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 KrlPrqlAcc&ProsICIRCIDateDuo.indd

 

COORDINATION OF DIVISION COMPLEXES ACROSS THE PLASTID
ENVELOPE MEMBRANES

By

Jonathan Matthew Glynn

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Genetics

2009

ABSTRACT

COORDINATION OF DIVISION COMPLEXES ACROSS THE PLASTID
ENVELOPE MEMBRANES

By
Jonathan Matthew Glynn
Chloroplasts are cellular organelles descended from a cyanobacterial
endosymbiont and house the photochemical machinery which powers synthesis of
reduced carbon compounds from carbon dioxide and water. In higher plants, the
chloroplast is also the site of synthesis for a select group of lipids, amino acids, and plant
hormones. This set of organelle-specific functions make the chloroplast essential to
survival of land plants, the most prominent group of terrestrial primary producers. The
replication and segregation of plastids within land plants occurs through binary fission
and is an important part of plant cell biology, as chloroplast size and number may impact
the efficiency of photosynthesis and the partitioning of chloroplasts to daughter cells

during plant cell division.

The apparatus that facilitates the scission of a single chloroplast into two daughter
chloroplasts is a complex macromolecular machine, partly composed of a host-derived
dynamin ring on the outside of the organelle and an endosymbiont-derived FtsZ ring (Z-
ring) inside the organelle. The activities of these two rings must be tightly coordinated
across the two envelope membranes of the chloroplast to ensure the timely progression

and completion of division.

Here, I show that ARC6, a known FtsZ assembly factor that promotes formation
of FtsZ filaments, also specifies the mid-plastid positioning of the paralogous outer
envelope proteins PDVl and PDV2, which have parallel functions in dynamin
recruitment. PDV2 positioning requires a direct interaction between ARC6 and PDV2
that may be regulated by post-translational modification of ARC6 within the
interrnembrane space. I also show that PARC6 (Paralog of ARC6), like ARC6, is a
multifunctional inner envelope chloroplast division protein. PARC6 acts downstream of
ARC6 to position PDVl at the division site, but is not required for PDV2 or ARCS
localization. Arabidopsis parc6 mutants exhibit compound chloroplast division
phenotypes and FtsZ filament morphology defects suggesting that PARC6 acts
antagonistically to ARC6 within the chloroplast stoma as an inhibitor of FtsZ assembly.
This FtsZ assembly-inhibiting activity of PARC6 may occur through interaction with
ARC3, a protein with functional similarity to bacterial MinC. PARC6-OPP localization
is dynamic, consistent with its complex role in division. Our findings indicate that
PARC6 and ARC6 play related but distinct roles in coordinating the internal and external

components of the chloroplast division complex.

DEDICATION

This work is dedicated to my mother and father, both of whom instilled the values and

ideals that are a critical component of any scientist’s success.

This work is dedicated to my grandfather, Elmer E. Lehman, whose patience,

thoughtfulness, and pragmatic approach to life were an inspiration to all who knew him.

This work is dedicated to Van McWilliams and Helmut Bertrand, two teachers who kept

me interested in science at times when I might have followed another path.

This work is dedicated to my family, whose understanding and encouragement over the

years have been irreplaceable.
Most importantly, this work is dedicated to Angie. Your attitude, patience, kindness,

understanding, support, and smile have helped make this possible. You are the love of

my life and I am thankful for every day we have together.

iv

ACKNOWLEDGEMENTS

I wish to acknowledge my advisor, Katherine W. Osteryoung, for her support and

approach to mentoring that fosters independent critical thinking.

I wish to thank the members of my guidance committee for their encouragement
and useful commentary:
Dr. Jianping Hu
Dr. Robert Last
Dr. Beronda Mongomery-Kaguri

Dr. Andreas Weber

I wish to thank Aaron Schmitz, Deena Kadirj an—Kalbach, Shin-ya Miyagishima, Qiang
Wang, and Yue Yang for helpful scientific discussions and friendship over the past

several years.

I wish to thank John Froehlich for insightful experimental advice, useful scientific

discussions, and technical support throughout this project.

TABLE OF CONTENTS

LIST OF FIGURES .................................................................................. ix
LIST OF ABBREVIATIONS ........................................................................ xiii
CHAPTER 1
EVOLUTION AND OPERATION OF THE CHLOROPLAST DIVISION
MACHINERY ...................................................................................................................... 1
Plastid Biology and Analysis of Chloroplast Division ............................................. 2
A Survey of the Molecular Biology of Plastid Division .......................................... 7
Regulating Plastid Fission: Why do Plastids Divide in Land Plants? .................... 24
Perspectives and Acknowledgements ...................................................................... 29
CHAPTER 2
IDENTIFICATION AND MODELING OF AN FTZ2-BINDIN G DOMAIN WITHIN
ARC6 .................................................................................................................................. 30
Abstract ................................................................................................................. 31
Introduction ............................................................................................................ 32
Results .................................................................................................................... 35
The F tsZ2-binding domain of ARC6 occupies amino acids 351 -503 ......... 35
The C-terminus of F tsZ2 is suflicient for ARC6-FtsZ2 interaction ........... 37
ARC6ZBD can be modeled onto the structure of ZipA ZBD ........................ 39
Mutations that aflect ZipAZBD-FtsZ interaction afl'ect ARC 6 ZB D-F tsZ2
interaction .................................................................................................. 44
Analysis of the ARC6ZBD-FtsZ2 interaction by pulldown assay ............... 46
In vivo analysis of the ARC 6 F44 2 D site-directed mutation ........................ 47
In vivo analysis of chloroplast-targeted ARC6 ZBD ................................... 53
Discussion ............................................................................................................. 58
Materials and Methods .......................................................................................... 64
CHAPTER 3
ARC6 BINDS AND POSITIONS PDV2 DURING PLASTID FISSION IN
ARABIDOPSIS ................................................................................................................... 70
Abstract ................................................................................................................. 71
Introduction ............................................................................................................ 72
Results .................................................................................................................... 75
PD V2 localization and topology are similar to PDVI ............................... 75
The [MS regions of ARC6 and PD V2 interact ............................................ 81
PD V2 family members possess a unique C-terminal domain .................... 83
The conserved terminal glycine of PD V2 is required for interaction with
ARC6 and for PDVZfimction in vivo .............................................................. 83

vi

The complete IMS region of ARC6 is not required for ARC6 localization

but is required for chloroplast division activity .......................................... 90
ARC6 acts upstream of PD V2 and is required for PD V2 activity ............. 92
ARC6 is required for positioning PD V2, PD V1, and ARC5 ...................... 94
Discussion ............................................................................................................ 100
Materials and Methods ......................................................................................... 104
CHAPTER 4
CHARACTERIZATION OF THE PDV2 BINDING DOMAIN OF ARC6 ................... 110
Abstract ............................................................................................................... 111
Introduction .......................................................................................................... 1 12
Results .................................................................................................................. 114
The PDV2 binding domain of ARC6 occupies amino acids 63 6- 759 ....... 114
Structural Analysis of ARC6 p31) and PD V2IMS ...................................... 1 16
A conserved serine within ARC 61231) influences ARC 6-PD V2
interaction ........................................................................ 120
A phosphomimetic mutation in ARC6 p31) causes chloroplast division
defects in vivo ........................................................................................... 124
PD V2, PD VI, ARC5, and F tsZ localize to the division site in ARC 63744 E
mutants ..................................................................................................... 126
Discussion ............................................................................................................ 1 30
Materials and Methods ......................................................................................... 136
CHAPTER 5
PARC6 INFLUENCES FTSZ ASSEMBLY AND PDVl RECRUITMENT IN
ARABIDOPSIS ...................................................................................................... ....... 140
Abstract ............................................................................................................... 141
Introduction ........................................................................................................... 142
Results .................................................................................................................... 145
PARC6 family members are distinct fiom ARC 6 and are unique to
vascular plants ......................................................................................... 145
Mutations in PARC6 cause aberrations in chloroplast morphology and
F tsZ filament morphology ........................................................................ 149
PARC6 is an inner envelope protein with localization similar
to PD VI .................................................................................................... 151
PARC6 inhibits F tsZ assembly and interacts with ARC3 ........................ 155
PARC6 acts downstream of ARC 6 ........................................................... 157
PARC6 is required for PD VI localization, but not for PD V2 or ARC5
localization ............................................................................................... 161
PARC 6 binds the cytosolic domain of PD V1 in two-hybrid assays ......... 163
PD V1 and PD V2 independently localize to the division site ................... 164
Discussion ............................................................................................................ 167
Materials and Methods ......................................................................................... 173

vii

CHAPTER 6
A CONSERVED ASPARTATE WITHIN ARC6 INFLUENCES PLASTID SIZE AND

Z-RING POSITION .......................................................................................................... 178
Abstract ................................................................................................................ 179
Introduction .......................................................................................................... 180
Results .................................................................................................................. 183

arc6 D 20 5N is a hypomorphic allele of ARC6 that is associated with
defects in chloroplast morphology and number ....................................... 183
ARC 6 D 20 5 is conserved in land plants, algae, and cyanobacteria ........ 186
arc6 D20 5N is associated with miniaturized and misplaced plastidic Z-
rings, as well as a decrease in ARC6 protein levels ................................ 186
AtMinE— YF P has a unique localization pattern in Arabidopsis and
ARC60205 mutants do not interact with AtMinE ....................................... 189
ARC 6 probably functions downstream of F tsZI/F tsZ2 assembly and
downstream of A tMinE ............................................................................. 194
Discussion ............................................................................................................ 196
Materials and Methods ......................................................................................... 200

CHAPTER 7

CONCLUSIONS AND FUTURE DIRECTIONS ........................................................... 205
Summary of ARC6 and PARC6 Functional Analysis and Future Directions ...... 206
Unanswered Questions, Additional Observations, and New Hypotheses ............ 208
Working Model of the Coordination of the Chloroplast Divisome in
Tracheophytes ....................................................................................................... 21 5

APPENDICES ................................................................................................................. 224

REFERENCES ................................................................................................................ 242

viii

LIST OF FIGURES
Images in this dissertation are presented in color.
Figure 1.1. overview of chloroplast morphology and features of a typical chloroplast. . .3
Figure 1.2. Comparison of dividing chloroplasts and cyanobacteria. . . . . . . . . . .5

Figure 1.3. F tsZ2-1 immunolocalization in wild type Arabidopsis chloroplasts and in
arc6 mutants... 15

Figure 1.4. Schematic of the Arabidopsis ARC6 protein ..................................... 16
Figure 1.5. FtsZ and ARC6 colocalize to continuous rings within the chloroplast. . .. 17
Figure 2.1. Mapping the FtsZ2-binding domain of ARC6 by yeast two-hybrid assay. . .36

Figure 2.2. Two-hybrid HIS reporter assays showing ARC6-FtsZ2 interaction ........... 38
Figure 2.3. Alignment of ARC6ZBD with ZipAZBD ........................................... 40

Figure 2.4. Comparison of the E. coli ZipAZBD crystal structure and the A. thaliana

ARC6ZBD homology model ....................................................................... 42
Figure 2.5. Secondary structure predictions of ARC6ZBD .................................... 43
Figure 2.6. ARC6ZBD (1:442) site-directed mutants ............................................ 45
Figure 2.7. Pulldown of FtsZ2 by ARC6 ........................................................ 48

Figure 2.8. Demonstration of ARC6 and ARC6 F44 2D transgene expression in the arc6
(SAIL_693_G04) background .................................................................... 50

Figure 2.9. Quantitative analysis of chloroplast number in ARC6 1:44 ZD-expressing
transgenics ............................................................................................ 51

Figure 2.10. FtsZ immunolocalization in ARC6F44ZD-expressing transgenics. . 5........2

Figure 2.11. Schematic of the plastid-targeting vector pCAMBIA-ZBD used for
ARC6ZBD domain analysis ....................................................................... 55

Figure 2.12. Demonstration of expression and targeting of ARC6ZBD-EYFP..............56

ix

Figure 2.13. Chloroplast and FtsZ morphology in ARC6ZBD-EYFP expressing lines...57

Figure 2.14. ARC6 A A1-509-GFP and ARC6AA1-331-GFP can immunoprecipitate
FtsZ2 ................................................................................................. 63

Figure 3.1. Schematic comparison of PDVl and PDV2 proteins from Arabidopsis. ....76

Figure 3.2. Localization of YFP-PDV2 in Arabidopsis ......................................... 78
Figure 3.3. Colocalization of ARC6-CFP and YFP-PDV2 in Arabidopsis. . . . . . 7.9
Figure 3.4. PDV2 fractionation and topology using isolated pea chloroplasts ............. 80
Figure 3.5. The IMS-localized regions of PDV2 and ARC6 interact ...................... 82
Figure 3.6. Alignment of C-termini of PDVl and PDV2 family members ................ 84

Figure 3.7. The terminal glycine of PDV2 is important for interaction with ARC6. ....85

Figure 3.8. PD V20 307D is a loss-of—function allele ............................................. 87
Figure 3.9. PDVZG307D localization in Arabidopsis .......................................... 89

Figure 3.10. ARC6A1M3-GFP is partially functional and localizes to the division site..91
Figure 3.11. ARC6 acts upstream of PDV2 ...................................................... 93
Figure 3.12. Overexpression of PD V2 does not abrogate the arc6 phenotype ............ 95
Figure 3.13. ARC6 is required for localization of YFP-PDV2 to the division site ........ 97
Figure 3.14. PD V2 is not required for ARC6-GF P localization to equatorial rings. .....98
Figure 3.15. ARC6 is required for localization of GFP-ARCS to the division site. . . .....99
Figure 4.1. The PDV2 binding domain of ARC6 occupies amino acids 636-759 ...... 115

Figure 4.2. Pairwise alignment between the N-terrninal domain of Rst and the IMS
region of PDV2 ........................................................................................ 118

Figure 4.3. Structural model of PDVZIMS ...................................................... 119

Figure 4.4. Identification of a plant-specific motif within the C-termini of ARC6/Fm2
family members ..................................................................................... 121

Figure 4.5. A phosphomimetic mutation in the IMS region of ARC6 decreases its affinity
for PDV2 .............................................................................................. 123

Figure 4.6. ARC6S744E is dysfimctional ........................................................ 125
Figure 4.7. Phenotypes observed in lines expressing ARC6S744 A and ARC 657445. .....127

Figure 4.8. Localization of PDV2, ARC5, PDVl, and FtsZ in ARC6S744E~expressing
transgenics ............................................................................................ 128

Figure 4.9. An unusual phenotype observed in ARC6S744E petioles phenocopies plastidic
Min system defects ................................................................................ 132

Figure 5.1. Summary of PARC6 protein features and similarity to related proteins. . ...146
Figure 5.2. Phylogenetic analysis of PARC6, ARC6, and Ftn2 family members ........ 148
Figure 5.3. Diagram of the PARC6 locus and phenotypic analysis of parc6 mutants... 150

Figure 5.4. PARC6 is a protein that localizes to the division site and polar spots. . . 1 52

Figure 5.5. PARC6 is a chloroplast membrane protein ....................................... 154
Figure 5.6. Analysis of FtsZ morphology and FtsZ protein levels in parc6-1 ............ 156
Figure 5.7. PARC6 does not interact with an FtsZ protein ................................... 158
Figure 5.8. PARC6 interacts with ARC3 ........................................................ 159
Figure 5.9. PARC6 acts downstream of ARC6 ................................................ 160
Figure 5.10. Localization of PDVl, PDV2, and ARC5 in parc6 mutants .................... 162
Figure 5.11. The C-terrninus of PARC6 binds the cytosolic domain of PDVl .......... 165
Figure 5.12. PDVl and PDV2 localize independently of each other ...................... 166

Figure 6.1. arc6 D 20 5N is associated with a defect in chloroplast size and number. . l 85
Figure 6.2. Multiple sequence alignment showing conservation of ARC69205 ......... 187

Figure 6.3. F tsZ localization and mini-chloroplasts in arc6 020 5 N ......................... 188

xi

Figure 6.4. Preliminary analysis of ARC6, Ftle, and FtsZZ levels in arc6Dzo5N. .....190

Figure 6.5. Localization of AtMinE-YFP in Arabidopsis ................................... 192
Figure 6.6. ARC6-GFP does not localize to rings in ftsZ2 or arc12 mutants ............. 195
Figure 7.1. Domain architecture of ARC6 and PARC6 proteins ............................ 207

Figure 7.2. Coordination of Z-ring dynamics and ARC5 recruitment through ARC6 and
PARC6 ............................................................................................... 222

xii

35$

ABRC

AD

ARC

ATP

BD

BiF C

BLAST

CD

CFP

Chl

Col

CT

C-terminal

DIC

EMPTY

EMS

F ts

ABBREVIATIONS

Deletion

Cauliflower Mosaic Virus 35S promoter
Amino acid

Arabidopsis Biological Resource Center
Activation domain

Accumulation and Replication of Chloroplasts
Adenosine-5'-Triphosphate

DNA binding domain

Bimolecular Fluorescence Complementation
Basic Local Alignment Search Tool

Celcius

Circular dichroism

Cyan fluorescent protein

Chlorophyll

Columbia

C-terrninus or C-terminal

Carboxy terminal

Differential interference contrast

Empty vector

Ethane methyl sulfonate

Filamentation temperature sensitive

xiii

FZL
FZO
GC
GDP
GFP
GST
GTP
His
Hiss
IEM
IMS
INT

IP
IPTG
Ler
MORN
NADPH
NMR
N-terminal
OEM
PARC
PBD

PBS

FZO-like

Fuzzy onions

Giant chloroplast
Guanosine-5'-diphosphate

Green fluorescent protein
Glutathione-S-transferase
Guanosine-5'-triphosphate

Histidine

8X Histidine tag

Inner envelope membrane

Intermembrane space

Intermediate

Immunoprecipitation
Isopropyl-beta-D-thiogalactopyranoside
Landsberg erecta

Membrane occupation and recognition nexus
Nicotinamide adenine dinucleotide phosphate (reduced)
Nuclear magnetic resonance spectroscopy
Amino terminal

Outer envelope membrane

Paralog of ARC

PDV2 binding domain

Phosphate-buffered saline

xiv

PD

PDB
PDV

pro

SD
SD/-ULT
SDI-ULTH
SOE-PCR
TBS
THM

TM

TP

Ws

YFP

ZBD

Zip

Plastid dividing

Protein Data Bank

Plastid division

Proline or promoter (subscript)

Synthetic dropout

Synthetic dropout media lacking uracil, leucine, and tryptophan
SD media lacking uracil, leucine, tryptophan, and histidine
Site overlap extension polymerase chain reaction
Tris-buffered saline

Thylakoid membrane

Transmembrane domain

Transit peptide

Wassilewskija

Yellow fluorescent protein

FtsZ (or FtsZZ) binding domain

FtsZ-interacting protein

XV

Chapter 1

Evolution and Operation of the Chloroplast Division Machinery

Plastid Biology and Analysis of Chloroplast Division

Chloroplasts are cellular organelles in land plants, responsible for photosynthesis
and a number of other life-supporting functions, such as amino acid biosynthesis and
lipid metabolism (Pyke 2009). In land plants like Arabidopsis thaliana, each mature leaf
mesophyll cell contains 60-120 chloroplasts (Figure 1.1), depending on the accession and
growth conditions (Aldridge, Maple, et al. 2005, Pyke 1999, Pyke 2009, Pyke and Leech
1994). Chloroplasts within Arabidopsis are typically ovoid-shaped structures, 5-7
microns (pm) in length, and harbor three distinct compartments: the thylakoid lumen, the
stroma, and the intermembrane space (Figure 1.1). Each of these compartments is
bounded by a biological membrane: the thylakoid lumen is surrounded by the thylakoid
membrane, the stroma by the inner envelope membrane, and the entire organelle is
bounded by the outer envelope membrane; the space between the inner and outer
envelopes is called the intermembrane space (IMS). The thylakoid membrane is the site
of photocollection and electron transport that drives photosynthesis, while the thylakoid
lumen stores hydrogen ions generated during the photolysis of water molecules. These
hydrogen ions are transferred to the stroma through an ATP synthase embedded within
the thylakoid membrane to generate ATP within the stroma. The transport of electrons
through the electron transport chain results in the reduction of NADPH. Within the
stroma, ATP is consumed during operation of the Calvin cycle, which generates reduced
carbon (triose phosphate) from carbon dioxide and NADPH. The inner and outer
envelope membranes separate the chloroplast stroma from the surrounding cytosol and

other organelles. These two membranes contain a variety of metabolite transporters

 

Figure 1.1. Overview of chloroplast morphology and features of a typical
chloroplast. The lefi panel shows a typical mesophyll cell from an expanded leaf of
Arabidopsis thaliana accession Col-0. Using DIC microscopy, wild type chloroplasts
typically appear as greenish spherical or ovoid bodies, approximately 5-7 pm in diameter.
The right panel is a schematic detailing the compartments and membranes that make up a
typical chloroplast. Outer Envelope Membrane (OEM); Intermembrane Space (IMS);
Inner Envelope Membrane (IEM); Thylakoid Membrane (THM); and Stromal
compartment (Stroma). Scale bar = 10pm.

and protein translocators that allow the plastid to efficiently interface with the rest of the

cell (Pyke 2009).

The chloroplasts of land plants are decendants of a cyanobacterial endosymbiont
that first inhabited a primitive protozoan about 1.2-1 .8 billion years ago (Dyall, Brown, et
al. 2004, Yoon, Hackett, et al. 2004). Chloroplasts replicate through binary fission
(Schimper 1883), dividing near their midpoint, similar to many bacteria (Figure 1.2).

The process of division may appear relatively simple, but is a highly complex process
that involves a number of protein components. The chloroplasts of land plants maintain a
minimal genome, typically harboring 110-120 genes derived from the cyanobacterial
endosymbiont (Cui, Veeraraghavan, et al. 2006); however, all known plastid division
factors are encoded by nuclear genes in Arabidopsis and other land plants (Yang, Glynn,
et al. 2008). Some of the genes encoding plastid division factors are inventions of the
host organism and arose after a stable endosymbiotic relationship was established, while
others are descendants of cyanobacterial genes that were transferred to the nucleus from

the endosymbiont genome during the evolution of photosynthetic eukaryotes.

Because of their common origin, several components of the chloroplast divisome
share a high degree of sequence similarity with cyanobacterial division proteins from
extant species (Yang, Glynn, et al. 2008). Due to the evolutionary history of the
chloroplast, the study of chloroplast division has benefited greatly from the study of

bacterial cell division. Surprisingly, genetic screens for cyanobacterial cell division

 

 

 

 

 

 

 

-E

Figure 1.2. Comparison of dividing chloroplasts and cyanobacteria. (A)
Micrographs of chloroplasts from Arabidopsis thaliana Col-0 and (B) cells of the
cyanobacterium Synechococcus elongatus PCC7942 as they progress from single
individuals (upper panels) to two new entities (lowest panel). Scale bars = 5pm.

1 v

 

 

 

 

 

components have uncovered only a few novel division genes (Koksharova and Wolk
2002, Miyagishima, Wolk, et al. 2005), but the division of cyanobacteria has been
understudied in comparison to that of Escherichia coli (Lutkenhaus 2007, Margolin 2003,
Margolin 2005, Rothfield, Taghbalout, et al. 2005). Consequently, models of plastid
division at a molecular level are often compared to E. coli cell division models. Many of
the components that make up the divisome of E. coli are also utilized by cyanobacteria
(Miyagishima 2005), but cyanobacteria possess division factors that are unique to
photosynthetic prokaryotes (Koksharova and Wolk 2002, Miyagishima, Wolk, et al.
2005) and are probably the most relevant prokaryotic system for comparative analysis
with regard to the chloroplast. Regardless, the first known component of the plastid
divisome was identified by a reverse genetic approach, parsing sequences of translated
expressed sequence tags (ESTs) from Arabidopsis thaliana for similarity to the amino
acid sequence of the E. coli cell division protein FtsZ (Osteryoung and Vierling 1995).
Later studies identified more components of the chloroplast divisome by reverse genetics,
with the initial query sequences being E. coli cell division proteins (Colletti, Tattersall, et

al. 2000, Itoh, Fujiwara, et al. 2001, Maple, Fujiwara, et al. 2004).

While reverse genetic screens have been useful in identifying broadly-conserved
division proteins common to bacteria and plastids, they fall short in their ability to
identify plastid division factors of host origin, as these factors are generally products of
genes that emerged after the intial endosymbiotic event (Gao, Kadirj an-Kalbach, et al.
2003, Miyagishima, Froehlich, et al. 2006, Miyagishima, Nishida, et al. 2003a). The

discovery of these host-derived components has benefited from forward genetic analysis,

where the leaf cells of mutagenized lines are analyzed for aberrations in chloroplast
number and/or morphology by simple light microscopy (Miyagishima, Froehlich, et al.
2006, Pyke and Leech 1992, Pyke and Leech 1994) and the causative mutation mapped
by routine molecular methods (Jander, Norris, et al. 2002). Forward genetic studies have
been instrumental to the discovery of plant-specific genes involved in plastid division,
uncovering factors that act both inside and outside the organelle. Surprisingly, forward
genetic approaches have only recently uncovered high-level regulators of plastid division,
such as transcription factors and/or hormone responsive factors (Okazaki, Kabeya, et al.
2009), whose existence and involvement in plastid division and differentiation has been
postulated for some time. However, the elucidation of the complete inventory of plastid
division genes will not come through one mode of analysis, but through a comprehensive
approach that incorporates both forward and reverse genetics, comparative genomics,
transcriptome/network analysis, proteomics/interactomics, and integrative approaches

that utilize all of these strategies in multiple experimental systems.

A Survey of the Molecular Biology of Plastid Division

F tsZ

FtsZ is a conserved division protein found in most bacteria and is a structural
homolog of eukaryotic tubulins (Erickson, Taylor, et al. 1996). In E. coli, FtsZ forms a
ring (Z-ring) at the mid-cell which demarcates the site of cell division (Bi and
Lutkenhaus 1991), probably serving as a scaffold for other division proteins and

providing nominal contractile force during membrane constriction (Ghosh and Sain 2008,

Lan, Daniels, et al. 2009, Lan, Wolgemuth, et al. 2007, Osawa, Anderson, et al. 2008).
The biochemistry of the cyanobacterial FtsZ has yet to be rigorously examined and most
of what we know of FtsZ assembly and biochemistry come from several detailed studies
of F tsZ from E. coli, though the similarity between these two proteins (Stokes and
Osteryoung 2003) would suggest that they behave similarly in vitro. In vitro, E. coli FtsZ
assembles into a simple linear polymer (protofilament) with a diameter of about 5 nm in
the presence of GTP, but can form mini-rings or other higher-order structures under
certain conditions (Erickson, Taylor, et al. 1996). Hydrolysis of GTP to GDP can lead to
curvature of the FtsZ polymer, and this induced curvature is thought to be one source of
contractile force during cell division (Lu, Reedy, et al. 2000). The addition of calcium or
low pH conditions facilitates lateral association of FtsZ protofilaments into bundles or
sheets approximately 30 nm or more in diameter (Erickson, Taylor, et al. 1996). It is
thought that the protofilaments serve as building blocks for the cytosolic Z-ring in vivo,
undergoing further lateral association and organization in the presence of accessory
factors such as FtsA (Jensen, Thompson, et al. 2005, Pichoff and Lutkenhaus 2002) and
ZipA (Hale, Rhee, et al. 2000, RayChaudhuri 1999), proteins that enhance polymer
bundling and anchor the developing Z-ring to the inner leaflet of the cell membrane
(Osawa, Anderson, et al. 2008, Pichoff and Lutkenhaus 2005). In addition to the
polymer curvature associated with GTP hydrolysis, the sliding of adjacent protofilaments
has also been proposed as an alternative or additional source of contractile force during
bacterial cell division (Lan, Daniels, et al. 2009, Li, Trimble, et al. 2007), but is

energetically unfavorable (Erickson 2009) and has yet to be experimentally proven.

In contrast to most bacteria, all plant genomes encode at least two
phylogenetically-distinct families of FtsZ proteins, termed Ftle and FtsZZ
(Miyagishima, Nozaki, et al. 2004, Stokes and Osteryoung 2003). Consistent with their
divergent sequences, Ftle and FtsZ2 have distinct roles in plastid division (Schmitz,
Glynn, et al. 2009, Stokes and Osteryoung 2003, Vitha, McAndrew, et al. 2001). Both
Ftle and FtsZ2 have been shown to be part of the Z-ring, but the exact arrangement of
the polymer is unclear. Ftle-FtsZ2 mixed polymers are able to form large protofilament
bundles similar to the large protofilament bundles observed using bacterial FtsZ,
suggesting that the plastidic Z-ring is probably a heteropolymer rather than several
closely-associated Ftle and FtsZZ homopolymers (Olson 2008). While bundling is
probably not critical for contractile activity (Erickson 2009), it is probably required for
the formation of a functional and stable Z-ring (Hale, Rhee, et al. 2000, Low, Moncrieffe,
et al. 2004, Margolin 2003). The factors that contribute to FtsZ bundling and stability in

plastids are discussed later.

