the PH-like domain, followed by 72 °C for 10 min. The
PCR products were purified by QIAquick PCR purification kit (QIAGEN), digested with
restriction enzymes Afl HI and XhoI. The pLWOl expression vector digested with NcoI
(Compatible with Afl III) and XhoI together with digested PCR products was transformed
into E. coli DHSa competent cells. Positive clones growing from LB plates containing
100 ug/ml arnpicillin were picked, and plasmid DNAs were isolated and sequenced.
Site-directed mutagenesis was performed using GeneEditor in vitro site-directed

mutagenesis system (Promega).

46

2.2.2 Protein Expression and Purification - The sequenced plasmid was transformed
into expression host E. coli C41 (DE3) competent cells. A fresh single colony frOm the
selection plate was inoculated into 100 ml LB media containing 100 ug/ml antibiotic at
37 °C with shaking at 200 rpm for overnight. Twenty ml of this culture was transferred
into 1 L fresh LB media and the cells were grown at 37 °C until the OD600 reached to
0.8-1.0. The cells were then induced by adding 0.05 mM IPTG and incubated with
shaking at room temperature for 17 hrs. Cells were harvested by centrifugation and
stored at —80 °C.

To purify DLP-l wild type and mutant proteins, the cell pellets were resuspended in
Buffer A (50 mM sodium phosphate, 300 mM NaCl, 250 mM sucrose, 10% glycerol, 10
mM B-mercaptoethanol, 0.1 mM EDTA, pH 8.0). After sonication, the crude cell extract
was centrifirged at 4 °C for 20 min at 12,000xg. The supernatant was loaded onto a
pre-equilibrated column containing 20 ml Ni-NTA agarose slurry. The column was
washed with buffer B (50 mM sodium phosphate, 300 mM NaCl, 10% glycerol, 10 mM
B-mercaptoethanol, 20 mM imidazole, pH 8.0). The protein bound column was eluted by
Buffer C (50 mM sodium phosphate, 300 mM NaCl, 10% glycerol, 10 mM
B-mercaptoethanol, 200 mM imidazole, pH 8.0). The protein eluates were pooled and

concentrated to 1 ml by Amicon ultra centrifugal filter molecular cutoff (Millipore).

2.2.3 [on Exchange ancL Size Exclusion Chromatography — Ion exchange
chromatography was performed to further purify target proteins and remove

47

contaminants. The pooled and concentrated eluates from Ni-NTA were loaded onto a 1
ml HiTrap Q ion exchanger (Amersham Biosciences) pre-equilibrated with Buffer D (20
mM Tris-HCl, pH 8.5). Protein was eluted off the column with a linear concentration
gradient of NaCl from 0 to 1 M, at a flow rate of 1 ml/min. The peak fractions containing
highly purified target protein were pooled and concentrated.

The oligomeric states of wild type DLP-1 and mutants were determined by analytical
size exclusion chromatography. A Superdex 200 10/30 GL or Superdex 75 10/30 GL
column was pre-equilibrated with bufler F (20 mM HEPES, 150 mM NaCl, 5% glycerol,
pH 7.2). The protein was loaded onto the column and eluted at a rate of 0.5 ml/min.
Fractions containing homogenous protein was combined and concentrated. Protein
concentration was estimated by Bradford assay using bovine serum albumin as standard.
For SDS-PAGE analysis, samples were electrophoresed on precast NuPAGE 4-12%
Bis-Tris polyacrylamide gels (Invitrogen) and visualized with Coomassie blue staining

and by Western blotting.

2.2.4 Meagarement of GT Pose Activity - The GTPase activities of DLP-l wild type, the
GTPase domain, the G350D, the R3658, the APH, and the K38A were measured using a
Malachite Green phosphate assay kit (Bioassay systems) at room temperature. The
reaction included 1 uM purified target protein with 0.05% bovine serum albumin, 16uM
GTP, and 2 mM MgC12 with Buffer F in an 800uL volume. At different time points,
200uL malachite green reagent was added to stop the reaction and the mixture was

48

incubated at room temperature for 10 minutes to allow color to develop. Absorbance at
650 nm was measured spectrophotometrically, and the amount of released inorganic
phosphate was determined by using a standard curve of known phosphate concentrations.

All data points represent an average of at least three independent measurements.

2.2.5 Lipiafl’rotein Interactions - FAT-Westem blot assay was used to investigate the
interaction of DLP-l with free lipids. Membrane strips (Invitrogen) pre-spotted with
phospholipids were blocked using TBS-T-BSA buffer (10 mM Tris-HCl, 150 mM NaCl,
0.1% Tween 20, 3% fatty acid-free BSA, pH 8.0) for 1 h at room temperature. Two
ug/mL of target protein were incubated with TBS-T-BSA buffer at 4 °C overnight. The
membrane strips were washed three times and soaked into TBS-T-BSA with an
anti-hexaHis mouse monoclonal antibody (Clontech) at a 1:1,000 dilution at 4 °C
overnight. The membranes were then washed three times and incubated with secondary
antibody at a 1:5,000 dilution in TBS-T-BSA for 1 h at room temperature. After another
three-time washing, the target protein was detected by using the standard Western
Lightning Chemiluminescence Reagent (PerkinElmer LAS Inc.).

Further protein-lipids interactions were analyzed by liposome-binding assays. The
lipids consist of phosphatidylcholine, phosphatidylserine, or phosphoinositides were
incubated in TBS buffer (50 mM Tris-HCl, 100 mM NaCl, pH 7.0) at 37 °C for 1 h. Afier
vortexing for 5 min, the solution was centrifuged at 20,000xg at 4 °C. The pellet
containing 200 pg liposomes was resuspended in 100 [4] TBS followed by adding 10

49

ug/mL of protein. The binding reaction was performed by incubating the mixture at 30 °C
for 30 min. The solution was centrifuged at 20,000x g at 4 °C and the distribution of target
proteins was determined by SDS-PAGE and Western blotting using an anti-hexaHis
mouse monoclonal antibody (at 1:5,000 dilution) and anti-mouse antibody (at 1:10,000

dilution).

2.3 Results and Discussion

2.3.] Protein Expression anal Purification - To produce high-yield recombinant proteins
for functional and structural analysis, several aspects regarding membrane-associated
protein expression in E. coli were considered. First, a small 3.4 kb, hexa-histidine tagged
vector with high copy numbers, pLW01, was selected as an expression vector. The
pLW01 was generated based on pET-23d and pBluescript 11 KS + vectors and was
successfully used to express membrane protein P450 in E. coli (Bridges, Gruenke et al.
1998). Second, the E. coli C41 (DE3) strain was chosen as expression host. Although
DLP-l exists primarily in the cytoplasm, it functions as a membrane-interacting protein
and probably binds to mitochondrial membranes. The C41 strain is a derivative of E. coli
BL21 (DE3) and can allow high-yield expression of membrane proteins, probably
because of the formation of internal membranes (Miroux and Walker 1996; Arechaga,
Miroux et al. 2000). Third, the growth conditions including temperature and
concentration of the inducer (IPTG) were optimized to reduce the rate of protein
synthesis and minimize formation of inclusion bodies. With these optimizations, about 10

50

mg DLP-1 protein was purified from 1 L bacterial culture.

Several DLP-1 mutant species with deletions or point mutations were also
recombinantly expressed and purified for investigation of impacts of specific domains or
key amino acids on the biochemical properties of DLP-1. I have aligned the DLP-l
protein sequence with its homologues from other organisms and analyzed the secondary
structure to determine possible essential residues and domain boundaries. Except the
putative PH-like domain, other regions of the protein are highly conserved and contain
regular secondary structures with combination of helices and strands (data not shown).
However, no obvious secondary structures or sequence conservation were identified in
the PH-like domain. Therefore, I isolated the PH-like domain and also created DLP-1
APH in which the PH-like domain was deleted, to investigate whether the domain has
effects on DLP-1 biochemical functions. In addition, the GTPase domain was isolated to
analyze whether other domains are required for the DLP-1 firll GTPase activity and
whether the GTPase domain is critical for protein oligomerization. The G350 and the
R365 are located in the middle domain and are conserved in dynamin and a number of
dynamin-related proteins (Fig. 2.1). These two residues in DLP-l homologues were
previously reported to be important for protein firnction including oligomerization and
GTPase. activities (Ingerman, Perkins et al. 2005; Bhar, Karren et al. 2006;
Ramachandran, Surka et al. 2007). The K38A was used as a control to evaluate the
GTPase activities of other DLP-l species. The K38 is conserved in dynamin and related
proteins (Fig. 2.1). It is located in the GTP/GDP binding motif of the GTPase domain and

51

is critical for the GTP hydrolysis. Mutation of K38A was previously reported to abolish
the DLP-1 GTPase activity (Yoon, Pitts et al. 2001; Zhu, Patterson et al. 2004). The
schematic demonstration of these mutants is shown in Fig. 2.2A. With considerations of
DNA cloning and protein expression methods as described above, all the recombinant
DLP-1 species were purified to near homogeneity by Ni-NTA column and ion exchange

chromatography (Fig. 2.2B).

HSDLP TSVLSLQSGKSSEV
HSDYN GAVLNFQSGKSSEV
ScDNM SSILTLQSGKSSEV
DmDRP SSVISVQSGKSSEV

CeDRP SQSALGKSSVLNEV

 
 
  
 
 
 
 
 

 
  
   
 

     
 

HSDLP ETAKYITELGESCGARI
HSDYN EGI'I'ELGGARSGDQDYSI
ScDN'M DGITELGGTSSDNKCARI
DmDRP EGITELTARNETCGGARM
CeDRP EGTILARNETCGGARI

Figure 2.1: Sequence alignments of partial GTPase domain (upper) and the middle
domain (lower) of dynamin related proteins. HsDLP, H. sapiens DLP-1; HsDYN, H.
sapiens Dynamin; ScDNM, S. cerevisiae Dnmlp; DmDRP, D. melanogaster dynamin
related protein; CeDRP, C. elegans dynamin related protein. The K3 8, G3 50, and R365 of
DLP-1 were indicated by arrows.

52

Table 2.1: Specific GTPase activity of DLP-1 WT and mutants. The specific activities
were measured based on the linear regions of the reactions.

 

 

 

 

 

Protein Specific Activity

DLP-1 WT 0.90 uM GTP / min/ pM protein
GTPase domain (1-307) 0.35 uM GTP / min / uM protein
R3658 0.12 uM GTP / min/ uM protein
G350D 0.10 uM GTP / min/ uM protein

 

 

 

 

53

APl-l
63500

R3658

GTPase Domain
PH-like Domain

 

 

     

MW WT APH 6350 R365 GTP PH

Figure 2.2: Purification of DLP-l WT and mutants. (A) Schematic illustration of
constructs. APH, DLP-1 lacking 501-607 aa; G350D and R3658, DLP-l with a point
mutation at G350 and R365, respectively; GTPase domain, DLP-l 1-307 aa; PH-like
domain, DLP-1 501-607aa. (B) SDS-PAGE analysis of purified recombinant DLP-1 WT
and mutants. From left to right: molecular weight markers; DLP-l WT; APH; G350D;
R3658; the GTPase domain; the PH-like domain.

54

Figure 2.3: GTPase activity of purified DLP-1 WT and mutants by malachite green
colorimetric assay. (A) GTP hydrolysis by DLP-1 WT. The data were plotted as time
(min) vs. released phosphate (uM). Square, DLP-1 WT; triangle, DLP-1 K38A; (B) GTP
hydrolyzed by DLP—1 WT and mutants. The data were plotted as time (min) vs.
hydrolyzed GTP/total GTP (%). Squares, WT; triangles, the GTPase domain; stars,
R3658; circles, G350D; rhomboids, APH. All data points represent an average of at least
three independent measurements.

55

 

r el eas ed, “M

Pi

GTP hydrolysis, %

 

 

K38A

 

 

Time,minutes

 

 

 

Tlme, minutes

56

1-307

R3655
G350D

APH

2. 3.2 GTPa_se Activities — The full-length DLP-l showed efficient GTPase activity (Fig.
2.3A and Table 2.1) as apposed to the GTPase null mutant K38A, which is consistent
with previous studies (Yoon, Pitts et al. 2001; Zhu, Patterson et al. 2004). The isolated
GTPase domain demonstrated lower activity than the full-length protein (Fig. 2.38 and
Table 2.1). Although the GTPase domain is supposed to provide enzymatic function, its
lower activity measured in this experiment suggests roles of other domains in the DLP-l
full GTPase activity. Presumably it is the lack of GED that is probably responsible for the
reduced GTPase domain activity based on previous reports about roles of the GED of
dynamins or related proteins in GTPase activity.

Although controversial, the GED of dynamin was considered to have GTPase
activating protein (GAP) activity (Muhlberg, Warnock et al. 1997; Sever, Muhlberg et al.
1999). Direct interactions between the GTPase domain and the GED of dynamin were
demonstrated by yeast two-hybrid studies (Smirnova, Shurland et al. 1999). Meanwhile,
the crystal structures of the GTPase domains of Dictyostlium dynamin A and rat dynamin
1 revealed a hydrophobic groove suggesting a putative GED interacting site (Niemann,
Knetsch et al. 2001; Reubold, Eschenburg et al. 2005). Although strong interactions
between the GTPase domain and the GED of DLP-1 was not detected directly by yeast
two-hybrid assay, point mutations in the GEDs of DLP-1 or yeast DNMl were reported
to cause a reduction of GTPase activity probably by interfering the interactions of GED
with the middle domain and further the GTPase domain (Shin, Takatsu et al. 1999;
Fukushima, Brisch et al. 2001; Zhu, Patterson et al. 2004). Therefore, The possible roles

57

of the C-terminus, particularly the GED in the DLP-1, on GTPase activity may explain a
relatively low activity of the GTPase domain. Our results from another aspect,
demonstrated that the GED is important for the full GTPase activity of DLP-1.

Both the G350D and the R3658 mutations showed much lower GTPase activities (Fi g.
2.3B and Table 2.1). Since these two mutations are not physically located in the GTPase
domain, it is possible that they affect the protein activity by interfering with the
intramolecular interactions, particularly between the middle domain and the GED so that
the GED could not reach effectively to the GTPase domain (Fig. 2.4B). The crystal
structure of a member of the dynamin superfarnily, human guanylate-binding protein 1
(hGBPl) showed that the GED folds back interacting with the middle domain and
extends to the GTPase domain (Prakash, Praefcke et al. 2000; Prakash, Renault et al.
2000). The three-dimensional map of dynamin fiom cryo-electron micrographs suggested
a similar structural property of a “stalk” region constituted by the middle domain and the
GED (Zhang and Hinshaw 2001). In addition, yeast two-hybrid assays of dynamins and
related proteins showed direct interactions between the GED and the middle domain
(Shin, Takatsu et al. 1999; Smirnova, Shurland et al. 1999; Zhu, Patterson et al. 2004).
Therefore, The two mutations may cause conformational changes in the middle domain
of DLP-l and interfere the intramolecular interactions and fiirther affect the GTPase
activity.

The results that the two point mutations have lower activities than isolated GTPase
domain indicated that besides affecting the interactions between the GED and the GTPase

58

domain, other local protein structures that are important for the activity may also be
affected. Given the fact that the two residues are close, the similar low activities that both
point mutations exhibited indicated that they may fall in the same functional or structural
group. Meanwhile, to my knowledge, for the first time, I demonstrated that the middle

domain is important to the DLP-1 GTPase activity.

59

 

 

Figure 2.4: Schematic model of DLP-1 intramolecular interactions. GTPase: the
GTPase domain; Middle: the middle domain; PL: the PH-like domain. Arrow: active site
the GTPase domain (A) wild type DLP-1. The lines show the interactions between the
middle domain and the GED. (B) G350D or R3658. (C) APH.

60

Surprisingly, deletion of the PH-like domain abolished the GTPase activity (Fig. 2.3B).
It is hard to interpret this result since the GTPase domain is intact in the APH. A possible
interpretation is that deletion of the PH-like domain forces the GED to fold back fiom the
very end of the middle domain instead of from the end of the PH-like domain as in
full-length DLP-1. This structure change makes the C-terminus of the GED reaching
further than normal to the GTPase domain and blocking the active site of the GTPase
domain (Fig. 2.4C).

The GTPase activities of dynamins can be largely stimulated by assembly into higher
order structures fi'om tetrarners at low ionic strength environments or in presence of
specific phospholipids (Warnock, Hinshaw et al. 1996; Barylko, Binns et al. 1998;
Stowell, Marks et al. 1999). I have attempted to measure the stimulated GTPase activity
of wild type DLP-1. At low ionic state of 20 mM NaCl, DLP-1 formed a structure larger
than its native size (data not shown). However, no stimulated GTPase activity was
observed either in low ionic state or in presence of phospholipids (data not shown). These
observations may be accounted for by the fact that particular components of the fission
complex or specific mitochondrial membrane structures may be required for the

stimulation of DLP-1 GTPase activity.

2.3.3 Oligomeric States — Since the GTPase domain, the G350D, the R3658, and the
APH demonstrated reduced GTPase activities, I was interested in studying the oligomeric
states of these species to analyze whether the decreased activities have relationships with

61

protein oligomerization. For the middle domain, intermolecular interactions of dynamin
were previously reported to be important for the protein oligomerization and fiirther
assembly (Smirnova, Shurland et al. 1999; Zhang and Hinshaw 2001). The yeast
mutation equivalent to DLP-1 G350D and the human dynamin mutation equivalent to
DLP-1 R3658 have been showed to impair the protein GTPase activities and disrupt the
multimeric or tetrarneric protein states to dimeric state (Ingerman, Perkins et al. 2005;
Ramachandran, Surka et al. 2007). Therefore, I sought to determine whether the reduced
GTPase activities of the G350D and the R3658 are caused by disruption of protein
oligomerization. For the GTPase domain and the PH-like domain, I am interested in their
oligomeric states for investigating whether the domains are critical for full-length protein
oligomerization.

DLP-l exists primarily as tetramers in the cytosol and forms higher order structures
once binding to membranes (Shin, Takatsu et al. 1999; Zhu, Patterson et al. 2004). Our
size exclusion chromatography results showed that the full-length DLP-1 can be isolated
as a size of about 350 kDa, corresponding to the tetrarneric state (Fig. 2.5A and Table 2.2),
consistent with previous reports (Shin, Takatsu et al. 1999; Zhu, Patterson et al. 2004).
Similarly, the G350D and the R3658 were eluted also at about 350 kDa (Fig. 2.5B, C and
Table 2.2), displaying stable tetrarneric forms that are almost indistinguishable from that
of the wild type protein. These results suggested that the G350D and the R3658
mutations do not seem to affect the intermolecular interactions, and the reduced GTPase
activities do not appear to be caused by loss of protein quaternary structures. Although

62

the middle domain may be important for not only protein activities, but also protein
oligomerization, mutations in these two positions may not be strong enough to alter the
intermolecular interactions. However, they are sufficient to show that the middle domain
of DLP-1 is important for protein activities, and the untouched protein oligomeric states
support our interpretation about the reduced GTPase activities of the G350D and R3658
that alteration of intramolecular interactions caused by possible conformational changes
in the middle domain may be responsible for decreased activities (Fig. 2.4A).

Unlike the DLP-1 G350D and R3658 which keep uninfluenced oligomeric states, the
equivalent mutations in yeast DNMl and human dynamin was disrupted to dimers
compared to the wild type multimers and tetramers, respectively (Ingerman, Perkins et al.
2005; Ramachandran, Surka et al. 2007). Although these proteins are functionally similar,
difference in protein natures may explain the distinct effects of equivalent mutations on
protein oligomerization. For example, compared to the tetrameric DLP-1 found in cytosol
or in vitro (Shin, Takatsu et al. 1999; Zhu, Patterson et a1. 2004), the Dnml aggregates in
cytosol and was isolated as multimers likely 8-12 mers (Ingerman, Perkins et al. 2005).
Therefore, it is possible that either DLP-l forms relatively stronger tetramers, or other
residues evolved in DLP-l that is critical for the dimer to tetramer transition.

In addition, the reduced GTPase activities with untouched tetrarneric forms that the
two middle domain mutations G350D and R3658 exhibited are similar with the
phenotypes reported previously for the mutation in the DLP-1 GED, K679A (Zhu,
Patterson et a1. 2004). Because of the intramolecular interactions between the middle

63

domain and the GED, mutations on either side would disrupt the similar interactions and
cause similar defective “open” state. The parallel phenotypes of mutations from different
locations may provide another evidence that the middle domain interacts with the GED.

The isolated GTPase domain (l-307aa) was characterized as 35 kDa, a monomeric size
by exclusion chromatography (Fig. 2.5D and Table 2.2), indicating that probably this
domain is not involved in DLP-1 oligomerization. Lack of oligomerization may be
another factor responsible for the reduced activity of the isolated GTPase domain. In
addition, during testing possible domain boundary of the GTPase domain, I also isolated
the 1-323aa and 1-340aa fragments because secondary structure prediction suggested a
long helix ending at residue 323 and sequence alignments favored residue 340 as the
domain boundary (data not shown). However, although these fragments were purified,
they aggregated when analyzed by size exclusion chromatography (data not shown). This
may suggest that the part from residue 308 to 340 is involved in the formation of middle
domain helix bundle and inclusion of this part into the isolated GTPase domain may
cause the hydrophobic region exposed to solvent.

Taking the activities and protein oligomerization results of the isolated GTPase domain,
the G350D, and the R3658, our findings support the idea that the GED plays roles in the
protein GTPase activity. And proper intramolecular interactions with the middle domain
are important for normal GED functions on DLP-1 activity.