Determining the site of F tsZ protofilament and ring assembly: The Min System

Mid-cell positioning of the Z-ring in bacteria is important for symmetrical
division and ensures equal partitioning of cellular contents, including chromosomes and
plasmids, during fission. In E. coli, two major mechanisms drive placement of the Z-ring
at the mid-cell: the Min system and the Nucleoid occlusion (Noc) system (Rothfield,
Taghbalout, et al. 2005). The Min system of E. coli is composed of three proteins: MinC,
MinD, and MinE (Lutkenhaus 2007). MinC binds F tsZ and inhibits Z-ring assembly by

preventing lateral associations (I-Iu, Mukherjee, et al. 1999, Scheffers 2008, Shen and

Lutkenhaus 2009). However, the FtsZ polymer-inhibiting activity of MinC is regulated
by MinD, which is tethered to the membrane and promotes MinC activity only in the
polar zones of the cell (Hu and Lutkenhaus 1999, Raskin and de Boer 1999a, Szeto,
Rowland, et al. 2002). MinD is regulated by MinE, through a mechanism in which MinE
binds to MinD and causes it to be released from the membrane (Hu and Lutkenhaus
2001). The maximum concentration of MinE occurs near the midcell adjacent to the
membrane, thereby causing the concentration of active MinC and MinD to be highest at
the poles (Hale, Meinhardt, et al. 2001). While the determinants for MinE localization
are unknown, the midcell zone created by MinE activity allows for FtsZ protofilament
assembly along the inner leaflet of the plasma membrane at the midcell (Rothfield,
Taghbalout, et al. 2005). In some bacteria that lack MinE, DivIVA tethers MinD at the
cell poles and inhibits F tsZ polymerization within the polar zone by maintaining a higher

concentration of active MinCD at the poles (Marston and Errington 1999).

In general, it has long been thought that the functions of the Arabidopsis orthologs
of MinD (AtMinD) and MinE (AtMinE) are similar to their bacterial counterparts with
respect to their functions during chloroplast division. However, recent studies have
shown that AtMinD is able to rescue the cell division defects of an E. coli minD minE
double mutant, indicating that AtMinD has probably acquired additional functions,
relative to its bacterial counterpart (Zhang, Hu, et al. 2009a). Additionally, the
localization patterns of AtMinD and AtMinE vary slightly from those of their bacterial
homologs. AtMinD localizes to polar spots and equatorial structures in Arabidopsis

(Nakanishi, Suzuki, et al. 2009), while AtMinE localizes to polar regions in tobacco

10

chloroplasts (Maple, Chua, et al. 2002). Intriguingly, orthologs of MinC have been lost
from the genomes of vascular plants (Yang, Glynn, et al. 2008) and recent work shows

that plants have acquired at least two additional factors to fine-tune the operation of the
Min system: ARC3 and MCDl (Glynn, Miyagishima, et al. 2007, Maple, Vojta, et al.

2007, Nakanishi, Suzuki, et al. 2009).

ARC3 is a chimeric protein, consisting of an FtsZ-like N-terminal domain, a C-
terrninal PIPSK-like domain, and a conserved middle domain of unknown function
(Shimada, Koizumi, et al. 2004). ARC3 is proposed as a functional replacement for
MinC in plastids (Maple, Vojta, et al. 2007), as arc3 mutants possess chloroplast
morphologies of varying size (Marrison, Rutherford, et al. 1999, Pyke and Leech 1992,
Pyke and Leech 1994), with the enlarged plastids containing multiple FtsZ rings (Glynn,
Miyagishima, et al. 2007) reminiscent of those observed within bacterial minC mutants
(Levin, Shim, et al. 1998). ARC3 exhibits both polar and equatorial localization within
the chloroplast (Maple, Vojta, et al. 2007, Shimada, Koizumi, et al. 2004). Furthermore,
ARC3 has been shown to interact with Ftle , MinD, and MinE (Maple, Vojta, et al.
2007), suggesting that it acts as a critical interface between the Z-ring and the plastidic
Min system. Based on these data, ARC3 has been hypothesized to inhibit FtsZ assembly,
but it is not yet known if ARC3 is sufficient to inhibit FtsZ assembly in vitro, similar to

bacterial MinC (Hu, Mukherjee, et al. 1999, Scheffers 2008).

MCDl is a plant-specific inner envelope protein that regulates division site

placement within the chloroplast (N akanishi, Suzuki, et al. 2009). Like ARC3 (Maple,

11

Vojta, et al. 2007, Shimada, Koizumi, et al. 2004) and AtMinD (Nakanishi, Suzuki, et al.
2009), MCDl localizes to polar zones and to equatorial structures within the chloroplast
(Nakanishi, Suzuki, et al. 2009). MCDl binds AtMinD and is required for AtMinD
localization in vivo, and thereby directs the position of the Z-ring and division site

(N akanishi, Suzuki, et al. 2009). However, it is unclear what pathways lie upstream of

MCDl to direct its localization and activity within the chloroplast.

Orthologs of characterized nucleoid occlusion factors are not encoded in
cyanobacterial or plant genomes, suggesting that chloroplasts do not utilize a Noe-like
mechanism to prevent scission of the plastid chromosome during fission (Glynn,
Miyagishima, et al. 2007). Moreover, Z-rings can form around nucleoids in
cyanobacteria, indicating that a Noe-like mechanism is probably absent from this lineage
(Miyagishima, Wolk, et al. 2005). Both plastids and cyanobacteria maintain multiple
copies of their chromosome (F alkow, Dworkin, et al. 2006, Pyke 2009), consistent with
chromosomal preservation during division being accomplished by maintaining multiple
chromosomal copies (Scott and Possingham 1980). Notably, plastids and cyanobacteria
possess an additional membrane system, the thylakoid membrane, that is not found in E.
coli or other bacteria (Falkow, Dworkin, et al. 2006). It has been proposed that some
mechanism of thylakoid partitioning might occur in plants, as the thylakoid membranes
might be an impediment to the contractile apparatus and thylakoids appear to be actively
redistributed during plastid fission (Boffey and Lloyd 1988, Leech, Thomson, et al. 1981,

Possingham and Lawrence 1983). It is possible that a thylakoid segregation mechanism

12

might be integrated into the plastidic Min system, but work on this aspect of plastid

division remains to be explored.

Assembly and stabilization of the Z-ring: ARC 6

The plastidic Min system only creates a zone conducive to F tsZ polymer
assembly, but mounting evidence suggests that additional factors are required to: (1)
stabilize the Z-ring by promoting protofilament bundling and (2) anchor FtsZ polymers to
the membrane. In E. coli, two proteins serve this purpose: ZipA and FtsA. Both ZipA
and FtsA proteins elicit their effect upon the Z-ring by binding to a short conserved motif
near the C-terminus of FtsZ called the core motif or core domain (Pichoff and
Lutkenhaus 2002). ZipA is a bitopic membrane protein of the plasma membrane thought
to enhance bundling of FtsZ filaments near the membrane (Hale, Rhee, et al. 2000,
RayChaudhuri 1999). FtsA is a peripheral membrane protein that has been shown to
anchor the Z-ring to membranes both in vivo and in vitro (Jensen, Thompson, et al. 2005,
Osawa, Anderson, et al. 2008, Pichoff and Lutkenhaus 2005). FtsA may also be involved
in protofilament bundling, as some functional redundancy between FtsA and ZipA is
evident in E. coli (Geissler, Elraheb, et al. 2003). Two other division factors are known
to bind the core motif in some gram positive organisms, Eer (Singh, Makde, et al.
2007) and ZapA (Low, Moncrieffe, et al. 2004); these proteins bind the C-terminus of
FtsZ and somehow modulate FtsZ dynamics, but their precise effects upon FtsZ
polymerization are still under investigation. Interestingly, engineering the short
membrane-tethering amphipathic helix of F tsA onto the C-terrninus of FtsZ is sufficient

to drive Z-ring formation along artificial membranes in vitro (Osawa, Anderson, et al.

13

2008). Cyanobacteria, green algae, and land plants do not encode orthologs of ZipA or
F tsA, but do encode FtsZ molecules that bear the conserved C-terminal core motif
(Miyagishima, Nishida, et al. 2003b). This suggests that other factors, possibly with an
amino acid sequence dissimilar from FtsA and ZipA, might provide a similar FtsZ

tethering and/or bundling function within plastids.

ARC6 was discovered by a forward genetic approach and encodes a protein
product that stabilizes the plastidic Z-ring, possibly by promoting FtsZ filament
formation (V itha, Froehlich, et al. 2003). ARC6 may also provide both FtsZ bundling
and tethering functions, despite the divergence of its sequence from ZipA or FtsA; ARC6
overexpressors have elongated FtsZ filaments and arc6 mutants have short, disorganized

FtsZ filaments (Figure 1.3) (V itha, Froehlich, et al. 2003).

ARC6 is a nuclear-encoded bitopic inner envelope protein of cyanobacterial
origin (V itha, F roehlich, et al. 2003); the cyanobacterial ortholog of ARC6 is called Ftn2
(Koksharova and Wolk 2002). The Arabidopsis ARC6 protein is 801 amino acids in
length (Figure 1.4), bears an N-terminal transit peptide that directs the protein to the
chloroplast, and a putative J -domain that may facilitate interaction with Hsp70/DnaK-
type chaperones. ARC6 family members also harbor a single transmembrane domain
that anchors the protein in the inner envelope membrane (V itha, Froehlich, et al. 2003).
Like the F tsZ proteins, ARC6 forms a continuous (non-punctate) ring at the division site

(Yang, Glynn, et al. 2008); this ring co—localizes with the plastidic Z-ring (Figure 1.5).

14

 

 

 

 

 

Figure 1.3. F tsZ2-1 immunolocalization in wild type Arabidopsis chloroplasts and in
arc6 mutants. (A) FtsZ forms equatorial rings (large arrowhead) in wild type
chloroplasts and (B) forms short fragments or spots (small arrowheads) in arc6 mutants,
suggesting that ARC6 stabilizes the Z-ring in vivo. Chlorophyll autofluorescence is
indicated in red. Scale bar = 5pm.

15

 

N-terminal Conserved Region C-terminal Conserved Region

/
TP JD TM/v

 

 

 

 

Figure 1.4. Schematic of the Arabidopsis ARC6 protein. The above panel highlights
the known features of ARC6 based on previous data (Vitha, Froehlich, et al. 2003). The
N-terminus of ARC6 resides in the chloroplast stroma and the C-terminus of the protein
resides in the intermembrane space. The N-terminal (amino acids 86-509) and C-
terminal (amino acids 683-793) conserved regions show very high similarity to other
ARC6/Ftn2 family members. Transit peptide (TP, amino acids 1-67); Predicted J-

domain (JD, amino acids 89-153); and Transmembrane Domain (TM, amino acids 615-
635)

16

 

 

Figure 1.5. FtsZ and ARC6 colocalize to continuous rings within the chloroplast.
Immunolabeling of FtsZZ-l (A, green) and ARC6 (B, bright red color) is shown.
Chlorophyll autofluorescence (B, dull red color) roughly marks the boundaries of the
chloroplast. Colocalization of the two proteins (arrowhead) is indicated in panel (C)
using an overlaid image from panels (A) and (B). Scale bar = 5 pm.

The N-terminus of ARC6 resides within the stroma, where it interacts with FtsZ2
family members; this interaction requires the C-terminus of FtsZZ (Maple, Aldridge, et
al. 2005, Schmitz, Glynn, et al. 2009). Presumably, the ARC6-FtsZ2 interaction is
analogous to that observed for FtsA-FtsZ and/or ZipA-FtsZ in E. coli (Pichoff and
Lutkenhaus 2002), but an FtsZ-tethering or FtsZ-bundling activity for ARC6 has yet to be
shown. In chapter 2 of this work, I describe refinement of the boundaries of the stroma]
domain of ARC6 responsible for binding FtsZZ, show that the C-terminus of F tsZZ is
both necessary and sufficient for binding ARC6, attempt to discern the effect of ARC6-
FtsZ2 interaction in vivo, and provide a homology model of ARC6 that points toward
structural and functional similarities between ARC6 and ZipA, despite their divergent

amino acid sequence.

The function of the IMS-localized C-terminal region of ARC6 was completely
unknown until a recent study that showed the C-terminus of ARC6 binds to PDV2 and is
responsible for positioning both PDVl and PDV2 (described below) at the division site
(Glynn, Froehlich, et al. 2008). In chapter 3, I present these results and some
additional data in support of this function for the C-terminus of ARC6. Chapter 4 builds

on these findings, where I present approximation of the boundaries of the PDV2-binding

domain of ARC6 (ARC6pBD), characterize a predicted site of post-translational

modification within this domain, and introduce a new hypothesis on the regulation of

division by ARC6pBD phosphorylation and dephosphorylation.

18

ARC6/Ftn2 family members are present as single-copy genes in cyanobacteria,
algae, and moss. In tracheophytes, ARC6 has undergone at least one gene duplication
event, resulting in a novel plastid division gene we have termed PARC6 (Paralog of
ARC6). Like ARC6, PARC6 is a multifunctional division protein that resides within the
inner envelope membrane, acting downstream of ARC6 to position PDV] and possibly
regulating FtsZ assembly through ARC3 — implicating PARC6 in the operation of the
plastidic Min system (Glynn, Yang, et al. 2009, Zhang, Hu, et al. 2009b). In chapter 5, I
summarize this work and provide further insights into the role of ARC6, PARC6, PDVI ,

and PDV2 during plastid division in vascular plants.

While it is unknown if ARC6 and PARC6 definitively interact with each other or
how they precisely influence each other’s activity in vivo, it is suspected that some
functional connection between these two paralogous proteins exists due to their high

degree of sequence similarity. In chapter 6, I highlight preliminary analysis of a novel

ARC6 allele, arc6 D 20 5 N, that exhibits division defects reminiscent of Arabidopsis min

mutants and propose a hypothesis that the arc6 D 20 5N division defect results from

aberrations in the ability of ARC6 to interact with a Min-system component.

Facilitating outer envelope constriction during division: ARC 5, PD VI, and PD V2

In addition to the division factors inherited from the ancestor of the chloroplast, a
few factors were invented by the host following symbiosis. Notably, three of these host-
derived factors operate within or upon the outer envelope membrane. ARC5 is a

dynamin-like protein that localizes to both cytosolic patches and an equatorial ring on the

19

outside of the plastid. Unlike the continuous ARC6 ring (Glynn, Froehlich, et‘ al. 2008,
Vitha, F roehlich, et al. 2003, Yang, Glynn, et al. 2008), the ARC5 ring has a
discontinuous, or punctate, appearance in Arabidopsis and other photosynthetic
eukaryotes (Gao, Kadirjan-Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006,
Miyagishima, Nishida, et al. 2003a, Yang, Glynn, et al. 2008). The significance of this
punctate pattern is still unclear. It is anticipated that these foci represent discrete sites of
membrane remodeling rather than a bounding contractile spiral (which would likely
appear as a continuous ring in vivo), which suggests that initial constriction of the outer
membrane might invove membrane removal from discrete foci rather than direct
mechanical constriction of the outer envelope by ARC5. Notably, the late stages of
dynamin constriction activity involve a significant change in the organization of ARC5
topology relative to the outer envelope membrane, in which ARC5 polymers come into
direct contact with the cytosolic face of the outer envelope (Yoshida, Kuroiwa, et al.
2006). This end-stage change in orientation may reflect a simultaneous change in the
pinchase activity of ARC5 polymers, but further investigation is required to confirm this

hypothesis.

ARC5 can be recruited to the division site by either PDVl or PDV2, two coiled-
coil proteins that arose during land plant evolution Miyagishima, Froehlich, et al. 2006).
Like ARC5, PDVl localizes to a punctate ring (Miyagishima, Froehlich, et al. 2006).
PDV2 localizes to a continuous ring similar to that observed for ARC6 (see chapter 3).
Interestingly, full ARC5 contractile activity requires both PDV] and PDV2, as pdvI and

pdv2 mutants possess a large proportion of dumbbell-shaped chloroplasts, similar to the

20

dynamin activity defect observed in arc5 mutants (Miyagishima, Froehlich, et al. 2006).
Overexpression of PDVl and/or PDV2 leads to an increase in the rate of chloroplast
division in vivo (Okazaki, Kabeya, et al. 2009). Taken together, these observations
indicate that the PDV proteins probably affect the rate of plastid division through
modulating ARC5 contractile activity, although the PDV proteins probably influence the
operation of other division components as well. It is still unclear if either of the PDV
proteins interacts directly with ARC5 to mediate its recruitment and/or activity along the

cytosolic surface of the chloroplast outer envelope.

Other factors with notable roles in maintaining chloroplast morphology and/or division.
A number of other factors have been discovered through genetic and cytological
analyses. The precise role of these proteins in the process of chloroplast division remains

unclear, but we will provide a brief treatment of each of these here.

GCl (Giant Chloroplast 1) is distantly related to E. coli SulA and was identified
by a reverse genetic approach (Maple, Fujiwara, et al. 2004). In proteobacteria, SulA
expression is part of the SOS response to DNA damage (Huisman, D'Ari, et al. 1984) and
delays cell division by directly binding FtsZ and inhibiting FtsZ assembly (Trusca, Scott,
et al. 1998). For Arabidopsis, it is unclear if GCl binds Ftle or F tsZ2 in vivo;
overexpression of GCI bypasses plastid division defects observed in AtFtsZ
overexpressors (Raynaud, Cassier-Chauvat, et al. 2004), but strangely GCl does not bind
F tle or FtsZZ in two-hybrid assays (Maple, Fujiwara, et al. 2004). However, GCl can

self-associate (Maple, Fujiwara, et al. 2004), similar to SulA (Cordell, Robinson, et al.

21

2003). GCl is localized to the chloroplast periphery and has not been shown to interact
with any other division components (Maple, Fujiwara, et al. 2004). While GCl
deficiency results in enlarged chloroplasts, similar to arc6 mutants, it is unclear if GCl
overexpression inhibits chloroplast division (Maple, Fujiwara, et al. 2004, Raynaud,
Cassier-Chauvat, et al. 2004) in the same way that SulA overexpression leads to division
arrest in cyanobacteria (Sakr, Jeanjean, et al. 2006). Further work is required to
understand the inputs that lead to GC 1 activity and the precise mechanism by which GCl

affects plastid morphology in vivo.

MSL2 and MSL3 belong to a family of mechanosensitive ion channels that reside
in the chloroplast (Haswell and Meyerowitz 2006). MSL2 and MSL3 appear to be
redundant genes with respect to chloroplast morphology, as only mle msl3 double
mutants exhibit plastid morphology defects (Haswell and Meyerowitz 2006). MSL3 has
been shown to colocalize with AtMinE in transient expression assays conducted using
tobacco leaf cells, but the significance of this finding remains unclear (Haswell and
Meyerowitz 2006). It has been suggested that mechanosensitive ion channels, like MSL2
and MSL3, are a means by which chloroplasts sense volume, thereby influencing the
decision to continue expanding or to initiate division (Pyke 2006). This seems
reasonable, since chloroplast volume is only slightly reduced in division-impaired
backgrounds like arc6 (Pyke 1998, Robertson, Pyke, et al. 1995). However, the precise
mechanism by which MSL2 and MSL3 influence plastid morphology, and presumably

division, remains to be elucidated.

22

F ZL (FZO-Like) was discovered based on its similarity to FZO (Gao, Sage, et al.
2006), a protein known to mediate mitochondrial membrane fusion in animals and fungi
(Hales and Fuller 1997). FZL is chloroplast-targeted protein that localizes to the
envelope and thylakoid membranes (Gao, Sage, et al. 2006). Unlike the other
characterized members of the dynamin superfamily within plants, FZL is suspected to be
a membrane remodeling GTPase that regulates organization of the thylakoid membranes,
as le mutants have thylakoid morphology defects and possess aberrant vesicular
inclusions within the stroma (Gao, Sage, et al. 2006). le mutants also show defects in
chloroplast morphology —— possibly implicating the organization or position of the
thylakoid membranes in division (Gao, Sage, et al. 2006). Curiously, FZL-GFP localizes
to spots or vesicles around the periphery of the chloroplast, which seems inconsistent
with its role in thylakoid biogenesis. However, it has recently been shown that tgd4 act]
double mutants have enlarged plastids with thylakoid defects similar to le mutants,
perhaps as a result of the loss of both sources of membrane material that partly compose
the thylakoid membrane in plants (Xu, Fan, et al. 2008). Based on its mutant phenotype
and localization, we suspect that FZL might be part of a vesicular transport system that
aids trafficking of membrane material from outside the chloroplast (i.e. the TGD4
pathway) to the thylakoid membrane (Xu, Fan, et al. 2008), which suggests that the
chloroplast morphology phenotype observed within fil mutants is a pleiotropic effect, but
further work is required to rigorously test this hypothesis. Regardless, the impact of
thylakoid morphology upon the process of chloroplast division is interesting and should

be investigated further.

23

Some factors that make up the chloroplast divisome remain to be identified. The
Plastid Dividing (PD) rings are electron-dense structures that appear after the Z-ring is
established; these structures are dynamic and may be involved in mediating constriction
(Miyagishima, Kuroiwa, et al. 2001). While the consituitve parts of the PD rings are
unknown, the Kuroiwa laboratory is in the process of dissecting these rings and
identifying their constituent parts using a synchronizable red algae, Cyanidioschyzon
merolae (Yoshida, Kuroiwa, et al. 2006). It is thought that land plants will possess the
same components, as the timing, size, and dynamics of the PD rings in land plants are

similar to those observed in the red algae (Kuroiwa, Kuroiwa, et al. 1998).

While several factors have been introduced here, much work is still required to
identify the complete function of each component and the pathways that feed into their
expression and activity. Furthermore, several factors are likely to be uncovered in the
coming years using comparative genomics, biochemical, and genetic approaches.
Throughout this text, I will attempt to overtly indicate where pieces are missing to

reiterate this idea, as a final model must account for all the components.

Regulating Plastid Fission: Why Do Plastids Divide in Land Plants?

It is currently theorized that at least four major inputs feed into the regulation of

plastid fission in higher plants: photosynthetic and photoprotective activity, metabolite

transport, cellular division, and cellular differentiation. Presumably, each of these inputs

24

contributes to the fitness of a particular organism. We will briefly review each of these

here.

Photocollection and Photoprotection

The viability of chloroplast division mutants with large chloroplasts raises the
question of why plant cells evolved to have multiple chloroplasts. The answer may have
to do partly with the ability of multiple small chloroplasts to redistribute more effectively
than fewer large chloroplasts inside the cell. Unlike motile unicellular algae with single
chloroplasts, land plants are sessile and cannot relocate in response to sudden changes in
light intensity. Chloroplasts in land plants avoid potentially damaging high light by
moving to the cell periphery and orienting in columns parallel to the plane of incoming
light; under low-light conditions, chloroplasts gather in periclinal layers to maximize
light absorption (Kasahara, Kagawa, et al. 2002). This movement is believed to be
largely performed through actin reorganization and myosin motor activity (Paves and
Truve 2007, Schmidt von Braun and Schleiff 2008). Consistent with the importance of
plastid division for efficient plastid movement, a mutant allele of Arabidopsis Ftle
shows diminished plastid movement in response to high light (Yoder, Kadirj an-Kalbach,
et al. 2007). Interestingly, the thylakoid morphology of big-plastid mutants is
reminiscent of that in high light-adapted plants (Austin 11 and Webber 2005), and
probably arises due to the impaired movement ability of the enlarged chloroplasts within
division-defective backgrounds (Jeong, Park, et al. 2002). As discussed earlier,
mutations in genes influencing thylakoid organization also impair chloroplast division

(Gao, Sage, et al. 2006, Xu, Fan, et al. 2008). The regulation of thylakoid morphology

25

and chloroplast movement is probably incorporated into a larger system that maximizes
photosynthetic output and minimizes light-induced damage in response to changing

environmental stimuli (Suetsugu and Wada 2007, Wada, Kagawa, et al. 2003).

Metabolic Exchange: The Possible Effect of Organelle Area-to- Volume Ratio

In addition to the potential role of photosynthesis in regulating plastid division,
the shuttling of reduced carbon and metabolites between the plastid and the cytosol might
also influence the division process. Presumably, an optimal ratio of organelle surface
area to internal volume is maintained in chloroplasts. A larger plastid, such as those in
the arc6 background, has a disproportionate surface to volume ratio relative to smaller
plastids (Pyke and Leech 1992, Pyke and Leech 1994, Pyke, Rutherford, et al. 1994), and
as a result may have insufficient transporter capabilities. Maintaining large numbers of
small plastids may allow for greater membrane surface area and prevent bottlenecks of
metabolic flux between the plastid and the cytosol or between the plastid and other
organelles. However, investigation of this potential regulatory input into the division

process has not been explored.

Cell Division and Cell Expansion

Chloroplast division is intimately coupled to cell expansion and cellular division
in algae, and to some degree in plants. In dividing meristematic cells and juvenile leaf
cells, the pace of plastid division must keep pace with cellular division to ensure that each

resulting cell has one of these essential organelles (Boffey and Lloyd 1988, Possingham

26

and Lawrence 1983), though aplastidic cells can reside within the leaves of plants (Chen,
Asano, et al. 2009, Forth and Pyke 2006, Pyke, Rutherford, et al. 1994). At least one
factor, CDTl , has been implicated in coordinating cell division with plastid division in
plant cells (Raynaud, Perennes, et al. 2005). CDTl proteins are known to be part of the
pre-replication complex, required for origin licensing during nuclear DNA replication in
eukaryotes (DePamphilis 2003, Thomer, May, et al. 2004), including Arabidopsis
(Castellano, Boniotti, et al. 2004, Masuda, Ramos, et al. 2004). However, in
Arabidopsis, RN Ai knockdown lines that diminish expression of CDTl paralogs have
profound defects in chloroplast morphology and CDTl has been shown to interact with
ARC6 (Raynaud, Perennes, et al. 2005). The precise details of how CDTl impacts
chloroplast division are unknown, but the dual function of CDTl proteins in nuclear

DNA replication and plastid division should be examined further.

As cells expand, the number and/or volume of chloroplasts increases
proportionately in mesophyll cells, so that ~70% of the cell volume is consistently
occupied by chloroplasts, even in division-impaired mutants (Pyke 1998, Robertson,
Pyke, et al. 1995). Cell expansion is part of leaf cell development in plants and is
influenced by brassinosteroids (Hu, Poh, et al. 2006, Nakaya, Tsukaya, et al. 2002),
auxin (Chen, Shimomura, et al. 2001, Jones, Irn, et al. 1998), and cytokinin (Baskin,
Cork, et al. 1995, Beemster and Baskin 2000). Of these, only cytokinins have been
shown to influence plastid differentiation (Reski, Wehe, et al. 1991) and plastid division
(Okazaki, Kabeya, et al. 2009). In Arabidopsis, application of exogenous cytokinin or

overexpression of CRF 2 (Cytokinin-Responsive Factor 2) leads to an increased rate of

27

plastid division as a result of increased expression of PDV] and PDV2 (Okazaki,
Kabeya, et al. 2009), two factors involved in modulating dynamin (ARC5) recruitment
and activity (Miyagishima, Froehlich, et al. 2006). Strangely, while cytokinin does
upregulate the rate of chloroplast division through the PDV proteins, expression levels of
other division factors such as FtsZ or ARC6 are unchanged in response to cytokinin
(Okazaki, Kabeya, et al. 2009) — suggesting that the protein level of PDVl and PDV2
can stimulate the activity of both cytosolic and stromal division factors from their
position within the outer envelope membrane (Okazaki, Kabeya, et al. 2009). While it is
clear that cell expansion and cytokinin certainly influence plastid differentiation and
division, the full scope of the gene regulatory network that controls plastid fission

remains to be determined.

Controls on Division During Cellular Development and Difikrentiation

In addition to binary fission of chloroplasts within leaf cells, plastids can undergo
developmentally-regulated division and differentiation during plant development. For
example, plastid differentiation during fruit development in tomato occurs through a
unique budding/fragmentation mechanism where even enlarged chloroplasts are capable
of differentiating into normal chromoplasts during fruit ripening (Forth and Pyke 2006).
The bypass of the block in plastid division within sufl'ulta mutants that occurs during
tomato fruit development is not observed with the enlarged chloroplasts of ftsZ, arc6, or
arc12 mutants (although Arabidopsis itself does not possess chromoplasts); once an
oversized chloroplast is generated, it is presumed to have lost its polarity and can no

longer divide using an FtsZ-based mechanism (Schmitz, Glynn, et al. 2009). It is

28

possible that the pathway that produces stromules, stroma-containing extrusions of the
plastid, might be involved in these fission events, as stromules are generated even in the
absence of a ftmctional Z-ring (Holzinger, Kwok, et al. 2008). The molecular
identification of suflulta and the study of plastid fission in the big-plastid backgrounds
(i.e. arc6, arc12, and others) will likely provide critical insights into FtsZ-independent

pathways of plastid partitioning.

Perspectives and Acknowledgements

Previous to this body of work, several issues in the field of plastid division were
unclear. Within the following chapters, I attempt to address at least three major
questions: (1) what discrete region of ARC6 binds FtsZ2 family members and what is the
purpose of this interaction in vivo? (2) How does ARC6 participate in the coordination of
the stromal and cytosolic plastid division machineries? (3) What is the function of
Arabidopsis ARC6 paralog (PARC6) and how might PARC6 and ARC6 link into the
plastidic Min system? I follow these chapters with some general conclusions and
suggested future directions for research. Additionally, this volume contains two
appendices that highlight peripheral projects undertaken during the last few years. I hope
it is a clear and fascinating read. Please note that portions of this chapter are reproduced
from (Glynn, Miyagishima, et al. 2007) (http://wwwiafficdkn and is Copyright
Blackwell Publishing. Blackwell grants authors the freedom to reuse their own articles in

new publications, provided that they are the editor of the new publication.