Because deletion of the PH-like domain abolished the GTPase activity, I was interested
in investigating the oligomeric state of the DLP-1 APH to determine whether the loss of

64

activity was caused by collapse of the overall protein structure without the PH-like
domain, and whether the PH-like domain is critical for the full-length protein
oligomerization. The chromatography results showed that the APH exists as about 300
kDa, stable tetramers without any sign of quaternary structure being disrupted (Fig. 2.5E
and Table 2.2), indicating that deletion of the PH-like domain does not affect the protein

oligomerization.

65

Figure 2.5: Size exclusion chromatography of DLP-1 WT and mutants. Panels A, B,
C, and E show chromatography on a Superdex 200 10/30 gel filtration column; panels D
and F show chromatography on a Superdex 75 10/30 gel filtration column. (A) DLP-l
WT; (B) G350D; (C) R3658; (D) the GTPase domain; (E) APH; (F) the PH-like domain;
(G) Kav versus molecular weight plot of protein standards. The standards were plotted as
Kav vs. molecular weight. Kav was obtained by the formula of Kav = (Vc-V0)/(V,-Vo),
where Ve is the elution volume of each molecular weight markers, V0 the void volume,
and V, total volume of the column.

66

 

Absorbance at 280nM (mAU)

Absorbance at 280nM (mAU)

180'
30‘

 

.10

m-
1001
mi
120-
100-

so-

‘01
20"
0d

 

 

 

 

10 1! 20 25

Retention volume (ml)

 

 

T I U I

10 15 20 25

Retention volume (ml)

67

Figure 2.5 continued
C
1001
80 4
00 ‘
40 'l

20"

Absorbance at 280nM (mAU)
O
I

 

I I t I T

-20 '0 5 10 15 20 25

 

Retention volume (ml)

80%

4° 0 r 10 15 20 25

Retention volume (ml)

 

 

Absorbance at 280nM (mAU)

 

 

 

 

68

Figure 2.5 continued

Absorbance at 280nM (mAU)

Absorbance at 280nM (mAU)

 

 

 

 

,2. t I I I 1
0 I 10 15 20 25

Retention volume (ml)

330-
m- n
230-

180‘

 

130'

0 5 10 18 20 26

Retention voltune (ml)

 

 

 

69

Figure 2.5 continued

0.6 '-
0.5 '
0.4 I

3
g 0.3+

0.2 -
Superdex 75

0.1 '

 

Superdex 200

 

I I I I I I I I j

0 50 100 150 200 250 300 350 400 450 500 550 600 650 700

um (x1000)

70

Table 2.2: Calculated molecular weight of DLP-1 WT and mutants by gel filtration.
Molecular weight was based on the protein standards (Fig. 4G).

 

 

 

 

 

 

 

 

Protein Calculated MW Oligomeric state
DLPl WT 350000 tetramer
DLPl_G350D 350000 tetramer
DLPl_R3658 350000 tetramer
DLPl_1-307 35000 monomer
DLPl_APH 300000 tetramer
DLPl_PH-like domain 45000 tetramer

 

 

 

 

A munber of research has described the GED of dynamins or related proteins as a key
player in protein oligomerization and higher order structure formation (Danino and
Hinshaw 2001; Praefcke and McMahon 2004). Also, the middle domain was considered
to be critical for these structures (Zhang and Hinshaw 2001; Ingerman, Perkins et al.
2005; Ramachandran, Surka et al. 2007). I have attempted to expressed DLP-1 fragments
with the middle domain or the GED deleted to directly study the effects of these two
domains on protein oligomerization. Unfortunately, they were aggregated in non-soluble
inclusion bodies in regardless of expression conditions optimized. However, this may
suggest that both the middle domain and the GED are critical for the intra- and
intermolecular interactions, and when lacking one of these domains, the hydrophobic
regions that are supposed for the interactions are exposed to solvent and the protein
aggregate. Since the GTPase domain itself is a monomer, and the PH-like domain is not
essential for the protein oligomerization, our results suggest important roles of the middle

domain and the GED in DLP-l oligrnerizations.

7l

The isolated PH-like domain showed a size of about 45 kDa, a tetramer by exclusion
chromatography (Fig. 2.5F and Table 2.2). At this point, although we could not exclude
the PH-like domain to be involved in the protein oligomerization, the fact that the APH is
in stable tetrarneric state suggested that the PH-like domain may not be the major
determinant of DLP-l tetramerization. The untouched quaternary protein structure
without major damage on the ability of oligomerization of the APH also supports our
previous interpretation that inappropriate intramolecular interactions may cause the
abolishment of the APH GTPase activity (Fig. 2.4B).

Taken together, our results support the potential roles of the middle domain and the

GED in DLP-1 intra- and intermolecular interactions.

N
k

Interactions: with Lipids and Membranes - The dynamins bind to lipids,
particularly to negatively charged phospholipids, primarily by their PH domains
(Klein, Lee et al. 1998; Lemon and Ferguson 2000). However, for the
dynamin-related proteins, there is no obvious domain or motif indicating a
typical PH domain. Usually the regions between the middle domains and the
GEDs of dynamin-related proteins are physically analogous to the dynamin PH
domain, and these regions also contain large content of positively charged
residues that have potential to interact with negatively charged membrane lipids.
For dynamin-related proteins, only Arabidopsis dynamin-like protein 2 (ADL2)
was reported to bind specially to phosphatidylinositol 4-phosphate (PI(4)P) in

72

vitro without knowing which region is responsible for the binging (Kim, Park et
al. 2001). DLP-1 was reported to bind and tubulate membranes (Yoon, Pitts et al.
2001). However, the specificities of the binding and what region(s) is responsible

for the binding remain unclear.

A B

0°“ 5’"! 2‘...“

 

GNOUIDwN-i

 

PH-Iike domain

Figure 2.6: Lipids-binding of DLP-1 WT and the PH-like domain. For phospholipids
that each number represents, see Tables 2.3.

73

Table 2.3 A: Illustration of binding strength of DLP-1 WT to phospholipids. “--”, no
binding; “+”, binding. The strength of binding is relative to the control spot (No. 16).

B'
' '0 acid --

1"?"

hati choline
. . l

ethanolamine
line

 

$899993"?

++++
4+1-
+4—1-

Table 2.3 B: Illustration of binding strength of the PH-like domain to phospholipids.
Please refer to A for legend.

B' . . 'ds
' acid -- . ' ' l-
hati choline -- . 3,4
' ' ' 1 -- . 3,5
++++ 4,5
++ . 3,4,5
++ 'dic acid

L
L

ethanolamine --
line --

1.
2.
3.
4.
5.
6.
7.
8.

 

Using Fat-Westem blot assay, I found that DLP-1 wild type protein bound to free lipids,
specifically to negatively charged phospholipids (Fig. 2.6A). The protein bound
dominantly to phosphoinositides (PIs) with minor interactions with phosphatidylserine
(PS) and phosphatidic acid (PA) (Table 2.3). Different with the ADLZ that the protein

bound exclusively to PI(4)P and the dynamins that bound to PI(4,5)P2 specifically,

74

DLP-1 interacted with all PIs similarly without distinguishable specificities to one or a
few of them. Although some reported membrane-interacting proteins seem to bind to
specific lipids, binding to P15 without specificities is not uncommon to proteins that
interact with membranes (DiNitto, Cronin et al. 2003; Lemmon 2007; Lemmon 2008).
For example, the Annexin, epsin N-terrninal homology (ENTH), AP180 N-terrninal
homology (ANT H), Bin/amphiphysin/Rvs (BAR), and Fer-CIP4 homology (F-BAR)
domain-containing proteins bind either general acidic phopholipids or PIs without
specific preferences (Lemmon 2008). However, research on a number of
membrane-binding proteins indicated that, for proteins that bind lipids without specificity,
if specificity exists, it could be acquired by induction of specific signals or environments
(Lemmon 2008). For example, increase of calcium concentration in cytosol is a signal for
some Annexin domain-containing proteins, and curved membrane environments seem to
be required for the BAR domain-containing proteins to acquire PIS-binding specificity
(Zimmerberg and McLaughlin 2004; Gerke, Creutz et al. 2005; Lemmon 2008).
Therefore, non-specific interactions of DLP-l with PIs do not exclude the possibilities
that it binds specifically to one of the PIs at specific mitochondrial fission time or
location. Meanwhile, since usually each individual PI is not abundant on the
mitochondrial outer membranes, the ability to bind promiscuously to P15 or other acidic
phospholipids may provide an advantage for DLP-1 to interact efficiently with
membranes. Moreover, during mitochondrial fission events, the P18 may aggregate to
specific fission sites to allow membranes recognition by DLP-1. However, lipid binding

75

does not exclude the possibility that DLP-1 interacts with other molecules such as hFisl
to form division complex onto the mitochondrial outer membranes. It is likely that the
interaction of specific (or some non-specific) lipids with protein is necessary for DLP-l
to function properly in the mitochondrial fission process. Also, DLP-1 binding to P18 may
serve as a signal to cytosol or to the inner mitochondrial space to further promote
mitochondrial fission or other events such as apoptosis. For example, even if the DLP-1
doe not directly provide mechanical force to break the membrane, binding to the
membrane may signal other molecules to facilitate the fission events. Generally speaking,
our results suggested that DLP-1 may interact with membranes partially, if not totally, by
binding acidic phospholipids in general and P15 preferentially.

Using the Fat-Westem blot assay, the isolated PH-like domain was found to bind to
lipids with a pattern that is similar with that of the wild type DLP-1, mostly to P18 with
similar binding strength (Fig. 2.63 and Table 2.3). These results, to some extent, indicate
that the PH-like domain may facilitate the interactions of DLP-1 with membranes, which
is consistent with the three-dimensional structure of dynamin fi'om cryo-electron
micrographs that the PH domain is oriented to face the membrane directly (Zhang and
Hinshaw 2001).

Although contains large content of positively charged residues, the protein sequence of
the DLP-1 PH-like domain was not found to be conserved when aligned with other
dyanmin-related proteins. This may suggest that function of the PH-like domains of
dynamin-related proteins may not be limited to binding to membranes. They could have

76

evolved and acquired other functions, possibly interacting with specific proteins to
facilitate the formation of protein complexes. This phenomenon is commonly observed
for some lipids-bind domains such as the ENTH, ANTH, BAR, and F-BAR, which
demonstrate inconspicuous preferences to specific acidic phospholipids, but are involved
in other cellular functions such as curving or deforming membranes (Itoh and De Camilli
2006; Lemmon 2008). Also, the PH domain-containing proteins often have protein
interacting domains such as Src homology 2, 3 (8H2, 8H3) to strengthen membrane
binding by interacting with other protein molecules (Schlessinger 2000). Since there is no
such a domain predicted or identified in the DLP-1, the PH-like domain may be bi- or

multifunctional to involve in both membranes and protein interactions.

    

”W51"? SIN. P 5’". P SM 3’", ,
DLP-1 API-l 63500513653 PH

Figure 2.7: Liposome-binding of DLP-1 WT and mutants. The liposomes used in the
assay were made from DOPS (see text). S/N refers to the supernatant fraction and P to the
pellet.

77

DLP-1 has been reported to bind and tubulate lipid membranes in liposome-binding
assays (Yoon, Pitts et al. 2001). I investigated whether the mutant DLP-l species
impaired the liposome-binding properties by using this assay, since liposome vesicles are
good approximation of the lipid membrane structure (DiNitto, Cronin et al. 2003). The
results showed that the G350D, the R3658, and the APH were dominantly fractionated in
the pellets as wild type DLP-1 was (Fig. 2.7), indicating that the mutations did not cause
significant changes on the protein liposome-binding properties. Since these mutants also
retained proper oligomeric states (Fig. 2.4), it is likely that the tetrameric structure is one
of the factors required for the membranes binding. Interestingly, APH still bound
liposomes, indicating that besides the PH-like domain, there must be other regions in
DLP-1 also are responsible for the membrane association. If the PH-like domain alone
binds too weakly to lipids to drive DLP-1 to membranes, combination of other
lipids-interacting regions would help cooperatively promote membranes targeting.
Unexpectedly, although bound to free lipids, the PH-like domain was ahnost exclusively
fractionated in the supernatant, indicating that it did not bind liposomes (Fig. 2.7). This
contradictory results could be explained by the unique structures of liposomes that more
approximate in vivo membrane structures than free lipids do, and the PH-like domain
may need specific structural scaffolds which is provided by other domains of DLP-1 to
interact with membranes. Unlike the free lipids, the membrane surface is usually curved
and global which may only be associated by specific protein structures. In vivo, the
PH-like domain may be oriented the way that one side interacts with membranes and

78

another side with other protein molecules. For the wild type protein, other domains may
help drive membranes associations by both interacting with membranes themselves and
stabilizing the PH-like domain orientation to enhance the binding. However, for isolated
PH-like domain, although has lipids-interacting abilities, without other domains’ support
as scaffolds, the membranes interactions may not be stable given the fact that the binding
may be weak or non-specific.

Taken together, the results suggested that other than the PH-like domain, DLP-l may
contain more membrane-interacting regions, and these regions may also function as a
scaffold for supporting and orienting the PH-like domain for appropriate membrane

targeting and protein interactions.

2.4 Conclusion

The biochemical and structural properties of recombinant full-length DLP-l and
selected mutants have been studied. The DLP-l WT, G350D, R3658, APH, GTPase
domain, and PH-like domain were expressed in E. coli and purified to near homogeneity
by a number of protein purification tools including Ni-NTA column, ion exchange, and
analytical gel filtration chromatography. GTPase activity, oligomeric state, and
lipid-binding properties of appropriate DLP-1 species were investigated.

Compared to the full-length DLP-1, the isolated GTPase domain exhibited a relatively
lower GTPase activity, suggesting a role of the C-terminus of DLP-1 probably the GED
in the full GTPase activity. The G350D and R3658 mutants in the middle domain

79

severely impair the GTPase activity, but have no significant effects on the protein
oligomeric state, indicating that these two mutations interrupted the intramolecular but
not intermolecular interactions, and therefore, the middle domain of DLP-1 is important
for the protein activity probably by facilitating appropriate connections between the
GTPase domain and the GED.

Size exclusion chromatography showed that deletion of the PH-like domain has no
significant effects on protein oligomeric state. Together with the results that the isolated
GTPase is a monomer, this indicated that the GTPase domain and the PH-like domain
may not be major determinants of DLP-l tetramerization, further suggesting important
roles of the middle domain and the GED in protein intermolecular interactions.

The DLP-l and the isolated PH-like domain bound free phosphoinositides suggesting
that DLP-1 may interact with membranes directly by binding acidic phospholipids
preferentially phosphoinositides, and the PH-like domain may be responsible for the
interactions. Although GTPase activity abolished, the APH bound to liposomes
suggesting in addition to the PH-like domain, other regions of DLP-l may function as
lipids-interacting enhancer as well as scaffolds for orienting the PH-like domain into

appropriate membrane targeting.

8O

References

Arechaga, I., B. Miroux, S. Karrasch, R. Huijbregts, B. de Kruijff, M. J. Runswick and J.
E. Walker (2000). "Characterisation of new intracellular membranes in

Escherichia coli accompanying large scale over-production of the b subunit of
F lFo ATP synthase." Febs Letters 482(3): 215-219.

Barylko, B., D. Binns, K. M. Lin, M. A. Atkinson, D. M. Jameson, H. L. Yin and J. P.
Albanesi (1998). "Synergistic activation of dynamin GTPase by Grb2 and
phosphoinositides." J Biol Chem 273(6): 3791-7.

Bhar, D., M. A. Karren, M. Babst and J. M. Shaw (2006). "Dimeric Dnml-G385D
interacts with mdvl on mitochondria and can be stimulated to assemble into
fission complexes containing Mdvl and Fisl." Journgl of Biological Chemisgy
281(25): 17312-17320.

Bridges, A., L. Gruenke, Y. T. Chang, I. A. Vakser, G Loew and L. Waskell (1998).
"Identification of the binding site on cytochrome P450 2B4 for cytochrome b5
and cytochrome P450 reductase." J Biol Chem 273(27): 17036-49.

Danino, D. and J. E. Hinshaw (2001). "Dynamin family of mechanoenzymes." Curr Qpin
Cell Biol 13(4): 454-60.

DiNitto, J. P., T. C. Cronin and D. G Larnbright (2003). "Membrane recognition and
targeting by lipid-binding domains." Sci STKE 2003(213): re16.

Fukushima, N. H., E. Brisch, B. R. Keegan, W. Bleazard and J. M. Shaw (2001). "The
GTPase effector domain sequence of the Dnmlp GTPase regulates self-assembly

and controls a rate-limiting step in mitochondrial fission." Mol Biol Cell 12(9):
2756-66.

Gerke, V., C. E. Creutz and S. E. Moss (2005). "Annexins: linking Ca2+ signalling to
membrane dynamics." Mt Rev Mol Cell Biol 6(6): 449-61.

Ingerman, E., E. M. Perkins, M. Marino, J. A. Mears, J. M. McCaffery, J. E. Hinshaw and
J. Nunnari (2005). "Dnml forms spirals that are structurally tailored to fit
mitochondria." Journ_al of Cell Biology 170(7): 1021-1027.

Itoh, T. and P. De Camilli (2006). "BAR, F-BAR (EFC) and ENTH/ANT H domains in

the regulation of membrane-cytosol interfaces and membrane curvature." Biochim
Biophys Acta 1761(8): 897-912.

81

Kim, Y. W., D. S. Park, S. C. Park, S. H. Kim, G W. Cheong and I. Hwang (2001).
"Arabidopsis dynamin-like 2 that binds specifically to phosphatidylinositol
4-phosphate assembles into a high-molecular weight complex in vivo and in
vitro." Plant Physiology 127(3): 1243-1255.

Klein, D. E., A. Lee, D. W. Frank, M. 8. Marks and M. A. Lemmon (1998). "The
pleckstrin homology domains of dynamin isoforms require oligomerization for
high affinity phosphoinositide binding." Journal of Biological Chemistry 273(42):
27725-27733.

Lemmon, M. A. (2007). "Pleckstrin homology (PH) domains and phosphoinositides."
Biochem Soc Svmp(74): 81-93.

Lemmon, M. A. (2008). "Membrane recognition by phospholipid-binding domains." N_at
Rev Mol Cell Biol 9(2): 99-111.

Lemmon, M. A. and K. M. Ferguson (2000). "Signal-dependent membrane targeting by
pleckstrin homology (PH) domains." Biochemical Journal 350: 1-18.

Miroux, B. and J. E. Walker (1996). "Over-production of proteins in Escherichia coli:
Mutant hosts that allow synthesis of some membrane proteins and globular
proteins at high levels." Journal of Moleflar Biology 260(3): 289-298.

Muhlberg, A. B., D. E. Warnock and S. L. Schmid (1997). "Domain structure and
intramolecular regulation of dynamin GTPase." Embo J 16(22): 6676-83.

Niemann, H. H., M. L. Knetsch, A. Scherer, D. J. Manstein and F. J. Kull (2001). "Crystal
structure of a dynamin GTPase domain in both nucleotide-free and GDP-bound
forms." Embo J 20(21): 5813-21.

 

Praefcke, G J. and H. T. McMahon (2004). "The dynamin superfamily: universal
membrane tubulation and fission molecules?" Nat Rev Mol Cell Biol 5(2):
133-47.

Prakash, B., G J. Praefcke, L. Renault, A. Wittinghofer and C. Herrmarm (2000).
"Structure of human guanylate-binding protein 1 representing a unique class of
GTP-binding proteins." Nature 403(67 69): 567-71.

 

Prakash, B., L. Renault, G J. Praefcke, C. Herrmarm and A. Wittinghofer (2000).
"Triphosphate structure of guanylate-binding protein 1 and implications for
nucleotide binding and GTPase mechanism." Embo J 19(17): 4555-64.

82

Ramachandran, R., M. Surka, J. S. Chappie, D. M. Fowler, T. R. Foss, B. D. Song and S.
L. Schmid (2007). "The dynamin middle domain is critical for tetramerization and
higher-order self-assembly." Embo Journa_l 26(2): 559-566.

Reubold, T. F., S. Eschenburg, A. Becker, M. Leonard, S. L. Schmid, R. B. Vallee, F. J.
Kull and D. J. Manstein (2005). "Crystal structure of the GTPase domain of rat
dynamin 1." Proc Natl Acad Sci U S A 102(37): 13093-8.

Schlessinger, J. (2000). "Cell signaling by receptor tyrosine kinases." Call 103(2):
211-25.

Sever, 8., A. B. Muhlberg and S. L. Schmid (1999). "Impairment of dynamin's GAP
domain stimulates receptor-mediated endocytosis." Nature 398(6727): 481-6.

Shin, H. W., H. Takatsu, H. Mukai, E. Munekata, K. Murakami and K. Nakayama (1999).
"Intermolecular and interdomain interactions of a dynamin-related GTP-binding
protein, Dnmlepslp-like protein." J Biol Chem 274(5): 2780-5.

Shin, H. W., H. Takatsu, H. Mukai, E. Munekata, K. Murakami and K. Nakayama (1999).
"Intermolecular and interdomain interactions of a dynamin-related GTP-binding
protein, Dnmlp/Vpslp-like protein." Journal of Biologi_cal Chemistg 274(5):
2780-5.

Smirnova, E., D. L. Shurland, E. D. Newrnan-Smith, B. Pishvaee and A. M. van der Bliek
(1999). "A model for dynamin self-assembly based on binding between three
different protein domains." J Biol Chem 274(21): 14942-7.

Stowell, M. H., B. Marks, P. Wigge and H. T. McMahon (1999). "Nucleotide-dependent
conformational changes in dynamin: evidence for a mechanochemical molecular

spring." Nat Cell Biol 1(1): 27-32.

Warnock, D. E., J. E. Hinshaw and S. L. Schmid (1996). "Dynamin self-assembly
stimulates its GTPase activity." J Biol Chem 271 (37): 22310-4.

Yoon, Y., K. R. Pitts and M. A. McNiven (2001). "Mammalian dynamin-like protein
DLPl tubulates membranes." Molecrfiar Biologyof the Cell 12(9): 2894-2905.

Zhang, P. and J. E. Hinshaw (2001). "Three-dimensional reconstruction of dynamin in the
constricted state." Nat Cell Biol 3(10): 922-6.

Zhu, P. P., A. Patterson, J. Stadler, D. P. Seeburg, M. Sheng and C. Blackstone (2004).

83

"Intra- and intermolecular domain interactions of the C-terminal GTPase effector
domain of the multimeric dynamin-like GTPase Drpl." Journal of Biological
Chemist_ry 279(34): 35967-35974.

Zimmerberg, J. and S. McLaughlin (2004). "Membrane curvature: how BAR domains
bend bilayers." Curr Biol 14(6): R250-2.

84

CHAPTER 3

Structural studies of DLP-1 and MFN-1

85

3.1 Introduction

Structural studies on DLP—l and MFNs, particularly X-ray crystal structure
determinations at high resolution, would provide a physical foundation for functions of
these proteins, shed light on the molecular mechanisms of mitochondrial fission and
fusion, and help understand some human diseases caused by mitochondria-related defects.
One of the rate-limiting steps in structural biology is protein crystallization. Depending
on a protein’s physical nature, some proteins are relatively easy to crystallize, while,
some others, such as large and complex proteins from higher organisms, are much more
challenging for crystallographer. For membrane-interacting proteins, particularly integral
membrane proteins, other factors such as protein expression and purification may also
become rate-limiting steps for structural studies instead of the crystallization. This
situation arises from the fact that membrane-interacting proteins contain hydrophobic
transmembrane or membrane-interacting regions, which can interfere with folding and
assembly during expression. Molecular protein engineering sometimes is a helpful tool
bypassing these problems and increasing the chance of protein crystallization.

In this chapter, I will describe structural studies of DLP-1 and MFNs from the protein
expression and purification through crystallization trials. Molecular engineering methods

for protein crystallization enhancement are also addressed.

3.2 Crystallization trials of DLP-1
The DNA cloning, protein expression, and purification procedures of wild type DLP-1
were as described in chapter 2. Fractions of the tetrarneric size of DLP-l by size

exclusion chromatography were collected and concentrated to about 10 mg/mL with a

86

50K Amicon Ultra centrifugal unit (Millipore) for crystallization studies. Initially,
Microbatch-under-oil method with an ORYX-4 crystallization robot (Douglas
Instruments) was used. With a total of at least 864 screening conditions (Hampton
Research I & II, Membrane Faction I & H, Lite Screen I & II, Cryo Screens I & II,
Wizard Screens I & II, Axygen Biosciences I, III, IV, & V) together with about 200 other
grid screens designed by myself, 0.75 uL of purified protein and 0.75 uL precipitant
reagent were mixed in each drop for crystallization.

Microbatch and hanging drop vapor diffusion are quite different methods that could
lead to different patterns of crystallization results. A report that compared the two
methods has shown that about 60 percent of crystals generated by the two methods have
overlapped conditions, while the other 40 percent are from unique conditions (Chayen
1998). In other words, crystals grow from one method may not show up in the same
condition by the other method. Therefore, hanging drop vapor diffusion was also used
with the above crystallization conditions for DLP-l. Two uL of protein and 2 uL of
precipitant were mixed to equilibrate against 1 mL reservoir solution.

The hGBPl was crystallized with protein concentrations of 50-100 mg/mL (Prakash,
Praefcke et al. 2000; Prakash, Renault et al. 2000) and is the only eukaryotic member in
dynamin superfarnily that has full-length protein crystal structure available for reference.
Since DLP-1 may structurally analogous to hGBPl, protein concentration of DLP-1 was
adjusted from 10 mg/mL to up to 50 mg/mL for crystallization. There was no obvious
decrease of protein solubility observed in most conditions. Also, since in about 40 percent
of conditions, the protein was precipitated, decreasing protein concentration to up to 2

mg/mL was attempted. The protein solubility was improved slightly with low protein

87

concentration.

The GTPase domain of DLP-l contains the four conserved GTP-binding motifs.
Previous structures of dynamins GTPase domains (Niemann, Knetsch et al. 2001;
Reubold, Eschenburg et al. 2005) and hGBPl (Prakash, Renault et a1. 2000) indicated
that the GTP/GDP binding region may be flexible if without ligands on. After GTP/GDP
binding, the region is highly organized and ordered. This flexible region without ligands
binding may be one of the energy barriers that hinder the protein crystallization. Thus,
co-crystallization of DLP-1 with 1 mM GDP was performed. Also, since GTP is not
stable substrate (which will be hydrolyzed to GDP), GTP analogues such as GppNHp
were used for co-crystallization. Although proteins in the dynamin family do not require
guanine nucleotide to maintain stability (unlike canonical small GTPases), I expected that
binding of GDP or GTP analogues could cause a conformational change that help
maintain an ordered local structure or generate new crystal contacts. Unfortunately, with
all these attempts, to date, no reproducible crystals of DLP-1 protein have yet been

obtained.

3.3 Protein engineering methods for crystallization enhancement
3.3.1 Generalgintroductioné and rm

It is a challenge to crystallize large complex proteins, particularly membrane-
interacting proteins from higher organisms like humans, because these proteins usually
contain multiple structurally independent regions or relatively flexible domains that need
to interact with unknown molecular partners in vivo. This complexity may give rise to an

energy barrier to prevent protein from forming highly ordered and aggregated crystals.

88

However, molecular protein engineering may help reduce this complexity and increase
the chance of crystallization. With other factors such as precipitant reagents fixed,
protein itself could serve as a variant in crystallization process (Dale, Oefner et a1. 2003).
Since the wild type DLP-1 has not yet crystallized by varying simply the solution
conditions, protein engineering was performed in an effort to enhance protein

crystallization.

3. 3. 1. 1 Limited proteolysis

Large proteins from higher organisms such as plants or humans usually are composed
of multiple functional regions (domains). The individual domains are relatively compact,
but the Whole protein is heterogeneous because of motions between domains caused by
flexible domain linkers. This conformational heterogeneity often results in difficulties in
protein crystallization (Koth, Orlicky et al. 2003). To solve this problem, it is effective to
identify stable and crystallizable firnctional domains of such proteins. The challenge in
this work is to identify the domain boundaries (Koth, Orlicky et al. 2003). Multiple
sequence alignments and secondary structure prediction are among the ways for
identifying individual domains. However, a more promising method is limited proteolysis
followed by mass spectrometry and N-terrninal sequencing. The rationale for limited
proteolysis is that the domain-domain connection regions are highly exposed to solvents
and can be recognized easier by proteases than regions within a compact domain (Koth,

Orlicky et al. 2003).

89

3.3.1.2 Surface-entropy reduction

The basis of the surface-entropy reduction method is to reduce the entropic cost of
protein crystallization by modifying the target protein using molecular protein
engineering. Typically, residues with large flexible side chains in solvent-exposed loops
are mutated to small amino acids (Derewenda and Vekilov 2006). For instance, highly
hydrophilic residues in loop regions such as lysine or glutamate will be replaced by less
hydrophilic amino acids such as alanine. Sometimes, flexible protein regions with a high
degree of conformational heterogeneity have a higher entropic state and a lower
propensity for crystallization. Thus, deleting or replacing the flexible regions may be
helpful strategy for crystallization as it would reduce the interfering effects of the
heterogeneity (Dale, Oefner et al. 2003; Schwartz, Walczak et al. 2004; Kim, Dobransky

et al. 2005).

3.3.1.3 Large fusion partners as protein expression and crystallization enhancer

Large fusion tags such as maltose-binding protein (MBP) and glutathione—S-transferase
(GST) have been proven to enhance the expression, improve the yield and stability, and
facilitate purification of the proteins to which they are fused (Sachdev and Chirgwin 2000;
Skerra and Schmidt 2000; Smith 2000). Recently, several protein crystal structures have
been reported as fusion proteins (Kobe, Center et al. 1999; Liu, Manna et al. 2001; Ke,
Mathias et al. 2002). The target proteins were fused in the C-terminus of E. coli MBP
and the whole fusion proteins were used for crystallization without cleaving off the MBP.
With a modified linker between the MBP and the proteins of interest, the presence of the

MBP did not interfere with the native structures of the target proteins, as indicated by the

90

crystal structures. Several advantages of co-crystallizing a protein with fusion partner
have been recognized. First, large fusion tags such as MBP enhance solubility and
stability of target proteins by avoiding formation of inclusion bodies in E. coli. Second,
since the crystal contacts of the fusion proteins are dominated by MBP/MBP or
MBP/protein interactions as indicated by the crystal structures, the fusion partner could
facilitate crystallization by increasing those contacts. Third, the conditions used to
crystallize the native MBP and the crystal contacts found in the native MBP crystals may
be used to guide the crystallization of the fusion protein. Last, but not least, the three-
dimensional structure of MBP can be used as a search model to solve the crystallographic
phase problem by molecular replacement.

One challenging aspect of co-crystallizing fusion proteins is modification of the linker
between the tag and the target protein. The characteristic that is shared by most
successfully crystallized firsion proteins is a short rigid connection such as three alanines,
instead of a long flexible linker. Shorter linkers may help avoid the conformational
heterogeneity introduced by flexible linkers. Since most proteins that crystallized by
fusion protein method are around or less than 10 kDa, there may be a size limitation of
co-crystallization with MBP. It is possible that the larger volume occupied by a large
target proteins may interfere with any crystal contacts made by MBP.

An alternative large fusion tag is the monomeric DsRed (Invitrogen), which is an
engineered mutant of a red fluorescent protein from Discosoma sp. reef coral. The
tetrarneric form of native DsRed is not a suitable crystallization carrier because fusion
with an oligomeric tag may result in a chimera protein with a non-native quaternary

structure of the target protein. However, monomeric form of an engineered DsRed is less

91

likely to affect the native quaternary structure of the target protein. A significant
advantage of DsRed is that the red color of the fusion protein can be visualized directly,
so that the protein expression and purification process can easily be monitored. Moreover,
the red color can also be used as an indicator to differentiate the protein crystals from
those of salts, which is a common problem associated with crystallization microbatch

evaporation method.

3.4 Experimental procedures, results, and discussion on DLP-l
3.4.1 Fragments conatruction

If fiill-length proteins have problems on expression or crystallization, a traditional
approach is to investigate the individual protein domains. Usually single domains or
domain combinations are more readily expressed and crystallized, given the fact that they
do not have the heterogeneity caused by motion between domains as in full-length
proteins. Crystal structures of single domains or domain combinations of a protein can
also give useful information on protein functional mechanisms. To determine the domain
boundaries, combined with previous literature, secondary structure prediction and
sequence alignments were used to ensure that conserved regions and secondary structures

around domain boundaries are not interrupted.

3.4.1.1 Secondary structure prediction

Protein sequence was analyzed online by PSI-PRED and SABLE protein secondary
structure predication severs, and the predictions were compared visually (Fig. 3.1). The
GTPase domain of DLP-l is consisted of combinations of helices and strands which is

consistent with previous crystal structures of other dynamin GTPase domains (Prakash,

92

Renault et al. 2000; Niemann, Knetsch et al. 2001; Reubold, Eschenburg et al. 2005).
After the GTPase domain, the structure appeared to consist of a collection of helices
connected by loops. The middle and the GED domains are comprised of consecutive long
helices which may needed for the intra- or intermolecular interactions. The PH-like
domain does not contain specific secondary structures but only loops connected by two

short helices, which indicate that it may be structurally disordered.

3.4.1.2 Sequence alignments

The protein sequences of Apis mellifera dynamin-like protein, Caenorhabditis elegans
dynamin-like protein, Homo sapiens dynamin-like protein, and Saccharomyces cerevisiae
dynamin-like protein were analyzed online by T-Coffee multiple sequence alignment
serve, and the results were visualized by the ESPript server. The GTPase domain, the
middle domain, and the GED are very conserved over the four species (Fig. 3.2).
However, no conservation was found in the PH-like domain, again suggesting that this

region may be flexible.

93

DLP 1 secondary

 

 

 

HELIX it. 11“! 1. sum: lira «a Law, ‘ll l 3 l1" 2 u n M ”.24” In I ' 1| fl
PflPRED l ’
SABLE . ‘
IWHleHWH“HWMltPHWWWHWH“HWPWHWHWWPHWH“HWHIPWNWli
l 50 100 200 300 400 500
STRAND _
PaPRED . ]
SABLE lg a
PWIWHWWVHWWWWWHWWNWHTWWIWWWHWIWIHFWWWHWWNWW
LOOP
glifit S FERRET-7F mum .47.: tam-bl» [granular are ’
PflPRED ‘ ’ ' 9 j
SABLE j

    

'l'l'll'll'l'l'l'l‘l‘l'f'l'l‘l'l'i'l‘l'l'l'l'l‘l‘ "I‘l'l'l'l'l‘l‘ 'l‘l'l'l‘l'lil'l'l'l‘l'l' 'I‘l'l'l'l'l l"lll'l'l'l'll'fll'l'l'l'lg

Figure 3.1: Comparison of secondary structure predictions of DLP-l from PSI-
PRED and SABLE severs by visualizing the predictions. Sequence numbers are
indicated below HELIX comparison. Red color, high possibility of being an indicated as
secondary structure; Blue color, low possibility of being an indicated as secondary
structure. Other colors, possibility between indicated by the red and blue colors.

94

Figure 3.2: Multiple-sequence alignments of DLP-1. Am DLP-1, Apis mellifera
dynamin-like protein; Ce DLP-1, Caenorhabditis elegans dynamin-like protein; Hs DLP-
l, Homo sapiens dynamin-like protein; 8c DNM-l, Saccharomyces cerevisiae dynarnin-

like protein; These sequence alignments were generated by using T-Coffee and ESPript
severs.

95

 

 

 

 

 

   

 

 

 

 

 

19 29 39 49 59 59
awn-1 .. ""',»v'r«1 . \ 19"? -519"??? L'r‘s'mr vr s: ’i - war!
CoDLP-l . . . ,1vvi .. Li: w g,Iv .v‘ s a. E'VL, a l r- ' ~13
nuns-r ... :vr 3 ~ wtn _ 91v .v r ,s: jVL s 3 L1 '15
sww1 3’,-...: TVIJJ . F fill-’1; . 15516334.: 25:33.; - . ‘l ’1.) a...“ g H) ., ramvi'l
rmonp<1
CoDLP-l
HsDLP-l
scmmka

 

 

 

   

 

   

 

 

  
       

 

299 299 299 399 319
“$653114 .r Efrain- ’L‘filifiWii f.- “E!
.u -L1VDAV’ .r Lassa, ~L aavsxu .AJo RI
xfsvrpsr .x ELAqu/ Ljrarnvn Ea‘pl
51411115555 ‘- r. .1535 .- . timSLLllfli 50
330 349 359
Cebu-11,1113 Atari. :{st 2 AI],
sane-1 '1 sarLi Lxrx aura! '71: AI 9
56030-1 -5M_.5V_L4§__v 455545.355 55.14.4455-
409
AmDLP-l 1r ,.a arr;'?7‘7i"""-
CoDLP-l - .‘ ‘ _:Af
HsDLP-l - ‘a.
scam-1 ‘ i$3.." MEWSWTM

 

 

 

AmDLP-l
CoDLP-l
flaDLP-l
ScDflM-l

' runs vs v KB
:uva,vr L,ax
L1 344.51 E3551.“

 

$39 $40
AmDLP-l “
CaDLP-l
flsDLP-l
ScDIl~1

 

s29
AmDLP-l
ClDLP-l
BIBLE-1
ScDNN-l

 

    

, 979
- . -' '..'it"i‘£’.“§'i
' m‘NLuszfifi

AnDLP-l
CaDLP-l
HsDLP-l
scnfll~1

    

 

Axiom-1
CoDLP-l
HIDLP-l
scum-1

 

96

3.4.1.3 Protein expression and purification

Based on the secondary structure predictions and the multiple sequence alignments,
several fragments of DLP-1 were cloned into pLW01 vector and expressed in E. coli.
These fragments are: the GTPase domain 1 (1-307aa), the GTPase domain 2 (1—323 a),
the GTPase domain 3 (1-340 aa), the PH-like domain 1 (PHI, 497-602 aa), the PH-like
domain 2 (PH2, 523-602 a), the GTPase + Middle domain (1-500 a), the Middle + PH-
like domain + GED (330-710 a), and the PH-likc domain + GED (500-710 aa) (Fig. 3.3).
The cloning, expression, and purification procedures are as described for DLP-1 full-
length protein. The isolated GTPase domain 1, 2, 3, the PHI, and 2, were expressed and
exhibited high or moderate solubility. However, other fragments were found exclusively
in the low-speed pellets during purification, presumably inclusion bodies, in spite of
various growth conditions. Attempts of solubilizing some of these fragments using
denaturants from inclusion bodies failed.

Among the three isolated GTPase domain fragments, only the 1-307aa forms stable
monomers (see chapter 2), while the other two fragments aggregate as measured by gel
filtration. As discussed in chapter 2, the region from residue 308 to 340 of DLP-1 may be
involve in the formation of helices bundle in the middle domain and inclusion of these
residues into the isolated GTPase domain may cause the hydrophobic regions exposed to

solvents and lead to protein aggregation.
3.4.1.4 Crystallization

Both PH-like domain fragments forms stable tetramers and remain soluble even

concentrated to 50 mg/mL. Crystallization of the GTPase domain 1 and the PH-like

97

domain 1 and 2 were performed with procedures described before. However, no crystals

have been obtained, to date.

3.4.2 Limited proteolysis

Since DLP-l contains multiple domains, and the fiill-length protein failed to crystallize,
limited proteolysis is applicable to this protein. The fastest way to perform limited
proteolysis is the in-drOp proteolysis (Gaur, Kupper et al. 2004; Johnson, Roversi et al.
2006). Different kinds of proteases such as trypsin, chymotrypsin, papain, and proteinase
K were added to the solution containing the purified full-length protein before
crystallization screening. Ideally, if there are crystals grown from the drops, they will be
re-solubilized and subjected to N-terminal sequencing and mass spectrometry for
sequence determination (Gaur, Kupper et al. 2004). However, there was no crystal
observed from any screening conditions. Although the in-drop method is easy to
manipulate, one of its disadvantages is that the heterogeneity resulting from the presence
of other digested fragments may inhibit crystallization. Therefore, larger-scale limited
proteolysis is necessary.

With larger-scale limited proteolysis, the full-length DLP-1 was digested with the
proteases listed above for up to 8 hrs. At each time point, digested products were sampled
for SDS-PAGE (Fig. 3.4). A common proteases-resistant band of about 50 kDa was
observed from the SDS-PAGE indicating that this fragment may be a compact individual
domain or domains combination. The protein band was transferred by electroboltting to
PVDF membrane and sent for N-terminal sequencing. The final digested product was

mixed with formic acid for mass determination using mass spectrometry.

98

The N-terminal sequencing showed that the first five amino acids of the fragment are
ENGVE which starts from the residue number 82 of the protein. The mass spectrometry
obtained a major peak of about 49 kDa (Fig. 3.5). Combined with both sequencing and
mass results, the fragment was identified to be from residue 82 to 516. The N-terminus is
around the second motif of the GTPase domain and the C-terminus is in the beginning of
the PH-like domain. Both regions may be flexible and solvent exposed such that they are

easily accessed by proteases.

 

C GTPase
C GTPase
C GTPase
< GTPase > 1-340

 

 

 

 

 

497-602
523-602

  
 

 

C GTPase M1 316
3 3 0-7 1 0 Mlddl l: ‘

 

Figure 3.3: Schematic illustration of DLP-1 constructs.

The identified fragment was cloned into pLW01 expression vector and recombinant
expressed in E. coli BL21 (DE3) cells as described for wild type DLP-l and other
fragments previously. Since the N-terminus of the fiagment starts from the middle of the
GTPase domain, truncation of the GTPase domain may destroy the protein structure.

Therefore, another fragment encoding residues from 1 to 516, which includes the entire

99

GTPase domain, was also cloned. However, both fragments were expressed in inclusion
bodies in regardless of application of methods to lower the protein synthesis rate during
expression.

One possible reason that these fragments are not soluble may be the disruption of the
intra- and intermolecular interactions. It is purported that the GED folds back to interact
with the middle domain; the results from chapter 2 support this idea. The cleavage in the
flexible PH-like domain removed all the C-terminus including the GED. Therefore, the
middle domain could no longer interact with the GED, leading to disrupted intra- and
intermolecular interactions and an increased exposure of the hydrophobic regions of the
middle domain. It is necessary to consider a way to balance the maintenance of critical

interactions, while removing some protein flexibility.

100

DLP_1 by Chymotrypsm DLP__1 by Trypsin

      
   

M 0 10 30 60 120 240 M 0 10 30 60 120 240

DLP_1 by Proteinase K Mrnutes

191

51
39
28
19
14

 

I, gag-5...: rhiai

M 0 10 30 60 120 240

Minutes

Figure 3.4: SDS-PAGE analysis of Limited proteolysis of DLP-l by different
proteases. Purified DLP-1 and protease were incubated at room temperature at a ratio of
100021 (wzw). The reaction was stopped at each time points by adding PMSF. Proteolytic
products were resolved by SDS-PAGE. Arrows indicate similar size of proteolytic
products.