29

Chapter 2

Identification and Modeling of an FtsZZ—Binding Domain within ARC6

30

Abstract

The plastidic Z-ring is not a static structure, but probably undergoes remodeling
in response to various input signals. The regulation of FtsZ filament dynamics is an
important part of remodeling the Z-ring during its assembly and constriction. ARC6 has
been shown to be required for assembly and/or stabilization of the Z-ring in vivo. This
stabilizing activity has been proposed to occur through F tsZZ family members, as ARC6
has been shown to bind FtsZZ, but not Ftle. However, the precise effect of ARC6 upon
Z-ring assembly through its interaction with FtsZ2 has not been demonstrated

experimentally, nor have the exact boundaries of the FtsZ2-binding domain (ZBD) of

ARC6 been identified. To provide insight into the function of ARC6ZBD, we fine-

mapped the boundaries of this domain using two hybrid assays and used this information

to guide analysis of the in vivo function of ARC6ZBD. Our results indicate that

ARC6ZBD occupies amino acids 351-503. This domain is conserved amongst the

Chlorophytes and cyanobacteria, but bears no sequence similarity to other characterized

protein sequences. We generated a homology model of ARC6ZBD based on the crystal

structure of the FtsZ-binding domain of E. coli ZipA, an FtsZ-bundling protein.

ARC6ZBD was fit onto the known structure of ZipAZBD, despite the large degree of

sequence divergence between ARC6ZBD and ZipAZBD, suggesting that ARC6ZBD might

aid bundling of F tsZ filaments through a ZipA-like mechanism, but rigorous in vivo

analysis and structural data is still needed to validate this model.

31

Introduction

Plastids are essential organelles. In plant cells, binary fission is a mechanism by
which each cell receives its required complement of plastids and likely is influenced by
development, environmental cues, and cellular metabolism (Boffey and Lloyd 1988,
Leech, Thomson, et al. 1981, Possingham and Lawrence 1983). Plastids evolved from a
cyanobacterium that took up residence within a primitive protozoan (Cavalier-Smith
2000). Like many free-living cyanobacteria and proteobacteria, chloroplasts divide by
binary fission in a process that requires FtsZ (Osteryoung, Stokes, et al. 1998, Strepp,

Scholz, et al. 1998, Vitha, McAndrew, et al. 2001).

FtsZ is a filament-forming GTPase with structural similarity to tubulin (Erickson,
Taylor, et al. 1996, Mukherjee, Dai, et al. 1993). E. coli FtsZ forms protofilaments that
undergo organized assembly into a structure at the mid-cell called the Z-ring, which acts
as a dynamic scaffold for other division components (Margolin 2005) and probably
generates the force that contributes to cellular constriction (Erickson 2009, Li, Trimble, et

al. 2007, Osawa, Anderson, et al. 2008).

Several systems contribute to controlling both FtsZ protofilament assembly and
Z-ring assembly. The Min system, composed of MinCDE, prevents protofilament
assembly at the poles, but allows bulk protofilament assembly to occur at the mid-cell,
ensuring that the cell divides near its midpoint (Lutkenhaus 2007). ZipA is a bitopic

membrane protein that binds a conserved C-terminal segment of FtsZ in E. coli and is

32

thought to be important for protofilament bundling and organization adjacent to the inner
leaflet of the cell membrane at the midcell (Hale, Rhee, et al. 2000, Mosyak, Zhang et al.
2000, RayChaudhuri 1999). The conserved C-terminal motif of bacterial FtsZ consists of
~12 amino acids and is required for FtsZ functidn (Ma and Margolin 1999). Besides
ZipA, the conserved C-terminus of FtsZ is a target for FtsA, another protein involved in
cell division (Pichoff and Lutkenhaus 2002, Pichoff and Lutkenhaus 2005, Pichoff and
Lutkenhaus 2007, Ricard and Hirota 1973). FtsA is a membrane-associated protein that
acts as a membrane anchor for the Z-ring (Osawa, Anderson, et al. 2008, Pichoff and
Lutkenhaus 2005) and binds the C-terminal tail of FtsZ through a domain that is
structurally distinct from the F tsZ-binding domain of ZipA (Mosyak, Zhang, et al. 2000,
van den Ent and Lowe 2000). While the coordinated action of MinCDE, ZipA, and F tsA
contribute to FtsZ ring formation in proteobacteria, the precise roles of proteins

influencing assembly and dynamics of the Z-ring in plastids is still somewhat unclear.

In chloroplasts, FtsZ assembles into ring-shaped structures near the middle of the
organelle, analogous to the bacterial protein (V itha, McAndrew, et al. 2001). In the case
of plants, however, FtsZ gene duplication and divergence have generated two distinct
families of FtsZ protein (termed F tle and FtsZ2), both of which are required for
chloroplast division (Osteryoung, Stokes, et al. 1998, Schmitz, Glynn, et al. 2009). The
most notable difference between these two families occurs at their C-terrnini, where
FtsZZ family members possess a C-terminal motif similar to the ZipA- and FtsA- binding
motif of bacterial FtsZ proteins; this motif is absent from plant F tle family members

(Miyagishima, Nishida, et al. 2003b, Stokes and Osteryoung 2003).

33

While ZipA is an essential gene in some bacteria, orthologs of ZipA are absent
from cyanobacteria and plants (Miyagishima, Nishida, et al. 2003b). However, we have
hypothesized the existence of an FtsA-like or ZipA-like protein in these lineages because
FtsZZ family members have retained the conserved C-terminal extension present in most
bacterial F tsZ proteins. Despite the inability to identify a ZipA-like protein in plants by
BLAST (Altschul, Madden, et al. 1997), analysis of ARC6 suggests that it may fulfill a
similar functional role within the plastid. ARC6 promotes Z-ring assembly, possibly by
bundling FtsZ filaments within the plastid since loss-of-function arc6 mutants have short
F tsZ filaments distributed throughout the stroma and transgenic lines overexpressing
ARC6 have elongated F tsZ filaments, some of which are spiraled or branched (V itha,
Froehlich, et al. 2003). A previous report revealed an interaction between ARC6 and
FtsZZ-l that requires the conserved C-terrninal extension of F tsZ2-l and proposed a
functional role analogous to that of ZipA or FtsA in E. coli; although the precise
boundaries for the FtsZ2 binding domain were not identified in this study (Maple,

Aldridge, et al. 2005).

The results we show here define the FtsZ2 binding domain (ZBD) of Arabidopsis

ARC6 and show that the ARC6ZBD shares features with the FtsZ-binding domain of E.

coli ZipA (Mosyak, Zhang, et al. 2000). We then used site-directed mutagenesis to

identify critical residues within the ARC6ZBD that are probably sites of intermolecular

contact between ARC6 and FtsZZ based on our structural model. Further, we introduced

transgenes into Arabidopsis that are predicted ARC6ZBD loss-of-function or gain-of-

34

function alleles based on our protein interaction analyses in yeast. Unfortunately, these

transgenes were uninformative with regard to ARC6ZBD function in vivo and further

analysis will be required to validate the boundaries of ARC6ZBD and provide insight into

its structure and function in vivo.

Results

The F tsZZ—binding domain of ARC 6 (ARC 6231)) occupies amino acids 351-503.

Previously, our colleagues discovered that a portion of the conserved stromal
region of ARC6 (AA 154-509) specifically bound FtsZZ-l in two-hybrid experiments
(Maple, Aldridge, et al. 2005, Maple and Moller 2006). Recently, our laboratory showed
that both Arabidopsis FtsZ2 isoforrns (FtsZ2—1 and FtsZ2-2) bind the same region of
ARC6 with similar affinity and are functionally redundant (Schmitz, Glynn, et al. 2009);

therefore, we used only FtsZZ-l (hereforth FtsZZ) in our analyses for simplicity.

To delineate the precise boundaries of the ARC6ZBD, we used a two-hybrid-

based domain-mapping approach, using exon junctions as the boundaries for our initial
constructs (Figure 2.1) since critical domains might be less likely to be interrupted by
introns during genome evolution (de Souza, Long, et al. 1996, Fodor and Aldrich 2009).
The conserved stromal region of ARC6 is encoded within four exons. We generated 9

constructs representing different portions of ARC6 and assayed these for interaction with

FtsZZ. The results of these assays indicate that ARC6ZBD is contained within AA 351-

35

 

TP JD TM

 

 

 

FtsZZ Interaction
AA 154-509

_— n.

AA 170-509

i

AA 245-509

E

AA 332-509
AA 154-331
1‘ - —
AA 344-509
— +++

 

AA 351-503

 

 

i

 

AA 358-509

AA 344-490

 

 

 

Figure 2.1. Mapping the FtsZ2-binding domain of ARC6 by yeast two-hybrid assay.
Constructs shown are GAL4-AD fusions of ARC6, with FtsZZ-l present as a GAL4-BD
fiision. Interaction strength is reported as the ratio of growth (-HIS : +HIS) of a
transformant on supplemented synthetic dropout media (+—++ growth ratio is greater than
or equal to 0.67; - growth ratio is less than or equal to negative control). The minimal
portion of ARC6 required for interaction with FtsZZ is boxed. Transit peptide (TP); J -
domain (JD); transmembrane domain (TM); N-terminus (N); and C-terminus (C).

36

503 (Figure 2.1); the degree of reporter activation from AA 351-503 was similar to that

of the original boundaries described previously (Maple, Aldridge, et al. 2005).

The C-terminus of F tsZZ is sufi’icient for ARC 6-F tsZZ interaction.

Previously it was shown that the C-terminus of FtsZ2 is necessary for ARC6-
FtsZZ interaction (Maple, Aldridge, et al. 2005), but it was unclear if this short C-
terrninal motif was sufficient for ARC6-FtsZ2 interaction, or if other features within
FtsZZ might be required for ARC6-FtsZZ interaction. To address this question, we
generated two-hybrid constructs encoding the C-terminal 19 amino acids of Arabid0psis

FtsZZ as a GAL4 fusion and transformed them into yeast carrying plasmids encoding
either ARC6154-509 01' ARC6ZBD. Our results indicate that the C-terminus of F tSZZ is
sufficient for ARC6-FtsZ2 interaction, as GAL4 fusions to both the mature FtsZZ protein

and 19 amino acid motif at the C-terminus of FtsZZ (F tsZ2459_473) generated comparably

high levels of reporter activation (Figure 2.2), though the presence of the full-length

FtsZZ protein was associated with slightly higher levels of HIS3 reporter activation.

To gain insight into the structure and function of the ARC6ZBD, we wanted to

generate a testable structural model of this protein domain. There is currently no
experimentally-tested structural data available for ARC6, so our initial approaches
involved the use of BLAST-based sequence queries (Altschul, Madden, et al. 1997) to
identify similar domains with solved crystal or NMR structures. Using this approach, we

were unable to identify similar sequences with solved three-dimensional structures.

37

 

INTERACTION PAIRING SD/-ULT SD/-ULTH

ARC6154-509 VS. FI82249-478

 

ARC6351-503 VS~ Ftszz49—478
ARC6154-509 VS- FtSZ7-459-478
ARC6351-503 VS- FtSZ2459-478
ARC6154-509 VS- FtSZ249-462
ARC6351_503 vs. FtsZZ49_462

Empty vector vs. FtsZ249-473

Empty vector vs. FtsZZ459_478

Empty vector vs. FtsZ249_462

  
 

ARC6154_ 509 vs. Empty vector

ARC6351_503 vs. Empty vector -
Empty vector vs. Empty vector

1/10 1/100 1/10 1/100

 

 

 

Figure 2.2. Two-hybrid HIS reporter assays showing ARC6-FtsZ2 interaction.
Yeast harboring pairings of AD-ARC6 and BD-FtsZZ are shown following 48 hours of
growth on the synthetic dropout (SD) media indicated. ARC6-FtsZ2 interaction requires
only ARC6351_503 (ARC6ZBD) and F tsZ2459_47g. Removal of the C-terminus (AA463-
478) of FtsZ2 abolishes ARC6-FtsZ2 interaction. Dilutions from a stationary phase

culture at OD600 = 1.0 are indicated 1, 1/10, and 1/ 100.

ARC 6231) can be modeled onto the crystal structure of ZipA ZBD

Because ARC6ZBD binds the short conserved C-terminal motif of F tsZ2 family

members and this motif shares conservation with an analogous motif found at the C-

terrninus of bacterial FtsZ sequences (hereforth referred to FtsZCT), we compared the

ARC6ZBD amino acid sequence to sequences of bacterial proteins known to bind F tsZCT,

as some of these have solved crystal structures. At least three proteins bind bacterial

FtsZCT: F tsA (Pichoff and Lutkenhaus 2007), ZipA (Hale, Rhee, et al. 2000), and Eer

(Singh, Makde, et al. 2007). The F tsZ-binding domain of FtsA is at least 288 amino

acids in length (Pichoff and Lutkenhaus 2007) and there is no crystal structure available

for Eer. The F tsZ-binding domain of ZipA (ZipAZBD) has a solved structure (Mosyak,

Zhang, et al. 2000) and is similar in length (144 amino acids) to the ARC6ZBD (153

amino acids) and was chosen to generate a structural model of ARC6ZBD. We generated

a sequence alignment between ARC6ZBD and ZipAZBD to identify regions of similarity

between these two domains (Figure 2.3). The homology-modeling server at

http://proteins.msu.edu/ was then used to superimpose the amino acid sequence of

ARC6ZBD onto the main chain structure of ZipAZBD (Protein Data Bank Identification

Ntunber 1F47) (Mosyak, Zhang, et al. 2000), using the identical residues (Figure 2.3)

between the two sequences as anchor points. Following this, we modeled the sidechains

of ARC6ZBD using SCWRL (Wang, Canutescu, et al. 2008) to get the lowest energy

conformation of ARC6ZBD with minimal conflict.

39

 

ARC6ZBD EEALALVAQSEGKKEHLLOEE QFQQLQ
ZipAZBD . DKPKRK IMN GSELNG
ARcézBD M EIPA L EIDFG RGL L
ZipAZBD NS QQAGFE HRHLSP GSGP F
AncszBD IG DECRMWLG ELEN
ZipAZBD s. . ANMVKPGT MIS<D EQVP
ARC6ZBD SN NDDLP ET A VFPRF
ZipAZBD sy LQNFK LQs QHI DE GGVVL

ARC6ZBD DKKFKLGDYYDDPI LSY RVEVV
ZiPAZBD RMMTPQKLREYQIIIREV ANA. .

 

 

 

Figure 2.3. Alignment of ARC6ZBD (AA 351-503) with ZipAZBD (AA185-328).

ARC6ZBD was aligned with ZipAZBD using a BLOSUM62 scoring matrix with
CLUSTALW2 (htt ://www.ebi.ac.uk/Tools/clustalw2 . Identical residues are shown

 

boxed with shading; similar residues are indicated with boxed lettering. The asterisks (*)

above the ARC6ZBD sequence indicate residues that are hypothesized to be critical for

intermolecular contact between ARC6 and FtsZZ, based on comparison with E. coli ZipA-

FtsZ interaction (Moreira, Fernandes, et al. 2006).

40

A comparison of ARC6ZBD and ZipAZBD is shown in Figure 2.4 and illustrates the
prospective structural similarity of these two functionally-similar protein domains,
despite their significant differences in amino acid sequence. If this model is correct, it

indicates that ARC6ZBD is a domain made up of both alpha helices and beta sheet

structures. Curiously, this model conflicts with the mostly helical secondary structure

predictions made for ARC6ZBD (Figure 2.5) by three EVA-rated prediction algorithms

(http://cubic.bioc.coltunbiaedu/evaf): PSI-PRED, PROFsec, and SABLE (Rost and
Eyrich 2001). However, these programs are not trained to predict secondary structure at
physiological pH of the stroma, which is approximately 8.0 under optimal light

conditions (McDonald 2003). This difference may explain the discrepancy between the
ARC6ZBD secondary structure inferred by comparison with ZipAZBD and those
predicted by PSI-PRED, PROFsec, and SABLE — as the pH of the surrounding medium
can significantly affect protein folding by altering the charge or polarity of amino acid

side chains (Cramer, Engelman, et al. 1992, Dobson 2003, Pagel and Koksch 2008). In

any event, these predictions and models need to be carefully and rigorously validated

with experimental data (X-ray or NMR-solved structures).

41

1190

 

Figure 2.4. Comparison of the E. coli ZipAZBD crystal structure and the A. thaliana

ARC6ZBD homology model. Cartoon representations of (A) the crystal structure of
ZipA (AA 185-328) and (B) the homology model of ARC6 (AA 351-503). Labels and
arrows in (A) indicate residues shown to be important for intermolecular contact with the
C -terminus of E. coli FtsZ (Moreira, F emandes, et al. 2006); the analogous residues in
ARC6 are indicated in panel (B). In panel (A). the side chain ofthe central phenylalanine
residue (F269) that stabilizes the ZipA-FtsZ interaction is shown projecting into the
hydrophobic pocket that is occupied by the C-terminus of FtsZ: the modeled side chain of
the corresponding phenylalanine in ARC 6 (F442) is shown for comparison in (B).

Amino termini are shown in blue and carboxy termini are shown in red.

42

 

A
PsiPRED

PROF see
SABLE

B
PsiPRED
PROF sec

SABLE

C
PsiPRED
PROF sec

SABLE

 

 

 

 

Figure 2.5. Secondary structure predictions of ARC6ZBD. Predictions for (A)
helical, (B) strand, and (C) loop secondary structures using three top-rated secondary
structure prediction algorithms. Red indicates a high propensity for the secondary
structure indicated and blue indicates low probablility for structure indicated. Large tick
marks = 10 AA and small tick marks = 5 AA.

43

Mutations that influence ZipA-FtsZ interactions influence ARC 6—F tsZZ interaction.

To test the validity of our model, I made mutations in ARC6ZBD using SOE—PCR

and then tested for interaction of ARC6ZBD with FtsZZ using yeast two-hybrid assays. A

previous study had shown that mutations in the following residues of ZipAZBD affect

binding of FtsZ: 1196, K250, and F269; of these ZipAp269 has the most significant
impact upon the strength of the ZipA-FtsZ interaction (Moreira, Femandes, et al. 2006).
l hypothesized that if ARC6ZBD and ZipAZBD do share a common structure and binding

scheme for F tsZ, then the analogous positions within ARC6ZBD should be similar to

those in ZipAZBD. Therefore, mutagenesis of those same residues within ARC6ZBD

should result in varied binding of FtsZZ, in a manner similar to that observed for

ZipAZBD. In comparing the two structures (Figure 2.4), I found residues in ARC6ZBD

analogous to F tsZCT contact-critical regions in ZipAZBD. I generated mutations in

ARC6ZBD, focusing on ARC6F442, as this residue occupies a position analogous to

ZipAp269. The following ARC6ZBD mutations were tested for interaction with FtsZ2:

F 442Y, F442A, F442K, and F442D (Figure 2.6). Site-directed mutation of F442 had
varying effects upon ARC6-FtsZZ binding, with charged residues causing major
disruption of the interaction between the two proteins (Figure 2.6). These results are

consistent with the effect of analogous mutations on ZipA

44

 

ARC6 Construct F tsZ2 Interaction

[. . , , ...an .. _ fl +++

 

 

[ Mir." .l +++

 

F4424 .. . .. +++

 

 

 

 

 

Figure 2.6. ARC6ZBD (F442) site-directed mutants. GAL4-ARC6ZBD fusions
carrying site-directed mutations for phenylalanine 442 (F442) were tested for their ability
to interact with GAL4-FtsZZ-l in yeast using a HIS reporter assay. Hydrophobic
substitutions (F442Y and F 442A) do not affect interaction strength between ARC6 and
FtsZ2, but charged substitutions (F 442K and F442D) disrupt ARC6-FtsZZ interaction
consistent with ARC6F442 residing within a hydrophobic pocket, similar to ZipAF269
(Mosyak, Zhang, et al. 2000). Interaction strength is shown as the ratio of yeast growth
on —HIS/+HIS synthetic dropout media. Growth ratio 3 0.67 (+++); 0.66 3 growth ratio
2 0.33 (++); 0.32 3 growth ratio > vector-only negative control (+); and growth ratio <
negative control (-).

45

(Moreira, Femandes, et al. 2006) and suggest that F tsZ2 might bind ARC6ZBD within a

hydrophobic pocket, similar to ZipA-FtsZ interaction (Moreira, F emandes, et al. 2006,

Mosyak, Zhang, et al. 2000). From this analysis, I conclude that the structures of the

ARC6ZBD and ZipAZBD might be similar, with the C-terminus of FtsZ2 binding to

ARC6ZBD within a hydrophobic pocket, but further analysis of ARC6ZBD site-directed

mutants and real structural analyses (i.e. X-ray crystallography or NMR) are required to

fully test this structural model.

Analysis of the ARC 6 ZBD‘F tsZ2 Interaction by Pulldown Assay.

To confirm our two-hybrid results, I attempted to use a pulldown approach to

compare the Ftszz-binding abilities of ARC6ZBD and ARC6ZBD (F442D)- I expressed
the C-terminus of F tsZZ (AA 459-478) as a C-terminal fusion to GST (GST-FtsZ2CT).
Both ARC6ZBD and ARC6ZBD (F442D) were expressed with a polyhistidine tag at their

N-termini (Hi83-ARC6ZBD and Hi83-ARC6ZBD (F442D) respectively). All fusions were

expressed under control of an IPTG-inducible promoter in E. coli. After immobilizing

and purifying the polyhistidine-ARC6 fusion proteins using Ni-sepharose beads, I treated

Hiss-ARC6ZBD or Hiss-ARC6ZBD (124429) with clarified cell extracts fi'om lines
expressing GST-FtsZZCT. After a series of washes, bound protein was eluted from the
beads using concentrated imidazole the eluates were probed for GST-FtsZZC—r. In

preliminary results, I observed similar GST-FtsZZC-r binding characteristics for both

46

Hiss-ARC6ZBD and Hiss-ARC6ZBD(F4421)) (Figure 2.7A). I repeated this protocol with

more stringent wash conditions and observed that neither ARC6 fusion was able to retain

GST-ZZCT (not shown). To gain insight into the root cause of this result, I performed a
pulldown between Hiss-ARC6ZBD and either GST-FtsZZCT or GST-FtsZ2CT (F466 A);

the FtsZ2F466 A mutation was previously shown to abolish interaction between ARC6 and
F tsZZ using a two-hybrid based approach (Maple, Aldridge, et al. 2005). Surprisingly,
both FtsZZCT and F tsZZCT(p466 A) interact with ARC6ZBD using pulldowns (Figure

2.7B). Curiously, the results of the in vitro pulldowns between site-directed mutants of

ARC6 and FtsZ2 (Figure 2.7) are inconsistent with two-hybrid assays (Maple, Aldridge,

et al. 2005) (Figure 2.6), leading us to question the overall importance of ARC6F442 and

FtSZZF466 in vivo.

In vivo analysis of the ARC6 F44 2 D site-directed mutation.

Because ARC6F442 was shown to be critical for ARC6-FtsZZ interaction in yeast,

but not for in vitro interaction, I aimed to test one of these results in vivo and clarify the

role of ARC6ZBD by analyzing a prospective ZBD loss-of-ftmction mutant. To do this, I

generated transgenes encoding the native ARC6 protein and an ARC6F44213 mutant

protein. Both of these transgenes were placed under control of the native ARC6 promoter
(V itha, Froehlich, et al. 2003). Constructs were introduced into Col-0 and arc6

(SAIL_693_G04) backgrounds by Agrobacterium-mediated

47

 

 

 

 

 

A 1 2 3 B 1 2 3
a-HIS a-GST
mam-m air-a, - -
a-GST a—GST

 

 

 

 

 

 

 

 

 

 

Figure 2.7. Pulldown of FtsZZ by ARC6. (A) Anti-HIS inununoblot (upper panel)
showing eluted bait protein and anti-GST (lower panel) showing eluted GST-FtsZZC-r
fusions. Pulldown eluates from immobilized Hiss-ARC6ZBD treated with GST-FtsZ2C-r
(lane 1); pulldown eluates from immobilized Hi83-ARC6ZBD(F44ZD) treated With GST-
FtsZZCT (lane 2); and pulldown eluates from beads treated with GST-FtsZZCT extract.
(B) Anti-GST immunoblot showing eluted GST—FtsZ2CT fusions (upper panel) and input
proteins (lower panel). Pulldown eluates from immobilized Hiss-ARC6ZBD treated with
GST-FtsZ2CT (upper panel, lane 1); pulldown eluates from Hiss-ARC6ZBD treated with
GST-FtsZ2C1-(p466 A) (upper panel, lane 2); and pulldown eluate fi'om beads treated with
only Hiss-ARC6ZBD (upper panel, lane 3). Pulldown inputs (0.5 pg total protein/lane)
for GST-FtsZ2C-r (lower panel, lane 1); GST-FtsZ2C-r (p466 A) (lane 2); and Hiss-
ARC6ZBD (lower panel, lane 3). The molecular weight of the Hiss-ARC6ZBD and Hiss-
ARC6ZBD (F442D) fusions are ~19 kD. The molecular weight of GST-FtsZZCT and
GST-FtsZZCT(F466 A) are ~28 kD.

48

transformation and transgenic individuals were selected using hygromycin. Once plants

were approximately 5 weeks old, T1 individuals and controls were examined for ARC6 or

ARC6 F44 2 D transgene expression by immunoblotting (Figure 2.8); only transformants

from the SAIL_693_G04 background (arc6) were analyzed for ARC6 protein levels, as it

would be impossible to differentiate the native and transgenic ARC6 protein products

using our polyclonal ARC6 antibody. Transformed (T1) arc6 individuals with

ARC6 F44 2 D transgene expression near the range of expression yielding wild type

chloroplast phenotypes were quantitatively assayed for chloroplast number and

morphology (Figure 2.9), though all T] lines were subjected to microsc0pic examination.
Consistent with pulldown results that hinted that ARC6ZBD (F442D) mutant proteins

might be fully functional (Figure 2.7A), the ARC6 F44 ZD transgene was able to

complement the arc6 mutant (Figure 2.9D-E). Further, I observed very few defects in

Col—0 plants transformed with ARC6 F44 2 D, (not shown) indicating that this transgene

probably does not impart any dominant negative effect on chloroplast morphology,

consistent with its ability to fully rescue an arc6 mutant (Figure 2.9).

To determine if the ARC6 F44 2 D-expressing lines have subtle defects in FtsZ

assembly or disassembly, we examined Z-ring morphology within chloroplasts of

juvenile leaves (Figure 2.10). Consistent with the ability of the ARC 6 F44 2D transgene to

complement the chloroplast morphology defect of arc6 mutants (Figure 2.9D-E), FtsZ

49

 

a-ARC6

 

195kD>

 
  
  

117kD>
97kD>

50 kD >
Ponceau

 

 

 

Figure 2.8. Demonstration of ARC6 and ARC6F442D transgene expression in the
arc6 (SAIL_693_G04) background. Exposure from an anti-ARC6 immunoblot (top
panel) and Ponceau stain of the corresponding membrane (lower panel). Whole cell
extracts from controls and transformed arc6 mutants: molecular weight marker (lane 1);
untransformed SAIL_693_G04 (lane 2); untransformed Col-0 (lane 3); ARC6 F44 2 D'

expressing transgenic line #1 (lane 4); ARC 6 F44 2 D-expressing transgenic line #2 (lane 5);
ARC 6-expressing transgenic line #1 (lane 6); ARC 6-expressing transgenic line #2 (lane
7); ARC6-expressing transgenic line #3 (lane 8) ; and ARC6-expressing transgenic line #4
(lane 9). Two milligrams of whole-cell extract from flower buds are loaded in each lane.

50

 

   

I...

E

Chloroplasts per cell

 

Col-0 arc6 arc6 arc6
+ARC6 (T1) + F442D (T1)

 

 

 

Figure 2.9. Quantitative analysis of chloroplast number in ARC6F442D-expressing
transgenics. Micrographs of mesophyll cells from: Col-0 (A); arc6 (B); ARC6-

expressing transgenic line #3 (C); and ARC6 F44 2D-expressing line #2 (D). Scale bars =
10 um. Quantitative analysis of chloroplast number within lines shown in A-D and
Figure 2.8. Chloroplasts were counted using single-plane images from mesophyll cells
with average areas of 3838 i 713 umz. Ten cells were analyzed for each line using leaf
tip samples from plants ~5 weeks post-germination. Error bars = standard deviation from
the mean.

51

A _
A

Figure 2.10. F tsZ immunolocalization in ARC6F4420-expressing transgenics. FtsZ2-
1 localization (green) in wild type Col-0 (A); arc6 (B); ARC6-expressing transgenic line

 

#3 (C); ARC 61:44 ZD-expressing line #2 (D). FtsZ localizes to equatorial rings in wild
type chloroplasts (large arrowheads, panels A, C, and D) and disorganized fragments
(small arrowheads, panel B) in arc6 loss-of—function mutants. Chlorophyll
autofluorescence (red) marks the approximate boundaries of the chloroplast. Scale bars =
5 pm.

52

morphology within the chloroplasts of ARC6F44ZD-expressing lines (Figure 2.10D) was

clearly different from that in untransformed arc6 mutants (Figure 2.10B) and was
indistinguishable from that in wild type Col-0 (Figure 2.10A) and transgene-

complemented arc6 mutants (Figure 2.10C).

From these results, I conclude that ARC6F442 is either not involved in mediating

ARC6-FtsZ2 interaction in vivo, or is only one of several residues that stabilize ARC6-

FtsZ2 interaction in vivo. These results are congruent with the comparable FtsZ2-binding

abilities of both ARC6ZBD and ARC6ZBD (F442D) in pulldown assays (Figure 2.7).
However, these results are inconclusive with regard to the function of ARC6ZBD, as the

prospective ARC6 F44 2 D transgene was fully functional and therefore did not provide

further insight into the role of this proposed domain.