101

 

 

 

 

 

 

Jana—us...“ .afimmmohuomaoaen 05 .8 03m 2: 353%.: de— av 2:55 .3 on: a 2.55
0.13:. 95 5 33a «8.33 2E. Stun—958% 82: .3 Evans: ash—83.:— defl 2: we nowaifiuofiu 8a: "Wm 0.5»:—
éoowm 880 803.... 880 800m 88... 88v 083 88* 88¢
82 ...P... .....p. ...._-..... .r»..p?_..u.—..Lul—-..._. ~p..._..»~—.u.»h... .y..._...._t...~....— ...—0
r , ,
_ ,a _ , _ m .4 m : v
38% . . _ .
3.8% m 8»? auto . ,. , , o mum:
. m 88 339. 388 .
ONDGMD 0.809 mdflnfl‘
934$
0.9.55. _ odor:
reto— ,
35$ 089m . Q88:
0.33m. z.
. r %
08%..
\
3.23 /
mNamaV

\
. Qwfibv
hm.» .9323 - .8.

 

 

102

.QmmmBuodfin38.52355} unv 53m ~532on souoscom ~38“:an oom‘tzw war“: Baacocow 33 5326on mg donnNEwao
5 Eu 9 mosEmoh .3095 32 9. 5:858 no.“ 332650 Eu #2.: $0520 .3955 nwE 8865 8:: 520 .2662 some B.“
888 .3095 £30 2% 2: .mca.> ”FED mo 9598:: 2668 £53m Amos: :8th .3955 3:892:328 :3: 5E 282:0
9:2on 5:2: “53:8 Ana @5453 52:3. 85.5m 2: 2:: ”326%.: 7.59 .3 E5269:— thuouoontam 6d «Earn

 

 

 

 

 

 

 

 

NWN 3.8 _.N Nod mmN

 

 

 

.3me

103

3.4.3 Surface-entropy reduction

Surface-entropy prediction of DLP-1 revealed that the high ratio of charged and
flexible residue clusters in the PH-like domain causes high surface entropy (Fig. 3.6).
Secondary structure predictions showed that the PH-like domain does not have a regular
secondary structure and probably forms a large solvent-exposed loop (Fig. 3.1). Sequence
comparison of DLP-1 with the crystallized hGBPl showed that hGBPl lacks the PH-like
domain between the conserved middle domain and the GED (data not shown). These
analyses indicated that the PH-like domain of DLP-1 may be highly flexible causing an
energy barrier for crystal formation. Therefore, combined with the previous limited
proteolysis results, replacement of the PH-like domain with a shorter, but less mobile
linker by protein engineering may be a way to enhance the likelihood of DLP-1
crystallization. Whatever our designed construct is, it should capture the domain-domain
interactions between the GED and the middle domain, and the GED and the GTPase
domain in the crystal structure.

The X-ray crystal structure of hGBPl showed that the GED forms a long helix and
interacts with the middle and GTPase domain (Prakash, Praefcke et a1. 2000; Prakash,
Renault et al. 2000). For DLP-1, biochemical analyses supported these interactions (Shin,
Takatsu et al. 1999; Zhu, Patterson et al. 2004). If the PH-like domain of DLP-1 is a
flexible region that is recognized by proteases, the fragments obtained by limited
proteolysis method would likely to be the GTPase domain plus the middle domain, and
the GED alone, but not a combination of both. Therefore, replacing the PH-like domain
of DLP-1 may be the best way to enhance the likelihood of crystallization as well as

obtain the structural information of the domain-domain interactions particularly that of

104

the middle domain and the GED. Furthermore, once the domain interactions are met, it is
also possible to reduce the formation of inclusion bodies as described for the limited
proteolysis fragment.

The DLP-1 APH was designed, cloned, expressed, and purified, as described in chapter
2. The protein was less soluble in vivo than the wild type mainly because of lacking of the
hydrophilic PH-like domain. The solubility problem was solved by expressing the protein
at low temperature (18 °C) and low IPTG (0.02 mM) induction. At least 1 mg purified
protein was obtained after the final step of purification from 1 L bacterial culture.
However, no crystals have yet been observed with the screening conditions as described

above.

105

Figure 3.7: Illustration of designed RMG series expression vectors. (A) pRMG-
pfuMBP; (B) pRMG-pfuMBP-cZX; (C) pRMG_ecoMBP; (D) pRMG_DsRed_M.

106

 

Figure 3.7 continued

 

:3: ”WWW , ms TGA

ru- \/~’\]’\

pRMG-pfuMBP

 

Basal vector:
pLW01

.—

107

Figure 3.7 continued

 

MCS TGA

 

 

 

B s'pR MG-PfuMBP_ czx'

Basal vector:
pMAL-c2X

(N I33)

108

Figure 3.7 continued

 

 

rTEV(BamHI) “639$ DTG

    
    

pRMG-EcoMBP

Basal vector:
pLW01

109

Figure 3.7 continued

 

ATG- MCS Cm. pro/x

 

 

«fax

D ‘ pRMG-DsRed_M

\

Basal vector:
pLW01

3.4.4 Fu_sion protein_s for crystallization aids
3.4.4.1 Development of expression vectors for P. fim'osus MBP

The Pyrococcus furiosus (pfu) MBP was showed to be a more efficient solubilizing
partner than the E. coli MBP (Fox, Routzahn et a1. 2003), although currently it is not
widely used for protein solubilization and co-crystallization. With the crystal structure
available, the pfu MBP would serve as an alternative choice of protein expression and
crystallization enhancer (Evdokimov, Anderson et al. 2001).

The pfir MBP was cloned into pLW01 and pMAL-c2X vectors creating
pRMG_pfuMBP and pRMG_pfuMBP_c2X expression vectors, respectively (Fig. 3.7A
and B). Briefly, the gene encoding the pfu MBP was amplified from genomic cDNA by

PCR. The reaction was comprised of 35 cycles of denaturing at 94 °C for 30 s, annealing

110

at 55 °C for 45 sec, and elongation at 72 °C for 1 min and 20 sec followed by 72 °C for
10 min. For the pRMG_pfuMBP vector, the PCR products were digested with restriction
enzymes Afl III and BamHI, and the pLW01 vector was digested with NcoI (compatible
with Afl III) and BamI-II. For the pRMG_pfuMBP_c2X vector, both PCR products and
the pMAL-c2X vector were digested with NdeI and BamHI. The digestion products were
ligated and transformed into E. coli DH5a competent cells. Positive clones growing from

LB plates containing antibiotics were picked and sequenced.

3.4.4.2 Development of expression vectors for E. coli MBP

The E. coli (eco) MBP was cloned into pLW01 vector creating pRMG_ecoMBP
expression vector (Fig. 3.7C). The detailed procedure for vector construction was similar
as described for the pfu MBP. In the C-terminus of MBP, a three-alanine linker was added
by modification of restriction enzyme NotI for co-crystallization purpose. Another
advantage of this vector is that an rTEV site with a BamHI site was inserted behind the
linker. If only the target protein is wanted for fimction-structure studies, the TEV protease
can be used to cleave off the MBP and the linker, leaving entire target protein. With this
vector, researchers can choose different restriction enzyme sites depending on whether
the purpose is to co-crystallization or protein solubilization, or both. For instance, if only
for solubilization enhancement was desired, the BamHI is utilized so that the MBP could
be cleaved after purification. For both co-crystallization and solubilization, NotI site
should be used because the linker between the MBP and target proteins will be a short

three-alanine stretch.

111

3.4.4.3 Monomeric DsRed

The monomeric DsRed was cloned into pLW01 vector using NcoI and BamHI sites
creating pRMG_DsRed_M expression vector (Fig. 3.7D). The gene encoding DsRed-
monomer was amplified fi'om stRed-Monomer vector (BD Biosciences). The

procedure for vector construction was similar as described for the pfii MBP.

3.4.4.4 More applications of the expression vectors

Besides the advantages described before of using the expression vectors such as
enhancing protein expression, solubility, and co-crystallization, guiding crystallographic
phase, and monitoring protein purification by color, fusion target proteins with MBPs or
DsRed can be applied to other research efforts such as identifying protein-protein
interactions in vitro or providing diagnostics for a protein’s oligomeric state.

Protein-protein interactions can be determined in vitro by immobilizing one protein
onto a column matrix and letting the putative partner flow through the column. If the two
proteins interact, the second protein will bind to the column. Upon elution of
immobilized protein, the putative partner can be detected by SDS-PAGE or Western
blotting. With our vectors, we can fuse one protein with MBP and a second one with
DsRed. The MBP fusion will be immobilized onto an amylose column, and binding of the
second DsRed fusion protein will turn the column red. For a more quantitative
measurement, the fluorescence of the elution fi'actions can be monitored to detect the
amount of bound partner.

Since both the MBPs and the DsRed are monomeric, they can also be used to diagnose

the target protein oligomeric state. Several crystallographic studies have showed that

112

MBP does not affect the target protein oligomeric state, and the target proteins are the
major factor drives fusion protein to oligomerization if the target proteins are oligomeric
(Kobe, Center et al. 1999; Liu, Manna et al. 2001; Ke, Mathias et al. 2002). Therefore,
for proteins with poor solubility that could not be purified without solubilizing aids, we
can fuse them with the MBPs or the DsRed to increase the solubility. And the oligomeric
states of the purified fusion proteins will reflect those of the target proteins.

Researchers who have used our expression vectors in their research have obtained
quite successful results. For example, when attempting to measure the ATPase activity of
a protein target in vitro, the Benning group could not obtain recombinant protein because
of the poor solubility. By fusion the target protein with MBP, the fusion protein was
soluble enough to be purified and the activities were measured (Lu, Xu et al. 2007). The
Benning group tried to examine phosphatidic acid (PA) binding by a putative lipid
transporter component, but solubility problems and non-specific lipid binding plagued
most fusion constructs. When the pRMG_DsRed_M vector was used to fuse DsRed with
the target protein, the resulting fusion protein had improved solubility and made the
determination of PA binding much more facile (Lu and Benning 2009). Another example
of a success application for these expression vectors is the use of the pRMG-ecoMBP by
Thines et al. (Thines, Katsir et al. 2007) to express recombinant plant proteins for

protein-protein interaction assays.

3.4.4.5 DLP-1' fusion proteins expression, purification, and crystallization

The DLP-1 GTPase domain, the PH-like domain, and the GED were cloned into the

pRMG-ecoMBP vector. Since the purpose of these experiments were to co-crystallize the

113

fusion proteins, the NotI and XhoI restriction enzyme sites were use for cloning by which
the linker between the MBP and the target proteins was the three-alanine. The GED
fusion protein appeared exclusively in inclusion bodies, while other two fusions are quite
soluble. As there is 6x-His tag in the C-terminus of the fusion proteins, the Ni-NTA
column was used for protein purification. Although the amylose column is often the
preferred means for purification tool, MBP-fusion proteins may sometimes display
weaker binding to the amylose column, which causes protein loss in the washing fi'action.
Moreover, MBP binding to the amylose column is also markedly diminished in the
presence of detergents. Thus, the C-terminal 6x-His tag ensures a means to rapidly purify
the fusion protein.

Both the PH-like domain and the GTPase domain fusion proteins were further purified
to near homogeneity by ion exchange chromatography. No crystals for the PH-like
domain and the GTPase domain fusion proteins have yet been observed in the initial
rounds of crystallization trials. The DsRed fusions of the same fragments exhibited
similar behavior in crystallization trial as those of MBP fusions. A number of factors may
have led to this situation. The GTPase domain may be a little too large for co-
crystallization with MBP, such that MBP may not dominate in the formation of ordered
crystal contacts. For the PH-like domain, which is of protein size suitable for co-
crystallization, the domain may be too flexible for crystallization. Another explanation is
that the linker lengths and conformation may not yet be suitable for crystallization.

Additional experiments are being attempted to explore this situation in greater depth.

114

3.4.5 Homoggv model of the GTPase dorm of Dlil
3.4.5.1 Experimental procedures and the model

If protein crystal structures are not available, one can predict the 3-D protein structure
of a target protein empirically using homology modeling. A homology model is useful for
fimctional predictions and the design of mutagenic experiments. The bottleneck for this
experiment is the availability of a suitable crystal structure that is highly homologous to
the target protein. Without high protein sequence homology, the model will not be
reliable, particularly in the non-conserved regions. The crystal structure of the GTPase
domain of rat dynamin was solved and the domain shares greater than 70% sequence
homology with the GTPase domain of DLP-1 (Reubold, Eschenburg et al. 2005). Thus, it
is suitable to make a homology model to get a general overview about the 3-D structure
of the DLP-1 GTPase domain.

The protein sequences of the GTPase domains of DLP-1 and the rat dynamin were sent
to hm;://proteins.msu.edWsewersmomologlrnodelingserve/construct homology PDB.
The initial model was modified by Pymol software to visualize conserved residues. The
GTPase domain model is a compact core containing seven helices and eight strands. The
sheets are inside the core and surrounded by the helices. Six sheets are parallel and two

are anti-parallel (Fig. 3.8).

115

   

A B

Figure 3.8: Homology model of the GTPase domain of DLP-l. This model was
created using the crystal structure of the GTPase domain of Rattus norvegicus dynamin 1
(PDB: 2AKA) as a template. (A) Front view. (B) Back view. In B, identical residues are
highlighted by showing the side chains.

3.5 MFNs experimental procedures and results
3.5.1 Protein expression and purification

The DNAs encoding for human MFN—1 and MFN—2 were amplified from cDNAs
IMAGE (clone ID 5270347 and 3901235, respectively) by PCR. The PCR amplification
was comprised of 35 cycles of denaturing at 94 °C for 30 sec, annealing at 55 °C for 45
sec, and elongation at 72 °C for 2 min and 30 sec followed by 72 °C for 10 min. The
PCR products were purified by QIAquick PCR purification kit (QIAGEN), digested with

restriction enzymes BamHI and XhoI for ligation with pRMG-N-F LAG vector, and Ncol

116

and BamHI for pRMG-C-F LAG vector. The digestion product was ligated and
transformed into E. coli DHSa competent cells. Positive clones growing from LB plates
containing 100 ug/mL ampicillin were picked, and plasmid DNAs were isolated and
sequenced.

The sequenced plasmids were transformed into expression host E. coli C41 (DE3)
competent cells. Fresh single colonies from selection plates were inoculated into 100 ml
LB media containing 100 ug/mL ampicillin at 37 °C with shaking at 200 rpm overnight.
Twenty mL of this culture was transferred into 1 L fresh LB media and the cells were
grown at 37 °C to an A600 of 0.8-1.0. The cells were then induced by adding 0.05 mM
IPTG and incubated with shaking at 18 °C for 36 hrs. Cells were harvested by
centrifugation and stored at —80 °C.

To purify MFN-1, cell pellets were resuspended in Bufl‘er A (50 mM sodium phosphate,
300 mM NaCl, 250 mM sucrose, 10% glycerol, 10 mM B-mecaptoethanol, pH 8.0), and
the crude cell extract was sonicated and centrifuged at 4 °C for 20 min at 12,000x g (low-
speed centrifugation). The supernatant was further centrifuged for one hour at 45,000x g
at 4 °C (high-speed centrifugation) to separate the cell membrane fragments from the
soluble proteins. The supernatant of the high-speed centrifugation was saved for SDS-
PAGE analysis, and the pellet of high-speed centrifugation were re-suspended and
incubated in Buffer A with 1% detergent (octyl glucoside or dodecyl maltoside) at 4 °C
for one hour. After another high-speed centrifugation, the supernatant was loaded onto a
Ni-NTA column. The protocol for column wash and elution was the same as described for
DLP-1 except that the all buffers contain 0.1% detergent. However, less then 0.1 mg of

MFN-1 was obtained from 1 L culture, which is an insufficient amount at this time for

117

structural studies. SDS-PAGE and Westem-blot analyses have shown that the MFN-l was
produced at high levels, but present in the pellet following low—speed centrifugation,
probably inclusion bodies (Fig. 3.9). Since most inclusion bodies were formed because of
high rate of protein synthesis, protein expression conditions such as growth temperature
and IPTG induction, went through more optimization. However, the systematic alteration
of growth and induction conditions failed to produce soluble MFN-1. Attempts of using
nonionic detergents to solubilize MFN-l from inclusion bodies were unsuccessful.

The low amount of MFN-1 purified from the low-speed supernatant was loaded to
analytic gel filtration for size determination. The procedure was similar as described for
DLP-1 except that the running buffer contains 0.1% detergent. Most protein was eluted in
the void volume, and some were in a later peak, which corresponds a tetrarneric size of
MFN-1. The aggregation problem can be solved by adding 10 mM DTT into elution
buffer and gel filtration running buffer (Fig. 3.10), indicating that the aggregation of
MFN-1 is partially caused by disulfide bonds. The MFN-2 expression was not detected

under a number of different expression conditions.

3.5.2 MFN fragment construction
3.5.2.1 Secondary structure prediction and sequence alignments

The secondary structure prediction (Fig. 3.11) and sequence alignments of MFN-l and
-2 were performed, as described for DLP-l. Like DLP-1, the secondary structure of
MFNs is comprised of combinations of helices and strands in the GTPase domain,
exclusively long helices in the HRl and HR2, and a region predicted to be

transmembrane segments. The major difference is that there is a long helix in the N-

118

terminus before the GTPase domain that is functionally~ unknown. The crystal structure
of a bacterial dynamin-like protein (BDLP) shows that this helix is involved in the
formation of helix bundles with the HRl and HR2 (Low and Lowe 2006).

The protein sequences of Saccharomyces cerevisiae dynamin-like protein, Homo
sapiens mitofusin-2, Danio rerio mitofusion—l; Homo sapiens mitofusin-1; Xenopus
laevis mitofusin; Caenorhabditis elegans fuzzy onions protein (homologue of human
mitofusins) were used for multiple sequence alignment (Fig. 3.12). The GTPase domain,
the HRl, and the HR2 are highly conserved over these species. The only less conserved

region is the transmembrane domains.

Purification of MFN-1

   

    

103 100
81 80 100
47 75
60 LP LS/N
MFN-1 MW 50
MW MFN] MW LS/N LP
(A) (B) (C) (D)

Figure 3.9: Expression and purification of MFN-1 in E. coli. The arrows indicate
MFN-1. (A) and (B) SDS-PAGE and Western-blot analysis of purified MFN-l. (A) SDS-
PAGE analysis. (C) and (D) SDS-PAGE and Westem-blot analysis showing that most
MFN-l is present in pellet of low speed-centrifugation, presumably inclusion bodies. (C)
SDS-PAGE analysis. MW, molecular weight markers; LS/N, supernatant of low-speed
centrifugation; LP, pellet of low-speed centrifugation; (D) Western-blot analysis using
anti-6Xhis antibody.

119

3.5.2.2 Protein expression and purification

Based on the secondary structure predictions and the multiple sequence alignments,
several fi'agments of MFN-1 were cloned into pLW01 vector and expressed in E. coli.
These fi'agments are: the GTPase domain (1-352 a), the GTPase domain + HRl (1-572
a), the HRl + TM + HR2 (354-741 a) (Fig. 13). The cloning, expression, and
purification procedures are as described for DLP-1 fragments. Unfortunately, all
fragments were expressed exclusively in inclusion bodies, in spite of various adjusted
growth conditions. Attempts of solubilizing some of these fragments using denaturants

from inclusion bodies were not successful.

120

Absorbance at 280 nl (mAU)

 

‘I I I I
0 5 10 15 20 25 30
Rotontlon volume (ml)

 

Figure 3.10: Size exclusion chromatography of MFN-1 by a Superdex 200 10/30
column. (A) chromatograph showing the tetrameric form of MFN-l in the major peak.
(B) Western blotting of the major peak fractions by anti-F LAG antibody. The protein
molecular weight was calculated as for DLP-l described in chapter 2.

121

Figure 3.11: Multiple-sequence alignments of MFN-l by T-Coffee and ESPript. Hs
MFN-2, Homo sapiens mitofusin-2; Dr MFN-1, Danio rerio mitofusion-l; Hs MFN-1,
Homo sapiens mitofusin-1; X1 MFN, Xenopus laevis mitofusion; Ce on, Caenorhabditis
elegans fuzzy onions protein (homologue of human mitofusions).