In vivo analysis of chloroplast targeted ARC6 ZBD

As a secondary approach to elucidating the function of the FtsZZ-binding domain
of ARC6, I generated a novel expression vector that targets the desired protein to the
chloroplast stroma using the transit peptide of RecA (Kohler, Cao, et al. 1997). In
addition to the N-terminal chloroplast transit peptide of RecA, the vector also encodes a

C-terminal EYFP to verify protein expression and localization to the chloroplast. We

placed the coding sequence of ARC6ZBD between that of the RecA transit peptide and

EYFP; this transgene is expressed under control of the CaMV 358 promoter (Figure

53

2.11). The transgene was introduced into Ws-2 and arc6-I backgrounds by

Agrobacterium-mediated transformation. To assess localization of the ARC6ZBD-EYFP

fusion, I examined mesophyll cells from young leaves of several lines by fluorescence
microscopy (Figure 2.12A-D). I observed EYFP signals that overlay with chlorophyll

autofluorescence in transformed Ws-2 and arc6-I backgrounds (Figures 2.12B and
2.12D), indicating that the ARC6ZBD-EYFP was successfully imported into the
chloroplast. Unlike ARC6, which forms a mid-plastid ring, I only observed plastid-

targeted ARC6ZBD-EYFP as a diffuse signal that was distributed throughout the

chloroplast stroma (Figures 2.123 and 2.12D). Following examination by fluorescence

microscopy, I verified transgene expression using an anti-YFP antibody. Protein extracts
from some individuals within this T1 population included a cross-reacting protein of ~47
kD (Figure 2.12B), which corresponds to the molecular weight of imported (processed)

ARC6ZBD-EYFP. To determine if the plastid-targeted ARC6ZBD-EYFP expressing lines

have defects in FtsZ assembly, I examined Z-ring morphology within dividing

chloroplasts of juvenile leaves (Figure 2.13, insets). In each case, FtsZ morphology was

similar to untransformed plants. To determine if the plastid-targeted ARC6ZBD-EYFP

had any gross effect upon chloroplast division, we examined fully expanded mesophyll

cells from the same lines. Consistent with our observations of FtsZ morphology (Figure

2.13, insets), expression of ARC6ZBD-EYFP had no effect upon chloroplast morphology

in Ws-2 nor arc6-I backgrounds (Figure 2.13A-D, large panels).

54

NcoI ApaI BgIII Xmal BstEII

     

 

pCAMBIA-ZBD

Figure 2.11. Schematic of plastid—targeting vector pCAMBIA-ZBD used for
ARC6ZBD domain analysis. To gain insight into the role of the FtsZZ binding domain

of ARC6, the coding sequence for ARC6ZBD was cloned into a modified plant
transformation vector derived from pCAMBIA-l302 (Hajdukiewicz, Svab, et al. 1994).
Expression of the integrated transgene is driven by the CaMV 35$ promoter (grey box,
left side). The inserted coding sequence for the protein or domain of interest (light blue,

ARC6ZBD) is directed to the chloroplast by the transit peptide of Arabidopsis RecA
(coding sequence shown in light green) (Kohler, Cao, et al. 1997). Protein expression
and localization is confirmed by fusion to EYFP (coding sequence is shown in yellow).
The nopaline synthase terminator sequence is used for transcriptional termination (grey
box, right side). The left and right border sequences of the T-DNA are indicated by the
large black arrowheads. Restriction sites are indicated above the schematic. Markers for
selection of transformants are not shown. Drawing is not to scale.

55

    

50 kD>

38 kD>

117kD>—- . .
97kD>- '5“ ‘1’.“

Figure 2.12. Demonstration of expression and targeting of ARC6ZBD-EYFP.
Merged fluorescence micrographs from young leaves of Ws-2 (A); Ws-2 expressing
ARC6ZBD-EYFP (B); arc6-I (C); and arc6-1 expressing ARC6ZBD-EYFP (D).
Exposure times in all four panels are identical. (E) a-YF P immunoblot (upper panel) and
Ponceau stain (lower panel) from arc6 lines expressing pCAMBIA-ZBD-derived
ARC6ZBD-EYFP. Molecular weight marker (lane 1); protein extracts from three
different transformant lines (lanes 2-4); Ws-2 (lane 5); and arc6-1 (lane 6). Extracts in
lane 3 were taken from the transgenic individual shown in panel D. Scale bar = 5 pm.

56

 

Figure 2.13. Chloroplast and FtsZ morphology in ARC6ZBD~EYFP expressing lines.
Terminal chloroplast phenotypes in expanded leaf cells from Ws-2 (A); arc6-1 (B); Ws-2

expressing plastid-targeted ARC6ZBD-EYFP (C); arc6-I expressing plastid-targeted

ARC6ZBD-EYFP (D). Insets in each panel show FtsZ2-1 localization (green) in
chloroplasts (red) within young leaves of the same lines shown in the large panels of A-
D. The inset micrographs are the same scale as the large panels. Scale bars = 10 um.

57

Taken together, from these results I conclude that either ARC6ZBD does not

function in plastid division or that ARC6ZBD cannot function outside of the context of

the full-length ARC6 protein.

Discussion

Here I have identified a putative FtsZ2-binding domain of ARC6 (ARC6ZBD) and

attempted to characterize its function in vivo. From our results, I conclude that the

conserved ARC6ZBD resides within the context of a larger stromally-localized conserved

region, occupying amino acids 351-503 (Figure 2.1). Multiple sequence alignment

between ARC6 orthologs of higher plants reveals that ARC6ZBD is conserved amongst
this group of organisms (Glynn, Yang, et al. 2009) and our pairwise alignment (Figure

2.3) suggests that ARC6ZBD is distantly related to the FtsZ binding domain of ZipA

(ZipAZBD) (Figure 2.3). Despite these differences in amino acid sequence, these

divergent domains are similar in length and all bind the core motif of FtsZ proteins.

Using sequence alignment sofiware, a homology-modeling platform, and side—chain

positioning algorithms, we generated a model of ARC6ZBD (Figure 2.4) based on the
known crystal structure of ZipAZBD (Mosyak, Zhang, et al. 2000). This model
highlights the potential importance of at least 3 key residues within ARC6ZBD and

provides a testable foundation for structural analysis of ARC6ZBD by crystallographic or

58

NMR-based methods. Similar to the ZipA-FtsZ interaction (Moreira, Femandes, et al.

2006), site-directed mutagenesis of a conserved phenylalanine residue abolished

ARC6ZBD interaction with FtsZZ in yeast two-hybrid assays (Figure 2.6), suggesting that
this residue of ARC6ZBD might secure the core motif of FtsZ2 within a hydrophobic
pocket. To test this, I generated a site-directed mutation in recombinant ARC6ZBD
protein, ARC6ZBD(F44215), that I predicted to be unable to bind FtsZ2 in pulldowns, but I
could not validate two-hybrid results using this assay; both ARC6ZBD and ARC6ZBD
(F442D) had similar affinity for FtsZZ in vitro (Figure 2.7). This result conflicts with the
structural model and hypothesis that ARC6F442 is critical for ARC6 function in vivo. To
determine the functional relevance of ARC6F442 and to assess which of the interaction
assays was yielding valid information, I made a construct that represents an ARC6ZBD
loss-of-function mutant (ARC6pr0-ARC6 F44 2 D), based on the two-hybrid results (Figure

2.6). ARC6pro-ARC 6 F44 2 D was able to fully complement an arc6 T-DNA insertion

mutation (Figure 2.9) and F tsZ ring morphology in complemented individuals was

indistinguishable from that in wild type chloroplasts (Figure 2.10), suggesting that the

two-hybrid tests between ARC6ZBD (F442D) and FtsZ2 may have yielded false negative
results. However, the in vivo functionality of the ARC6F442D protein is completely

consistent with our pulldown assays that show that ARC6ZBD (F442D) is able to

coprecipitate the C-terminus of FtsZZ (Figure 2.7). It is possible that site-directed

59

substitutions of ARC6 residues 1362 and R422 (both individually and in combination

with ARC6F442 substitutions) might provide sufficient evidence to validate our structural

model (Figure 2.4) if those substitutions significantly disrupt ARC6-FtsZ2 interaction

both in vitro and in vivo. However, in the absence of these additional site-directed

mutants and the lack of an experimentally-detennined structure for ARC6ZBD, our ZipA-

based structural model should be considered preliminary.

While our work highlights a prospective functional domain within ARC6, the

conflicting conclusions drawn from the two-hybrid and in vivo results does lead to some

uncertainty regarding the structure and function of ARC6ZBD. Our results here reiterate

the previous findings of our colleagues (Maple, Aldridge, et al. 2005), who used yeast
two-hybrid assays to identify the ARC6-FtsZ2 interaction. They claimed to confirm this
two-hybrid result using BiFC between ARC6 and FtsZZ in tobacco (Maple, Aldridge, et
al. 2005, Maple and Moller 2006), using C-terminal BiFC tags for ARC6 and F tsZ2 to
reconstitute YFP in vivo. The bitopic topology of ARC6 (V itha, Froehlich, et al. 2003)
and the stromal localization of FtsZ2 (McAndrew, Froehlich, et al. 2001) conflict with
this result, because the C-terminus of ARC6 has been shown to reside within the
intermembrane space; BiFC-based assays will not work across membrane boundaries, as
the two halves of the fluorescent moiety cannot physically interact when separated by a
membrane barrier (Bracha-Drori, Keren, et al. 2004). In this case, the topology of ARC6
in these assays was probably not the native topology (perhaps due to overexpression or

the presence of the C-terminal tag) or some FtsZZ was present within the intermembrane

60

space. We also analyzed a mutation in FtsZZ (FtsZ2F466 A) that was shown to impede

ARC6-FtsZ2 interaction by yeast two-hybrid (Maple, Aldridge, et al. 2005), but in our

hands FtsZZF466 A was coprecipitated by ARC6ZBD (Figure 2.7), suggesting that

previous results (Maple, Aldridge, et al. 2005) might be two-hybrid false negatives.

Strangely, expression of plastid-targeted ARC6ZBD did not modify the chloroplast

morphology of F tsZ immunolocalization patterns of wild type or arc6 mutants (Figure
2.13), leading us to question the overall relevance of ARC6ZBD in the process of
chloroplast division; perhaps the FtsZ2-binding domain of ARC6 has some other function
in vivo. Further confounding the issue, another group has demonstrated that interaction
between the cyanobacterial ortholog of ARC6 (Ftn2) and FtsZ requires the J -domain of
Ftn2, suggesting that an interaction between ARC6 and FtsZZ might occur directly

through the putative J-domain of ARC6 — or may be indirect, perhaps occuring through

a DnaK-like chaperone (Mazouni, Domain, et al. 2004). When combined with our

negative results obtained using a prospective ARC6ZBD gain-of-fimction mutant (Figure

2.13), we believe more work is needed to show that ARC6ZBD directly binds FtsZ2

within the stroma and to identify the functional implications of that interaction, if it does,

in fact, occur.

Regardless, the overall effect of ARC6 upon Z—ring formation is profound, as

arc6 mutants possess serious defects in Z-ring assembly (V itha, Froehlich, et al. 2003).

While recent data clearly provide evidence of a complex containing ARC6 and FtsZZ, it

61

does not resolve the issue of direct ARC6-FtsZZ interaction (McAndrew, Olson, et al.
2008). Moreover, preliminary data suggest that ARC6 can immunoprecipitate F tsZZ-l ,
even in the absence of the FtsZZ-binding domain of ARC6, indicating that there may an
indirect linkage between these two proteins in vivo (Figure 2.14). Further work will be
required to sort out the details of the molecular arrangement within the ARC6-FtsZ2

complex and determine how ARC6 might modulate Z—ring formation in vivo.

62

50 kD>

3 7 kD> a-FtsZZ

 

Figure 2.14. ARC6 A A 1-509-GFP and ARC6“ 1-331-GFP can immunoprecipitate
FtsZ2. Immunoblot showing eluates from immunoprecipitation reactions using an
immobilized anti-GFP antibody incubated with whole-cell extracts from arc6 lines
expressing a 35S—ARC6AA 1_509-GFP transgene (lane 1); 35S—ARC6 AA 1-331-GFP (lane
2); or 35S-GFP (lane 3). Blot was probed with an FtsZ2-1-specific antibody. The
molecular weight of processed F tsZZ-l is predicted by ChloroP (Emanuelsson, Nielsen,
et al. 1999) to be ~45 .3 kD (upper band). The lower molecular weight band in lane 1 (*)
is probably derived from FtsZ2-l, due to the high specificity of the FtsZ2-1 antibody. It
is not clear if this lower molecular weight band represents an FtsZ2-l breakdown product

or if ARC6 AA 1-509 might somehow affect FtsZZ-l translation or processing in vivo.

63

Materials and Methods

Yeast two-hybrid vector construction and reporter assays.

All GAL4-ARC6 fusion-encoding plasmids were generated by PCR using primers
that correspond to the desired coding sequences, with Ndel and Xmal adapters to
facilitate directional ligation. PCR products were generated using the ARC6 cDNA clone
available from ABRC, cut, and ligated into Ndel-Xmal digested pGADT7 (Clontech).
All GAL4-ARC6 site-directed mutants (F442A, F442D, F 442K, and F 442Y) were
generated by SOE-PCR (Heckman and Pease 2007) and cloned into pGADT7 as
described above. All GAL4-FtsZ2 fusions were generated using primers corresponding
to the desired coding sequence, with Ndel and BamHI adapters to facilitate the desired
directional ligation of the insert. PCR products were generated from an FtsZZ-l clone,
cut, and ligated into Ndel-BamHI digested pGBKT7 (Clontech). All clones were
sequenced prior to transforming yeast to ensure that the inserts were free of coding errors.
HIS3 reporter assays were performed in line with the manufacturer’s (Clontech)
recommendations using an established protocol (Maple, Aldridge, et al. 2005). Relative

interaction strengths were computed by comparison of the ratio of yeast growth on SD/-

ULTH media to yeast grth on SD/-ULT media after ~48 hours of growth at 28-30 0C.

Multiple sequence alignments and homology model.
Multiple sequence alignments were performed using ClustalW (EBI) using the
BLOSUM62 substitution matrix with all other parameters set to their default values. The

alignment was formatted using ESPript (http://esnrint.ibcp.fr/ESPript/cgi-

64

bin/ESPript.cgi/). For homology modeling, a pairwise alignment between ZipA A A 135-

323 and ARC6A A 3 5 1-503 was performed and used to make a primary homology model

from PDB file 1F47 (Mosyak, Zhang et al. 2000) (http://wwwpdbcgh with all
identical residues from the ZipA-ARC6 pairwise alignment as anchor points using a
homology modeling program (http://proteins.msu.edu/). We performed refinement of
side chain positions using SCWRL3 (httn://dunbrack.fccc.edu/SCWRL3.php). Cartoon
representations of the homology model were generated using Pymol

(http://pymolsourcefogenefl).

In vitro pulldowns.

All fusion proteins were expressed in E. coli BL21 (DE3) Codon Plus cells

(Stratagene) induced at OD600 = 0.8 with 2 mM IPTG for 2 hours at 37°C. 750 ug of

protein from induced cell extract was used for each bait/prey combination. Pulldowns

between Hi83-ARC613D and GST-FtsZZ-ICT or Hi83-ARC6ZBD (F442D) and GST-

F tsZZ-ICT were performed as in previous experiments (Glynn, Froehlich, et al. 2008),

except Triton X-100 was present at 0.1% in all wash buffers to prevent clumping of the

sepharose beads.

Analysis of chloroplast morphology and number.
Light micrographs depicting chloroplast morphology in expanded leaf cells were
taken using DIC Optics on a Leica DMIBOOOB Inverted Microscope outfitted with a

Leica DFC3 20 Camera. Samples for chloroplast morphology and quantitation were

65

prepared and analyzed using established protocols (Pyke and Leech 1991). Fluorescence
micrographs were taken using a Leica DMRA2 using Q-Capture Camera Control
Software (Q-Imaging) and the filter sets indicated (Leica) as previously described (V itha,
Froehlich, et al. 2003). Image analysis and RGB composites were made using Image]

v1.37 (NIH) (Bearer 2003).

F tsZ immunofluorescence
Tissue preparation, fixation, and immunofluorescence analysis were carried out as
described previously (Miyagishima, Froehlich, et al. 2006, Vitha, Froehlich, et al. 2003,

Vitha, McAndrew, et al. 2001).

Construction of pCAMBIA-ZBD.

pCAMBIA-ZBD was constructed from pCAMBIA-1302 (Hajdukiewicz, Svab, et
al. 1994) by removing an NcoI-BstEII fiagment from pCAMBIA-1302 and performing a
four-point ligation (see Figure 2.11) with: (1) NcoI/ApaI digested PCR products
amplified from a RecA cDNA using primers TTTT'ITCCATGGATTCACAGCTAGTCT
TGTC and TTT'I‘TTGGGCCCTCTGTCATCGAATTCAGAACTGAT'T; (2) ApaI/BgllI
digested PCR products amplified from an ARC6 cDNA using primers TTTTTTGGG
CCCGAAGTTGCACTTGCTCTTGTGGCT and TTTI‘TTAGATCTAACTACCTCCA
CTCTTTCCAAGT; and (3) BglII/BstEII digested PCR products amplified from an
EYFP-coding plasmid (F. Brandizzi Laboratory, MSU DOE-PRL) using primers
TTTTTTAGATCTCCCGGGATGGTGAGCAAGGGCGAGGAGCT and

I I I I I IGGTTACCTTACTTGTACAGCTCGTCCATGCC. Following transformation,

66

clones were selected on LB plates containing kanamycin (50 ug/mL) and screened by
PCR. Clones containing an insert of the expected size were miniprepped (Promega) and
sequence-verified at the MSU-RTSF facility before transforming Agrobacterium and

Arabidopsis.

Fluorescence microscopic analysis of ARC 6 ZBD'E YF P expressing plants.

Fluorescence micrographs were taken using a Leica DMRA2 using Q-Capture
Camera Control Software (Q-lmaging) and the filter sets indicated (Leica) as previously
described (V itha, Froehlich, et al. 2003). Image analysis and RGB composites were

made using Image] v1.37 (NIH) (Bearer 2003).

Protein extraction and immunoblotting.

Proteins from whole-cell extracts (equivalent to ~2-3 mg liquid-nitrogen ground
tissue) were prepared as previously described (Wiegel and Glazebrook 2002), separated
by SDS-PAGE, and transferred to nitrocellulose. Anti-GF P Irnmunoblots were
performed using Clontech JL-8 anti-GFP monoclonal antibody at 1:1000 in 5% nonfat

dry milk in TBS-T, pH 7.4 (Miyagishima, Froehlich, et al. 2006).

Transgenic plant material and microscopic analyses.

Untagged ARC6 and ARC6 F44 2 D transgenes were generated by SOB-PCR and

ligated into a modified pCAMBIA-l 302 (Hajdukiewicz, Svab, et al. 1994) using NcoI
and BstEII; BstEII removes the mGFP tag from pCAMBIA-1302 and retains the NOS

terminator sequence. All clones were sequence-verified and transformed into

67

Agrobacterium tumefaciens GV3101 by electroporation. Overnight cultures of

A grobacterium carrying the T—DNA construct were used to transform Ws-2 and arc6-1

plants (Clough and Bent 1998). Following hygromycin selection, putative transgenic T1

plants were screened for transgene expression by Western blotting of whole-cell extracts
from various individuals and probing with an anti-GFP antibody (BD-Biosciences J L-8
monoclonal antibody) and by fluorescence microscopy. Immunofluorescence analysis
was carried out as previously described (V itha, McAndrew, et al. 2001) using expanded

leaf tissue.

Immunoprecipitation from Homogenized Arabidopsis Leaf Tissue.

Transgenic arc6 plants expressing GFP-fusion proteins were grown in soil for 21 days.
Approximately 150-200 mg of tissue was collected from whole leaves and ground to a
powder in liquid nitrogen. Four (4) volumes of IP buffer (25 mM Tris, pH 7.50; 150 mM
NaCl; 0.1% Triton X-100; and 1X Roche Protease Inhibitor Cocktail) were added to the
frozen powder and gently mixed by pipeting on ice over several minutes. The
homogenized slurry was spun through a Miracloth filter to remove large debris. 2.5 pg of
BD Living Colors A.v. Peptide Antibody (Cat. # 632377) was gently mixed into to each
filtered extract and incubated overnight. 50 uL of pre-conditioned protein A beads were
added to each sample and incubated for 2 hours at room temperature with gentle mixing
by inversion every 10 minutes. The beads were washed 4 times with 500 uL of IP buffer
(each wash was 500 uL). Immune complexes were eluted from the beads using 100 uL
of elution buffer (0.2M Glycine, pH 1.85) and then buffered with 20 uL of l M Tris, pH

8.5. 29 uL of 6X SDS-PAGE buffer were added to each buffered eluate, samples were

68

vortexed briefly, boiled, collected by centrifugation, and loaded for SDS-PAGE analysis.
Transfer to nitrocellulose was performed using a GENIE Blotting Apparatus (Idea
Scientific) according to the manufacturer’s recommendations. Rabbit-derived anti-

FtsZZ-l antibodies were used as described previously (Vitha, McAndrew, et al. 2001 ).

Acknowledgements

I thank Bill Wedemeyer for instruction and assistance in generating the structural

model of ARC6ZBD and for all the excellent structural analysis tools he has freely

provided at his website.

69

Chapter 3

ARC6 Binds and Positions PDV2 During Plastid Fission in Arabidopsis.

70

Abstract

Chloroplasts arose from a free-living cyanobacterial endosymbiont and divide by
binary fission. Division involves the assembly and constriction of the endosymbiont-
derived, tubulin-like F tsZ ring on the stromal surface of the inner envelope membrane
and the host-derived, dynamin-like ARC5 ring on the cytosolic surface of the outer
envelope membrane. Despite identification of many proteins required for plastid
division, the factors coordinating the internal and external division machineries are
unknown. Here, we provide evidence that this coordination is mediated in Arabidopsis
by an interaction between ARC6, an F tsZ-assembly factor spanning the inner envelope
membrane, and PDV2, an ARC 5 recruitment factor spanning the outer envelope
membrane. ARC6 and PDV2 interact via their C-terminal domains within the
intermembrane space of the chloroplast, consistent with their in vivo topologies. ARC6
acts upstream of PDV2 to localize PDV2 (and hence ARC5) to the division site. We
present a model whereby ARC6 relays information on stromal FtsZ ring positioning
through PDV2 to the chloroplast surface to specify the site of ARC5 recruitment.
Because orthologs of ARC6 occur in land plants, green algae, and cyanobacteria whereas
PDV2 occurs only in land plants, the connection between ARC6 and PDV2 represents the
evolution of a plant-specific adaptation to coordinate the assembly and activity of the

endosymbiont- and host-derived plastid division components.

71

Introduction

The plastids of plant cells arose from cyanobacteria by endosymbiosis and, like
cyanobacteria, replicate by binary fission. This process requires the coordinated action of
at least two macromolecular complexes, one composed of the tubulin-like cytoskeletal
protein FtsZ and the other of the dynamin-related protein ARC5. These proteins
assemble into mid-plastid ring-shaped structures on opposite sides of the two envelope
membranes to mediate constriction of the organelle (Osteryoung and Nunnari 2003).
Two nuclear-encoded plant F tsZ paralogs, Ftle and FtsZ2, function within the
chloroplast stroma. Both evolved from cyanobacterial FtsZ and have unique non-
overlapping functions in plastid division (Schmitz, Glynn, et al. 2009, Stokes and
Osteryoung 2003, Vitha, McAndrew, et al. 2001). Analogous to-bacterial FtsZ (Bi and
Lutkenhaus 1991, Goehring and Beckvvith 2005), the plastidic FtsZ proteins assemble at
an early step in division to form an equatorial ring, the Z-ring, on the stromal face of the
inner envelope membrane (V itha, McAndrew, et al. 2001). In bacteria and presrunably in
plastids, the Z-ring probably functions both as a scaffold for the recruitment of other
division proteins and to provide the contractile force needed to pull the membrane inward
during constriction (Ghosh and Sain 2008, Lan, Daniels, et al. 2009, Lan, Wolgemuth, et

al. 2007, Osawa, Anderson, et al. 2008).

In contrast, ARC5 was a post-symbiotic adaptation of the eukaryotic host and

functions outside the chloroplast (Gao, Kadirjan-Kalbach, et al. 2003, Miyagishima,

Nishida, et al. 2003a). ARC5 is a member of the dynamin family of membrane

72

“pinchases,” best characterized for their roles in endocytic vesicle-budding and
mitochondrial fission in eukaryotes (Cerveny, Tamura, et al. 2007, Hoppins, Lackner, et
al. 2007, McNiven 1998, Shaw and Nunnari 2002, Ungewickell and Hinrichsen 2007),
though dynamins probably evolved from a prokaryotic ancestor (Low and Lowe 2006).
ARC5 in Arabidopsis and its orthologue in the red alga Cyaniodioschyzon merolae
function late in chloroplast division by assembling on the cytosolic surface of the outer
envelope membrane, where they are thought to perform the final squeeze that aids
partitioning of the two daughter organelles (Yoshida, Kuroiwa, et al. 2006). Both
plastidic FtsZ and ARC5 arose early in the evolution of the chloroplast division
machinery, as indicated by their occurrence in both the red and green lineages

(Miyagishima 2005).

The assembly of the division machinery appears to occur in a linear order, with
F tsZ assembly initiating the process and ARC5/dynamin mediating late-stage organelle
constriction (Miyagishima, Nishida, et al. 2003a). However, the mechanisms
coordinating these evolutionarily and compartrnentally disparate events are unknown.
Previous studies in Arabidopsis have implicated three plastid division proteins as
candidate mediators of this coordination: ARC6, PDVl and PDV2 (Miyagishima,
Froehlich, et al. 2006, Vitha, Froehlich, et al. 2003). ARC6 is a bitopic transmembrane
protein of the inner envelope membrane with its larger N-terminus protruding into the
stroma and smaller C-terminus residing within the intermembrane space (IMS). ARC6
was inherited from the cyanobacterial endosymbiont and is localized to the mid-plastid .

division site. The chloroplasts of Arabidopsis arc6 mutants, which have one or two

73

oversized chloroplasts per mesophyll cell (Pyke, Rutherford, et al. 1994), possess many
short disorganized filaments, while ARC6 overexpressors have excessively long F tsZ
filaments (V itha, F roehlich, et al. 2003). These findings, along with the fact that the N-
terminus of ARC6 interacts specifically with FtsZ2 (Maple, Aldridge, et al. 2005),
suggest that ARC6 facilitates F tsZ polymer assembly and regulates FtsZ ring dynamics
through FtsZZ (Maple, Aldridge, et al. 2005, McAndrew, Olson, et al. 2008, Vitha,
Froehlich, et al. 2003). The function of the C-terminal IMS region of ARC6 is unknown,
but it probably has a significant role in plastid replication based on the breadth of its
conservation amongst plants, algae, and cyanobacteria (Glynn, Yang, et al. 2009, Vitha,
Froehlich, et al. 2003). ARC6 is one of the few division proteins known to connect the
stromal Z-ring to the IMS, suggesting it could play a critical role in coordinating the inner

and outer subassemblies of the division machinery.

PDVl and PDV2 are paralogous plastid division proteins identified based on the
similarity of the pdvl mutant phenotype to that of arc5; in both mutants, chloroplasts are
enlarged and dumbbell-shaped (Miyagishima, Froehlich, et al. 2006). PDVl is a bitopic
outer envelope membrane protein that localizes to the division site, with its N-terminus
residing in the cytosol and its C-terrninus extending into the IMS. PD VI and PD V2 have
partially redundant functions in recruiting ARC5 to the chloroplast. In pdvl and pdv2
mutants, ARC5 localizes to the central constriction in the enlarged chloroplasts, but pdvI
pdv2 double mutants fail to recruit ARC5 to the chloroplast (Miyagishima, Froehlich, et
al. 2006). PDVl and PDV2 proteins share some degree of sequence similarity and

domain arrangement, though the localization and detailed function of PDV2 was not

74

determined previously (Miyagishima, Froehlich, et al. 2006). PDV] and PDV2 have no
significant sequence similarity to known proteins and are only evident in land plants,
suggesting they represent an evolutionary development in the transition of plants to

terrestrial habitats.

Here, we show that PDV2 has a localization and topology similar to that of PDVl
in the outer envelope membrane. However, PDV2 family members have a unique C-
terminal extension in their IMS regions that is lacking in PDVl proteins. We further
show that the C-terminal IMS regions of ARC6 and PDV2 interact and that this
interaction is required for full chloroplast division activity in Arabidopsis. Using genetic
analysis, we demonstrate that ARC6 is required for positioning of PDV2 and ARC5, but
PDV2 is not required for mid-plastid localization of ARC6. Our results establish a
physical link across the envelope membranes at the division site and suggest that ARC6,
through interaction with PDV2 within the IMS, coordinates Z-ring and ARC5 activity to
synchronize scission of the envelope membranes. We present a model for the physical

arrangement and interaction of these proteins in the chloroplast membranes.

Results

PD V2 localization and topology are similar to PD VI .

The similarity of PDV2 to PDVl in sequence and domain arrangement (Figure

3.1) suggested its localization and topology might be similar to those of PDVl

75

 

 

 

 

 

 

 

PDV]
N l D c
— — TM i
272 AA
PDV2
N l D c
— _ TM I 7
307 AA

 

 

 

Figure 3.1. Schematic comparison of PDVl and PDV2 proteins from Arabidopsis.
Shaded lines below each protein indicate regions of high similarity between the two PDV
paralogs. A comprehensive alignment and phylogeny of these two proteins that includes
sequences from several plant species is published elsewhere (Glynn, Froehlich, et al.
2008). Amino terminus (N); carboxy terminus (C); transmembrane domain (TM); and
amino acids (AA).

76

(Miyagishima, Froehlich, et al. 2006). To verify localization of PDV2 in vivo, we

expressed a YFP-PDV2 fusion protein from the PDV2 promoter (PD V2pm- YFP-PD V2)

in wild-type Arabidopsis plants. Similar to PDVl , the YFP signal localized to the mid-
plastid in young emerging leaves. The YFP-PDV2 signal appeared as a continuous ring
(Figure 3.2) rather than as a series of equatorial spots as was previously noted for PDVl
(Miyagishima, Froehlich, et al. 2006). When YFP-PDV2 was co-expressed with an
ARC6-CFP fusion protein, their fluorescence signals colocalized within mesophyll
chloroplasts (Figure 3.3). To confirm the predicted topology of PDV2, we carried out in
vitro chloroplast import and protease-protection assays (Figure 3.4, left panel).
Following incubation with isolated pea chloroplasts, radiolabled PDV2 produced by in
vitro translation (Figure 3.4, lane 1) was retained in the membrane fraction (Figure 3.4,
lane 2). The protein was susceptible to degradation by therrnolysin, consistent with outer
envelope localization. However, a portion of PDV2, roughly corresponding to the size of
the predicted C-terminal IMS domain (predicted size ~10 kD), was protected from
therrnolysin degradation (Figure 3.4, lane 4, arrowhead), but was sensitive to trypsin,
which can penetrate the outer envelope membrane and enter the IMS (Jackson, Froehlich,
et al. 1998, Tranel, Froehlich, et al. 1995). ARC6, used as an inner envelope control in
these experiments (Figure 3.4, right panel), behaved as shown previously (V itha,
Froehlich, et al. 2003). Protease protection assays on wild type Arabidopsis chloroplasts
using antibody-based detection of PDV2 showed the N-terminus of PDV2 to be
susceptible to thermolysin and trypsin treatment (Ronit Knopf and Zach Adam, personal
communication), consistent with in vitro assays. These results confirm that PDV2, like

PDV] , is a mid-plastid-localized, bitopic protein of the chloroplast outer envelope

77

 

Figure 3.2. Localization of YFP-PDV2 in Arabidopsis. A PD VZPrO
transgene was expressed in Arabidopsis Col-0 plants and tissue samples from emerging
leaves were examined for YFP-PDV2 localization. Signals observed from: Chlorophyll
autofluorescence (Chl, panel A); (B) Yellow Fluorescent Protein (YF P, panel B); and
merged image (panel C). YFP-PDV2 was typically observed as a continuous ring at the
mid-chloroplast (arrowhead, panel C). Scale bars = 10 um.

- YFP-PD V2

78

 

. f; .- B . l) a
.3 .31}
a ,

Figure 3.3. Colocalization of ARC6-CFP and YFP-PDV2 in Arabidopsis. ARC6-
CF P and YFP-PDV2 were expressed under control of their native promoters and
examined in emerging leaves. Micrographs from brightfield (A); CFP channel (B); YFP
channel (C); chlorophyll autofluorescence (D); and a merged image from CFP, YF P and
chlorophyll channels (E). Scale bar = 10 um.

79

--++-- Thermolysin - - + +

   

----++ Trypsin ----++
PS PSPS P S P S P S
50kD> 100m>
37“» 75kD>
25kD>
20kD>
kD>
‘ 50
1234567 891011121314
PDV2 ARC6

Figure 3.4. PDV2 fractionation and topology using isolated pea chloroplasts. In vitro
chloroplast import of [3H-leucine]-labeled PDV2 or [ 5S-methionine]-labeled ARC6 and
fractionation of chloroplasts following protease treatment and hypotonic lysis. In vitro-
transcribed translation products are shown for PDV2 (lane 1) and ARC6 (lane 8). Pellet
(P) fractions are shown in lanes 2, 4, 6, 9, 11, and 13. Supernatant (S) fractions are
shown in lanes 3, 5, 7, 10, 12, and 14. The arrowhead points to the IMS-localized C-
terminal fragment of PDV2 that remains following thermolysin treatment.

80

membrane, with its N-terminus exposed to the cytosol and its smaller C-terminus residing

within the intermembrane space.

The IMS Regions of ARC6 and PD V2 Interact.

The localization of PDVl (Miyagishima, Froehlich, et al. 2006), PDV2 (Figure
3.2) and ARC6 (V itha, Froehlich, et al. 2003) at the division site, along with their
demonstrated topological orientations in the envelope membranes, suggested the
possibility that ARC6 might interact with PDVl and/or PDV2 in the IMS. To test this,
we carried out yeast two-hybrid assays with constructs encoding the C-terminal IMS

domains of ARC6 and PDVI or PDV2. PDV2 strongly and specifically activated the

HIS3 reporter in the presence of ARC61MS (Figure 3.5A, third row). No interaction was

observed between PDVIIMS and ARC61MS using two hybrid assays (Figure 3.5A, first

row). To confirm the ARC6IMs-PDV21MS interaction we carried out a pulldown assay

(Figure 3.5B). A GST-PDVZIMS fusion protein was precipitated from crude E. coli

extracts with Ni-Sepharose beads coated with His-ARC61MS (Figure 3.5B, lane 1), but no

pulldown of GST-PDV21MS was observed using uncoated Ni-Sepharose beads (Figure

3.5B, lane 2). These results support an ARC6-PDV2 protein-protein interaction within

the intermembrane space of dividing chloroplasts.

81

A SD/-ULT SD/—ULTH

PDVIIMS vs. EMPTY e

PDV2,M S.vs EMPTY

B 1/10 1/100 1/10 1/100

     

 

 

or-HIS

 

or-GST

 

Figure 3.5. The IMS-localized regions of PDV2 and ARC6 interact. (A) Two-hybrid
assays between PDV proteins and ARC6. The IMS regions of the proteins indicated
were fused to either GAL4-BD (PDVl and PDV2) or GAL4-AD (ARC6). Dilutions

(indicated at bottom of panel) from a starting culture of OD600 = 1.0 are spotted onto
synthetic dropout (SD) media containing (left) or lacking (right) histidine. (B)
Immunoblots of eluates from pulldown assays using either immobilized Hiss-ARC6IMS
treated with GST-PDVZIMS (lane 1) or naked beads treated with GST-PDVZIMS (lane 2).

82

PD V2 Family Members Possess a Unique C-Terminal Domain.

To gain insight into features that distinguish the C-terminal IMS domains of
PDVl and PDV2 from one another and therefore might be important for the PDV2-
specific interaction with ARC6, we generated sequence alignments between PDVl and
PDV2 proteins from several land plants using a CLUSTALW identity scoring matrix
(Larkin, Blackshields, et al. 2007) to clearly define PDVl-specific and PDV2-specific
features (Glynn, Froehlich, et al. 2008); the C-terminal segment of this alignment is
shown in Figure 3.6. As noted previously, all PDV] proteins have a conserved C-
terrninal glycine residue, and mutation of this residue impairs PDVl function in
Arabidopsis (Miyagishima, Froehlich, et al. 2006). A conserved terminal glycine is also
found in all known PDV2 family members (Figure 3.6, red asterisk). However, PDV2
(294 at 15 amino acids, n = 7 sequences) proteins are generally longer than PDVl (261 :t
14 amino acids, n = 7 sequences); PDV2 family members harbor a conserved extension at
their C-terminus. This C-terminal extension, which includes the terminal glycine residue,

might be involved in mediating PDV2 interaction with ARC6.

The Conserved Terminal Glycine of PD V2 is Required for Interaction with ARC6 and
PD V2 Function in vivo.

To ask whether the C-terminal glycine of PDV2 is important for the ARC6-PDV2

interaction, we engineered a missense mutation in the PDVZIMS two-hybrid plasmid that

changes this glycine to aspartate. Following co-transformation of the resulting construct

(PDV21MS(G307D)) and ARC6IMS into yeast and selection for both plasmids, we

observed no grth on medium lacking histidine (Figure 3.7, row 3), indicating that

83

 

AtPDVl MA ............................
GhPDVl LA ............................
HtPDVl SA ............................
OsPDVl LA ............................
PtPDVl LA ............................
VvPDVl YA ............................
ZmPDVl LA ............................
AOPDVZ E . E QPRCLVKERVEI PFDLDVSAPKINYGFG
ABPDVZ E . EARCLVKERLEI PFDPVVRTPNVNYGCG
AtPDVZ E EARCLVKERVEI PFDSVVAKRDVTYGYG
LBPDVZ E . ESRCLVKERVKIPFKSVVTLPDVNYGCG
OsPDVZ E . RAHCVVKERVEIPFDTNLASPNASYGLG
PtPDV2 E. EVRCVVKERVAVPFNSVAGKPDVNYGSG
TaPDVZ GGD RAHCVVKERVEIPFGSSLDAPNASYGLG

— * **** ** *4: *

Figure 3.6. Alignment of C-termini of PDVl and PDV2 family members. All PDV2
proteins possess a C-terminal extension that is not found in PDVl family members. The
dark line (bottom) corresponds to to the C-terminal conserved segment shown in Figure
3.1. Boxed regions indicate similarity between PDVl and PDV2; boxed shaded regions
represent identity between PDVl and PDV2. Asterisks (*) represent identical residues
within the C-terminal extension of all PDV2 family members. The red asterisk highlights
the conserved glycine that ends all PDV2 proteins. Arabidopsis thaliana PDVl
(AtPDVl), Gossypium hirsutum PDVl (GhPDVl), Helianthus tuberosus PDVl
(HtPDVl), Oryza sativa PDVl (OsPDVl), Populus trichocarpa PDVl (PtPDVl), Vitis
vinifera PDVl (VvPDVl), Zea mays PD V1 (ZmPDVl), Arabidopsis thaliana PDV2
(AtPDV2), Asparagus ojj‘icinalis PDV2 (AoPDV2), Aquilegia species PDV2 (AsPDV2),
Lactuca sativa PDV2 (LsPDV2), Oryza sativa PDV2 (OsPDV2), Populus trichocarpa
PDV2 (PtPDV2), and Triticum aestivum PDV2 (TaPDV2). The full-length alignment is
published elsewhere (Glynn, Froehlich, et al. 2008).

84

A SD/-ULT SD/—ULTH

PDV2 1M, vs. ARC6 ,M, m _ 1
PDV2 IMSvs. EMPTY _ _ 2
PDV2IMS (mmvs. ARC6,MS n - 3
PDV21M3(G3O7D,VS. EMPTY _ -.
EMPTY vs. ARC6,MS m -5

 
 

 

1/10 1/100 1/10 moo
B 1 2 3 4

..-ns

r: a—GST

 

Figure 3.7. The terminal glycine of PDV2 is important for interaction with ARC6.
(A) Yeast two-hybrid assays between PDV2 proteins and ARC6. The IMS regions of the

proteins indicated were fused to either GAL4-BD (PDV2 and PDV20307D) or GAL4-AD

(ARC6). Dilutions (indicated at bottom of panel) from a starting culture of OD600 = 1.0
are spotted onto synthetic dropout (SD) media containing (left) or lacking (right)
histidine. (B) Immunoblots of eluates from pulldown assays using either immobilized

Hiss-ARC6IMS treated with GST-PDVZIMS (lane 1); naked beads treated with GST-
PDVZIMS (lane 2); Hi83-ARC6IMS treated with GST-PDVZIMS (6307p) (lane 3); or
naked beads treated with GST-PDV2IMS ((33071)) (lane 4).

85

mutation of the terminal glycine in PDV2 diminishes the ARC6IMs-PDV21MS interaction
in yeast. We incorporated the same mutation into PDV2IMS to create GST-PDVZIMS

(G 307D) for use in pulldown assays and observed that the terminal G307D mutation
reduced the affinity of PDVZIMS for ARC6IMS (Figure 3.7B, lane 3). The higher affinity

of GST-PDV21MS for His-ARC61MS (Figure 3.7B, lane 1) was not a result of biased prey

input into the reaction, as equal amounts of total protein were applied to each pulldown

reaction and we verified equal input levels of GST-PDVZIMS and GST-PDV2IMS

(03071)) in the reactions by anti-GST immunoblotting. We conclude from these results

that the IMS-localized domains of ARC6 and PDV2 interact with each other and that the

conserved terminal glycine of PDV2 is an important mediator of this interaction.

To assess the potential importance of the terminal missense mutation in vivo, we

compared the relative abilities of PD V2pm—PD V2 and PD VZPm-PD V20 307D transgenes

encoding full-length proteins to complement pdv2-l , a line carrying a T-DNA insertion
allele of PD V2 in which chloroplasts are frequently constricted and larger than in wild
type (Miyagishima, Froehlich, et al. 2006) (Figure 3.8B). We reasoned that a PDV2
protein incapable of interaction with ARC6 would not be able to rescue the pdv2
phenotype. The wild type transgene fully complemented pdv2 in the majority of selected

lines (Figure 3.8C). In contrast, no modification of the pdv2 phenotype occurred in lines

transformed with PD V2pro-PD V20 30 7D (Figure 3.8D) despite evidence based on

86

 

 

  

 

. £3
"a Ponceau

 

Figure 3.8. PD V2030”) is a loss-of function allele. Chloroplast phenotypes from leaf
cells of Col-0 (A); pdv2-I (B); pdv2-I transformed with a PD V2pro-PDV2 transgene (C);
and pdv2-I transformed with a PD V2pm-PD VZG 307D transgene (D). An anti-PDV2
immunoblot is shown in (E) carrying recombinant PDV2 (lane 1); extract from plant
expressing a PD VZPm-PDVZG 307D transgene (lane 2); extract from Col-0 (lane 3); and
extract from pdv2-1 (lane 4). Chloroplast phenotype from a plant expressing 3 5S-PD V2
(F). An immunoblot showing PDV2 expression levels is shown in (G) carrying extract

from plant expressing 35S-PDV2 (lane 1); extract from Col-0 (lane 2); and extract from
pdv2-I (lane 3). Scale bars = 10 um.

87

 

immunoblotting with a PDV2-specific antibody that PDV20307D protein levels were

equivalent to or greater than PDV2 levels in wild type (Figure 3.8E). Lack of

complementation by PD V2pm-PD V20 307D was not a consequence of overexpression, as

overexpression of PD V2 does not impede chloroplast division, but rather accelerates this

process (Figure 3.8F-G) (Okazaki, Kabeya, et al. 2009).

To determine if the defect in PDV2g307D might be due to mislocalization or
perhaps a loss in activity, we examined YFP-PDVZG307D localization in Arabidopsis.

The YFP-PDV20307D signal was observed as scattered fragments around the periphery

of the chloroplast in the wild type background (Figure 3.9B), in contrast to the rings

observed in YFP-PDV2 expressing lines (Figure 3.9A), likely due to the inability of

PDVZG307D and ARC6 to interact with each other. While YFP-PDV2 was observed

forming rings in the wild type plastids, we did not observe YFP-PDV23307D associated

with chloroplasts in the pdv2-I background, but it did localize to small epidermal plastids

(compare Figures 3.9C-D); the significance of this observation is not yet clear.

We conclude that PDV2 interaction with ARC6 occurs through the conserved C-
terminal glycine of PDV2 and that the major role of this protein-protein interaction is to
localize PDV2 to the division site in vivo; loss of this interaction leads to a defect in

plastid division that is reminiscent of pdv2 and arc5 loss-of-function mutants.

88

 

Col-ll ("HI-ll

ptli'Z-I

b

 

 

 

 

Figure 3.9. PDV2G307D localization in Arabidopsis. YFP-PDV2 and YFP-

PDVZG307D were expressed under control of the native promoter in both Col-0 (A, B)
and pdv2-I (C, D) backgrounds. YFP-PDV2 localizes to continuous rings around
plastids (arrowheads) in Col—0 (A) and pdv2-1 mutants (C). In contrast, YFP-
PDV20307D fails to localize to organized structures in either background (B, D). Scale
in all images is identical. Scale bar = 2 pm.

89

The Complete IMS Region of ARC6 is not required for ARC6 Localization, but is
required for Chloroplast Division Activity.

To determine if the C-terminal IMS-localized region of ARC6 is required for
plastid division in vivo, as suggested by its interaction with PDV2, we expressed a

truncated ARC6-GFP fusion protein lacking much of its C-terminal IMS region

(ARC6A1MS-GFP) in the arc6 background under control of the ARC6 promoter (Figure

3.10A). A full-length ARC6-GFP protein, shown previously to complement the severe
plastid division defect in arc6 mutants (V itha, Froehlich, et al. 2003, Vitha, Holzenburg,
et al. 2005) was expressed as a control. Accumulation of the fusion proteins in
transgenic individuals was confirmed by immunoblotting (Figure 3.10B), and chloroplast
morphology was examined in fixed leaf cells. Similar to previous results (V itha,
Froehlich, et al. 2003), we observed rescue of the arc6 phenotype (Figure 3.10D) in lines
expressing full-length ARC6-GFP (Figure 3.10B), as indicated by the increase in ntunber
and decrease in size of the chloroplasts (25-60 cps/cell; Figure 3.10D) relative to those in

the parent arc6 plants (1-2 cps/cell; Figure 3.10C). A change in chloroplast number and

size was also observed in lines expressing ARC6A1Ms-GFP, but the extent to which the

arc6 mutation was complemented (4-10 cps/cell; Figure 3.10E) was less than observed in
arc6 lines expressing the full-length ARC6-GFP control. These results show that the

IMS region of ARC6 is required for full plastid division activity in vivo. Moreover, arc6

plants expressing either ARC6pm-ARC6AIMS-GFP or 35Spm-ARC6 A IMS-GFP exhibit

GFP rings localized to sites of constriction in chloroplasts of young leaves (Figure

3.10F), indicating that the C-terminal region of ARC6 is not required for mid-plastid

90

ARC6-GFP -;
TP

 
  

ARC6A1Ms-GFP -i

 

Figure 3.10. ARC6AIMS-GFP is partially functional and localizes to the division
site. (A) Schematic of GFP fusions used for analysis. (B) Anti-YFP immunoblot

showing expression of ARC6-GFP (lane 1) and ARC6AIMS-GFP (lane 2) in Arabidopsis
arc6 mutants. Chloroplast phenotypes of arc6 (C); arc6 complemented with ARC6!”-
ARC 6-GFP (D); and arc6 expressing ARC6pr0—ARC6AIWGFP (E) in leaf mesophyll
cells. Localization of ARC6-GFP (arrowheads) in young leaf cells of an arc6 mutant
expressing 35S-ARC 6 A [MS-GFP (F) suggests that only stromal components are required

to localize ARC6 to sites of constriction. ARC6AIMS-GFP encodes amino acids 1-682 of
ARC6 fused to GFP. Transit peptide (TP); transmembrane domain (TM); green
fluorescent protein (GFP); kilodaltons (kD). Scale bars = 10 pm.

91

localization of ARC6 or for constriction of the plastid during division.

ARC6 acts upstream of PD V2 and is required for PD V2 activity.

Because ARC6A1Ms-GFP localizes to constrictions at the plastid division site and

arc6 plants expressing ARC6A1MS-GF P phenocopy the terminal chloroplast number

within mesophyll cells of pdv2 mutants (Miyagishima, Froehlich, et al. 2006), we
hypothesized that ARC6 acts upstream of PDV2. Because the phenotypes of arc6 (1-2
chloroplasts/cell) and pdv2 (4-8 chloroplasts/cell) mutants are easily distinguished by

microscopic observation (compare Figures 3.11C-D), we made reciprocal crosses to test

for epistasis between the arc6-I and pdv2-1 alleles. F1 individuals were allowed to self-

fertilize, and the chloroplast phenotypes in leaves of 105 F2 individuals were examined

(Figure 3.11). We observed all the expected phenotypes in the F2 population, including

individuals with wild-type (WT) phenotypes, intermediate (INT) phenotypes consistent
with PD V2/pdv2 heterozygotes (Miyagishima, Froehlich, et al. 2006), pdv2-like
phenotypes, and arc6-like phenotypes in a ~9:3:4 ratio. Genotype analysis of individuals

with arc6-like phenotypes confirmed the presence of arc6 pdv2 double mutants within

the F2 population (Figure 3.11F). The overrepresentation of arc6 mutant phenotypes in

the F 2 population and the arc6-like phenotype of the arc6 pdv2 double mutant are

consistent with ARC6 acting upstream of PD V2.

92

 

pdv2-l

F2 Phenotype Obs.

WT/IN T: 63
arc6 25
pdv2 17

TOTAL 105

 

PD was pdv2 arc6

arc6-I

Exp.
59
26
20
105

Figure 3.11. ARC6 acts upstream of PDV2. Chloroplast phenotypes of Col-0 (A); Ws-
2 (B); pdv2-I (C); arc6-I (D); PDV2/pdv2 heterozygotes (E); and pdv2 arc6 double
mutants. Phenotypic distribution amongst an F2 population originating from arc6-I x
pdv2-I crosses (G) with observed (Obs.) phenotypes and expected (Exp.) values for
ARC6 acting upstream of PDV2 (x2 E 0.7; d.f. = 2; P 2 0.9). Scale bars = 10 um.

93

Interestingly, overexpression of PD V2 alone leads to an increase in the rate of
chloroplast division (Figure 3.8) (Okazaki, Kabeya, et al. 2009). Because ARC6 acts

upstream of PD V2, we suspected that PDV2 function might be fully dependent upon the

presence of ARC6. To test this, we introduced a T-DNA carrying 35Spm-PD V2 into the

arc6 mutant to ask whether overexpression of PD V2 would abrogate the arc6 phenotype.

Following selection, we assayed T1 individuals for PDV2 protein levels by

immunoblotting and identified multiple lines with elevated PDV2 levels (Figure 3.12,
lower right). In none of these PD V2-overexpressing lines did we observe significant
modification of the arc6 phenotype (Figure 3.12, upper right); mesophyll cells from all
PD V2-overexpressors still contained only 1-2 oversized chloroplasts (Figure 3.12, lower
left). This result indicates that ARC6 is required for PDV2 function in vivo. Moreover,
overexpression of PD V2 in pdvl mutants (J .M. Glynn and K.W. Osteryoung,
unpublished) does not bypass the division defect in the pdvl background, consistent with

non-overlapping fimctions for PDV2 and PDVl.

ARC6 is Required for Equatorial Positioning of PD V2, PD V1, and ARC5.
We showed that ARC6 acts upstream of PDV2 (Figure 3.11 and Figure 3.12) and that
ARC6 is probably responsible for positioning PDV2 at the division site by binding the C-

terminus of PDV2 (Figure 3.7, Figure 3.8, and Figure 3.9). To confirm this prediction,

we introduced a transgene carrying PDVZPm- YFP-PD V2 into Col-0 and into an arc6 T-

DNA insertion mutant (SAIL_693_G04) and examined YFP localization in young leaves

94

 

(it-PDV2

 

Figure 3.12. Overexpression of PDV2 does not abrogate the arc6 phenotype.
Chloroplast phenotypes within leaf mesophyll cells of Ws-2 (A); arc6-l (B); and arc6-I
expressing a 35Spm-PD V2 transgene (C). An anti-PDV2 immunoblot (D) demonstrating
relative PDV2 in the PD V2 overexpressor shown in panel C and the arc6 mutant shown
in panel B. Scale bar = 10 um.

95

of T1 individuals. In the wild type background, we observed equatorial rings of YFP-

PDV2 in 34% of chloroplasts (Figure 3.13A; n = 200 chloroplasts from 20

independent T1 individuals). In contrast, arc6 lines expressing YFP-PDV2 exhibited

diffuse YFP localization around the periphery of the chloroplasts, but no YFP-PDV2

rings were observed (Figure 3.138; n = 200 chloroplasts from 16 T1 individuals); this

result is consistent with the diffuse localization of the dysftmctional YFP-PDV20307D

fusion around plastids (Figure 3.9). In the converse experiment, we examined ARC6-
GFP localization in wild-type and pdv2 backgrounds (Figure 3.14). In both cases, ARC6-
GFP was observed in equatorial rings, indicating that PDV2 does not influence ARC6
localization, even in the enlarged chloroplasts of a pdv2 mutant. We conclude that PDV2
is targeted to the outer envelope in an ARC6-independent manner, but absolutely requires

ARC6 for its localization to the division site.

To extend this analysis and further elucidate the role of ARC6 in mediating
dynamin recruitment through PDVl-PDV2, we examined GFP-ARC5 localization in
wild-type cells and arc6 mutants (Figure 3.15). In young wild-type leaf cells, GFP-
ARCS was frequently observed in a punctate pattern around the division site (Figure
3.15A). In contrast, we observed GFP-ARC5 in patches within the arc6 mutant that do
not appear to be associated with chloroplasts (Figure 3.15B), similar to reported GFP-
ARCS localization in the pdvl pdv2 double mutant; these patches are probably cytosolic
(Miyagishima, F roehlich, et al. 2006). Because GFP-ARC5 still localizes to the division

site in a pdv2 mutant (Miyagishima, Froehlich, et al. 2006), we conclude that ARC6

96

 

Figure 3.13. ARC6 is required for localization of YFP-PDV2 to the division site.
YFP-PD V2 was expressed from its native promoter in Col-0 (A) and arc6 mutant (B)
backgrounds. YFP-PDV2 localizes to central rings within chloroplasts of wild type cells
(arrowheads), but exhibits diffuse localization around the periphery of the plastid in arc6
mutants. A merged image showing signals fiom both chlorophyll autfluorescence (red)
and YFP (green) is shown in each panel. Scale bars = 10 um.

97

(0L0

 

Figure 3.14. PDV2 is not required for ARC6-GFP localization to equatorial rings.
ARC 6-GF P was expressed from its native promoter in Col—0 (A) and pdv2-I (B)
backgrounds. ARC6-GFP localizes to central rings within chloroplasts of wild type cells
and in pdv2 mutants (arrowheads), suggesting that equatorial localization of ARC6 is
independent of PDV2. A merged image showing signals from both chlorophyll
autfluorescence (red) and GFP (green) is shown in each panel. Scale bars = 10 um.

98

( 'uI-II

 

Figure 3.15. ARC6 is required for localization of GFP-ARC5 to the division site.
GFP-ARC5 was expressed from its native promoter in Col-0 (A) and arc6 (B)
backgrounds. GFP-ARC5 localizes to punctate equatorial rings within chloroplasts of
wild type cells (arrowheads), but is only observed in cytosolic patches in arc6 mutants.
A merged image showing signals from both chlorophyll autofluorescence (red) and GF P
(green) is shown in each panel. Scale bars = 10 um.

99

mediates dynamin recruitment and patterning through both PDVl and PDV2, but ARC6

might mediate PDVl function through an indirect mechanism.

Discussion

Chloroplast division involves the consecutive formation and simultaneous
constriction of the mid-plastid F tsZ and ARC5/dynamin rings on the stromal and
cytosolic surfaces of the envelope membranes, respectively. ARC6 in the inner envelope
interacts directly with FtsZ via its stroma-exposed N-terminus and is required for Z-ring
assembly, while PDV2, shown here to be a transmembrane protein of the outer envelope,
mediates recruitment of ARC5 to the chloroplast surface (Miyagishima, Froehlich, et al.
2006). Our results place ARC6 upstream of PDV2 in the chloroplast division process,

identify a physical linkage between the C-terminal IMS domains of ARC6 and PDV2,

and reveal that an important function of the ARC6IMS domain is to direct localization of

PDV2 to the division site. These findings establish a role for a key inner envelope plastid
division protein in organizing components of the outer envelope division machinery.
Because the N-terminus of ARC6 plays a role in organizing FtsZ in the stroma (V itha,
Froehlich, et al. 2003), these findings suggest a model wherein interaction between
ARC6 and PDV2 within the IMS links FtsZ assembly with ARC5 recruitment, thereby

promoting coordinated fission of the two envelope membranes.

Our observation that ARC6 is able to localize to the division site without most of

its IMS region (Figure 3.10) suggests that stromal factors are (sufficient to organize ARC6

100

at the division site within chloroplasts. The FtsZ ring is a strong candidate as an ARC6-
positioning factor. The recently demonstrated ability of recombinant E. coli FtsZ to
assemble into rings inside liposomes suggests that FtsZ proteins, which are highly
conserved across kingdoms, require only a membrane tether and GTP for ring assembly
in vivo (Osawa, Anderson, et al. 2008). We suspect that self-assembly of the Z-ring at
the mid-plastid division site in plants, controlled by MinD, MinE, and ARC3 (Colletti,
Tattersall, et al. 2000, Fujiwara, Nakamura, et al. 2004, Glynn, Miyagishima, et al. 2007,
Maple, Chua, et al. 2002, Maple, Vojta, et al. 2007, Reddy, Dinkins, et al. 2002),
establishes the site for ARC6 localization, presumably via direct interaction with FtsZ2
(Maple, Aldridge, et al. 2005). From our results, we propose that ARC6 transduces
positional information from the Z-ring to the intermembrane space and serves as a

landmark for PDV2 and hence ARC5 recruitment (Figure 3.13 and Figure 3.15).

Consistent with this idea, arc6 plants expressing ARC6A1MS-GFP (Figure 3.10)

phenocopy pdv2 and arc5 mutants with respect to their terminal chloroplast number
(Gao, Kadirjan-Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006), suggesting an
equivalent block in the division process in these three genetically-distinct backgrounds.
Additionally, the loss of PDV2 or ARC5 results in the formation of multiple adjacent Z-
rings on the stromal side of the inner envelope membrane (Miyagishima, Froehlich, et al.
2006), suggesting that information may also travel inward from the outer envelope to the
Z—ring through ARC6. Based on the collective data, we hypothesize that ARC6 has
evolved to organize and coordinate the stromal and cytosolic components of the division

complex and relay information about the status of each across compartment boundaries.