 

Helix-2
DrflrN-I
BsHFN-l

Gotta

HAHFN-Z
DrHFN-l
BIIIN-l
XIHFN
C0320

Hsflffl-Z
DrMFN-l
Benin-1

CGFIO

fisMFN-Z
DIMFN-l
Benin-1
xluru
Garza

asHFN-Z
DrMFN-l
HBMFN-l
XII?!
Ccrzo

BBHFN-Z
DzHFN-l
nears-1
XlflPfl
Cetzo

Annex-2
annex-1
BAHFH-l
11!!!
Gotta

HsMFN-Z
DrHFN-l
Bantu-1
X1HFN
CoFto

asHFN-Z
DININ-l
HBHFN-l
x1urn
Cetzo

BAKER-2
DrHFn-l
HAMFN-l
XIMFN
Cbon

usurn-z
DrMFN-l
aeMFN-l
XIHFN
Cero

 

 

 

   

 

1 19 29 39 59 c9
-\; ............ a/\/ i . \;§Tfii' . . “ioz-“ £i*i
.............................. n-. 5_=x ‘ . ‘3'2-vo; ‘ VA H
.............................. pan.-.a ,A ~ VI“ 5* MA A
.............................. I IAL”8 A “Vie‘v"WI a
:g- A: :~ ..351 :x a x xx:« A..n‘\gQQn 2 ug_ ....fig 3
79 89 100 110 120 130
9173 in? .i :W FA‘ A“, 'rvu'A’“, pn’ms'rrm It;
I: a ugmut .1. II N; IV A .3 9A .
I! In 'IATI.‘ pg: .sm 8v A. DA 3 I I5
Q! :: \vsr ._-/:I . A 3 HIV 8,1.DK .r:
149 159 149 179
a . 2.532%! $3 5 ixAhAgrv unaffifl .5
x ; RA -IA 3 33:31 TV QLA 9* 'b
a "!A 'IA 3 nxxsv rv QLA i *n~ ...
a " BN‘ 'tA g: axxsv rv out1F 3 *P'-'
'!Q 'r a
209 219 229 239 249 239 269
9, j _ “., .9

 

 

 

 

 

 

 

 

 

     

 

         

   
     
   
 

. 299 W309 319 329 339
_ i '1 1W.” .ifi'dfir._. -im ‘9
«i. In an light I jun, 91H:%".:lff9‘:: ..... aL, 5 .
H , ‘Ej-dlb‘lj.‘ El EAR? ISAj' IL 9”“ 1L “1
o; is,” ...... . m o .... 1
lg- buunaai$§%nd3‘iig -.AJ§313 J "Luué-
349 360 370 380 3.9 609
- jifififi '. WP”““Q”“ A3 6?? by i Iunsfl ? fl fifirfi;r§§<
f"' A .... Q IIA 91: av: Innorx v5.3 In .‘ A1
.A- Q .-L Q rIAfi 91L IVA .,Innsv. -‘=xl t: q 1:“
L—:K A 9.} ,9 rIr or: AVA Iuan. .<,rw xgx; u:
2.3.2.5 - -. i m m u- .rruu I v9.92
419 429 449 459
1’6- mil ~-+ 323va iii-" _ .. .
xg.xmnfi, Mell’ .va 1‘ njsvzré I ;.,:. :
'“0 Kit ..Vfl”h¥’ .V *1 Ast v -:-x«’
a “9 an 9 . a _9, q sLLv»r;;a’
VI'-h1:?97g';£‘ Lfirwmb I-'£i-
499 490 509
"Mini! ~ I m $655 3313‘ '5"
’ucxzmx ,‘s sMuhsu. 5999 II 11 A,
non LA ,-r -M ALA 99A II.L' L
ucn~IA Eusxanasn. can 1&‘l I?
an AA . r way» an to a “3.2

  
    

Vii? Efisniifi AWFmfivix
LIAV. IdsvrsA nsvrr; jvvanrv.
Irnv A svrsA ‘AGIII .VI If! ,
havu L§ snrsA ,ucvrv; -;vv irV‘ ALI
ms aflazm - mm . v‘ w

609 739 749
mm“. flail-n m
Itnquzxq .. « unvo- a van I‘
ulnauxvnr can; a A29: .3 Inn A
oxnxurva. ~ =

 

198.:I‘VLI ..
aaunavana:. :1 EIIAL bar: :5

Figure 3.11 continued

I:

3

I”

H
aaavw
qubA
Hwani

The full-length MFN-l and MFN-2 proteins were also cloned into the pRMG-ecoMBP
vector using BamHI and XhoI sites. The major purpose of this experiment was to
increase the solubility of MFNs. The solubility was increased but not good enough to get
purified proteins. Since MBP is mostly used with soluble proteins, membrane proteins

may not be good candidates for MBP fusion.

MFN_I _Secondary

HELIX ll rapi'ubnjiu it; a: with“. E :a ’l h-ma‘! Mic II '-.' t.

PflPRED

SABLE

 

l‘l‘l'l‘l'l l‘i‘l'n’l'l'l‘l'l‘lri'l‘l'l'.‘l‘l‘l'.‘l‘l'l'l'l'l'l‘l‘l'l‘l'l'l'l'l l'l‘l'i l‘l‘l'i'Wl'I‘l'l'l'i'l'l'l l'l'l'l‘l'l‘l'lll'l'l',
l 50 100 200 300 400 500 600 700 74]

STRAND

 

PflPRED

SABLE

   

LOOP

 

"Itnli ’ 7' 7 . ... Hi Amman"

 
 

lkWRED

SABLE . . 3'
'PI'FH‘I'FI l'l'l‘Plrl‘l‘l'l‘l‘lr'l‘l‘lrjl‘l‘l‘I'l‘l'l‘l‘l’l‘lrl'l‘l’l‘l'l'l’n‘l’l'l‘l‘l‘l'i‘l'l'l‘ lI'l‘l lW‘l'lTl‘l'l'l'l’Wl'l

Figure 3.12: Comparison of secondary structure predictions of MFN-1 from PSI-
PRED and SABLE severs by visualizing the predictions. Sequence numbers are
indicated below HELIX comparison. Probabilities that each color represents are as in Fig.
3.1.

124

 

 

 

 

 

 

 

 

 

 

 

Figure 3.13: Schematic illustration of MFN-l constructs.

3.5.3 Protein re-naturation

Previous results have shown that recombinant MFN-1 tended to form inclusion
bodies. To obtain large amounts of well-folded MFN-1 for structural studies, protein re-
naturation methods were explored. However, MFN-1 aggregated in the middle of the re-
naturation process with either guanidine hydrochloride or urea as the denaturation
reagents.

With failure of classical renaturation methods, we generated non-classical inclusion
bodies of MFN-1 to try protein renaturation using ionic detergents. Expressing proteins
at low temperature has been shown to produce readily solubilized “non-classical”
inclusion bodies (J evsevar, Gaberc-Porekar et al. 2005) containing large amounts of
partially folded protein. Instead of using strong denaturants such as guanidine
hydrochloride or urea, the non-classical inclusion bodies can be solubilized by non-
denaturing solvents or low concentration of ionic detergents such as N-lauryl-sarcosine

(sarcosyl). Another reason that sarcosyl was selected as a solubilizing agent is that it was

125

reported to successfully solubilize integral membrane proteins from inclusion bodies
(Bruckner, Gunyuzlu et al. 2003).

MFN-l was expressed in E. coli at 15 °C and the inclusion body was re-solubilized by
0.2 % sarcosyl. For further biochemical analysis, the protein was purified by Ni-NTA
column and sarcosyl was replaced with non-ionic detergent dodecyl maltoside by gel
filtration column. A wide peak in the void volume and a peak of the size of tetrarneric
MFN—1 size appeared in the gel filtration chromatogram. Although the renatured MFN-1
had no detectable GTPase activity, there is no report in the literature of recombinant
MFN-1 exhibiting GT Pase activity in in vitro assays. With the current preparation of

renatured MFN-l, no crystals have yet been observed.

3.6 Summary

A total of at least 1064 screening conditions of precipitants were used for DLP-1 with
microbatch and hanging drop vapor diffusion crystallization methods. Molecular protein .
engineering was designed and performed using limited proteolysis, surface-entropy
reduction, and large fusion tags, to increase the likelihood of protein crystallization. A
total of at least 32 protein fragments of DLP-1 were cloned, expressed, and purified for
crystallization purpose. However, no crystals have yet been obtained. Homology model
of the GTPase domain of DLP-1 were made for structural overview.

MFN-1 was expressed mostly in inclusion bodies. Protein renatured using ionic
detergent method yielded solubilized protein, but in vitro GTPase activity was not
observed. A total of at least 13 full-length or fragments of MFN-1 and -2 were cloned

and expressed with different DNA vectors or E. coli cell lines. However, no crystals have

126

yet been obtained after the initial rounds of crystallization screening.

Nonetheless, several expression vectors containing large fusion partners were designed
and constructed. Enhancing protein expression, solubility, and co-crystallization, guiding
crystallographic phase, monitoring protein purification by color, detecting protein-protein
interactions, and diagnosing protein oligomeric state, are among the advantages of using

these vectors.

127

References

Bruckner, R. C., P. L. Gunyuzlu and R. L. Stein (2003). "Coupled kinetics of ATP and
peptide hydrolysis by Escherichia coli FtsH protease." Biochemistry 42(36):
10843-52.

 

Chayen, N. E. (1998). "Comparative studies of protein crystallization by vapour-diffusion
and microbatch techniques." Acta Crystallogr D Biol Crystallogr 54(Pt 1): 8-15.

Dale, G E., C. Oefner and A. D'Arcy (2003). "The protein as a variable in protein
crystallization." J Struct Biol 142(1): 88-97.

Derewenda, Z. S. and P. G Vekilov (2006). "Entropy and surface engineering in protein
crystallization." Acta Cgstallogr D Biol Cgstallogr 62(Pt 1): 116-24.

Evdokimov, A. G, D. B. Anderson, K. M. Routzahn and D. S. Waugh (2001). "Structural
basis for oligosaccharide recognition by Pyrococcus furiosus maltodextrin-
binding protein." J Mol Biol 305(4): 891-904.

Fox, J. D., K. M. Routzahn, M. H. Bucher and D. S. Waugh (2003). "Maltodextrin-
binding proteins from diverse bacteria and archaea are potent solubility
enhancers." FEBS Lett 537(1-3): 53-7.

Gaur, R. K., M. B. Kupper, R. Fischer and K. M. Hoffmann (2004). "Preliminary X-ray
analysis of a human V(H) fragment at 1.8 A resolution." Acta Crystallogr D Biol

Cgstallogr 60(Pt 5): 965-7.

J evsevar, S., V. Gaberc-Porekar, I. Fonda, B. Podobnik, J. Grdadolnik and V. Menart
(2005). "Production of nonclassical inclusion bodies from which correctly folded
protein can be extracted." Biotechnol Prog 21(2): 632-9.

Johnson, S., P. Roversi, M. Espina, J. E. Deane, S. Birket, W. D. Picking, A. Blocker, W.
L. Picking and S. M. Lea (2006). "Expression, limited proteolysis and preliminary
crystallographic analysis of IpaD, a component of the Shigella flexneri type III
secretion system." Acta Crystallogr Sect F Struct Biol Cryst Commun 62(Pt 9):
865-8.

Ke, A., J. R. Mathias, A. K. Vershon and C. Wolberger (2002). "Structural and
thermodynamic characterization of the DNA binding properties of a triple alanine
mutant of MATalpha2." Structure 10(7): 961-71.

Kim, A. R., T. Dobransky, R. J. Rylett and B. H. Shilton (2005). "Surface-entropy
reduction used in the crystallization of human choline acetyltransferase." Acta
Crystallogr D Biol Crystallogr 61(Pt 9): 1306-10.

Kobe, B., R. J. Center, B. E. Kemp and P. Poumbourios (1999). "Crystal structure of
human T cell leukemia virus type 1 gp21 ectodomain crystallized as a maltose-

128

binding protein chimera reveals structural evolution of retroviral transmembrane
proteins." Proc Natl Acad Sci U S A 96(8): 4319-24.

Koth, C. M., S. M. Orlicky, S. M. Larson and A. M. Edwards (2003). "Use of limited
proteolysis to identify protein domains suitable for struCtural analysis." Methods
Enzmol 368: 77-84.

Liu, Y., A. Manna, R. Li, W. E. Martin, R. C. Murphy, A. L. Cheung and G Zhang (2001).
"Crystal structure of the SarR protein from Staphylococcus aureus." Proc Natl
Acad Sci U S A 98(12): 6877-82.

Low, H. H. and J. Lowe (2006). "A bacterial dynamin-like protein." Nature 444(7120):
766-9.

 

Lu, B. and C. Benning (2009). "A 25-amino acid sequence of the Arabidopsis TGD2
protein is sufficient for specific binding of phosphatidic acid." J Biol Chem
284(26): 17420-7.

Lu, B., C. Xu, K. Awai, A. D. Jones and C. Benning (2007). "A small ATPase protein of
Arabidopsis, TGD3, involved in chloroplast lipid import." J Biol Chem 282(49):
35945-53.

Niemann, H. H., M. L. Knetsch, A. Scherer, D. J. Manstein and F. J. Kull (2001). "Crystal
structure of a dynamin GTPase domain in both nucleotide-free and GDP-bound
forms." Embo J 20(21): 5813-21.

Prakash, B., G J. Praefcke, L. Renault, A. Wittinghofer and C. Herrmarm (2000).
"Structure of human guanylate-binding protein 1 representing a unique class of
GTP-binding proteins." Nature 403(6769): 567-71.

 

Prakash, B., L. Renault, G J. Praefcke, C. Herrmann and A. Wittinghofer (2000).
"Triphosphate structure of guanylate-binding protein 1 and implications for
nucleotide binding and GTPase mechanism." Embo J 19(17): 4555-64.

Reubold, T. F., S. Eschenburg, A. Becker, M. Leonard, S. L. Schmid, R. B. Vallee, F. J.

Kull and D. J. Manstein (2005). "Crystal structure of the GTPase domain of rat
dynamin 1." Proc Natl Acad Sci U S A 102(37): 13093-8.

Sachdev, D. and J. M. Chirgwin (2000). "Fusions to maltose-binding protein: control of
folding and solubility in protein purification." Methods Enzymol 326: 312-21.

Schwartz, T. U., R. Walczak and G Blobel (2004). "Circular permutation as a tool to
reduce surface entropy triggers crystallization of the signal recognition particle
receptor beta subunit." Protein Sci 13(10): 2814-8.

Shin, H. W., H. Takatsu, H. Mukai, E. Munekata, K. Murakami and K. Nakayama (1999).

129

"Intermolecular and interdomain interactions of a dynamin-related GTP-binding
protein, Dnmlp/Vpslp-like protein." J Biol Chem 274(5): 2780-5.

Skerra, A. and T. G Schmidt (2000). "Use of the Strep-Tag and streptavidin for detection
and purification of recombinant proteins." Methods Enzymol 326: 271-304.

Smith, D. B. (2000). "Generating fUSlOIlS to glutathione S-transferase for protein studies."
Methods Enzmol 326: 254-70.

Thines, B., L. Katsir, M. Melotto, Y. Niu, A. Mandaokar, G Liu, K. Nomura, S. Y. He, G
A. Howe and J. Browse (2007). "JAZ repressor proteins are targets of the
SCF(COIl) complex during jasmonate signalling." Nature 448(7154): 661-5.

Zhu, P. P., A. Patterson, J. Stadler, D. P. Seeburg, M. Sheng and C. Blackstone (2004).
"Intra- and intermolecular domain interactions of the C-terminal GTPase effector
domain of the multimeric dynamin-like GTPase Drpl." J Biol Chem 279(34):
35967-74.

130

CHAPTER 4

Studies on the (pro)renin receptor and the promyelocytic zinc finger protein

131

4.1 Introduction

The cytoplasmic tail of the (P)RR is short, but fitnctionally important for the pivotal
roles of the (P)RR in blood pressure and cardiovascular regulation. Since the tail is the
only region that is exposed to the cytosol, it should be the direct mediator for the signal
transduction pathways that activated by binding of (pro)renin to (P)RR. Structural studies
on the cytoplasmic tail might shed light on the fiinctions of the (P)RR in the downstream
signal transduction.

PLZF has been identified as one of the protein molecules that interact with the (P)RR
and transduce (pro)renin signals. When (pro)renin binds (P)RR, the PLZF is activated
and translocated to the nucleus to be recruited to the cis element of the (P)RR promoter
(Schefe, Menk et al. 2006). The transcription of the (P)RR is then repressed by the
activated PLZF, creating a short negative feedback loop. Direct interaction of PLZF with
the cytoplasmic domain of the (P)RR was confirmed by the yeast two-hybrid study and
coimmunoprecipitation with truncated (P)RR proteins (Schefe, Menk et al. 2006).
However, which region(s) of the PLZF is responsible for the interaction remains unclear.

Since the cytoplasmic tail of the (P)RR is short (19 residues), the MBP fusion method
was chosen for simplifying the protein expression and purification, to generate a unique
chimeric protein for binding studies, and to enhance the potential for protein
crystallization. This chapter describes the structural determination of the cytoplasmic tail
of the (P)RR with MBP firsion method. Heterologous expression of PLZF truncated

proteins was perform for structural studies as well as investigation of which specific

132

region(s) of the protein is responsible for the interaction with the (P)RR.

4.2 (P)RR wild type protein
4.2.1 Secondmtructure and hydropathy prediction

The secondary structure of the (P)RR was predicted using the PSI-FRED online server.
The outcome showed that the protein is comprised of a mixture of helices and stands
connected by loops. Two major loops are located in the residues from 50 to 100, and fiom
160 to 175 which may be potential active sites for the (pro)renin binding. The
cytoplasmic tail forms a loop structure. The hydropathy prediction was performed by the
TMHMM online serve 2.0. Two major hydrophobic regions were revealed by the
predication. One is in the first 20 amino acids, the signal peptides. The other is between

residues 300 and 330, the transmembrane domain.

4.2.2 Expression and purification

The DNA encoding for the (P)RR was cloned into the pLW01 vector. Sequenced
plasmid was transformed into E. coli C41 (DE3) competent cells for expression. The
expression procedure was similar with that for MFN-1 described in chapter 3. In short,
single colonies were inoculated into 20 ml 2YT media with shaking at 37 OC overnight.
The culture was transferred into 1 L 2YT media and the cells were grown at 37 °C until
the OD600 reached to 1.0. IPTG was added into the culture at final concentration of 1

mM for induction and the cells were grown at 18 °C for 20 hrs.

133

To purify the (P)RR, cell pellets were re-suspended in buffer A (50 mM Tris, 300 mM
Sucrose, pH 8.0), and the crude cell extract was sonicated and centrifuged at 4 °C for 20
min at 12,000 g (low-speed centrifugation). The supernatant was further centrifuged for
one hour at 45,000 g at 4 °C (high-speed centrifugation) to separate the cell membranes
from the soluble lysate. The supernatant of the high-speed centrifugation was saved for
SDS-PAGE analysis. And the pellet of high-speed centrifugation were re-suspended in
buffer B (50 mM sodium phosphate, 300 mM NaCl, 10% glycerol, pH 8.0) and incubated
with 0.75% dodecyl maltoside at 4 °C for one hour. After another high-speed
centrifugation, the supernatant was added with 10 mM imidazole and loaded onto a
Ni-NTA column. The gradients of lysis, washing, and elution buffer were same as those
for DLP-l purification (chapter 2) except that all buffers contained 0.05% dodecyl
maltoside and the concentration of imidazole in the washing buffer was 38 mM.

The purified protein was loaded to analytic gel filtration for size determination. The
procedure was similar with that for DLP-1 (chapter 2) except that the column running
buffer was 30 mM Tris, 300 mM NaCl, 0.2% decyl maltoside. Most protein was eluted in
the void volume, indicating a size greater than the limit of the column, 1300 kDa. Since
there are detergents present in the buffer, conclusion could not be drawn whether the
large size was caused by protein aggregation or whether it is higher order structure of

(P)RR formed when binding to detergent micelles.

134

4.2.3 Crystallization

The purified (P)RR protein was concentrated with a 30K Amicon Ultra centrifugal unit
(Millipore) to about 10 mg/mL. Initial screens were performed by microbatch-under-oil
method with an ORYX crystallization robot (Douglas Instruments). A total of 198
conditions (Hampton Research screen I & II; Cryo screen I & II) were applied by
combining 0.75 uL of purified protein with 0.75 uL screen solutions. Within 30 days,
small crystals grew from conditions of 28% PEG 400, 0.2 M CaClz, 0.1 M HEPES, pH
7.5. Since there was detergent in the protein solution, sitting drop vapor diffusion method,
which is less affected by the lower surface tension of detergent solutions, was used for
crystal optimization. A matrix including different concentration of major precipitates,
additives, and pH ranges were designed in order to obtain large and single crystal suitable
for X-ray diffraction. With this designing matrix, putative crystals grew from most of the
conditions. However, the crystals are small and not in perfect shapes (Fig. 4.1). Further

efforts to optimize the crystallization conditions are underway.

135

 

Figure 4.1: Putative crystals of the (pro)renin receptor in hanging drops. The crystals
growing condition was 28% PEG 400, 0.2 M CaClz, 0.1 M HEPES, pH 7.5.

4.3 Structural determination of the cytoplasmic tail of the (P)RR

4.3.1 Experimental procedure

4.3.1.1 Cloning desim - The DNA encoding the cytoplasmic domain of (pro)renin
receptor was engineered into pRMG-ecoMBP vector with three alanines as a linker. In
short, two oligonucleotides encoding the 19 amino acids flanking with designed
restriction enzyme sites were synthesized. The sequences of the complementary
oligonucleotides are:
5 ’ -GGCCGCCGATCCTGGATATGATAGCATCAT'TTATAGGATGACAAACCAGAAG
ATTCGAATGGATTGA -3’ and 5’-
TCGAGTCAATCCATTCGAATCTTCTGGTTTGTCATCCTATAAATGATGCTATCATA
TCCAGGATC -3’. The oligonucleotides were annealed by slowly cooling down afier
incubating at 95 °C, and then phosphorylated by polynucleotide kinase at 37 °C. The

product was ligated to pre-cut pRMG-MBP vector into the Not I and Xho I sites and

136

transformed into E. coli DH5a competent cells. Positive clones growing from LB plates
containing 100 pg/mL ampicillin were picked, and plasmid DNA were isolated and

sequenced.

4.3.1.2 Protein expression M purification - The sequenced plasmids encoding the
fusion protein (MBP-(P)RR19) were transformed into the expression host E. coli BL21
(DE3) competent cells. A fresh single colony from a selection plate was inoculated into
100 m1 LB media with 100 pg/mL ampicillin. After shaking at 37 °C with 200 rpm
overnight, 20 ml of this culture was transferred into 1 L fi'esh LB media with antibiotics.
When an OD600 reached to about 0.8-1.0, IPTG was added into the culture with a final
concentration of 0.05 mM. The growing temperature was reduced to 24 °C and the cells
were continuously shaken for 17 hours. The cells were harvested by centrifugation and
stored at —80 °C.