101

Though PDVl and PDV2 have partially redundant functions in ARC5 recruitment
(Miyagishima, Froehlich, et al. 2006), their exact functional relationship in the plastid
division complex is not yet clear. In contrast to ARC6 and PDV2, we did not detect an
interaction between ARC6 and PDVl in yeast two-hybrid assays (Figure 3.5, row 1).
However, preliminary experiments suggest that ARC6 is required for localization of
PDVl (Glynn, Froehlich, et al. 2008) and PDV2 (Figure 3.13), consistent with the lack of
ARC5 recruitment in an arc6 mutant (Figure 3.15). We do not believe that ARC6-
dependent localization of PDVl is mediated directly by PDV2 because GFP-ARC5
localizes properly in pdv2-I (Miyagishima, Froehlich, et al. 2006). Rather, the data
suggest that another factor acts downstream of ARC6 to position PDV] , and hence
ARC5, independently of PDV2 (Glynn, Yang, et al. 2009). In this context, it is
interesting to note that PDVl and PDV2 appear to have distinct localization patterns at
the division site: GFP-PDVl is observed in discrete foci (Miyagishima, Froehlich, et al.
2006, Yang, Glynn, et al. 2008), whereas YFP-PDV2 is observed as a continuous ring
(Figure 3.2). The significance of this difference is unknown, but it could reflect their
interactions with different positioning factors — ARC6 in the case of PDV2 and another

factor in the case of PDV].

The PDV2-interacting IMS region of ARC6 is mostly conserved with the
corresponding region of the ARC6 cyanobacterial ortholog Ftn2, yet PDV2 is not found
in cyanobacteria or green algae. Sequence alignments suggest the presence of land-plant-
specific motifs within the IMS region of ARC6 that may have evolved in parallel with the

emergence of PDVl and PDV2 in land plants (Glynn, Yang, et al. 2009). Ftn2 is

102

localized to the division site in cyanobacteria (Mazouni, Domain, et al. 2004) and
presmnably has a topology similar to that of ARC6 (i.e. its C—terminus protrudes into the
cyanobacterial periplasm). This leads us to ask: did PDV2 replace a periplasmic
component of the cyanobacterial divisome (for example, a component of the
peptidoglycan synthesis machinery or the murein layer itself) or does the ARC6-PDV2
interaction represent a completely new plant-specific function for ARC6? Identification
of the precise boundaries of the PDV2-binding domain of ARC6 and rigorous sequence
comparisons of ARC6 and Ftn2 family members will likely provide further insight into

the evolution and operation of the plastid division machinery in land plants.

103

Materials and Methods

Plant transformation vectors and analysis

ARC6-GFP and ARC6AIMS-GFP were cloned into pCAMBIA-l302

(Hajdukiewicz, Svab, et al. 1994) or a derivative of pCAMBIA-1302 containing the

ARC6 promoter region (Vitha, Froehlich, et al. 2003) using NcoI-BglII sites for

insertion. PD V2pr0- YFP-PD V2 was generated by removing the 35S promoter and

mGFP-His DNA sequences from pCAMBIA-1302 and replacing them with a fragment
carrying the PD V2 promoter region and E YFP (Clontech) fused to the PD V2 coding
sequence. For ARC6-CFP, ARC6 was cloned into a derivative of pCAMBIA-l 302

containing the ARC6 promoter region (V itha, F roehlich, et al. 2003) and ECFP coding

sequence (Clontech). Lines expressing proteins from both PD V2pro- YFP-PD V2 and

ARCopm-ARC6-CF P were generated by crossing independently transformed T1 lines

(with demonstrated transgene expression) and selecting progeny with both YFP and CFP

expression by epifluorescence microscopy. PD V2pro-PD V2 and PD VZPm-PD V20307D

were generated by PCR and cloned into a derivative of pCAMBIA-l 302 that lacks the
35S promoter and mGF P coding sequence. All plant transformations were performed as

previously described (Clough and Bent 1998) using A grobacterium tumefaciens GV3101.

Selection of T1 individuals was performed on MS containing hygromycin (20-25ug/mL).

104

Microscopy and Image Analysis

Light micrographs depicting chloroplast morphology in expanded leaf cells were
taken using DIC Optics on a Leica DMI3000B Inverted Microscope outfitted with a
Leica DFC320 Camera. Samples for chloroplast morphology and quantitation were
prepared and analyzed using established protocols (Pyke and Leech 1991). Fluorescence
micrographs were taken using a Leica DMRA2 using Q-Capture Camera Control
Software (Q-Imaging) and the filter sets indicated (Leica) as previously described (V itha,
Froehlich, et al. 2003). Image analysis and RGB composites were made using ImageJ

v1.37 (NIH) (Bearer 2003).

T wo-Hybrid Analysis
Yeast strain AH109 (Clontech) was cultured and transformed as recommended by

the manufacturer using standard Synthetic Dropout (SD) Media (Clontech) as indicated.

ARC6 A A 637-801 was cloned into pGADT7 using Ndel-Xmal and PDV2 A A233-307 was

cloned into pGBKT7 using Ndel-Xmal sites. PDV2 A A 233-307 ((33079) was generated by

PCR-based mutagenesis and cloned into pGBKT7 using Ndel-Xmal sites.

Pulldown Assays
Recombinant His-ARC6 (encoding ARC6 A A 637-801 with an N-terminal 8X His-

tag) was generated in pHIS8-3 (Salk Institute, La Jolla, CA) and expressed in BL21

(DE3) Codon Plus Cells (Stratagene). 2mM IPTG was applied to the cells at OD600 ~

0.8 and incubated for 4h to generate the His-ARC6IMS fusion protein. Total protein was

105

extracted from the induced cells by sonication in Buffer A (1X TBS, 40mM imidazole,
and 1 Roche EDTA-free protease inhibitor cocktail tablet #11836170001 per 100 mL
Buffer A) and treatment of the sonicated material with 0.5% Triton X-100 for 30 minutes
at room temperature. Following centrifugation at 18000 x g, the supernatant was
collected and analyzed by Bradford assay. 750 ug of total protein was applied to 50 uL
of Ni-Sepharose beads equilibrated in Buffer A (N i-Sepharose 6 Fast Flow, GE

Healthcare, Uppsala, Sweden). Beads were washed 4 times in Buffer A to enrich for His-

ARC61MS. Negative controls (beads only) were simply washed in Buffer A following
equilibration. Production of crude GST-PDV2 (encoding PDV2 A A 233-307 with an N-

terrninal GST tag) and GST-G307D (encoding GST-PDV2 A A 233-307(G307D)) was

performed using the methods described above, except pGEX-4T-2 (GE Healthcare) was
used as the parent vector for cloning and expression of recombinant PDV2 proteins. The
cell extracts containing the GST fusions were prepared by sonication and treatment with
Triton X-100 and then centrifuged to yield a crudely purified sample. Total protein was

measured by Bradford assay and 750 ug of total protein was applied to the naked Ni-

Sepharose beads or Ni-Sepharose beads coated with His-ARC61MS. Following a 2.5 h

incubation on a rocking platform, the samples were washed several times with Buffer A
and then eluted into 250 pL Buffer B (Buffer A containing 1M imidazole).
Approximately 240 uL of eluate was recovered following a brief spin and total eluted
material was precipitated in acetone and reconstituted in 6X sample buffer prior to SDS-
PAGE. Approximately 12.5% of the eluted material was applied in each lane for

separation on 15% SDS—Polyacrylarnide gels. Separated proteins were transferred to

106

nitrocellulose (GE Water and Process Technologies) and blotted with anti-His (Pierce,
#15165) or anti-GST (Sigma, #G-l417) antibodies as recommended by the manufacturer
and detected by standard chemilluminescence methods using the manufacturer’s HRP-

based detection protocol (Pierce).

Immunoblotting of Plant Material

Proteins from whole-cell extracts (equivalent to ~2-3 mg liquid-nitrogen ground
tissue) were prepared as previously described (Wiegel and Glazebrook 2002), separated
by SDS-PAGE, and transferred to nitrocellulose. Anti-GFP Immunoblots were
performed using Clontech JL-8 anti-GFP monoclonal antibody at 1:1000 in 5% nonfat
dry milk in TBS-T, pH 7.4 (Miyagishima, Froehlich, et al. 2006). Anti-PDV2

Immunoblots were performed using 1 mg/mL Protein-A purified IgG’s from New
Zealand White rabbit sera (CRP, Inc.) directed against PDV2AA1_212 at 1:15000 in 5%
nonfat dry milk in TBS-T, pH7.4. Recombinant PDV2 (rPDV2) used to generate the
immunogen used for anti-PDV2 antibody production was generated by cloning the
coding sequence for PDV2 A A 1-212 into pHIS8-3 (Salk Institute, La Jolla, CA),
expressing in BL21 (DE3) Codon Plus Cells (Stratagene), and purifying the resulting

fusion protein from cell extracts using Ni Sepharose 6 Fast Flow (GE Healthcare,

Uppsala, Sweden) according to the manufacturer’s recommendations.

Phylogenetic Analysis
ClustalW2 was used to generate a multiple sequence alignment using PDVl and

PDV2 sequences from the organisms shown. Scoring was based on an identity (ID)

107

matrix with default settings: Protein Gap Open Penalty = 10.0; Protein Gap Extension
Penalty = 0.2; Protein ENDGAP = -1; Protein GAPDIST = 4. Phylogeny of the PDV
proteins (Glynn, Froehlich, et al. 2008) was calculated using MEGA 4 (Tamura, Dudley,

et al. 2007).

Chloroplast Import, Fractionation, and Protease Protection Assays

The cDNA for PDV2 was subcloned into pDEST14 (Invitrogen) according to the
manufacturer’s protocol prior to in vitro transcription translation. In vitro-produced
protein was generated using a Promega TnT® Coupled Reticulocyte Lysate System
(Promega, Madison, WI) according to the manufacturer’s protocol using 3H—labeled
leucine (Perkin-Elmer) as a marker for PDV2 and 35S-labeled methionine (Perkin-Elmer)
to track ARC6 protein. Pea chloroplasts were isolated, and import assays and
fractionation were performed according to previously established protocols (Jackson,

Froehlich, et al. 1998, Tranel, Froehlich, et al. 1995, Vitha, Froehlich, et al. 2003).

Accession Number

Genbank accession numbers used for multiple sequence alignments and
phylogenetic analysis are as follows: Arabidopsis thaliana PDVl (AAM64850),
Gossypium hirsutum PDVl (DW511385 and DW512027), Helianthus tuberosus PDVl
(EL464676), Oryza sativa PDVl (NP_001042451), Populus trichocarpa PDVl
(ABK94742), Vitis vinifera PDVl (CAO69353), Zea mays PDVl (EE161990 and
DR959634), Arabidopsis thaliana PDV2 (NP_028242), Asparagus oflicinalis PDV2

(CV287540), Aquilegia species PDV2 (DT728284 and DT749563), Lactuca sativa PDV2

108

(DY983 634), Oryza sativa PDV2 (EA202618), Populus trichocarpa PDV2 (a
GENSCAN based prediction (http://genes.mit.edu/GENSCAN.html) of Populus
trichocarpa Scaffold LG__IX 3439000-3442500), and Triticum aestivum PDV2

(CK209373 and BJ311269).

Acknowledgements

I thank John Froehlich and Shin-ya Miyagishima for their advice dining the early
phases of this project. I thank John Froehlich for persevering through the difficulties we
experienced in translating PDV2 in vitro and for his outstanding work that helped us
confirm PDV2 topology in isolated chloroplasts. I also thank Ronit Knopf and Zach
Adam of The Hebrew University in Israel for sharing data with us on native PDV2

topology in Arabidopsis prior to publication.

Most of this chapter and the figures therein are reproduced from (Glynn,
Froehlich, et al. 2008) (http://www.plantcell.orgD and is Copyright American Society of
Plant Biologists. ASPB grants to authors whose work has been published in The Plant
Cell the royalty-free right to reuse images, portions of an article, or full articles in any

book, book chapter, or journal article of which the author is the author or editor.

109

Chapter 4

Characterization of the PDV2 Binding Domain of ARC6.

110

Abstract

During plastid division, interaction between ARC6 and PDV2 is required to
position PDV2 within the outer envelope membrane. PDV2 helps recruit the dynamin-
like protein, ARC5, to the division site. Previously, we showed that a carboxy-terminal
glycine of PDV2 resides within a conserved motif and is required for ARC6-PDV2
interaction and PDV2 function in vivo. However, the specific domains or motifs within

ARC6 that contribute to PDV2 binding were not determined. Here, we isolate the PDV2-

binding domain of ARC6 (ARC6pBD) and perform comparative sequence and structural

analysis of both PDV2IMS and ARC6pBD. Multiple sequence alignments and structural

models highlight a plant-specific serine residue within ARC6pBD whose modification

might be important for controlling operation of the chloroplast divisome in vivo. Further,

we show that single amino acid substitution at this conserved serine within ARC6 affects

its ability to bind PDVZIMS, leads to arc5-like chloroplast morphology, and upregulates

Z-ring assembly within the stroma. Our results hint at a post-translational modification
of ARC6 that fine-tunes coordination of the stroma] and cytosolic plastid division

machineries in land plants.

111

Introduction

The posttranslational modification of proteins can have profound consequences
on their function (Huber and Hardin 2004). Several types of modification can occur in
plants, including ubiquitylation (Dreher and Callis 2007), acetylation (He and Amasino
2005), methylation (Fischer, Hofrnann, et al. 2006), lipoylation(Ewa1d, Kolukisaoglu, et
al. 2007), and phosphorylation (Christie 2007). Amongst these types of modification, the
phosphorylation and dephosphorylation of serine residues is a common theme amongst
proteins involved in signal-transduction (Braun and Walker 1996, Michniewicz, Zago, et

al. 2007, Sokolovski, Hills, et al. 2005).

Chloroplast division in plants and algae is mediated by two ring-like structures,
both of which are composed of polymer-forming GTPases: (1) the FtsZ ring (Z-ring) on
the stromal face of the inner envelope (McAndrew, Froehlich, et al. 2001, Vitha,
McAndrew, et al. 2001) and (2) the ARC5 ring on the cytosolic face of the outer
envelope membrane (Gao, Kadirjan—Kalbach, et al. 2003). The Z-ring is made up of two
non-redundant FtsZ proteins, Ftle and FtsZ2 (McAndrew, F roehlich, et al. 2001,
Schmitz, Glynn, et al. 2009, Vitha, McAndrew, et al. 2001). ARC5 is a dynamin-like
protein involved in the final squeeze that separates the two daughter organelles during
plastid division (Gao, Kadirjan-Kalbach, et al. 2003, Miyagishima, Nishida, et al. 2003a,
Yoshida, Kuroiwa, et al. 2006). In land plants, ARC5 is recruited to the surface of the

plastid by the outer envelope proteins PDVl and PDV2. While either PDV] or PDV2

112

can recruit ARC5 to the surface of the plastid, both PDVl and PDV2 are required for full

ARC5 contractile activity in vivo (Miyagishima, Froehlich, et al. 2006).

ARC6 is a bitopic chloroplast inner membrane protein (V itha, Froehlich, et al.
2003) that coordinates operation of the internal and external plastid division complexes
(Glynn, Froehlich, et al. 2008). In the stroma, the N~terminus of ARC6 facilitates FtsZ
ring (Z-ring) assembly (Vitha, Froehlich, et al. 2003); the Z-ring is a critical component
of the division apparatus, acting as a scaffold for other division proteins and providing
nominal contractile force during division (Erickson 2009, Osawa, Anderson, et al. 2008).
The intermembrane space (IMS)-localized portion of ARC6 binds PDV2 directly and
positions PDV2 at the division site; PDV2 is required for full dynamin (ARC5) activity

during the final steps of chloroplast division (Glynn, Froehlich, et al. 2008, Miyagishima,

Froehlich, et al. 2006).

Here, we characterize the PDV2 binding domain of ARC6 (ARC6pBD) and

identify a predicted serine phosphoacceptor within this domain that may modulate ARC6

activity in response to serine modification. We show that ARC6pBD occupies amino

acids 636-759 and that serine 744 might play a critical role in both mediating ARC6-

PDV2 interaction within the IMS and in regulating Z-ring assembly within the stroma.

ARC6S744 and the surrounding motif is found only in ARC6 orthologs from

embryophytes, the lineage in which PDV2 first emerged (Glynn, Froehlich, et al. 2008,

Miyagishima, Froehlich, et al. 2006). Phosphomirnetic mutation of serine 744 (S744E)

113

weakens ARC6-PDV2 interaction, leads to arc5-like plastid division defects, and
perturbs stromal Z-ring structure in vivo — but strangely, this mutation does not affect

localization of PDVl, PDV2, or ARC5 at the division site. Our results suggest that

persistent phosphorylation of ARC6S744 slows recruitment of PDV2, but leads to

profound upregulation of Z-ring assembly within the stroma. We propose that

phosphorylation of ARC6S744 is a regulatory feature that retards the recruitment of

PDV2 until the Z-ring is fully established and ready for envelope constriction to occur.

Results

The PD VZ-Binding Domain of ARC 6 (ARC 61031)) Occupies Amino Acids 636- 759.

We previously showed that the C-terrninal IMS-localized regions of ARC6 and
PDV2 interact; this interaction serves to position PDV2 within the outer envelope
membrane during division (Glynn, F roehlich, et al. 2008). While we showed that this
interaction required the conserved terminal glycine residue of PDV2, the discrete domain
within ARC6 that binds PDV2 was not defined. To gain further understanding of ARC6

function and its constitutive domains, we generated a series of two-hybrid deletion

constructs for ARC61MS (Figure 4.1). Following yeast transformation and selection on
appropriate medium, we assayed for growth on dropout medium containing or lacking
histidine; growth on medium lacking histidine is consistent with the presence of a

protein-protein interaction between bait and prey leading to activation of the HIS3

reporter gene. Based on the growth ratios, we observed that full interaction

114

INTERACTION PAIR

ARC6 AA636-801 VS- PDV2 AA233-307

S—D/-ULT S___D/-ULTH

   
   

 

._.

ARC6 AA636-759 VS- PDV2 AA233-307

N

ARC6 AA636-728 VS- PDV2 AA233-307

U.)

ARC6 AA648-801 VS- PDV2 AA233-307

 
 

A

ARC6 AA661-801 VS- PDV2 AA233-307

\1

kn

ARC6 AA636-801 VS. EMPTY

 

ON

ARC6 AA636-759 vs. EMPTY

ARC6 AA636-728 VS. EMPTY

0°

ARC6 AA648-801 vs. EMPTY

\O

ARC6 AA66l-801 VS. EMPTY

 
  
 

O

EMPTY vs. PDV2 AA233-307

._r
._n

EMPTY vs. EMPTY

 

l2

1 1/10 1/100 1 1/10 1/100

Figure 4.1. The PDV2 binding domain of ARC6 (ARC6pBD) occupies AA 636-759.
Yeast two-hybrid HIS reporter assays showing interaction strength between the IMS
region of PDV2 (AA 233-307) and fragments of the IMS region of ARC6. Left hand
column shows growth on synthetic dropout (SD) medium supplemented with histidine
and right hand column shows growth on medium lacking histidine. For a schematic of
the ARC6 protein, refer to Figure 1.4. Dilutions from a starting culture of OD600 = 1.0
are indicated at the bottom of panel. EMPTY = Empty vector control. AA = amino
acids.

115

between ARC6 and PDV2 only requires ARC6 A A636-759 (Figure 4.1). We call this

region of ARC6 the PDV2-binding _d_omain (ARC6pBD). The reporter activation of

ARC6pBD is comparable to that observed for the full length C-terminus of ARC6,

ARC6 A A636-801 (Figure 4.1). PDV2 does not interact with ARC6AA661-301 nor

ARC6AA536-723, but weakly interacts with ARC6 A A648-801 (Figure 4.1), suggesting that

the residues adjacent to the transmembrane domain of ARC6 are involved, but not critical

for ARC6-PDV2 interaction.

Structural analysis of ARC6 1931) and PD VZIMS-

To gain additional insights into the interaction between ARC6-PDV2 and guide

site-directed mutagenesis, we attempted to generate a model of ARC6pBD. ARC6pBD

was best-matched to the known structure of the Tim44 C-terminal domain (2CW9) using

the meta-server at http://bioinfo.pl (Ginalski, Elofsson, et al. 2003). However, attempts

to model ARC6pDB onto the structure of 2CW9 were unsuccessful, yielding several gaps

between residues in the main chain of ARC6pBD (data not shown), suggesting that this is

probably not a valid structural comparison. From this, we conclude that ARC6pBD

probably has a structure unique from any of the structures currently present within the

Protein Data Bank (http://wwwpdbcrgD.

116

We also attempted to generate a homology model for PDVZIMS, hoping that it

might reveal insights into both PDV2 and ARC6 function. Using the meta-server at
http://bioinfo.pl/, we determined that the best structural match for PDVZIMS is 2J6Z
(1 W53), the N-terminal domain of a phosphoserine phosphatase that is involved in stress

response in Bacillus subtilis and other gram-positive bacteria (Delumeau, Dutta, et al.

2004, Hardwick, Pane-Farre, et al. 2007). We made a pairwise comparison between

RSbUNT and PDVZIMS (Figure 4.2) and used this alignment to generate a homology

model of PDV21MS, using identical residues as anchor points for the PDVZIMS model.

The PDVZIMS sequence was highly similar to RSbUNT and was well-fit to 2J6Z

(Figure 4.3, inset), suggesting that these two proteins might have similar functions. The

N-terrninus of Rst is required for binding the Rst stress response serine kinase (Dutta

and Lewis 2003, Kang, Vijay, et al. 1998, Yang, Kang, et al. 1996), but RSbUNT does

not appear to have kinase or phosphatase activity itself— this activity is thought to be
exclusive to the C-terminal domain of Rst (Delumeau, Dutta, et al. 2004, Kang,_Vijay,
et al. 1998). In this case, the C-terminus of ARC6 might be analogous to Rst—
perhaps even possessing a kinase or phosphatase activity itself. From this rudimentary
structural analysis, we hypothesize that PDV2 may have a structure akin to a known
protein phosphatase or may influence the activity of its binding partner through a simple

protein-protein interaction, analogous to the Rst-Rst signal transduction cascade in

117

 

RstNT I HQL S YI . ETS QAQ
PDV2IMS S GPE R ‘ LF PHR
RstNT KFS TIE HRK
PDV2IMS ATR RTP Euvl EDG
RstNT YPSLP D DFLI MI
PDV2IMS VKER I VAK T

 

 

 

Figure 4.2. Pairwise alignment between the N-terminal domain of Rst and the
IMS region of PDV2. The N-terminal domain of Rst used for crystal structure of
2J6Z (RstNT, amino acids 7-81) and IMS region of PDV2 (PDVZIMS, amino acids
233-307) are aligned using a BLOSUM62 scoring matrix. Boxed positions indicate
sequence similarity and shaded positions highlight sequence identity.

118

 

Figure 4.3. Structural model of PDVZIMS. The IMS region of PDV2 (amino acids

233-307, main panel) was modeled using the known structure of RSbUNT (2J6Z/1 W53,
inset panel). Shown above are cartoon representations ofthe two structures. The side
chains of conserved residues within the C-terminal extension of PDV2 proteins (see

Figure 3.6) are highlighted in the PDVZIMS structural model.

119

some gram positive bacteria (Delumeau, Dutta, et al. 2004, Kang, Vijay, et al. 1998,
Yang, Kang, et al. 1996). This IMS-localized interaction between ARC6 and PDV2
might modulate posttranslational modification of ARC6, thereby altering the activity of

ARC6 within the IMS or chloroplast stroma.

A Conserved Serine within ARC 6 pm) Influences ARC6-PD V2 Interaction.

Because of the possible connection to protein kinase/phosphatase activity, we

closely examined alignments between tracheophyte ARC6pBD and the corresponding
region of cyanobacterial F tn2 proteins for embryophyte-conserved residues that might act
as phosphoacceptors (Figure 4.4), as PDV proteins are found only in land plants. We
identified a plant-specific motif (Figure 4.4, red bar) within the ARC6pBD that was
conserved amongst higher plants, but missing from or poorly aligned with algal ARC6
and cyanobacterial F tn2 sequences; all ARC6 orthologs fi'om higher plants possessed a
serine near the middle of this motif (Figure 4.4, arrowhead indicates S744 in AtARC6);

serine residues are common sites of posttranslational modification in eukaryotes.

Because algae and cyanobacteria do not encode orthologs of PDV2, this embryophyte-

specific serine residue (ARC63744), or other residues within the surrounding motif, might

be targets of ARC6 posttranslational modification that have coevolved with emergence of

PDV2 in land plants.

120

 

*
. SVTVSADGT

 

AtARC6 L . S
OsARC6 . SITISLDGR T. A
PtARC6 . SVTVSVDGLS V . S
AlARC6 . SVTVSADGT L . S
BnARC6 . SVTVSADGT L . S
RsARC6 . SVTVSADGT L . S
GrARC6 . SVTLSLDGQ V . S
lnARC6 . SVTVSVDGR I . S
CiARC6 . SVTISLDGR V . S
CcARC6 . SVTVSVDGQ I . L
CsARC6 . SVTLSQEGR W . S
OlARC6 .SVQVIG.TE F. V
CrARC6 LKRVTHKGA F S
GthnZ . QVADRR . PD A . V
NthnZ . VEKIGLFAD A . V
SeFtn2 .VQLSDG .DQ V. V
CsFth . FKPDSNNPN V. V
NsFth .FDQK. . .SD V. V
LsFth .VQIDSSNPN F . T

 

 

 

 

 

Figure 4.4. Identification of a plant-specific motif within the C-termini of

ARC6/F m2 family members. A segment of a CLUSTALW alignment comparing the
C-termini of ARC6 and Ftn2 family members is shown above and corresponds to
Arabidopsis thaliana ARC6 amino acids 729-759. The red bar above the sequences
indicates an amino acid motif unique to higher plants that may be involved in ARC6-
PDV2 interaction, based on two hybrid results (Figure 4.1). The black arrowhead points
to a serine (S744) predicted to be a phosphoacceptor (see text). Arabidopsis thaliana
ARC6 (AtARC6); Oryza sativa cv. japonica (OsARC6); Populus trichocarpa ARC6
(PtARC6); Arabidopsis lyrata (AlARC6); Brassica napus ARC6 (BnARC6); Raphanus
sativus (RsARC6); Gossypium raimondii (GrARC6); Ipomoea nil (InARC6); Cichorium
intybus (CiARC6); Coffea canephora (CcARC6); Citrus sinensis (CsARC6);
Ostreococcus lucimarinus CCE9901 (OlARC6); Chlamydomonas reinhardtii (CrARC6);
Gloeobacter violaceus PCC7421 (Gthn2); Nostoc punctiforme PCC 73102 (NpF m2);
Synechococcus elongatus PCC 6301 (SeFtn2); Cyanothece sp. CCYOl 10 (CsFtn2);
Nostoc sp. PCC 7120 (NsFtn2); and Lyngbya sp. PCC 8106 (LsFtn2). Label colors in the
lefthand column indicate groups of organisms: red = land plants, green = green algae, and
blue = cyanobacteria. Boxed regions indicate sequence similarity in >70% of the
lineages shown. Shaded regions indicate conserved residues.

121

To determine if ARC6S744 was a potential phosphoacceptor, we queried two
phosphorylation-prediction programs, PhosPhAt (Heazlewood, Durek, et al. 2008) and
NetPhos (Blom, Garnmeltoft, et al. 1999) to see if ARC6S744 might be a target for

phosphorylation. Both of these programs gave significant scores to ARC6S744

(PhosPhAt, 0.496; NetPhos, 0.932), suggesting that ARC6S744 may be phosphorylated in

vivo. Notably, both S740 and T742 also yielded significant scores using one, but not

both, of these prediction algorithms — suggesting that other residues within the plant-

specific motif of ARC6pBD (Figure 4.4) might also be targets for posttranslational

modification.

To determine if phosphorylation of ARC6S744 might affect binding of PDV2, we
generated site-directed mutants of ARC61MS, targeting the serine at position 744 in two-
hybrid vectors. We named these mutant proteins ARC61MS ($744 A) and ARC61MS

(S744E), which mimic the unphosphorylatable (S744A) and constituitively

phosphorylated (S744E) forms of the protein, respectively. Our results showed binding

strengths consistent with PDV2 affinities as ARC6IMS > ARC63744 A > ARC6S744E

(Figure 4.5A), indicating that PDV2 has low affinity for the phosphomimetic form of

ARC6, ARC6S744E. We validated this result using pulldown assays with recombinant

Hiss-ARC61MS and GST-PDV2IMS proteins, which confirmed that ARC61MS (874415)

122

 

A

INTERACTION PAIR
ARC61MS VS. PDVZIMS

SD/-ULT SD/-ULTH

   
 

 
 

._‘

EMPTY vs. PovleS

N

ARC6IMS vs. EMPTY

DJ

ARC61MS (S744A) VS. PDVZIMS

4:-

ARC6IMS (S744A) VS. EMPTY

 
 
 

ARC6IMs (S744E) VS- PDVZIMs

O\

ARC61MS (S744E) VS. EMPTY

\l

EMPTY vs. EMPTY

(I!

00

1 1/10 1/100 1 1/10 1/100

1 2 3 4

l... .- - Ia-HIS

I" ‘. .: lot-GST

 

 

 

 

 

 

Figure 4.5. A phosphomimetic mutation in the IMS region of ARC6 (ARC6S744E)
decreases its affinity for PDV2. (A) Site-directed mutations in ARC61MS were tested
for their ability to interact with PDV21MS by yeast two-hybrid. Interaction strengths are
as follows: ARC6-PDV2 (0.64); ARC61MS (S744A)-PDV2 (0.54); ARC61MS (57445)-
PDV2 (0.14); all negative controls (5 0.01). Relative interaction strength was determined

by ratio of growth on —HIS/+HIS synthetic dropout (SD) media. (B) Confirmation of two
hybrid results by pulldown assay using immobilized ARC6 proteins treated with equal

amounts of GST-PDVZIMS. Immunoblots of eluates from pulldowns are shown: Hiss-
ARC61MS (lane 1), H153-ARC6IMS (S744A)a H188-ARC61MS (37445), or uncoated beads
treated with GST-PDV21MS (lane 4).

123

has significantly lower affinity for PDVZIMS than the wild type (ARC6IMS) protein
(Figure 4.58). ARC6IMS (S744A) had slightly decreased affinity for PDVZIMS in two-

hybrid assays and slightly increased affinity for PDVZIMS in pulldown assays, suggesting

this mutation may not perturb ARC6 function.

A Phosphomimetic Mutation in ARC 61231) Causes Chloroplast Division Defects in vivo.

Because of the profound effect of our phoshphomimetic mutation (ARC6S744E)

on ARC6-PDV2 interaction, we wanted to know if this mutation was detrimental to the

division process in vivo. Therefore, we generated a series of transgenes encoding wild
type (ARC6), unphosphorylatable (ARC6S744A), and phosphomimetic (ARC6S744E)

versions of ARC6; all of these were driven by the native ARC6 promoter sequence. We

transformed wild type (Col-0) and arc6 (SAIL_693_G04) plants with each of these

transgenes using Agrobacterium (Clough and Bent 1998) and selected T1 individuals

using hygromycin. We examined chloroplast morphology in independent T1 lines that

were expressing the transgene within the range that should complement the arc6 mutant,

based on the expression of the control (wild type) transgene (Figure 4.6). Both the wild

type (ARC6) protein and the unphosphorylatable (ARC6S744A) proteins were able to

complement the arc6 mutant when expressed at similar levels. However, ARC6S744E

was unable to fully complement arc6 mutants (Figure 4.6) and imposed a dominant-

negative effect upon Col-0 transformants (data not shown), suggesting that this transgene

124

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A l 2 3 4 5 6 7 8 9 10
“7“»:- * tag—..-s. i
97kD>“? [i a-ARCO
117kD> =nlah'xv.:.s‘...'.'..$‘ ninth,»

B 97kD>Ei _ ' j HI Ponceau

70
60 l
=

Q

O

In

0 -

Q-

m

H

m

N

—

Q:

G

i-

O

_

i:

U
Col-0 arc6 arc6 + arc6 + arc6 +

 

ARC6 ARC6S744A ARC65744E

 

 

Figure 4.6. ARC6S744E is dysfunctional. (A) Sample of immunoblot data used to
quantitate ARC6 protein levels in transformed arc6 plants. Marker (lane 1); arc6 (lane
2); Col-0 (lane 3); four independent arc6 T1 transformants complemented by ARC6
transgene (lanes 4-7); 3 independent arc6 T1 ARC6S744E transformants (lanes 8-10).
Protein levels in ARC6S744 A lines were quantitated on a separate immunoblot (not
shown). (B) Analysis of chloroplast number in expanded leaves of Col-0, arc6, and
transgenic lines from 10 cells with average areas of 3109 d: 443 umz. Lines used for
quantitative analysis in panel B are marked with asterisks (*) in panel A. Error bars =
standard deviation.

125

was not functional in vivo, consistent with its low affinity for PDV2 in pulldowns and

two-hybrid assays (Figure 4.5). Interestingly, many of the lines expressing ARC 657441.;

contained elongated and partially-constricted chloroplasts, indicative of a defect in PDV

or ARC5 protein function (Figure 4.7E, arrowheads) (Gao, Kadirjan-Kalbach, et al. 2003,

Miyagishima, Froehlich, et al. 2006). Taken together, we conclude that ARC6S744 is

probably important for the ARC6-PDV2 interaction in vivo, but cannot yet conclude that

this residue is actually subject to post-translational modification in vivo.

PD V2, PD VI, ARC5 and F tsZ localize to the division site in ARC 657441.; mutants.

To determine the effect of our phosphomimetic mutation upon the characteristics
of other division components, we examined our transgenic lines and controls for

localization of PDV2, PDVl , ARC5, and F tsZ proteins.

Surprsingly, despite their lower affinity for PDV2 and perturbation in chloroplast

morphology, ARC6S744E mutants were still able to localize YFP-PDV2 to rings at sites

of chloroplast constriction (Figure 4.8C). This conflicts with the reduced affinity of

ARC6S744E for PDVZIMS in binding assays (Figure 4.5), but might reflect the ability of

ARC6S744E to bind PDV2, albeit with less stringency than the wild type protein (Figure

4.5).

126

 

   

...-

 

 

 

Figure 4.7. Phenotypes observed in lines expressing ARC6S744A and ARC6S744E.
Mesophyll cells from expanding leaves are shown for Col-0 (A); arc6 (B); arc6

expressing transgenic ARC6 (C); arc6 expressing transgenic ARC 65744 A (D); and arc6

expressing transgenic ARC6S744E (E). Arrowheads in (E) indicate elongated and
partially constricted chloroplasts. Scale bars = 10 pm.

127

 

PDV2

ARC5

PDVl

FmZZ

 

 

 

 

Figure 4.8. Localization of PDV2, ARC5, PDV], and FtsZ in ARC6S744E-expressing
transgenics. All fluorescently-labeled proteins were expressed using their respective
native promoters. Merged images showing chlorophyll autofluoresence (red) and the
indicated fluorescent marker (green) are shown. YFP-PDV2 localization in (fol-l) (A);

arc6 (B): and ARC657445 (C). GFP-ARC5 in Col—0 (D); arc6 (E); and ARC6S744E (F).
GFP-PDVl in Col—0 (G); arc6 (H); and ARC7657445(1). Immunolabeling of FtsZ2-l in
Col-U (J); arc6 (K); and .4RC6S7445 (L). Scale bars = 5 pm.

128

To determine if the division defect observed in ARC6S744E transgenics was a
result of an indirect effect upon ARC5 or PDVl localization, we examined GFP-ARC5
and GFP-PDVl localization in lines ARC6S744E transgenics. Both GFP—ARC5 (Figure
4.8F) and GFP-PDV] (Figure 4.81) were localized to the division site, though GFP-PDVl

was not observed in a ring in lines expressing ARC6S744E as it is in wild-type plants

(Miyagishima, Froehlich, et al. 2006).

Because Z-ring morphology is affected in the dumbbell-shaped plastids observed

in pdvl , pdv2, and arc5 mutants, we examined Z-ring morphology in ARC6S744E mutants

using an FtsZ2-specific antibody (Figure 4.8L). In mutants with defects in dynamin
activity, we typically observe multiple Z-rings at the division site, presumably as a
feedback response to the defect in dynamin activity at the outer envelope (Gao, Kadirjan-
Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006, Yang, Glynn, et al. 2008).

Consistent with a defect in dynamin activity (Gao, Kadirjan-Kalbach, et al. 2003,

Miyagishima, Froehlich, et al. 2006), ARC6S744E mutants have multiple Z-rings at the

division site (Figure 4.8L). Taken together, we conclude that ARC6S744 is important for

ARC6-PDV2 interaction, indirectly affecting dynamin (ARC5) pinchase activity at the

outer envelope through PDV2, as ARC6S744E mutants have defects in chloroplast and

FtsZ filament morphology similar to pdvl , pdv2, and arc5 mutants (Gao, Kadirjan-

Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006).

129

Discussion

Here, we identify and characterize the PDV2-binding domain of ARC6

(ARC6pBD). ARC6pBD occupies only a portion (AA 636-759) of the IMS-localized C-
terminal region of ARC6. While our structural analysis of ARC6pBD was not insightful,

the IMS-localized C-terminus of PDV2 (PDVZIMS) modeled onto the known structure of

Rst (Delumeau, Dutta, et al. 2004, Hardwick, Pane-Farre, et al. 2007), a bacterial

phosphoserine phosphatase (Delumeau, Dutta, et al. 2004, Hardwick, Pane-Farre, et al.

2007). Because of the structural similarity of PDV2IMS to Rst and the prospect of

posttranslational modification of ARC6, we examined alignments (Figure 4.4) within

ARC6pBD for plant-specific motifs containing predicted phosphoacceptor sites. This
analysis revealed that serine 744 (ARC6S744) as a high-probability phosphoacceptor

within a plant-specific motif of ARC6pBD, based on NetPhos and PhosPhAt prediction

programs. Our site-directed mutagenesis of this residue revealed that a phosphomimetic

mutation (ARC6S7445) causes reduced affinity for PDV21MS; this same mutation leads to

chloroplast division defects (Figures 4.6-4.8) similar to those seen in mutants with defects

in dynamin (ARC5) activity (Gao, Kadirjan-Kalbach, et al. 2003, Miyagishima,

Froehlich, et al. 2006). However, despite the reduced affinity for ARC6S744E for PDV2,

arc6 mutants expressing an ARC6S744E transgene still recruit YFP-PDV2 and other

division factors to the division site (Figure 4.8).

130

Our identification of the rough boundaries of the PDV2-binding domain is a first
step in understanding how ARC6 and PDV2 interact with each other and how this

interaction might be regulated. Further work is required to identify the precise

boundaries of ARC6pBD, which might allow ftu'ther refinement of the current structure-

function model for this domain. While structural similarity to Rst (2J6Z/1W53) is

apparent, it is not yet clear if PDVZIMS possesses phosphatase or kinase activity and if

ARC6pBD is truly a substrate for any posttranslational modification activity in vivo.

While ARC6S744 might act as a phosphoacceptor in vivo, PDV2 itself probably

does not modify this residue and likely uses ARC6pBD as a binding site, as the structural
similarity to the Rst phosphatase is limited to the Rst-binding domain of Rst.
However, modification of the plant-specific motif within ARC6pBD clearly reduces

affinity for PDV2, suggesting that a posttranslational modification at serine 744 might be
a mechanism by which PDV2 recruitment to the division site is modulated.

Alternatively, serine 744 of ARC6 might be required for structural intergrity of

ARC6pBD and our mutant proteins are simply misfolded, leading to reduced affinity for

PDV2. Either of these scenarios might explain the prevalence of dumbbell-shaped

chloroplasts in lines expressing ARC6S744E (Figures 4.7-4.8). Curiously, we

occasionally observed chloroplasts within petiole cells that possessed several

constrictions along their length in ARC6S7445-expressing lines (Figure 4.9), a phenotype

131

A R C 657445

l’etiole

 

Figure 4.9. An unusual phenotype observed in ARC6S744E petioles phenocopies
plastidic Min system defects. Multiple constrictions were observed along the length of
elongated chloroplasts within large petiole cells of ARC6S744E transgenics. This
phenotype has not been observed in arc5 mutants and may reflect ARC6-mediated
downregulation of AtMinD (Colletti, Tattersall, et al. 2000, Fujiwara, Nakamura, et al.
2004) and/or upregulation of AtMinE activity within the chloroplast (Fujiwara,
Hashimoto, et al. 2008, Maple, Chua, et al. 2002). Scale bar = 10 pm.

132

that resembles both atminD/arcI 1 loss-of-function mutants (Colletti, Tattersall, et al.
2000, Fujiwara, Nakamura, et al. 2004) and AtMinE overexpressors (Fujiwara,
Hashimoto, et al. 2008, Maple, Chua, et al. 2002). The significance of this finding is not
yet clear, nor is it understood why these phenotypes appear prominently in petiole cells,
but the connection between ARC6 activity and the operation of the plastidic Min system

is intriguing.

The high proportion of elongated dumbbell-shaped plastids in the ARC6S744E

transgenic lines caused us to look the localization of ARC5, PDVl, PDV2, and F tsZ.
Surprisingly, the localization of each of these proteins to the mid-plastid was mostly

unperturbed. However, PDV] was present exclusively as single foci associated with the

chloroplast (Figure 4.81) in ARC6S744E transgenics, but never as a complete punctate

ring, as is occasionally observed in wild type GFP-PDVl expressing lines (Miyagishima,
Froehlich, et al. 2006, Yang, Glynn, et al. 2008). The reason for this is unclear, but the
GFP-PDVl punctate ring is the most uncommon PDV] localization pattern observed;
PDVl is typically observed as a single spot in wild type cells and more observation of

these transgenic lines may be required. Additionally, while FtsZZ was localized to the

mid-plastid in ARC6S744E lines, several adjacent Z-rings were observed (Figure 4.8L),

reminiscent of the FtsZ2 localization pattern in pdvl, pdv2, and arc5 (Gao, Kadirjan-
Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006); all these mutants have defects
in dynamin activity, which presumably generates a signal leading to the formation of

several central Z-rings (Yang, Glynn, et al. 2008). Based on the Z-ring morphology in

133

our ARC6S744E transgenics, we propose that this signal is transduced through ARC6.

Because of this arc5-like Z-ring phenotype in ARC6S744E lines, we propose that the

dumbbell-shape chloroplasts in these lines result from a decrease of dynamin pinchase
activity at the outer envelope, perhaps resulting from a decrease in ARC6-PDV2 affinity

due to the phosphomimetic mutation in ARC6. Alternatively, upregulation of stromal Z-

ring assembly in ARC65744 E mutants (Figure 4.8L) might impede dynamin-mediated

envelope constriction, as the Z-ring must presumably either constrict or disassemble to

allow the division process to conclude.

How and why would ARC6S744 phosphorylation within the IMS affect Z-ring

assembly and/or dynamin activity? One highly speculative possibility is that S744

phosphorylation occurs to slow or regulate the division process. Perhaps ARC6S744 is

phosphorylated as the Z-ring is being assembled— this leads to controlled recruitment of
PDV2 until the Z-ring is fully assembled within the stroma and ready to constrict. At this
point, ARC6 molecules would then be progressively dephosphorylated at position 744,

causing increased ARC6-PDV2 affinity and hence increased dynamin (ARC5) activity at

the outer envelope with concurrent disassembly of the Z-ring in the stroma. If this is the

case, it is unclear why we did not observe smaller chloroplasts in ARC6S744 A lines;

perhaps this phosphoregulatory mechanism may only be relevant when PDV2 is

overexpressed, functions only under a particular set of growth conditions, or the

modification of additional residues within the plant-specific motif of ARC6pBD (such as

134

S740 and/or T742) might also contribute to regulating ARC6-PDV2 interaction.
Regardless, this hypothesis is in-line with observations that the rate of plastid division is
affected by the amount of PDV2 protein: less PDV2 leads to a slower rate of division and
more PDV2 leads to a faster rate of division (Okazaki, Kabeya, et al. 2009). ARC6 is
absolutely required for PDV2 localization at the division site (Figure 4.8B) and ARC6 is
absolutely required for PDV2-mediated division activity, as overexpression of PDV2 in
the arc6 background does not modify the number of chloroplasts observed within arc6
mutants (Glynn, Froehlich, et al. 2008). Presumably, the amount of PDV2 present at the

division site is a rate-limiting step, with regard to organelle constriction —— and

phosphorylation of ARC65744 might be a mechanism by which individual chloroplasts

regulate PDV2 recruitment to the division site, providing a checkpoint for proper Z-ring
assembly prior to dynamin-driven constriction of the organelle. Further work on this
aspect of the division process might lead to the discovery of a novel auto-
kinase/phosphatase activity for ARC6, or a separate novel kinase/phosphatase, that aids
fine-tuning of the division process by controlling the recruitment of PDV2 to the division

site.

135

Materials and Methods

Yeast two-hybrid analysis.

ARC6 A A 637-801 was cloned into pGADT7 using Ndel-Xmal and PDV2 A A233-

307 was cloned into pGBKT7 using Ndel-Xmal sites. Site-directed alleles of ARC6 were

made by SOE-PCR (Warrens, Jones, et al. 1997) and cloned into pGBKT7 using Ndel-
XmaI sites. Yeast strain AH109 (Clontech) was cultured and transformed as
recommended by the manufacturer using standard Synthetic Dropout (SD) Media
(Clontech) as indicated. Growth assays were conducted using previously-established

protocols.

Pulldown assays.

All fusion proteins were expressed in E. coli BL21 (DE3) Codon Plus cells
(Stratagene) induced at OD600 = 0.8 with 2 mM IPTG for 2 hours at 37°C. 750 ug of
protein from induced cell extract was used for each bait/prey combination. Pulldowns

between Hi83-ARC61MS and GST-PDVZIMs, ARC61MS (3744 A) and GST-PDVZIMs, or

ARC61MS (874413) and GST-PDVZIMS were performed as in previous experiments

(Glynn, Froehlich, et al. 2008), except Triton X-100 was present at 0.1% in all wash

buffers to prevent clumping of the sepharose beads.

136

Structural modeling.

A structural model of PDVZIMS was generated by comparison of PDV2 A A23 3-

307 to all current PDB structures (http://wwwpdborg) using the MetaServer at

http://bioinfo.pl. The best-matched protein (in terms of its secondary structure) was
aligned with the relevant query protein using CLUSTALW (Larkin, Blackshields, et al.
2007) and a PDB file containing data for a three dimensional model was generated using
the homology modeling server at http://prOteinsmsuedu. The final images used in

figures herein were rendered using PyMol 0.99rc6 (Delano Scientific).

Generation of transgenic lines.

ARC6S744 A and ARC6S744E transgenes were generated using SOE—PCR and

placed into a derivative of pCAMBIA-l 302 containing the native ARC6 promoter using
AvrII and BstEII restriction sites. Clones were sequence-verified and transformed into
Agrobacterium tumefaciens GV3101 prior to transformation of Arabidopsis (Clough and
Bent 1998). Trangenic lines were selected on Linsmaier-Skoog medium containing
hygromycin (25 ug/mL). Only hygromycin-resistant individuals were transplanted to soil

and used for further analysis.

Immunoblotting.
Plant extracts for determination of ARC6 protein level were taken from floral bud
tissue. Flower buds were ground in liquid nitrogen and prepared for 10% SDS-PAGE

using 6X sample buffer according to previously-established protocols (Wiegel and

137

Glazebrook 2002). 2.5 mg of homogenized tissue was loaded per lane. Immunoblotting
of plant extracts was performed using an ARC6-specific antibody at 1:2500 in blocking
buffer containing 5% nonfat dry milk (McAndrew, Olson, et al. 2008). The blot was
washed several times in TBS-T before applying the secondary antibody. The anti-rabbit
HRP-conjugated secondary was used at 1:5000 in TBS-T containing 5% nonfat dry milk
and the blot was washed several times in TBS-T before applying the HRP

chemilluminescent substrate (Therrno Scientific, Inc.) and exposing to film.

Quantitative analysis of chloroplast number.

Tissue preparation was carried out as described previously (Miyagishima,
Froehlich, et al. 2006, Pyke and Leech 1991, Vitha, Froehlich, et al. 2003). Cell size and
chloroplast number were measured using ImageJ v1.37 (NIH) (Bearer 2003).

Quantitative analysis of phenotypes was performed using 10 cells of similar size.

Microscopy.

Light micrographs depicting chloroplast morphology in expanded leaf cells were
taken using DIC Optics on a Leica DMI3000B Inverted Microscope outfitted with a
Leica DF C320 Camera. Samples for chloroplast morphology and quantitation were
prepared and analyzed using established protocols (Pyke and Leech 1991). Fluorescence
micrographs were taken using a Leica DMRA2 using Q-Capture Camera Control
Software (Q-Imaging) and the filter sets indicated (Leica) as previously described (V itha,
Froehlich, et al. 2003). Image analysis and RGB composites were made using ImageJ

v1.37 (NIH) (Bearer 2003).

138

Acknowledgements
I thank Aaron Schmitz, Yue Yang, and Katherine Osteryoung for helpful

discussions during the development and progression of this project.

139

Chapter 5

PARC6 Influences FtsZ Assembly and PDVl Recruitment in Arabidopsis.

140

Abstract

Chloroplast division in plant cells is accomplished through the coordinated action
of the tubulin-like F tsZ ring inside the organelle and the dynamin-like ARC5 ring outside
the organelle. This coordination is facilitated by ARC6, an inner envelope protein
required for both FtsZ assembly and ARC5 recruitment. Recently, we showed that ARC6
specifies the mid-plastid positioning of the outer envelope proteins PDVl and PDV2,
which have parallel functions in dynamin recruitment. PDV2 positioning involves direct
ARC6-PDV2 interaction but PDVl and ARC6 do not interact, indicating an additional
factor functions downstream of ARC6 to position PDVl. Here, we show that PARC6
(Paralogue of ARC6), an ARC6-like protein unique to vascular plants, fulfills this role.
Like ARC6, PARC6 is an inner envelope protein with its N-terminus exposed to the
stroma and Arabidopsis parc6 mutants exhibit chloroplast and FtsZ filament morphology
defects. However, whereas ARC6 promotes FtsZ assembly, PARC6 appears to inhibit
FtsZ assembly, suggesting ARC6 and PARC6 function as antagonistic regulators of FtsZ
dynamics. The FtsZ inhibitory activity of PARC6 may involve its interaction with the
FtsZ-positioning factor ARC3. A PARC6-GFP fusion protein localizes both to the mid-
plastid and to a single spot at one pole, reminiscent of the localization of ARC3, PDVl
and ARC5. We show that PARC6 positions and binds PDVl, but PARC6 is not required
for localization of PDV2 or ARC5. Our findings indicate that PARC6, like ARC6, plays
a role in coordinating the internal and external components of the chloroplast division

complex, but that PARC6 has evolved distinct functions in the division process.

141

Introduction

The chloroplasts of vascular plants replicate by binary fission, similar to their
cyanobacterial relatives, but utilize a machinery that is a composite of cyanobacterial and
host-derived components (Aldridge, Maple, at al. 2005, Gao, Kadirj an-Kalbach, et al.
2003, Kuroiwa, Kuroiwa, et al. 1998, Maple, Vojta, et al. 2007, Miyagishima, Takahara,
et al. 2001, Miyagishima, Wolk, et al. 2005, Osteryoung 2001, Osteryoung and Nunnari
2003, Osteryoung and Vierling 1995). The proteins that make up this dynamic complex,
all of which are nuclear—encoded, occupy the stroma, intermembrane space (IMS), and
cytosol, with several key factors acting at or across the envelope membranes (Gao,
Kadirjan-Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006, Miyagishima,

Nishida, et al. 2003a, Vitha, Froehlich, et al. 2003, Yang, Glynn, et al. 2008).

Within the stroma, two tubulin-like proteins, F tle and FtsZ2, both derived from
cyanobacterial FtsZ via endosymbiosis (Osteryoung, Stokes, et al. 1998, Osteryoung and
Vierling 1995, Stokes and Osteryoung 2003), are central players in the division process,
comprising a mid-plastid contractile ring (Z-ring) (McAndrew, F roehlich, et al. 2001,
Stokes, McAndrew, et al. 2000, Vitha, McAndrew, et al. 2001). In bacteria, the Z-ring
marks the division site, acts as a scaffold for other division components, and may provide
contractile force driving membrane constriction (Lan, Wolgemuth, et al. 2007, Li,
Trimble, et al. 2007, Margolin 2005, Osawa, Anderson, et al. 2008). A third FtsZ-
derived chloroplast division protein, ARC3, is a chimera of FtsZ and other host-derived

functional domains (Shimada, Koizumi, et al. 2004). ARC3 has likely replaced MinC, an

142

inhibitor of FtsZ polymerization and a component of the Min system, which positions the
Z-ring in bacteria (Glynn, Miyagishima, et al. 2007, Maple, Vojta, et al. 2007, Rothfield,

Taghbalout, et al. 2005).

On the cytosolic surface of the chloroplast, factors of host origin mediate
organelle constriction and separation (Gao, Kadirjan-Kalbach, et al. 2003, Glynn,
Miyagishima, et al. 2007, Miyagishima, Froehlich, et al. 2006, Miyagishima, Nishida, et
al. 2003a, Yang, Glynn, et al. 2008). ARC5 is a dynamin-like protein required for the
late stages of division, acting as a pinchase (Gao, Kadirjan-Kalbach, et al. 2003,
Miyagishima, Nishida, et al. 2003a, Yoshida, Kuroiwa, et al. 2006). Dynamins are best
known for their roles in vesicle-budding and endocytosis in eukaryotes (Hinshaw 2000,
Hinshaw and Schmid 1995, Praefcke and McMahon 2004, Wiejak and Wyroba 2002),
though the origins of dynamins can be traced to prokaryotes (Low and Lowe 2006).
ARC5 is recruited to the division site from cytosolic patches by PDVl and PDV2, two
coiled-coil transmembrane outer envelope proteins of host origin that are unique to land
plants (Glynn, Froehlich, et al. 2008, Miyagishima, Froehlich, et al. 2006). The finding
that ARC5 localizes to the division site in pdvl and pdv2 mutants but not in the double
mutant indicates that PDVl and PDV2 function independently in recruiting ARC5 to the
chloroplast, but both PDV proteins are required for full ARC5 contractile activity

(Miyagishima, Froehlich, et al. 2006).

The stromal F tsZ and cytosolic dynamin complexes that drive division of the

organelle must be coordinated across the two envelope membranes to ensure their

143

synchronous operation. A key factor coordinating these complexes is ARC6, a
transmembrane protein of the inner envelope that evolved from the cyanobacterial cell
division protein Ftn2 (Koksharova and Wolk 2002, Vitha, Froehlich, et al. 2003). The
stromal region of ARC6 binds FtsZ2 (Maple, Aldridge, et al. 2005) and mediates Z-ring
assembly (V itha, Froehlich, et al. 2003), while the IMS region interacts with and
positions PDV2 at the division site to facilitate ARC5 recruitment (Glynn, Froehlich, et
al. 2008). ARC6 is also required for PDVl positioning at the division site, but our finding
that ARC6 and PDVl do not interact suggested that an additional factor might function

between ARC6 and PDVl to localize PDVl (Glynn, F roehlich, et al. 2008).

Here we show that PARC6, an ARC6-like protein unique to vascular plants,
performs this function. Mutant analysis reveals that PARC6 is required for wild-type
chloroplast replication and accumulation. Like ARC6, PARC6 resides in the inner
envelope membrane (IEM) at the division site with its N-terminus facing the stroma.
However, PARC6 acts downstream of ARC6 to mediate PDVl localization, consistent
with a previous model (Glynn, Froehlich, et al. 2008, Yang, Glynn, et al. 2008) and show
that PARC6 interacts with the cytosolic domain of PDV]. In further contrast to ARC6,
PARC6 localizes at one pole of some chloroplasts, similar to the localization patterns of
PDVl and ARC3 (Maple, Vojta, et al. 2007, Miyagishima, Froehlich, et al. 2006).
Analysis of FtsZ filament morphology in parc6 suggests that PARC6 inhibits FtsZ
assembly, in contrast with ARC6, which promotes F tsZ assembly (V itha, Froehlich, et al.
2003). We further show that PARC6 interacts with ARC3, and hypothesize that PARC6

inhibits FtsZ assembly via this interaction. Collectively, these findings suggest that

144

PARC6 and ARC6 may function as antagonistic regulators of Z-ring dynamics, and that
PARC6 plays a role in coordinating Z-ring activity in the stroma with dynamin activity at
the outer envelope. Additionally, the interaction of PARC6 with a component of the Min
system and its localization pattern hint at the possibility that PARC6 may play a role in

establishing polarity within newly divided chloroplasts.

Results

PARC6 Family Members are Distinct fi'om ARC6 and are Unique to Vascular Plants.
We identified PARC6 by BLAST (Altschul, Madden, et al. 1997), shortly after
the identification of ARC6 by a map-based approach (V itha, Froehlich, et al. 2003). The
PARC6 locus (At3g19180) encodes a predicted protein product that shares ~21% identity
with Arabidopsis ARC6 (V itha, Froehlich, et al. 2003) and is predicted by ChloroP
(Emanuelsson, Nielsen, et al. 1999) to bear a chloroplast transit peptide at its N-terminus
(Figure 5.1). Consistent with this prediction, the N-terminus of PARC6 (amino acids
(AA) 1-76) targets a YFP fusion to the chloroplast stroma in tobacco (Glynn, Yang, et al.

2009)

In contrast to ARC6, which contains one transmembrane helix that anchors it in
the IEM (V itha, Froehlich, et al. 2003), Arabidopsis PARC6 has two predicted
transmembrane domains (Figure 5.1) based on output from Aramemnon (Schwacke,
Schneider, et al. 2003). Multiple sequence alignment between PARC6, ARC6, and Ftn2

proteins from several organisms revealed two major regions of similarity, one near the N-

145

 

TP 1-76 TMl 357-377 TM2 574-596
/ /

 

 

 

 

 

 

 

 

/
AtPARC6 E] [1 | 819AA
TP 1-67 TM 615-635
/ /
AtARC6 It. I in ] 801AA
TM 535-557
/
|720AA

 

 

 

Figure 5.1. Summary of PARC6 protein features and similarity to related proteins.
Lines under each diagram indicate regions of high similarity between all three proteins.
A multiple sequence alignment comparing sequences of PARC6, ARC6, and Ftn2 from
several species is shown elsewhere (Glynn, Yang, et al. 2009). Transit peptide (TP);
transmembrane domain (TM); and amino acids (AA).