Cell pellets were re-suspended in Buffer A (20 mM Tris-HCl, pH 7.4, 0.2 M NaCl, 1
mM EDTA, 10 mM B-mercaptoethanol). The crude cell extract obtained by sonication
was centrifuged at 24 °C at 12,000x g for 20 min. The supernatant was mixed with 10 ml
amylose resin (New England Biolabs) pre-equilibrated with buffer A and incubated at 4
°C for 5 hrs with shaking. The mixture was loaded onto a column to allow unbound
proteins to flow through. The column was then washed with 10 column volumes of buffer
A. These washing eluants were considered as the fusion protein-containing fractions and

were pooled and concentrated to 1 ml using an Amicon Ultra centrifugal unit (Millipore).

137

The protein was loaded onto a 1 m1 HiTrap Q ion exchanger (Amersham Biosciences),
eluted with a gradient generated by using buffer B (20 mM Tris-HCl, pH 8.5) and buffer
C (20 mM Tris-HCl, 1 M NaCl, pH 8.5) at a flow rate of 1 ml/min. The peak fractions
were pooled and concentrated. The buffer was exchanged with buffer D (20 mM Tris-HCl,
pH 7.4, 0.1 M NaCl) by passing two 5-ml spin columns with sephadexTM G-25 coarse
(GE healthcare). The protein was analyzed by SDS-PAGE and the concentration was
determined using a BCA assay (Thermo Scientific).

The oligomeric state of the MBP-(P)RR19 was determined by size exclusion
chromatography. A Superdex 200 10/30 GL column was pre—equilibrated with buffer D.
The purified fusion protein was loaded onto the column and eluted at a rate of 0.5 ml/min.

Fractions containing fusion protein was analyzed by SDS-PAGE and Western blot.

4.3.1.3 Ma_ss spectrometry - The molecular weight of the purified MBP-(P)RR19 was
analyzed by a Waters LCT Premier time-of-flight mass spectrometer which is coupled
with Shimadzu LC-20AD HPLC pumps and a SIL-5000 autosampler. Separation was
performed using a Thermo BetaBasic cyano column (1 x 10 mm) with a gradient
generated by 0.15% aqueous formic acid and 75% acetonitrile for online desalting and
elution. Instrument control, data acquisition, and processing were provided by the
MassLynx data system (Waters Ltd, Manchester, UK, version 4.1). Molecular weight was

calculated based on spectrum deconvolution using MaxEntl software.

138

4.3.1.4 Crystallization - Prior to crystallization, the apo-fusion protein was adjusted to
20 mg/mL. Preliminary crystallization screen was performed at 20°C by
microbatch-under-oil method using an ORYX crystallization robot (Douglas Instruments).
With crystal screens (HR2-110 and HR2-112, Hampton Research), 0.75 uL protein
solution and 0.7 5 uL crystallization reagent were mixed in each drop. After finding initial
crystallization conditions, crystal growth was scaled up by hanging-drop vapour diffilsion
method at 20°C (2 pl protein solution and 2 pl reservoir solution equilibrated against 1
ml reservoir solution). The precipitant concentration and the pH were optimized and the
best crystals were obtained using solutions consisting of 20% (w/v) PEG 4000, 0.2 M
MgClz and 0.1 M Tris, pH 8.5. Final crystal dimensions were 0.4 X 0.3 X 0.04 mm. To
occupy the ligand-bound sites with maltose, 0.5 mM maltose were added into the fusion
protein and the mixture was incubate at 4 °C for at least 3 hrs before crystallization trials
as described above for the apo-fusion protein. The crystals of the ligand-bound protein
was optimized to a size of 0.2 X 0.2 X 0.08 mm with the condition of 26% (w/v) PEG

4000, 0.2 M MgClz, and 0.1 M Na Cacodylate, pH 6.5.

4.3.1.5 Cmogrotection and data collection - Crystals for X-ray diffraction studies were
transferred stepwise into cryoprotectant solutions with increasing concentrations of
glycerol. The ape-fusion protein crystals were flash-cooled in the final cryoprotectant
solution consisted of 20 mM Tris-HCI (pH 7.4), 0.1 M Tris-HCl (pH 8.5), 0.1 M NaCl,

20% PEG 4000 (w/v), 0.2 M MgClz, and 15% glycerol (v/v). And the ligand-bound

139

protein crystals were in the final cryoprotectant solution consisted of 20 mM Tris-HCl
(pH 7.4), 0.1 M Na Cacodylate (pH 6.5), 0.1 M NaCl, 26% PEG 4000 (w/v), 0.2 M
MgClz, and 15% glycerol (v/v). X-ray diffraction data were collected at —l73 °C on
21-[D beamline (LS-CAT) using 3 MAR CCD detector at Advanced Photon Sources
(Argonne, IL). Complete datasets were collected from single crystals with a
crystal-to-detector distance of 250 mm and an exposure time of 1 sec per 1° oscillation
under the wavelength of the synchrotron radiation at 0.98 A. All diffraction images were
processed using DENZO and integrated intensities were scaled using the SCALEPA CK

from the HKL-2000 program package (Otwinowski and Minor 1997).

4.3.1.6 Structural determination and refinement - The structure of the fusion protein
was determined by molecular replacement using the programs from CCP4 suite (1994),
with the known structures of E. coli MBP as search models. Structure with PDB code of
1JW4 (Residues 1-363) (Duan and Quiocho 2002) was used for apo-fusion protein and
1ANF (Residues 1-363) (Quiocho, Spurlino et al. 1997) was for ligand-bound protein.
Models building were performed in Coot (Emsley and Cowtan 2004) using the 2F o-F c
and F o-F c electron-density maps. Translation/Libration/Screw (TLS) (Winn, Isupov et al.
2001) motion determination using both domains of MBP as TLS group and
non-crystallographic symmetry (NCS) were used for model refinement. The quality of
the models were evaluated using the program PROCHECK. The graphical figures were

visualized using the program PyMOL.

140

4.3.2 Results and discussion
4.3.2.1 Protein expression and purification
4. 3. 2. 1. I Rationale of strategies

Like many effector regions of receptors on membranes, the cytoplasmic tail of the
(P)RR is short and has only about 19 amino acids. We initially thought about synthesizing
the 19-residue peptides for structural studies. However, such peptides may not be well
behaved enough or be in the appropriate conformation for crystallization. Therefore, I
used the traditional method of recombinant expression in E. coli, but considered that 19
residues may be too short to be observed in both agarose gel and SDS-PAGE, which
would cause problems during DNA cloning and protein purification. Also, if we try to
express the 19 amino acids alone in E. coli, the protein may have solubility problem,
which is common for truncated protein fiagrnents. Therefore, an expression and
purification aid is necessary to create the recombinant fusion protein with the cytoplasmic
tail. Maltose binding protein (MBP) fi'om E. coli is commonly used to enhance the
expression, improve the yield and stability, and facilitate the purification and
crystallization. Most importantly, with an appropriate linker, MBP is a good
crystallization aid without affecting the native structure of target proteins (for details,
please refer to chapter 3).

Due to the short DNA sequence encoding for the 19 amino acids, it would be
extremely difficult to clone the gene into expression vectors by PCR. For traditional DNA

cloning using PCR, sequence shorter than 100 bp would cause a decrease of the success

l4l

rate. If shorter than 70 bp, PCR method may not be the best because of the problem of
recovering DNA from agarose gels. Thus, I used the oligonucleotide annealing method,
which allows the manipulation of relatively short DNA from 10 bp up to 100 bp. The
trick is that each oligonucleotide should be designed with sticky ends so that after
annealing, they resemble the structure that is cut by restriction enzymes and are able to
ligate to pre-cut vectors. This method can also be used for designing and constructing
DNA vectors particularly when require inserting short DNA sequences such as a
promotor sequence or a new restriction enzyme site (Brummelkamp, Bemards et al.
2002)

Another possible problem due to the short target protein is that it would be hard to
separate and distinguish the endogenous E. coli MBP with the MBP-(P)RR19 during
protein purification. Both MBPs would bind to the amylose column and they would also
be difficult to separate by ion exchange chromatography because the fusion of 19
residues in the C-terminus may not cause a significant change on MBP biophysical and
biochemical properties. I assumed that the MBP-(P)RR19 would affect the binding
affinity of MBP to amylose column that it would not bind the amylose resin as tight as
that of the endogenous native MBP that requires 10mM maltose to be eluted from the
column. Thus, I used large volume of washing buffer (10 column volume) without
maltose in it to wash the fusion protein off the amylose column and considered this
fraction as relatively pure MBP-(P)RR19. The elution fraction that was eluted with 10

mM maltose was considered as the mixture of fusion and endogenous MBPs and was

142

discarded.

  

é;

Wthu'LP FT w1 W2

 

Figure 4.2: SDS-PAGE of MBP-(P)RR19 purification by amylose column. Incubating
for 5~6 hrs with amylose and washing with about 10 column volumes would yield
highest amount and pure protein. MW, molecular weight standard; Wh, whole cells; LSN,
supernatant of low-speed centrifugation; LP, pellets of low—speed centrifugation; FT, flow ‘
through; W1, wash fraction 1; W2, wash fraction 2.

4. 3. 2. I . 2 Protein purification

I have screened different incubation time of the MBP-(P)RR19 with amylose and
different volume of washing fractions and found that incubating at 4 °C for 5~6 hrs and
washing with about 10 column volumn would be a balance between yield and purity of
the MBP-(P)RR19 (Fig. 4.2). Shorter incubation time or larger volume of washing caused

impurity of the fusion protein because the endogenous MBP did not bind the column tight

143

enough. While, longer incubation time or less volume of washing caused low yield of
fusion protein because more fusion protein would bind the column and would not eluted

until with maltose.

A 2450 .

1950 '

1450 '

 

950 '

Absorbance at 280 nM (mAU)

 

A
01
O

 

-50 II I I l I
0 5 10 15 20 25 30 35 40 45
Retentlon volume (ml)

 

Figure 4.3: Purification of the MBP-(P)RR19 by ion exchange chromatography. (A)
In the ion exchange chromatogram, the protein was eluted in the major peak at an ionic
strength between 0.25 and 0.35 M NaCl. (B) SDS-PAGE of fractions corresponding to
the major peak in panel A.

144

The theoretical isoelectric point of the MBP-(P)RR19 is 4.99. Therefore, anion
exchange is appropriate for further protein purification. The MBP-(P)RR was eluted in a
single peak at the ionic strength between 0.25 to 0.35 mM NaCl. Based on previous
chromatographic experience from the current laboratory, endogenous MBP usually is
eluted behind but closed to the peak of fusion proteins. In the MBP-(P)RR19 case, there
were no other peaks showing up in the chromatograrn (Fig. 4.3) indicating the firsion
protein was homogeneous. Meanwhile, it supports our assumption that the fusion protein
may have less affinity to the amylase resin than the endogenous MBP and the strategy of
collecting the washing fraction during purification. Judging from SDS-PAGE, the
MBP-(P)RR19 was greater than 98% pure after ion exchange purification (Fig. 4.3).

About 8 mg of purified protein was obtained from 500 ml culture for structural studies.

4. 3. 2. 1. 3 Mass spectrometry

Occasionally, with the fusion method, target proteins can be unexpectedly “cleaved”
off the MBP, leaving truncated fusions. Although the reasons are unclear, it may be due to
interrupted protein translation in vivo or protease cleavage during purification. Since the
MBP-(P)RR19 has similar molecular weight with that of MBP, it is hard to determine
expression of the full-length firsion protein by SDS-PAGE or Western blotting. The
ESI-TOF mass spectrometry has the accuracy to measure a protein’s molecular mass to
within one amino acid. The mass spectrometry results on purified MBP-(P)RR19 showed

that the fusion protein mass is 42488.5 Da (Fig. 4.4), with an error range of 10 Da. The

145

theoretical mass of MBP-(P)RR19 is 42497 Da.

146

.mQ SVN... £38 #83285 05 we an 3 mo owqfi Hobo 829$ 05 GEE» mm £033 «Q Wwwvmv £ 32: £289 2: $5
33265 anon SHE 2E. «0508958.: 38: .3 62:32: agar—m: 35.59 no 399$ 3:522: 8.2.53fl "vi unsure

. o8? , . 88v . 085w _ 0.83. _ 88v . 083 _ 8o? .80? . 89v . o2 9.
J J :4 1.51! 4.11.1 0
89809.

 

Smfiogv

 

 

 

839$ -09

147

4.3.2.2 Crystallization
4. 3.2.2. I Apo-MBP-(P)RR1 9

Before crystallization, the MBP-(P)RR19 was buffer exchanged with 20 mM Tris, 100
mM NaCl, pH 7.4 to remove excess salts from ion exchange chromatography. Initial
screens were performed by microbatch-under-oil method with an ORYX crystallization
robot (Douglas Instruments). A total of 198 conditions (Hampton Research screen I & II;
Cryo screen I & II) were applied by combining 0.75 pL of purified protein with 0.75 pL
screen solutions. Within 30 days, clustered large crystals grew from conditions of 30%
PEG 4000, 0.2 M MgClz, 0.1 M Tris pH 8.5 and 25% PEG monomethyl ether (PEGmme)
550, 0.01 M ZnSO4, 0.1 M MES pH 6.5. Hanging drop vapor diffusion was used to
optimize the conditions of the initial hits. A matrix including different concentration of
major precipitates, additives, and pH ranges were designed in order to obtain large and
single crystal suitable for X-ray diffraction. With this designed matrix, crystals
precipitated by PEGmme 550 were still highly clustered and no single crystal could be
picked for data collection. Whereas, with a concentration gradient of PEG 4000 from
18% to 30%, nice and single crystals grew about four days after the setup from 20% PEG

4000, 0.2 M MgClz, and 0.1 M Tris 8.5 (Fig. 4.5).

148

 

Figure 4.5: Crystals of apo-MBP-(P)RR19 from hanging drop vapor diffusion. The
optimized condition is 20% PEG 4000, 0.2 M MgClz, and 0.1 M Tris 8.5.

Glycerol was added to the mother liquor as a cryoprotectant. As adding too much
cryoprotectant increases the chances that the crystals would crack. It is necessary to find
the lowest amount of glycerol that would avoid the formation of ice but not cause crystal
damage. About 15% of glycerol was determined to be the best concentration for the
MBP-(P)RR19 crystals. Because there was no glycerol in the crystallization conditions,
the procedure of introduction was in a stepwise manner. Adding a high concentration of
cryoprotectant at one time can cause a sudden change of solution environment and
damage crystals. Therefore, crystals were first transferred into stabilizing buffer
containing 2% glycerol to let them adapt to the new glycerol environment. Then, the
glycerol concentrations was increased to 5%, 10%, and finally to 15%. There was no

obvious damage observed on surface of the crystals in the final cryoprotectant solution.

149

4. 3. 2. 2.2 Ligand-bound MBP-(P)RR19

Same crystallization and freezing procedures were applied to the ligand-bound
MBP-(P)RR19. Before screenings, 0.5 mM maltose was added into the protein solution
and the protein was kept on ice for at least three hours to allow binding of ligands. Highly
clustered and needle-shape crystals grew the next day of the setup by microbatch (Fig.
4.6A). The conditions are 30% PEG 4000, 0.2 M NH4AC, 0.1 M NaAC pH 4.6 and 30%
PEG 8000, 0.2 M (NH4)2SO4, 0.1 M Na cacodylate pH 6.5. Initial optimization using
hanging drop method generated similar clustered crystals. However, conditions by matrix
designing yielded much better crystals that are suitable for data collection with the

condition of 26% PEG 4000, 0.2 M MgClz, 0.1 M Na Cacodylate, pH 6.5 (Fig. 4.63).

150

 

Figure 4.6: Crystals of ligand-bound MBP-(P)RR19. (A) Crystals from microbatch
with conditions of 30% PEG 4000, 0.2 M NH4AC, 0.1 M Na acetate pH 4.6 (lefi) and
30% PEG 8000, 0.2 M (NH4)2SO4, 0.1 M Na cacodylate pH 6.5 (right). (B) Crystals from
hanging drop with optimized condition of 26% PEG 4000, 0.2 M MgClz, 0.1 M Na
Cacodylate, pH 6.5.

151

 

Figure 4.7: Structure of MBP-(P)RR19 with maltose bound. (A) Front view. (B) Side
view.

152

 

Figure 4.8: Structure of MBP-(P)RR19 without maltose. (A) Front view. (B) Back

view.

153

II. 373

Mot 376

    

Thr 377

Figure 4.9: The (P)RR19 in a 2Fo-Fc electron-density map contoured at one
standard deviation above the mean density. Residues 366-378 (from left top to bottom)
of molecule B with maltose bound are shown in sticks representation.

154

Figure 4.10: Dimeric interface of apo-MBP-(P)RR19 is mediated by hydrogen bonds
formed by residues from (P)RR19 peptide. Hydrogen bonds are highlighted by red
dashes. (A) Overview of hydrogen bonding pattern. (B) Detailed hydrogen bonds.

155

 

156

Figure 4.10 continued

 

157

Figure 4.11: Dimeric interface of ligand-bound-MBP-(P)RR19 is predominated by
hydrogen bonds formed by residues from (P)RR19 peptide. Hydrogen bonds are
highlighted by red dashes. (A) Overview of hydrogen bonding pattern. (B) Detailed
hydrogen bonds.

158

 

159

 

160

2450 '

1950 "

1450 '

950 '

 

450 '

Absorbance at 280 nM (mAU)

 

 

‘50 I I I I l l

0 5 10 15 20 25 30

Retention volume (ml)

 

Figure 4.12: Size exclusion chromatography of MBP-(P)RR19. (A) Chromatography.
The major peak is corresponding to a size of 42 kDa. (B) SDS-PAGE of fractions of the
major peak showed in A.

161

4.3.2.3 Crystal structure
4.3.2.3.] Overall structure

The MBP-(P)RR19 crystals with and without maltose bound have the space group
symmetry P212121. The MBP-(P)RR19 apo-crystals have the unit-cell parameters a =
47.78, b = 112.70, c = 175.11 A; the MBP-(P)RR19 crystals with maltoses have the
unit-cell parameters a = 41.95 , b = 96.78, c = 191.45. A close examination of the unit
cell parameters clear suggests that the two crystal forms are probably not identical,

perhaps arising from different molecular packings, despite the fact that they have the

same space group symmetry.

162

Table 4.1: Data-collection and processing statistics.

 

Crystals without maltose with maltose
Space group P212121 P212121
Unit-cell parameters
a(A) 47.78 41.95
b(A) 112.70 96.78
c(A) 175.11 191.45
a = B = r (°) 90 90
No. of molecules per ASU 2 2
Data collection
Wavelength (A) 0.979338 0.979338
Resolution (A) son—2.0 50.0-1.996
Unique reflections 61694 50067
Multiplicity 6.2 6.1
Completeness (%) 97.8 (98.9) 95.3(85.6)
Refinement statistics
R work 0.19 0.22
R free 0.26 0.29
B factor 26.79 34.52
R.m.s.d. bond lengths (A) 0.018 0.024
R.m.s.d. bond angles (°) 2.13 1.58

 

Both structures were solved by molecular replacement and refined to 2.0 A resolution
(Fig. 4.7 and 4.8). The statistics of the crystallographic data were summarized in Table
4.1. There are two chains in the asymmetric unit for both structures. In the starting model,
residues 1 to 363 are the MBP, and residues 364 and 365 are the short alanine linker. The
last 19 amino acids of the (P)RR cytoplasmic tail start from residue 366. The final model
of the apo-MBP-(P)RR19 contained 748 amino-acid residues including 373 amino acids
in chain A and 375 amino acids in chain B, 712 water molecules, and 11 magnesium ions
per asymmetric unit. The electron-density for residue 1 of MBP was poor in both chains

such that these residues could not be modeled into the structure. The final model of the

163

ligand-bound MBP-(P)RR19 is similar and contains 751 amino-acid residues including
375 amino acids in chain A and 376 amino acids in chain B, 683 water molecules, and 3
magnesium ions per asymmetric unit. The first residue in chain A and the first two
residues in chain B of the N-terminus of MBP were not included in the model due to
weak electron-densities. The ligand-bound model contained one maltose molecule in each
of the active sites of MBP per asymmetric unit. Only the first thirteen and eleven residues
of the (P)RR cytoplasmic tail were observed in the chain B and A, respectively, of the
ligand-bound protein, and the first eleven and nine residues were observed in chain B and
chain A of the apo-protein, respectively (Fig. 4.9). The rest residues are disordered and
could not be included in either model. It is likely that a portion of the C-terminus of the
(P)RR cytoplasmic tail is unstructured because the mass spectrometry experiment clearly
shows the presence of the 19 amino acids in the fusion protein. Since the (P)RR
cytoplasmic tail may interact with multiple proteins in different signal transduction
pathways, having an unstructured C-terminus may provide an advantage of
interconvertibility for adapting to multiple receptor/signaling molecules. However, the
unstructured C-terminus observed from this study does not exclude the possibility that
this region may become ordered when stimulated by specific signals such as (pro)renin

binding, protein modifications, and protein interactions.

4. 3. 2. 3.2 The molecular packing interaction: evidence for a (P)RR induced dimer

What was unusual about the two MBP-(P)RR19 crystal structures is that the dimeric

164

arrangement of the MBP molecules within the asymmetric unit. The formation of
“symmetric oligomers” in crystals is not uncommon and may arise coincidently fi'om the
crystal symmetry. To verify whether the “symmetric oligomers” arise from interactions
between the molecules in the asymmetric unit and not fi'om crystal symmetry, a closer
look at the molecular packing was done.

Using the LSQKAB program from CCP4 (1994), two monomers in asymmetric unit
were superimposed by rotating one monomer through a series of angles. When using
spherical polar coordinates omega, phi, chi to achieve the superposition, the chi angle can
be used to detect near perfect rotations (e.g., ~180° is a 2-fold or ~120° is a 3-fold). For
the ape-protein model, when chain A was superimposed onto chain B, the chi angle was,
176.31°, a value near 180°. For the ligand-bound model, the superimposition of chain A
onto chain B yielded a chi angle of 179.7°. Since the angle between rotation axis and
centroid vector is near to 90° (88.69° and 90.92° for apo- and ligand-bound models,
respectively), the superimposition most likely represented a pure rotation. These analyses
support the contention that the two molecules in the asymmetric unit are forming a
molecular dimer.

MBP exist as a monomer, in solution and in crystal structures, unless the assembly into
higher order oligomers is induced by a protein fusion. Interfacial contacts in asymmetric
unit of some dimeric or trimeric forms of MBP-fusions were predominately driven by
target proteins instead of MBPs (Kobe, Center et al. 1999; Liu, Manna et al. 2001). Since

the MBP-(P)RR19 forms dimer, it led to the hypothesis that the 19 amino acid tail

165

promotes the dimerization of the fusion protein, and the major dimeric contacts should be
found in the peptide regions. Using the PISA online sever
(http://www.ebi.ac.uk/msd-srv/prot_int/pistart.htrnl), protein interfaces analyses showed
that hydrogen bonds formed by the peptide amino acids, especially by the Tyr369 and
Tyr3 74, predominated in the dimer interface. No water-mediated hydrogen bonds, salt
bridges, disulfide bonds, or covalent bonds were observed in the PISA interface analyses.

For the apo-MBP-(P)RR19, the dimeric interface resulted in 756.3 A of buried surface
area, which corresponds to 4.7% of monomeric surface. The solvation free energy gain
upon formation of the dimeric interface is —11.2 kcal/M, indicating an energy favorable
state. Nine potential hydrogen bonds were found between the two monomers at the
interface, all of which are mediated by residues in (P)RR19 peptide with the Tyr 369 and
Tyr 374 involved in seven of them (Fig. 4.10 and Table 4.2A). The side chain OH group
of Tyr 369 fi'om each chain protruded from the peptide main chain and interact with the
main chain 0 atom of the linker Ala 364 from the other chain. The side chain OH group
of Tyr 374 in chain B is hydrogen bonding with the main chain 0 atom of Gly 368, the
main chain N atoms of Asp 370, Ser 371, and Ile 372, respectively. Since the side chain
of Tyr 374 in chain A was not included in the model due to weak electron-density map,
similar interaction was not observed in chain B. In the interface, contact is also made
between the main chain 0 atom of Ile 372 and the N82 atom of Gln 335, and between the
main chain N atom of Gly 368 and the main chain 0 atom of Pro367.

For the ligand-bound MBP-(P)RR19, the dimeric interface caused a 1148.9 A of buried

166

area which corresponds to 7.3% of monomeric surface. The solvation free energy gain
upon formation of the dimeric interface is —17.1 kcal/M, indicating an energy favorable
state. The ligand-bound structure had the similar overall interface arrangement where the
major potential contacts were mediated by the (P)RR19 peptide residues, particularly by
the Tyr 369 and Tyr 374 (Fig. 4.11 and Table 4.2B). Similar to the ape-protein structure,
the side chain OH group of Tyr 369 from each chain protruded from the peptide main
chain and interacted the main chain 0 atom of the linker Ala 364 from the other chain.
Residue Thr53 in chain A is 29 A away from its counterpart in chain B in the “open”
apo-protein structure. With ligands bound, the MBP represents a “closed” structure,
which brings the two Thr53 residues together to form a hydrogen bond.

The (P)RR19 peptide residues have more contact with the MBP in the more compact
“closed” form than in the “open” form. Several hydrogen pairs between the (P)RR19
peptide and MBP occur in dimer interface: the O atom of Thr377 and the main chain 0
atom of Gln72, the main chain 0 atom of Arg 375 and the N82 atom of Gln 72, the main
chain 0 atom of Met376 and the main chain N atom of the Ser73, and the main chain 0
atom of Tyr 369 and the side chain N82 atom of Gln 335. Although the Tyr 374 is
involved in formation of hydrogen bonds, the bonding pattern is different than that of the
apo-protein. In chain A, the OH group of Tyr374 interacts with the 07 atom of Ser337 of
chain B. Surprisingly, the main chain 0 atom of Tyr 374 in chain B hydrogen bonds with
the main chain N atom and side chain N82 atom of Gln 335, but the hydroxyl of Tyr 374

makes no obvious interactions with other side chains. The distances between the OH

167

group of Tyr 374 in chain B with other side chains no shorter than 5 A. However, the
electron density around Tyr374 in chain B is very weak compared to that of other
tyrosines observed in the structure. It is therefore possible that these interactions were not
observed due to the poor quality of the electron density in this region. The compact
“closed” MBP structure then yields contradictory observations: the MBP-(P)RR19 dimer
is more “symmetric” overall, but “closed” MBP structure provides less space in between
two monomers, which may cause some localized disorder blurring the detailed
conformation of the peptide.

The size exclusion chromatography of the MBP-(P)RR19 showed a peak that was
estimated to about 42 kDa (Fig. 4.12), indicating that the protein exists as a monomer in
solution. Since the protein concentration for crystallization is much higher than that in
solution for gel filtration, it is possible that the dimerization of the MBP-(P)RR19 is
concentration-dependent. In addition, if the hydrophobic interactions are the major force
bringing two monomers together, some fi'action of dimeric MBP-(P)RR19 should exist in
solution. Therefore, the monomeric form in solution excludes the possibility that the
dimers observed in the crystal structures were caused by the pure hydrophobic

interactions in the peptide region.

168

Table 4.2 A: Potential hydrogen bonds formed between two MBP-(P)RR19
monomers of apo-MBP-(P)RR19. The online server EBI Pisa
(http://www.ebi.ac.uk/msd-srv/prot_int/pistart.html) and software PyMOL was used for
find these contacts.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Molecule 1 Distance (A) Molecule 2

B: Gln 335 (N82) 3.2 A: Ile 372(0)
B: Gln 335 (N82) 3.3 A: Ala 374(0)
B: Tyr 369 (OH) 2.6 A: Ala 364 (O)
B: Tyr 374 (OH) 2.6 A: Gly 368 (0)
B: Ala 364 (0) 2.4 A: Tyr 369 (OH)
B: Tyr 374 (OH) 3.5 A: Asp 370 (N)
B: Tyr 374 (OH) 3.2 A: Ser 371 (N)
B: Tyr 374 (OH) 3.8 A: Ile 372 (N)
B: Pro 367 (0) 3.1 A: Gly368 (N)

 

Table 4.2 A: Potential hydrogen bonds formed between two MBP-(P)RR19
monomers of maltose-bound-MBP—(P)RR19. Please refer A for legend.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Molecule 1 Distance (A) Molecule 2

B: Thr 53 (0)7) 3.7 A: Thr 53 (07)
B: Thr 377 (Q1) 3.8 A: Gln 72(0)

B: Tyr 369 (OH) 2.7 A: Ala 364 (0)
B: Ser 337 (07) 3.8 A: Tyr 374 (OH)
B: Arg 375 (0) 3.2 A: Gln 72 (N22)
B: Met 376 (0) 3.7 A: Ser 73 (N)

B: Tyr 374(0) 3.4 A: Gln 335 (N82)
B: Ala 364 (O) 2.6 A: Tyr 369 (CH)
B: Tyr 374(0) 3.8 A: Gln 335 (N)
B: Tyr 369(0) 3.7 A: Gln 335 (N82)

 

4. 3. 2. 3.3 Description of the (P)RR cytoplasmic tail structure
The MBP-(P)RR19 contains the full 19 amino acids of the cytoplasmic tail as
determined by the mass spectrometry sequencing analysis (Fig. 4.4). In the models, the

first 13 and 11 residues out of 19 were observed in the ligand-bound structure chains B

169

and A, respectively; in the apo-structure, the first 11 and 9 residues were clearly observed
in chains B and chain A, respectively. Absence of the electron density for the remaining
residues indicates that this region of the cytoplasmic tail may be disordered. The
observed residues have a structure of relatively flexible loop without obvious secondary
structure (Fig. 4.13). The linker region of the Ala364 and Ala365 is involved in formation
of the C-terminus of the helix of the MBP. The loop region from residue Asp366 to
Tyr369 is relatively smooth and straight. A turn occurs between the Tyr369 and the Asp
370 and leads into a spiral-like structure from residue Asp 370 through Arg 375. A few
hydrogen bonds were formed among the residues in this region. Then the loop straightens
out again and extends into the space in between the two monomers until no more electron

density is observed.

Figure 4.13: Structure of the (P)RR cytoplasmic tail. The residual 366-378 is shown
from top to bottom.

The crystal structures of the MBP-(P)RR19 both in with and without maltose provide a

170

overview of the protein structure of the (P)RR cytoplasmic tail. The dimeric nature of the
complex in the crystal, for both forms, indicate that the (P)RR19 is the driving force for
the dimerization of monomeric MBP. Although in the maltose-bound form, the Thr 53
was also involved in interactions at interface, the predominate interactions were made by
hydrogen bonds fiom residues in the (P)RR19 peptide to MBP. These results suggested a
possible role of the (P)RR19 in the dimerization of full-length (P)RR protein, since the
(P)RR was reported to exist as a dimer (Schefe, Menk et al. 2006). Without additional
experimental evidence on other regions of the (P)RR, we cannot not yet conclude about
structural roles of the N-terminus and transmembrane domain. However, the structural
results of the (P)RR19 suggested that besides interacting with other signaling molecules,

the cytoplasmic tail may at least partially involved in protein oligomerization.

4.3.2.3.4 N-terminalfitsion

Since no more than 13 out of 19 amino acids of the (P)RR cytoplasmic tail were
observed in the MBP C-terminus fusion, it is necessary to consider other filSlOIl methods
to get a better structure for (P)RR19. Although it is likely that some of the residues in
(P)RR19 are flexible and do not form an ordered structure, fusing the 19 amino acids to
the N-terrrrinus of the MBP may be an alternative way to test it.

The annealing DNA cloning method was used for the 19 amino acids fusion to the
MBP N-terminus. Since there is 6Xhis tag in the C-terminus of the MBP in the expression

vector, the Ni-NTA column was used for initial protein purification. Further ion exchange

171

purification showed the protein was eluted at the same ionic strength range of 0.25 M to
0.35 M NaCl and the protein was purified to homogeneity.

Unfortunately, there was no crystals have yet been obtained in conditions with or
without maltose. One possible reason may be that the C-terminus of the 19 amino acids is
flexible and causes an energy barrier for crystallization when it is in between its
N-terminus and the MBP N-terminus. Since the N-terminus of the MBP is crucial to the
protein folding, it is also possible that the N-terminus fusion disrupt the MBP protein
structure and firrther impact the crystallization. Currently, to my knowledge, no crystal
structure of fusion protein in N—terminus of MBP was reported indicating that N-terminus

fusion may not be an applicable method for protein crystallization aid.

4.4 Cloning and expression of the promyelocytic zinc finger protein

(PLZF)

4.4.1 Experimental procedures

4.4.1.1 DNA Cloning of PLZf and truncated proteins - The DNA encoding for the
full-length PLZF, the BTB domain (1-132aa), the center domain (l37-377aa), and the
zinc fingers domain (378-673aa) were amplified from cDNA by PCR. The PCR
amplification was comprised of 35 cycles of denaturing at 94°C for 30 sec, annealing at
55°C for 45 sec, and elongation at 72°C for 2 min and 10 sec for the full-length PLZF; 40

sec for the BTB domain; 1 min for the center domain; and l min 30 sec for the zinc

172

fingers domain, followed by 72°C for 10 min. The PCR products were purified by
QIAquick PCR purification kit (QIAGEN), and digested with restriction enzymes BamHI
and XhoI for ligation with pRMG-ecoMBP vector, and NcoI and XhoI for pLW01 vector.
Ligation products were transformed into E. coli DHSu competent cells. Positive clones
growing from LB plates containing 100 pg/ml ampicillin were picked, and plasmid

DNAs were isolated and sequenced.

4.4.1.2 Protein Expression and Purification - The protein expression procedures were
same as described for that of DLP-1 in chapter 2, except that the E. coli BL21 (DE3) was
used as expression host cells. To purify the PLZF full-length and truncated proteins, the
cell pellets were re-suspended in Buffer A (50 mM sodium phosphate, 300 mM NaCl, 0.1
mM EDTA, pH 8.0). After sonication, the crude cell extract was centrifuged at 4 °C for 20
min at 12,000X g. The supernatant was loaded onto a pre-equilibrated column containing
20 ml Ni-NTA agarose slurry. The column was washed with buffer B (50 mM sodium
phosphate, 300 mM NaCl, 20 mM imidazole, pH 8.0). Protein bound column was eluted
by Buffer C (50 mM sodium phosphate, 300 mM NaCl, 200 mM imidazole, pH 8.0). The
protein eluates were pooled and concentrated to 1 ml by Amicon ultra centrifugal filter
molecular cutoff (Millipore).

Ion exchange chromatography was performed to further purify target proteins and
remove contaminants. The pooled and concentrated eluates from Ni-NTA were loaded

onto a 1 ml HiTrap Q ion exchanger (Amersham Biosciences) pre-equilibrated with

173

Buffer D (20 mM Tris-HCl, pH 8.5). Protein was eluted off the column with a linear
concentration gradient of NaCl from 0 to 1 M, at a flow rate of 1 ml/min. The peak
fractions containing highly purified target protein were pooled and concentrated. Protein

concentration was estimated by Bradford assay using bovine serum albumin as standard.

174

Figure 4.14: Purification of PLZF fragments by Ni-NTA column. (A) Full-length
PLZF. (B) BTB domain. (C) RD2 domain. (D) Zinc finger domain. Protein molecular
weigh marker was labeled at left (kDa). MW, Molecular weight standard; Wh, Whole
cells; LSN, Supernatant of low-speed centrifugation; FT, Flow through fraction; Wa,
Wahsing fraction; El and E2, Elution fractions.

175

 

MW Wh LSN F'I' Wa E1 E2

    

WhLSN FT Wa E1 E2

176

Figure 4.14 continued

C

75
50

  
    

 

MWWh LSN FT Wa E1 E2

'M'wwnsu i=1- Wa ' E1 E2

177

4.4.2 Results and discussion
4.4.2.1 Protein expression and purification

The full-length PLZF, the BTB, RD2, and zinc finger domains were cloned into the
pRMG-ecoMBP vector with restriction enzyme sites of BamHI and XhoI. With
expression and purification conditions optimized as described in chapter 2 for DLP-1, all
fusion proteins were purified by Ni-NTA column (Fig. 4.14). The BTB domains had the
highest yield with greater than 20 mg protein purified from 1 L culture. However, for the
full-length PLZF, although the protein was in soluble fraction (supernatant fraction of
low-speed centrifugation), most protein did not bind the Ni-NTA column, indicating that
the protein may be moderately aggregated and/or the 6Xhis is not accessible. Only about
less than 1 mg full-length PLZF protein was purified from 1 L culture. Similar problem
occurred for the RD2 and zinc finger domain but was not as severe as the full-length

protein and about 3-4 mg purified protein can be obtained for these two fragments.

4.5 Conclusion

The C-terminal 19 amino acids of the (pro)rennin receptor corresponding to the
cytoplasmic tail were fused into the C-terminus of E. coli maltose binding protein (MBP),
creating the MBP-(P)RR19 fusion protein. The chimera was expressed in E. coli and
purified to homogeneity. Protein crystals that are both in presence and in absence of the
MBP ligand, maltose, were obtained and X-ray diffraction data were collected. The

crystals were diffracted to a resolution of up to 1.996 A and belong to the space group

178

P212121. Depending on the presence or absence of maltose, the crystals have significantly
different unit-cell dimensions and molecular packing arrangements for the MBP-(P)RR19
fusion protein.

Structures in both forms were determined by molecular replacement using the available
MBP structures as phasing models. Despite of significantly different unit-cell dimensions
and molecular packing arrangements, there are two monomers in asymmetric unit for
both structures. The first 13 and 11 residues of the (P)RR cytoplasmic tail were included
in chain B and chain A in the ligand-bound model, respectively. And the first 11 and 9
residues were observed in the apo-structure chain B and chain A, respectively. Absence of
the strong electron density for the remaining residues suggested that this region of the
cyt0plasmic tail may be disordered. The available residues showed a structure of
relatively flexible loop without obvious helices or stands presented. The major
non-crystallographic interactions were predominated by the residues in the cytoplasmic
tail, particularly by the Tyr 369 and Tyr 374, suggesting roles of the cytoplasmic tail in
protein oligomerization.

The PLZF full-length and individual domains have been cloned into expression vectors
by fusing the E. coli MBP in their N-terminus as expression aids. These PLZF species
have been expressed in E. coli and have been purified. The protein-protein interactions of
the PLZF individual domains with the (P)RR cytoplasmic tail have been investigated.

Please refer to “Appendix” for preliminary binding results.

179

References
(1994). "The CCP4 suite: programs for protein crystallography." Acta Crystallogr D Biol
Cmstallogr 50(Pt 5): 760-3.

Brummelkarnp, T. R., R. Bemards and R. Agami (2002). "A system for stable expression
of short interfering RN As in mammalian cells." Science 296(5567): 550-3.

 

Duan, X. and F. A. Quiocho (2002). "Structural evidence for a dominant role of nonpolar
interactions in the binding of a transport/chemosensory receptor to its highly polar
ligands." Biochemistm 41(3): 706-12.

Emsley, P. and K. Cowtan (2004). "Coot: model-building tools for molecular graphics."
Acta Crystallogr D Biol Crystallogr 60(Pt 12 Pt 1): 2126-32.

Kobe, B., R. J. Center, B. E. Kemp and P. Poumbourios (1999). "Crystal structure of
human T cell leukemia virus type 1 gp21 ectodomain crystallized as a
maltose-binding protein chimera reveals structural evolution of retroviral
transmembrane proteins." Proc Natl Ag Sci U S A 96(8): 4319-24.

Liu, Y., A. Manna, R. Li, W. E. Martin, R. C. Murphy, A. L. Cheung and G Zhang (2001).
"Crystal structure of the SarR protein from Staphylococcus aureus." Proc Natl
Acad Sci U S A 98(12): 6877-82.

Otwinowski, Z. and W. Minor (1997). "Processing of X-ray diffraction data collected in
oscillation mode." Macromolecular Crystallography. Pt A 276: 307-326.

Quiocho, F. A., J. C. Spurlino and L. E. Rodseth (1997). "Extensive features of tight
oligosaccharide binding revealed in high-resolution structures of the maltodextrin
transport/chemosensory receptor." Structure 5(8): 997-1015.

Schefe, J. H., M. Menk, J. Reinemund, K. Effertz, R. M. Hobbs, P. P. Pandolfi, P. Ruiz, T.
Unger and H. Funke-Kaiser (2006). "A novel signal transduction cascade
involving direct physical interaction of the renin/prorenin receptor with the
transcription factor promyelocytic zinc finger protein." Circ Res 99(12): 1355-66.

 

Winn, M. D., M. N. Isupov and G N. Murshudov (2001). "Use of T LS parameters to
model anisotropic displacements in macromolecular refinement." Acta
Crystallogr D Biol Crystallogr 57(Pt 1): 122-33.

180

CHAPTER 5

Future directions

181

The biochemical and structural properties of DLP-1 and selected mutants have been
characterized, and the major functions of individual domains on GTPase activities, intra-
and intermolecular interactions, and membrane targeting have been identified or
confirmed in this thesis. However, more direct evidence from high-resolution crystal
structures is needed to elucidate functions of DLP-1 in more detail and reveal the
molecular mechanisms of mitochondrial fission and fusion. This need for structural
information of DLP-1 and MFNs has become a much more critical issue. As
protein-engineering methods have not yet yielded successful routes to solve DLP-1 and
MFN-1 structure, alternative strategies and methods will need to be explored in the future

to obtain protein crystals.

5 .1 Alternative protein expression system_s_gr_t_d crystallization screening methods

DLP-1 and MFNs, which are both eukaryotic proteins, were expressed in E. coli
mostly for easy manipulation and low-cost purposes. Although it is common to express
eukaryotic proteins for structural studies in prokaryotic expression systems, alternative
eukaryotic expression systems may be more appropriate for DLP-l and MFNs,
particularly for ensuring proper folding and membrane insertion. As MFN-1 was
expressed mainly as inclusion bodies in E. coli, other expression systems may improve
the yield of folded and soluble protein. However, since the overexpression of DLP-l or
MFN-1 in yeast, insect, or mammalian cells might seriously disrupt mitochondrial fission

or fusion, it may markedly compromise cell viability. Baculoviral-driven expression in

182

insect cells might be a suitable alternative expression system as the infected insect cells
need to survive for 24-72 hours during a normal protein expression experiment.

DLP-1 is a membrane-interacting protein, which probably requires lipids to stabilize its
local structures such as the PH-like domain. Therefore, crystallization screening with a
much wider range of detergents may be necessary to trap an ordered DLP-l protein
conformation that promotes crystallization. Although non-ionic detergents such as
dodecyl maltoside, decyl maltoside, and octyl glucoside have been attempted, there are
many more nonionic and zwitterionic detergents that might lead to DLP-1 crystallization.
Usually detergents screening is performed after initial hits were found to optimize further
the crystal grth conditions. This scenario is the most efficient as it minimizes the
screening strategies, which in turn reduces the workload. Since there are currently no
potential crystallization condition found for DLP-1 or DLP-1 fragments, and DLP-1
fusions, large scale brute-force screening with 10-20 different detergent types may need
to be performed. The up-coming acquisition of a new crystallization robot will make this
rather daunting task much more tractable.

For the purified MFNs, an alternative membrane protein crystallization method, lipidic
cubic phase, also called “in meso” method, can be used. Several membrane proteins have
been crystallized for hi gh-resolution structure determination by using this method (Kolbe,
Besir et al. 2000; Luecke, Schobert et a1. 2001; Gordeliy, Labalm et al. 2002; Katona,
Andreasson et al. 2003; Joharrsson, Wohri et al. 2009). In the cubic phase, the lipidic

compartments are interpenetrated by a freely communicating system of aqueous channels

183

(Landau and Rosenbusch 1996). Although the exact mechanism of “in meso”
crystallization remains unclear, the cubic phase may provide a lipid bilayer that is an
environment similar to the biological membranes. The membrane protein may
reconstitute into the bilayer and crystals nucleate and grow upon addition of precipitants
(Caffrey 2003). In addition, several recently crystallized membrane proteins have led the
lipidic bicelle as another alternative tool for crystallizing MFNs (F aharn, Boulting et al.
2005; Rasmussen, Choi et a1. 2007; Luecke, Schobert et a1. 2008; ijal, Cascio et al.
2008). The bicelle method could be considered a combination of the cubic phase and the
traditional detergent crystallization method. The bicelles are generated by mixing lipids
dimyristoyl phosphatidylcholine (DMPC) or ditridecanoyl phosphatidylcholine (DTPC)
with the detergent CHAPSO or nonyl maltoside (Johansson, Wohri et al. 2009). Like the
cubic phase, the bicelles provide a more bilayer-like environment for membrane proteins
than detergents. Given the relatively large amount of refolded, recombinant MFN-1 that
we have in hand, we intend to use the lipidic cubic phase and the bicelle methods to

screen for crystallization conditions for MFN-1.

5.2 Further structural studies on (P)RR and roles oftyrosines on the cytoplasmic tail in
protein oligomerization and signal transduction

The crystal structure of the cytoplasmic tail of (P)RR has been determined using MBP
fusion method and part of its flmction has been suggested by the structure. However, the

cytoplasmic domain is only a small portion of (P)RR. We need to determine more of the

184

(P)RR structure to elucidate more function and roles of (P)RR in blood pressure
regulation and organelle functions. Therefore, optimization of the crystal growth
conditions for the full-length (P)RR aiming to obtain better crystals suitable for X-ray
diffraction would be one of the major future directions. Since the (P)RR is an integral
membrane protein, screening different detergents may help shield hydrophobic regions
and produce less aggregated species and also promote better crystals. Meanwhile, the
lipidic cubic phase and bicelle method are applicable for (P)RR. In addition, if the poor
crystals of (P)RR are due to disordered or flexible regions, protein engineering may be
used to explore compact functional domains for structural studies. Currently, the most
promising fi'agments we have been considering would be the N-terminal extracelluar
domain.

Although some structural information was obtained for the MBP-(P)RR19 crystals, the
last six residues of the cytoplasmic tail were not observed in the electron-density maps.
To resolve the disordered regions in the MBP-(P)RRl9icrystals, we are now exploring
the use of alternative crystal freezing methods such as dehydration and reannealing to
reduce conformational flexibility of the protein termini in the MBP-(P)RR19 fusion. In
addition, since the two tyrosines were observed in the crystal structures of the
MBP-(P)RR19 to be important for dimeric interactions, more evidence from firnctional
and genetic studies such as site-directed mutagenesis may be pursued to support the
functions of the two tyrosines in protein oligomerization. For example, the protein

oligomeric state can be investigated be size exclusion chromatographic and

185

crystallographic studies after the tyrosines are mutated to functional-unrelated residues.
Also, (P)RR with these two tyrosines mutated can be transformed into mammalian cells
to study their roles in signal transduction by measuring the levels of activations of the

downstream signal pathways.

186

References

Caffrey, M. (2003). "Membrane protein crystallization." J Struct Biol 142(1): 108-32.

Faham, S., G L. Boulting, E. A. Massey, S. Yohannan, D. Yang and J. U. Bowie (2005).
"Crystallization of bacteriorhodopsin from bicelle formulations at room
temperature." Protein Sci 14(3): 836-40.

Gordeliy, V. I., J. Labahn, R. Moukhametzianov, R. Efiemov, J. Granzin, R. Schlesinger,
G Buldt, T. Savopol, A. J. Scheidig, J. P. Klare and M. Engelhard (2002).
"Molecular basis of transmembrane signalling by sensory rhodopsin H-transducer
complex." Nature 419(6906): 484-7.

J ohansson, L. C., A. B. Wohri, G Katona, S. Engstrom and R. Neutze (2009). "Membrane
protein crystallization from lipidic phases." Curr Opin Struct Biol 19(4): 372-8.

Katona, G, U. Andreasson, E. M. Landau, L. E. Andreasson and R. Neutze (2003).
"Lipidic cubic phase crystal structure of the photosynthetic reaction centre from
Rhodobacter sphaeroides at 2.35A resolution." J Mol Biol 331(3): 681-92.

Kolbe, M., H. Besir, L. O. Essen and D. Oesterhelt (2000). "Structure of the light—driven
chloride pump halorhodopsin at 1.8 A resolution." Science 288(5470): 1390-6.

 

Landau, E. M. and J. P. Rosenbusch (1996). "Lipidic cubic phases: a novel concept for
the crystallization of membrane proteins." Proc Naftl Acad Sci U S A 93(25):
14532-5.

Luecke, H., B. Schobert, J. K. Lanyi, E. N. Spudich and J. L. Spudich (2001). "Crystal
structure of sensory rhodopsin H at 2.4 angstroms: insights into color tuning and
transducer interaction." Science 293(5534): 1499-503.

 

Luecke, H., B. Schobert, J. Stagno, E. S. Irnasheva, J. M. Wang, S. P. Balashov and J. K.
Lanyi (2008). "Crystallographic structure of xanthorhodopsin, the light-driven
proton pump with a dual chromophore." Proc Natl Acad Sci U S A 105(43):
16561-5.

Rasmussen, S. G, H. J. Choi, D. M. Rosenbaum, T. S. Kobilka, F. S. Thian, P. C.
Edwards, M. Burghammer, V. R. Ratnala, R. Sanishvili, R. F. Fischetti, G F.
Schertler, W. I. Weis and B. K. Kobilka (2007). "Crystal structure of the human
beta2 adrenergic G-protein-coupled receptor." Nature 450(7168): 383-7.

187

ijal, R., D. Cascio, J. P. Colletier, s. Faham, J. Zhang, L. Toro, P. Ping and J.
Abrarnson (2008). "The crystal structure of mouse VDACl at 2.3 A resolution

reveals mechanistic insights into metabolite gating." Proc Natl Acad Sci U S A
105(46): 17742-7.

188

CHAPTER 5

Future directions

189

The biochemical and structural properties of DLP-1 and selected mutants have been
characterized, and the major fimctions of individual domains on GTPase activities, intra-
and intermolecular interactions, and membrane targeting have been identified or
confirmed in this thesis. However, more direct evidence from high-resolution crystal
structures is needed to elucidate functions of DLP-l in more detail and reveal the
molecular mechanisms of mitochondrial fission and fusion. This need for structural
information of DLP-l and MFNs has become a much more critical issue. As
protein-engineering methods have not yet yielded successful routes to solve DLP-1 and
MFN-l structure, alternative strategies and methods will need to be explored in the future

to obtain protein crystals.

5.1 Alternative protein eLpression systems and crystallization screeningetlgalq

DLP-1 and MFNs, which are both eukaryotic proteins, were expressed in E. coli
mostly for easy manipulation and low-cost purposes. Although it is common to express
eukaryotic proteins for structural studies in prokaryotic expression systems, alternative
eukaryotic expression systems may be more appropriate for DLP-l and MFNs,
particularly for ensuring proper folding and membrane insertion. As MFN-1 was
expressed mainly as inclusion bodies in E. coli, other expression systems may improve
the yield of folded and soluble protein. However, since the overexpression of DLP-1 or
MFN-1 in yeast, insect, or mammalian cells might seriously disrupt mitochondrial fission

or fusion, it may markedly compromise cell viability. Baculoviral-driven expression in

190

insect cells might be a suitable alternative expression system as the infected insect cells
need to survive for 24-72 hours during a normal protein expression experiment.

DLP-1 is a membrane-interacting protein, which probably requires lipids to stabilize its
local structures such as the PH-like domain. Therefore, crystallization screening with a
much wider range of detergents may be necessary to trap an ordered DLP-l protein
conformation that promotes crystallization. Although non-ionic detergents such as
dodecyl maltoside, decyl maltoside, and octyl glucoside have been attempted, there are
many more nonionic and zwitterionic detergents that might lead to DLP-1 crystallization.
Usually detergents screening is performed after initial hits were found to optimize further
the crystal growth conditions. This scenario is the most efficient as it minimizes the
screening strategies, which in turn reduces the workload. Since there are currently no
potential crystallization condition found for DLP-1 or DLP-1 fragments, and DLP-1
fusions, large scale brute-force screening with 10-20 different detergent types may need
to be performed. The up-coming acquisition of a new crystallization robot will make this
rather daunting task much more tractable.

For the purified MFNs, an alternative membrane protein crystallization method, lipidic
cubic phase, also called “in meso” method, can be used. Several membrane proteins have
been crystallized for high-resolution structure determination by using this method (Kolbe,
Besir et al. 2000; Luecke, Schobert et al. 2001; Gordeliy, Labalm et al. 2002; Katona,
Andreasson et al. 2003; Johansson, Wohri et al. 2009). In the cubic phase, the lipidic

compartments are interpenetrated by a freely communicating system of aqueous channels

191

(Landau and Rosenbusch 1996). Although the exact mechanism of “in meso”
crystallization remains unclear, the cubic phase may provide a lipid bilayer that is an
environment similar to the biological membranes. The membrane protein may
reconstitute into the bilayer and crystals nucleate and grow upon addition of precipitants
(Caffrey 2003). In addition, several recently crystallized membrane proteins have led the
lipidic bicelle as another alternative tool for crystallizing MFNs (Faham, Boulting et al.
2005; Rasmussen, Choi et al. 2007; Luecke, Schobert et al. 2008; ijal, Cascio et al.
2008). The bicelle method could be considered a combination of the cubic phase and the
traditional detergent crystallization method. The bicelles are generated by mixing lipids
dimyristoyl phosphatidylcholine (DMPC) or ditridecanoyl phosphatidylcholine (DTPC)
with the detergent CHAPSO or nonyl maltoside (Johansson, Wohri et al. 2009). Like the
cubic phase, the bicelles provide a more bilayer-like environment for membrane proteins
than detergents. Given the relatively large amount of refolded, recombinant MFN-1 that
we have in hand, we intend to use the lipidic cubic phase and the bicelle methods to

screen for crystallization conditions for MFN-1.

5.2 Further structural studies on (P)RR and roles of tyrosines on the cytgplasmic tail in
protein oligomerization and signal transduction

The crystal structure of the cytoplasmic tail of (P)RR has been determined using MBP
fusion method and part of its function has been suggested by the structure. However, the

cytoplasmic domain is only a small portion of (P)RR. We need to deternrine more of the

192

(P)RR structure to elucidate more function and roles of (P)RR in blood pressure
regulation and organelle functions. Therefore, optimization of the crystal grth
conditions for the full-length (P)RR aiming to obtain better crystals suitable for X-ray
diffraction would be one of the major future directions. Since the (P)RR is an integral
membrane protein, screening different detergents may help shield hydrophobic regions
and produce less aggregated species and also promote better crystals. Meanwhile, the
lipidic cubic phase and bicelle method are applicable for (P)RR. In addition, if the poor
crystals of (P)RR are due to disordered or flexible regions, protein engineering may be
used to explore compact functional domains for structural studies. Currently, the most
promising fi'agments we have been considering would be the N-terminal extracelluar
domain.

Although some structural information was obtained for the MBP-(P)RR19 crystals, the
last six residues of the cytoplasmic tail were not observed in the electron-density maps.
To resolve the disordered regions in the MBP-(P)RR19 crystals, we are now exploring
the use of alternative crystal freezing methods such as dehydration and reannealing to
reduce conformational flexibility of the protein termini in the MBP-(P)RR19 fusion. In
addition, since the two tyrosines were observed in the crystal structures of the
MBP-(P)RR19 to be important for dimeric interactions, more evidence from filnctional
and genetic studies such as site-directed mutagenesis may be pursued to support the
functions of the two tyrosines in protein oligomerization. For example, the protein

oligomeric state can be investigated be size exclusion chromatographic and

193

crystallographic studies after the tyrosines are mutated to functional-unrelated residues.
Also, (P)RR with these two tyrosines mutated can be transformed into mammalian cells
to study their roles in signal transduction by measuring the levels of activations of the

downstream signal pathways.

194

References

Caffrey, M. (2003). "Membrane protein crystallization." J Struct Biol 142(1): 108-32.

Faham, S., G L. Boulting, E. A. Massey, S. Yohannan, D. Yang and J. U. Bowie (2005).
"Crystallization of bacteriorhodopsin fiom bicelle formulations at room
temperature." Protein Sci 14(3): 836-40.

Gordeliy, V. I., J. Labahn, R. Moukhametzianov, R. Efremov, J. Granzin, R. Schlesinger,
G Buldt, T. Savopol, A. J. Scheidig, J. P. Klare and M. Engelhard (2002).
"Molecular basis of transmembrane signalling by sensory rhodopsin II-transducer
complex." Nature 419(6906): 484-7.

 

J ohansson, L. C., A. B. Wohri, G Katona, S. Engstrom and R. Neutze (2009). "Membrane
protein crystallization from lipidic phases." Curr Orrin Struct Biol 19(4): 372-8.

Katona, G, U. Andreasson, E. M. Landau, L. E. Andreasson and R. Neutze (2003).
"Lipidic cubic phase crystal structure of the photosynthetic reaction centre from
Rhodobacter sphaeroides at 2.35A resolution." J Mol Biol 331(3): 681-92.

Kolbe, M., H. Besir, L. O. Essen and D. Oesterhelt (2000). "Structure of the light-driven
chloride pump halorhodopsin at 1.8 A resolution." Science 288(5470): 1390-6.

 

Landau, E. M. and J. P. Rosenbusch (1996). "Lipidic cubic phases: a novel concept for
the crystallization of membrane proteins." Proc Natl Acad Sci U S A 93(25):
14532-5.

Luecke, H., B. Schobert, J. K. Lanyi, E. N. Spudich and J. L. Spudich (2001). "Crystal
structure of sensory rhodopsin II at 2.4 angstroms: insights into color tuning and
transducer interaction." Science 293(5534): 1499-503.

 

Luecke, H., B. Schobert, J. Stagno, E. S. Irnasheva, J. M. Wang, S. P. Balashov and J. K.
Lanyi (2008). "Crystallographic structure of xanthorhodopsin, the light-driven
proton pump with a dual chromophore." Proc Natl Acad Sci U S A 105(43):

‘ 16561-5.

Rasmussen, S. G, H. J. Choi, D. M. Rosenbaum, T. S. Kobilka, F. S. Thian, P. C.
Edwards, M. Burghammer, V. R. Ratnala, R. Sanishvili, R. F. Fischetti, G F.
Schertler, W. I. Weis and B. K. Kobilka (2007). "Crystal structure of the human
beta2 adrenergic G-protein—coupled receptor." Nature 450(7168): 383-7.

195

ijal, R., D. Cascio, J. P. Colletier, S. Faham, J. Zhang, L. Toro, P. Ping and J.
Abrarnson (2008). "The crystal structure of mouse VDACl at 2.3 A resolution
reveals mechanistic insights into metabolite gating." Proc Natl Acad Sci U S A
105(46): 17742-7.

196

Appendix: Preliminary mapping of PLZF fragments binding to the

(P)RR cytoplasmic tail

A.1 Methods

Protein—protein interaction assays - The purified PLZF hill-length or individual domains
were buffer exchanged to buffer A (50 mM sodium phosphate, 300 mM NaCl, pH 8.0)
and mixed with purified MBP fusion protein containing the C-terminus 19 amino acids of
the (P)RR (MBP-(P)RR19). The mixture was incubated at room temperature for 8 hrs
with shaking to allow maximal protein-protein interactions. The 1 ml solution was loaded
onto a column containing 2 ml Ni-NTA agarose slurry pre-equilibrated with Buffer B (50
mM sodium phosphate, 300 mM NaCl, 0.1 mM EDTA, pH 8.0) to allow unbound protein
to flow through. The column was then washed with buffer C (50 mM sodium phosphate,
300 mM NaCl, 20 mM imidazole, pH 8.0) for over 20 column volume. Proteins bound
column was eluted by Buffer D (50 mM sodium phosphate, 300 mM NaCl, 200 mM
imidazole, pH 8.0) and the composition of the eluants was analyzed by SDS-PAGE and

Westem-blot assays.

A.2 Results and discussion
A.2.I Preliminary results g” PLZF binding to the cytoplasmic tail of the (P)RR — The
PLZF fragments were tagged with both MBP and 6Xhis and the (P)RR19 was tagged with

only MBP. Therefore, when incubating the binding mixture onto a Ni-NTA column, the

197

PLZF fi'agments can be immobilized onto the column, and the MBP-(P)RR19 will flow
through unless binds PLZF fragments. The binding can be detected by SDS-PAGE and
Western blotting using anti-MBP antibody. Since the full-length PLZF protein was
purified to low level that was not sufficient for binding assays, I directly tested the
available individual domains for binding to the (P)RR cytoplasmic domain. The
SDS-PAGE and Western blotting showed considerable amount of MBP-(P)RR19
presented in the elution fraction of the binding assay together with the PLZF RD2 domain
(Fig. A1). Much less amount of MBP-(P)RR19 was observed together with the BTB
domain (Fig. A2) and negligible amount was present together with the zinc finger domain
(Fig. A3). To eliminate the possibilities that the RD2 domain interact with MBP in the
MBP-(P)RR19, MBP protein without any tags on was expressed and purified (Fig. A4).
And the MBP did not show any binding to the RD2 domain (Fig. AS). All these results
indicated that the RD2 domain is probably the major region of PLZF that is; responsible

for interactions with the (P)RR cytoplasmic domain.

A.2.2 Future directions — Since there is a Proline-rich region in the PLZF RD2 domain,
a motif that is commonly involved in protein-protein interactions, it is possible that this
region plays important roles in interacting with the (P)RR cytoplasmic domain. Therefore,
in the future, investigating roles of this motif by binding assays with the MBP-(P)RR19
may be necessary.

In addition, co-structure of the PLZF RD2 domain with the cytoplasmic tail of the

198

(P)RR would provide structural information for the molecular mechanisms of the
functions of both the PLZF and the (P)RR. Therefore, co-crystallization of the PLZF
RD2 domain (or possibly the BTB domain) with the (P)RR19 peptide may be another

future dictions for studies on the (P)RR-PLZF signal transduction pathway.

 

MW E 108E MW E 108E

Figure A1: Binding of the PLZF RD2 domain to the MBP-(P)RR19. (A) SDS-PAGE.
(B) Westem-blot using anti-MBP as primary antibody (New England Biolabs). MW,
Molecular weight standard; E, Elution from the binding assay; 10XE, 10 times
concentrated elution form the binding assay. Arrows indicate the MBP-(P)RR19.

199

80
60
50

._ 4o
30

       

20

 

Amw”: aging E 5XE 10XE

Figure A2: Binding of the PLZF BTB domain to the MBP-(P)RR19. (A) SDS-PAGE.
(B) Westem-blot using anti-MBP as primary antibody (New England Biolabs). MW,
Molecular weight standard; E, Elution from the binding assay; 5XE, 5 times concentrated
elution form the binding assay. 10XE, 10 times concentrated elution form the binding
assay. Arrows indicate the MBP-(P)RR19.

200

 

mw E 28E 48E 533E 108E

Figure A3: Binding of the PLZF zinc finger domain to the MBP-(P)RR19. (A)
SDS-PAGE. (B) Westem-blot using anti-MBP as primary antibody (New England
Biolabs). MW, Molecular weight standard; E, Elution from the binding assay; 2XE, 4XE,
5XE, and 10XE, 2, 4, 5, and 10 times, respectively, concentrated elution form the binding
assay. Arrows indicate the MBP-(P)RR19.

201

 

Figure A4: Purification of E. coli maltose-binding protein (MBP) by ion exchange
chromatography. The elution from amylose column was concentrated and used for
further purification on an anionic exchange column. The left lane shows molecular
weight standard. Other lanes show the fractions containing purified MBP.

   

MW E 108E

Figure A5: Binding of the PLZF RD2 domain to the MBP. (A) SDS-PAGE. (B)
Westem-blot using anti-MBP as primary antibody (New England Biolabs). MW,
Molecular weight standard; E, Elution from the binding assay; 10XE, 10 times
concentrated elution form the binding assay. Arrows indicate the size of MBP.

202

   

m11111111111111|1111111111“

239