146

terminus and the other near the C-terminus (Figure 5.1), suggesting that these regions of
PARC6, ARC6, and Ftn2 may have related functions. Several small regions conserved
within, and unique to, PARC6 family members were also evident upon close examination
of the alignment (Glynn, Yang et al. 2009); these regions may harbor functions that are
specific to PARC6 proteins. Notably, the conserved proline that resides within the
central motif of the predicted J -domain of most ARC6/Ftn2 family members (V itha,
Froehlich, et al. 2003) is not evident in PARC6 sequences (Figure 5.1), suggesting that

PARC6 does not function as a DnaJ-like co-chaperone.

While ARC6-like sequences are found in cyanobacteria, algae, and moss, PARC6
sequences were only detected in vascular plants (tracheophytes) suggesting that PARC6
emerged as a chloroplast division factor in this group of organisms. Consistent with this
observation, phylogenetic analysis shows that PARC6 and ARC6 sequences from
tracheophytes cluster into distinct clades and that both of these proteins may have
diverged from a common ARC6-like ancestor (Figure 5.2). As postulated previously
(Vitha, Froehlich, et al. 2003), ARC6 is more closely related to cyanobacterial Ftn2 than
is PARC6 (Figure 5.2), suggesting that ARC6 and Ftn2 share greater functional similarity
than PARC6 and Ftn2. These results, in combination with the sequence similarities and
differences highlighted above, suggest that PARC6 arose as a result of gene duplication
and divergence in primitive vascular plants and that it probably has evolved functions

distinct from those of ARC6 within the tracheophyte lineage.

147

'1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

64 ~——VvPARC6
54_[ PtPARC6
98 MtPARC6
'00 AtPARC6
OsPARC6
46 1 oo ZmARC6
__‘22{ OsARC6
AtARC6
‘ °° PpARC6
OlARC6
TeFtn2
‘00 . SeFtn2
41L NsFth
93 —---CwFt112
100 - CsFtn2
|-—-—l
0.1

 

 

 

Figure 5.2. Phylogenetic analysis of PARC6, ARC6, and Ftn2 family members.
Neighbor-joining tree showing the relationship between PARC6, ARC6, and Ftn2 family
members from various species. Numbers at the nodes represent bootstrap values and
scale bar corresponds to the number of amino acid substitutions per site. Arabidopsis
thaliana (At); Medicago truncatula (Mt); Oryza sativa cv. japonica (Os); Populus
trichocarpa (Pt); Vitis vinifera (V v); Ostreococcus lucimarinus (Ol); Physcomitrella
patens (Pp); Zea mays (Zm); Thermosynechococcus elongatus BP-l (Te); Synechococcus
elongatus PCC 7942 (Se); Nostoc sp. PCC 7120 (N s); Crocosphaera watsonii WH 8501
(Cw); and Cyanothece sp. ATCC 51142 (Cs).

148

Mutations in PARC6 cause aberrations in chloroplast morphology and F tsZ filament
morphology.

The similarity of PARC6 to ARC6 led us to investigate its role in plastid division.
We identified a T-DNA insertion (SALK_100009) in the first exon of PARC6 (parc6-1,
Figure 5.3A) and characterized homozygous lines for chloroplast morphology and

number (Figure 5.3C).

parc6-I mutants are less drastically impaired in chloroplast division than arc6
mutants (compare Figure 5.3C and Figure 5.3F), but mesophyll cells in parc6-I mutants
contain nearly 10-fold fewer chloroplasts (6.7 i 2.4 chloroplasts per cell, 11 = 15 cells)
than those in wild type (59.7 :t 11.2 chloroplasts per cell, n = 15 cells). parc6-I
mesophyll cell chloroplasts possess characteristics of chloroplasts in both arc3 and arc5
mutants (Pyke and Leech 1994). Like arc3 mutants, parc6-I mutants exhibit a
heterogeneous mixture of chloroplast sizes within individual cells (Marrison, Rutherford,
et al. 1999, Pyke and Leech 1992, Pyke and Leech 1994) (Figure 5.3C). Like arc5
mutants, parc6-1 mutants have some chloroplasts with prevalent constrictions (Figure
5.3C, arrowheads), suggesting a block in dynamin (ARC5) function (Gao, Kadirjan-
Kalbach, et al. 2003, Miyagishima, Froehlich, et al. 2006). We also identified two
additional alleles of PA RC6, parc6-2 and parc6-3, in a forward genetic screen (Figure
5.3D-E). These alleles confer phenotypes similar to those in parc6-I . parc6-2 behaves
as a dominant-negative mutation while parc6-I and parc6-3 are recessive to the wild-

type allele. These phenotypes lead us to conclude that PARC6 is a bona fide plastid

149

(SALK_100009) (Q526‘)
parc6-I parc6-3 TC'iA
At3g1 91 80
0 0
ATG parc6~2 4549 bP

(D112N)

 

Figure 5.3. Diagram of the PARC6 locus and phenotypic analysis of parc6 mutants.
(A) Gene structure and locations of characterized mutations in the PARC 6 locus. White
boxes indicate untranslated regions. Black boxes indicate exons. Introns are denoted by
thin grey lines. The transcribed region of At3gl9180 is 4549 base pairs (bp). ATG, start
codon; TGA, stop codon. (B-F) Chloroplast phenotypes in mesophyll cells of: Col-0 (B);
parc6-1 (C); parc6-2 (D); parc6-3 (E); and arc6-1 (F). Arrowheads denote sites of
chloroplast constriction in the parc6-I mutant. Scale bar = 10 pm.

150

division protein and suggest that PARC6 could influence operation of the division

machinery through ARC3, ARC5, or the ARC5 recruitment factors PDVl and PDV2.

PARC6 is an inner envelope protein with localization similar to PD VI.
To examine the subcellular localization of PARC6, we generated a transgene

encoding PARC6 fused to GFP expressed under control of the CaMV 35S promoter

(35SprosPARC6-GFP). This transgene was capable of complementing the division defect

in parc6-I mutants, indicating that it generates a functional protein product (Glynn,

Yang et al. 2009). Following selection of T2 individuals, we examined epidermal cells

in young leaves for expression of the fusion protein. PARC6-GFP localized to the
middle of ovoid, partially constricted, and deeply constricted plastids (Figure 5.4), though
whether it forms a complete ring during the early stages of division was not clear, as the
GFP signal in these plastids was very weak. The GFP signal was most evident in deeply
constricted plastids as single foci (Figure 5.4B), suggesting concentration of the fusion
protein at the isthmus connecting plastids just prior to separation. In some plastids,
PARC6-GFP was localized at the organelle surface in a single spot at one pole (Figure
5.4C), perhaps representing the persistence of the fusion protein at the pole following
separation of the daughter plastids. A similar pattern of localization has been observed in
Arabidopsis for both PDVl and ARC5 (Miyagishima, Froehlich, et al. 2006), but not for
ARC6 or PDV2 (Glynn, F roehlich, et al. 2008, Vitha, Froehlich, et al. 2003). In

addition, we observed both polar and medial localization of PARC6-YFP at the

151

 

Figure 5.4. PARC6 is a plastid protein that localizes to the division site and to polar
spots. PARC6-GFP localizes to mid—plastid puncta (A); mid-plastid spots (B); or
polar spots (C) in Arabidopsis. Scale bar 5 pm.

152

chloroplast periphery in transiently transfected tobacco leaf cells (Glynn, Yang et al.
2009), reminiscent of ARC3 localization under similar experimental conditions

(Maple, Vojta, et al. 2007). In tobacco, the dual localization in the same plastid may be a
consequence of overexpression as we did not observe both medial and polar localization
in the same plastid in Arabidopsis parc6-I mutants complemented with the PARC6-GE P
transgene (Figure 5.4). In both Arabidopsis and tobacco, PARC6-GFP was associated

with the chloroplast periphery, suggesting it is localized in the envelope membrane.

To verify that PARC6 is an envelope membrane protein, we examined the
fractionation of native PARC6 protein in isolated pea chloroplasts using an antibody
generated against the region of Arabidopsis PARC6 residing between the transit peptide
and the first predicted transmembrane domain (AA 77-356, Figure 5.1). The antibody
detected a protein migrating at ~1 16 kD in whole-cell extracts from pea (Figure 5.5, lane
1) and a competition assay in which the antibody was preincubated with the immunizing
antigen confirmed that this protein is native pea PARC6 (Glynn, Yang, et al. 2009). This
protein was enriched in fractions containing isolated intact chloroplasts (Figure 5.5, lane
2) and chloroplast membrane fractions, (Figure 5.5, lane 3), suggesting that PARC6 is a
chloroplast envelope protein. Preliminary analysis of PARC6 topology from pea
chloroplast fractions indicates that the N-terminus of the protein resides in the stroma, but
these experiments lacked controls and the orientation of the C-terminus of the protein
remains unclear (Glynn, Yang. et al. 2009). Further work is needed to fully verify the

topology of PARC6 in vivo.

153

1234 5678
<116kD>

<54kD>

   

OL-PARCG a-Ticl 10

Figure 5.5. PARC6 is a chloroplast membrane protein. Whole-cell extract (lanes 1
and 5); isolated chloroplasts (lanes 2 and 6); chloroplast membrane fractions (lanes 3 and
7); and chloroplast soluble; fractions (lanes 4 and 8) were probed with an anti-PARC6
antibody (left) or an anti-Tic110 antibody (right) (Davila-Aponte, Inoue, et al. 2003).
SDS-PAGE migration of PARC6 is slower than would be predicted based on its
molecular weight, but we observed similar behavior with ARC6 (McAndrew, Olson, et
al. 2008). Asterisk (*) indicates an unknown non-specific anti-PARC6 cross-reacting
protein. Kilodaltons (kD).

154

PARC6 inhibits F tsZ assembly and interacts with ARC3.

Because ARC6 acts as a positive regulator of FtsZ assembly (V itha, Froehlich, et
al. 2003), we examined FtsZ organization in homozygous parc6-I mutants by
immunostaining to determine if PARC6 also affects F tsZ assembly in vivo. In contrast to
the short FtsZ filaments observed in the enlarged chloroplasts of arc6 mutants (Vitha,

F roehlich, et al. 2003), we observed long Ftle filaments that appeared to be multiple
rings or spirals within the larger chloroplasts of parc6~1 (Figure 5.6B). Additionally, we
observed tubular chloroplasts containing clusters of FtsZ rings (Figure 5.6C,
arrowheads), reminiscent of the multiple Z-rings observed in arc3 mutants (Glynn,
Miyagishima, et al. 2007). The elongated FtsZ filaments observed in parc6-I were not
due to a change in Fle or FtsZZ levels, which were similar to those in wild type (Figure
5.6D). These results suggest that PARC6 might act as an antagonist of ARC6, with
PARC6 inhibiting FtsZ assembly, and also implicate PARC6 in the fimctioning of the
plastidic Min system (Fujiwara, Hashimoto, et al. 2008, Fujiwara, Nakamura, et al. 2004,
Itoh, Fujiwara, et al. 2001, Maple and Moller 2007, Maple, Vojta, et al. 2007, Nakanishi,

Suzuki, et al. 2009).

Arabidopsis ARC6 has been shown to interact with both FtsZ2 family members,
but not Ftle (Maple, Aldridge, et al. 2005, Schmitz, Glynn, et al. 2009). The profound
effect of PARC6 depletion on Z-ring morphology led us to test for interaction between
the putative N-terrninal stromal region of PARC6 (AA 77-357) and both families of

Arabidopsis FtsZ using two-hybrid assays. However, neither Ftle nor FtsZ2 interacted

155

   

123123

I.

Ponceau a-Ftle a—FtsZZ

  

50kD >

 

37kD >

Figure 5.6. Analysis of FtsZ morphology and FtsZ protein levels in parc6-1 .
Immunolocalization of Ftle in Col-0 (A); and parc6-1 mutants (B, C). Arrowheads
point to structures formed by F tsZ. Chlorophyll autofluorescence (red) and Ftle
localization (green) is indicated. FtsZ2 filament morphology is also perturbed in parc6-1
mutants, as both FtsZ2-1 and FtsZ2-2 do not form regular rings, but instead form highly
elongated filaments (Glynn, Yang et al. 2009) that are unlike the short FtsZ filaments
observed within the chloroplasts of arc6 mutants (Vitha, Froehlich, et al. 2003).
Immunoblots showing FtsZ levels are shown in panel (D). Ponceau stain of the
membrane used for immunoblotting with Ftle and FtsZ2-1 specific antibodies (left);

F tle immunoblot (center); and FtsZ2—1 immunoblot (right). After Ponceau staining,
blot was probed for Ftle, stripped, and then reprobed for FtsZ2-1. Molecular weight
marker (lane 1); Col-0 whole cell extract (lane 2); and parc6-1 whole cell extract (lane 3).
Scale bars = 5 pm.

156

with PARC6 (Figure 5.7). This result was somewhat surprising given the high affinity of

ARC6 for FtsZ2 (Maple, Aldridge, et al. 2005).

Based on the chloroplast size heterogeneity in parc6-1 (Figure 5.3) and arc3 (Marrison,
Rutherford, et al. 1999, Pyke and Leech 1992, Pyke and Leech 1994), the affect of
PARC6 on FtsZ ring and filament morphology (Figure 5.6) and the ARC3-like
localization of PARC6 in tobacco (Glynn, Yang et al. 2009, Maple, Vojta, et al. 2007),
we decided to test for interaction between PARC6 and ARC3 by two-hybrid. In contrast
to ARC6 (Maple, Aldridge, et al. 2005), the N-terminus of PARC6 interacted strongly
and specifically with the predicted mature ARC3 protein (Figure 5.8); these results were
confirmed by another group (Zhang, Hu, et al. 2009b). Interestingly, the PARC6-ARC3
interaction requires the Membrane Occupation and Recognition Nexus (MORN) domain
(Shimada, Koizumi, et al. 2004) (Figure 5.8), a region of ARC3 that was shown
previously to inhibit a specific interaction between ARC3 and Ftle (Maple, Vojta, et al.

2007)

PA RC6 acts downstream of ARC6.

‘ To investigate the functional relationship between ARC6 and PARC6, we
generated arc6 parc6 double mutants and compared their phenotypes with those of both
parental mutants and wild-type plants (Figure 5.9A-D). arc6 parc6 plants (Figure 5.9D)
possess mesophyll chloroplast phenotypes that are more like those of arc6 mutants
(Figure 5.9C) than parc6 mutants (Figure 5.9B), suggesting that PARC6 acts downstream

of ARC6 during chloroplast division.

157

INTERACTION PAIR SI)/-ULT SD/-ULTI'I

PARC6AA77_ 357 vs. Ftle- 1 --
PARC6AA77_ 357 vs. FtsZ2- l -
EMPTY vs. Ftle—l -_
EMPTsz. 14.2.2-1 --

 

 
 

  
 
   

EMPTY vs. EMPTY

 

Figure 5.7. PARC6 does not interact with an FtsZ protein. Unlike ARC6, PARC6
does not interact with an FtsZ protein, as growth was not observed on synthetic dropout
(SD) media lacking histidine (right column). Only constructs representing processed
stromal portions of each protein were used to test for interaction. Dilutions at bottom of

each column refer to dilution from a starting culture of OD600 = 1.0.

158

 

 

 

 

 

 

TP 1-76 TM] 357—377 TM2 574-596
/ / /
PARC6 El El 1
TP 141 MORN 598-741
/ /
ARC3 -
B
INTERACTION PAIR SD/-ULT SD/-ULTH

 

PARC6AA77_357 vs. ARC3
PARC6AA77-357 vs. ARC3-M
PARC6AA77_357 vs. EMPTY
EMPTY vs. ARC3
EMPTY vs. ARC3_M
EMPTY vs. EMPTY

  

    

1 1/10 1/100 1 1/10 1/100

Figure 5.8. PARC6 interacts with ARC3. A schematic of the proteins used for
interaction assays is shown in (A). The underline indicates the portion of the protein used
for assay. Results of the two-hybrid assay are shown in (B). Dilutions at bottom of each
column refer to dilution from a synthetic dropout (SD) starting culture of OD600 = 1.0.
For interaction test between PARC6 A A77-3 57 and ARC3, a growth ratio of 0.67 was
observed (growth on SD/-ULTH: growth on SD/-ULT) ~48h after spotting; all other
interaction pairs had growth ratios of 0.07 or less. Transit peptide (TP); transmembrane
domain (TM); ARC3-MORN (ARC3_M).

159

I 7 lipttrC6~l

.xncnxirv
’ \Rt‘tntil-‘l’ '

 

Figure 5.9. PARC6 acts downstream of ARC6. Chloroplast phenotypes of Col-0 (A);
parc6-1 (B); arc6-1 (C); and a parc6-I arc6-1 double mutant are shown. ARC6-GFP
localizes to mid-plastid rings in wild type Col-0 (E) and parc6-I mutants (F), indicating
that PARC6 is not required for ARC6 localization in Arabidopsis. Scale bar = 5 pm.

160

If PARC6 acts downstream of ARC6, we expected that ARC6-GFP should
localize normally to chloroplast constriction sites in parc6 mutants. To test this, we

examined ARC6-GFP localization in Col-0 and parc6-l mutants using an

ARC6prosARC6-GF P transgene that was used previously to examine ARC6 localization

in Arabidopsis (Glynn, Froehlich, et al. 2008). Consistent with ARC6 acting upstream of
PARC6, we observed localization of ARC6-GFP to sites of constriction in young leaf
cells of both wild-type (Figure 5.9E) and parc6 mutants (Figure 5.9F). From these data,
we conclude that PARC6 acts downstream of ARC6 and that PARC6 is not required for

localization of ARC6 to sites of chloroplast constriction.

PARC6 is required for PD VI localization, but not for PD V2 or ARC5 localization.

We previously reported that ARC6 interacts directly with PDV2 and positions
PDV2 at the division site (Glynn, Froehlich, et al. 2008). We also showed that ARC6 is
required for equatorial positioning of GFP-PDV] , though direct interaction between
ARC6 and PDVl could not be confirmed by two-hybrid. At that time, we hypothesized
the presence of at least one factor that acts as an intermediary between ARC6 and PDVl.
Because PARC6 acts downstream of ARC6 and because parc6 mutants (Figure 5.9B)
possess a subpopulation of chloroplasts that resemble the partially-constructed
chloroplasts observed in pdvl mutants (Miyagishima, Froehlich, et al. 2006), we
hypothesized that PARC6 might be the factor that acts between ARC6 and PDVl. To
test whether PARC6 is required for PDVl localization, we examined GFP-PDVl signals

(Miyagishima, Froehlich, et al. 2006) in Col-0 and parc6-1 mutants (Figure 5.10). We

161

 

 

1.

A .

GFP-PDVI

    

purctfil

 

 

purc'O— I

 

 

( 'ol—(l purco-I

 

Figure 5.10. Localization of PDVl, PDV2, and ARC5 in parc6 mutants. GFP-PDVI
localizes to mid-plastid constrictions in Col-0 (A), but is unable to localize to the division
site in parc6-I mutants (B). YFP-PDV2 (C, D) and GFP-ARC5 (E, F) localize to sites of
constriction in Col-0 (C. E) and parc6-1 mutants (D, F). Scale bar = 5 pm.

162

were unable to observe any GFP-PDV] signal at sites of chloroplast constriction in
parc6-I mutants (Figure 5.103), though GFP-PDV] localized properly to dividing
chloroplasts in wild-type (Figure 5.10A), was expressed in both wild type and parc6-1
backgrounds based on immunoblotting (Glynn, Yang et al. 2009), and can functionally
complement the pdvl mutant (Miyagishima, Froehlich, et al. 2006). Taken together, the
PDVl-like localization of PARC6 in Arabidopsis (Figure 5.4) and the loss of PDVl
localization in parc6 mutants (Figure 5.10B) indicate a major role for PARC6 in

positioning PDV] at the division site.

Consistent with PARC6 acting downstream of ARC6 (Figure 5.9) and ARC6-
dependent positioning of PDV2 (Glynn, Froehlich, et al. 2008), we observed YFP-PDV2
at sites of chloroplast constriction in parc6-I mutants as well as in wild type (Figure
5.10C-D). Further, in agreement with previous fmdingings that either PDV] or PDV2
are capable of recruiting ARC5 to the division site (Miyagishima, Froehlich, et al. 2006),
we also observed GFP-ARC5 positioned correctly in parc6 mutants despite loss of PDVl

positioning (Figure 5.10E-F).

PARC6 binds the cytosolic domain of PD VI in two—hybrid assays.

Because PARC6 is required for positioning PDVl at the division site, we aimed
to determine if PARC6 bound PDVl in two hybrid assays. The C-terminus of PDVl was
previously shown to be localized to the IMS (Miyagishima, F roehlich, et al. 2006). We
initially tested for interaction between the C-termini of PARC6 and PDVl , assuming that

the C-terrninus of PARC6 (AA 597-819, Figure 5.1) might reside within the

163

intermembrane space, with PARC6 having a topology similar to ARC6 (V itha, Froehlich,
et al. 2003). However, no interaction was detected between these two C-terminal
domains, nor did the middle region of PARC6 (AA 378-573, Figure l) interact with the
C-terminus of PDVl (M. Hemmes, J. Glynn, and K. Osteryoung, unpublished). Because
PARC6 possesses two predicted transmembrane domains (Figure 1), possibly allowing it
to span both envelope membranes (thereby placing the C-terminus (AA 597-819, Figure
5.1) in the cytosol), we also tested for interaction between the C-terminus of PARC6 and
the cytosolic N—terminus of PDVl. To our surprise, a weak interaction between these
two proteins was detected in yeast (Figure 5.11). This interaction is consistent with
PARC6-dependent positioning of PDVl , but the precise topology of PARC6 remains to
be determined. However, it must be seriously considered that PARC6 might have a more

dynamic or complex topology relative to other division proteins.

PD VI and PD V2 Independently Localize to the Division Site.

If PARC6 and ARC6 are responsible for positioning PDVl and PDV2, at the
division site, respectively, we predicted that PDVl and PDV2 should localize
independently of each other. To confirm this, we expressed GFP-PDVl in pdv2-1
mutants and YFP-PDV2 in pdvl-1 mutants. GFP-PDV] localized to the mid-plastid in
the pdv2-I background (Figure 5.12B), indicating that PDV2 is not required for PDVl
localization. Similarly, YFP-PDV2 localized to the mid-plastid in the pdvl-1 background
(Figure 5.12D). We conclude that PDVl and PDV2 localize to the division site
independently of one another — PDVl through PARC6 and PDV2 through ARC6 —

though PARC6 localization and/or activity probably depends upon ARC6 (Figure 5.9).

164

 

 

 

 

 

 

TP/1-76 TM 1/357-377 Th/AZ 574-596
PARC6 I
TM 207-225
PDV] I
B
INTERACTION PAIR SD/—ULT SDl-ULTH

 

 

 

PARC6m-vs. EMPTY

   

EMPTY vs. PDV] N,

EMPTY vs. EMPTY

 

1 1/10 1/100 1 1/10 1/100

Figure 5.11. The C-terminus of PARC6 binds the cytosolic domain of PDVl.
Schematic showing proteins tested in yeast two-hybrid interaction assays (A). The
underline indicates the segment of each protein used in the interaction assays. Results of
two-hybrid reporter assays on synthetic dropout (SD) media are shown in (B), showing
growth on media supplemented with histidine (left panels) and media lacking histidine

(right panels). Growth ratio (growth on SD/-ULTH: growth on SDl-ULT) for PARC6CT
and PDVlN—r was 0.37, while all other interactions tested were 0.1 or less. Dilutions at
the bottom of each column indicate dilution from a starting culture of OD600 = 1.0.

165

(ilil:‘l’-l’l)Vl

(bho

YFP—PDVZ

q

pdvl - I

 

Figure 5.12. PDVl and PDV2 localize independently of each other. GFP-PDV]
localization (A, B) and YFP-PDV2 localization (C, D) in Col-0 (A, C), pdv2-1 (B), and
pdvI-I (D) mutants. PDV2 is not required for localization of PDVl (B) nor is PDVl
required for localization of PDV2 (D). Scale bar = 5 pm.

166

Discussion

Here we characterize the new chloroplast division protein PARC6, a protein of
tracheophyte origin that, like ARC6, is a multi-functional integral inner envelope protein
that aids the coordination of the FtsZ and dynamin rings across the envelope membranes.
However, our results show that PARC6 acts downstream of ARC6 in the division process
and that its function differs significantly from that of ARC6: (1) PARC6 and ARC6 have
distinct localization patterns in Arabidopsis. In incompletely constricted chloroplasts,
ARC6 appears as a continuous ring at the division site (Glynn, Froehlich, et al. 2008,
Vitha, F roehlich, et al. 2003, Yang, Glynn, et al. 2008) whereas PARC6 localizes to
puncta. PARC6 also localizes to foci at the poles of some plastids, which has not been
observed for ARC6. (2) PARC6 acts as an inhibitor of FtsZ assembly whereas ARC6
promotes FtsZ assembly. (3) ARC6 is required for recruitment of PDVl , PDV2, and
ARC5 to the division site whereas PARC6 is required only for recruitment of PDVl. A
working model depicting the roles of ARC6 and PARC6 in the coordination and activity

of the FtsZ and dynamin rings at the division site is shown in a later chapter.

Despite the overall sequence identity between ARC6 and PARC6 (~21%), there
are many differences that may contribute to their functional divergence. PARC6 lacks
the J -like domain present in ARC6, suggesting it also lacks the co-chaperone activity
hypothesized for ARC6 (V itha, Froehlich, et al. 2003). Additionally in contrast with
ARC6, which has one transmembrane domain, multiple PARC6 sequences bear two

predicted transmembrane domains (Glynn, Yang et al. 2009), one of which (WI;

167

Figure 5.1) interrupts a segment conserved in ARC6 family members that resides in the
stroma (Glynn, Yang. et al. 2009, Vitha, Froehlich, et al. 2003); the divergence in ARC6-

PARC6 similarity in this region coincides with absence of the FtsZ2-binding domain

(ARC6 A A 3 51-503) in PARC6 proteins (Figure 5 .1). Preliminary protease protection

assays in pea support the TM] prediction (Glynn, Yang et al. 2009) and, along with the

interaction between PARC6 A A77-3 57 and ARC3 (Figure 5.8), are consistent with a

topology placing the N-terminus of PARC6 in the stroma. Though the TM2 prediction
remains to be confirmed, it is predicted in several PARC6 sequences and is aligned
closely with the ARC6 transmembrane domain (Figure 5.1) (Glynn, Yang, et al. 2009).
Moreover, we show here that the C-terminus of PARC6 interacts with the cytosolic
domain of PDVl (Figure 5.1 1), suggesting that PARC6 might traverse both envelope
membranes. While preliminary protease protection assays indicate that PARC6 does not
traverse the outer envelope (Glynn, Yang, et al. 2009), PARC6 may have a more
complicated or dynamic topology in vivo. If such an orientation is validated, it is curious
as to why the C-termini of ARC6 and PARC6 would be compartmentally separated,
given the sequence similarity near their C-termini. Full topological analysis will be

critical to confirm these results and for further dissection of PARC6 ftmctional domains.

The long FtsZ filaments observed in the enlarged chloroplasts of parc6-I mutants
(Figure 5.6) suggest that, in vivo, PARC6 inhibits FtsZ assembly. This activity may be
indirect because PARC6 does not interact with Ftle or FtsZ2 in two-hybrid assays
(Figure 5.7). The inhibitory effect of PARC6 on FtsZ assembly may rather be a

consequence of its interaction with ARC3 (Figure 5.8) (Zhang, Hu, et al. 2009b), because

168

ARC3 has been proposed as a functional replacement for MinC (Maple, Vojta, et al.
2007), a prokaryotic protein known to promote disassembly of bacterial F tsZ (Hu,
Mukherjee, et al. 1999, Margolin 2003, Pichoff and Lutkenhaus 2001). The similar
chloroplast and FtsZ morphologies in parc6 (Figure 5.3 and Figure 5.6) and arc3 mutants
(Glynn, Miyagishima, et al. 2007 , Marrison, Rutherford, et al. 1999, Pyke and Leech
1992, Pyke and Leech 1994), are consistent this hypothesis. The opposing activities of
ARC6 and PARC6 in promoting and inhibiting FtsZ assembly, respectively, suggest that
these two proteins may fimction as antagonistic regulators of Z-ring dynamics within the
stroma. We do not yet know whether this or other activities of PARC6 involve direct
interaction with ARC6, though preliminary data suggests that they may interact through

their stromal domains (Y. Yang and K. Osteryoung, unpublished).

The requirement of the ARC3 MORN domain for PARC6-ARC3 interaction
(Figure 5.8) contrasts with the finding that the MORN domain inhibits interaction
between ARC3 and Ftle (Maple, Vojta, et al. 2007). These observations suggest that
FtsZ disassembly may be controlled, at least in part, by the availability of the ARC3
MORN domain. Interaction of ARC3 with PARC6 could sequester the MORN domain,
allowing ARC3 to interact with Ftle to promote F tsZ filament disassembly, perhaps
through disruption of the Ftle -FtsZZ heteropolymer. Dynamic interaction between
ARC3 and PARC6 could thereby regulate FtsZ dynamics in vivo, though other factors,
including MinD, MinE and MCDl, certainly contribute (Aldridge and Moller 2005,
Colletti, Tattersall, et al. 2000, Fujiwara, Hashimoto, et al. 2008, Fujiwara, Nakamura, et

al. 2004, Itoh, Fujiwara, et al. 2001, Maple, Chua, et al. 2002, Maple and Moller 2007,

169

Nakanishi, Suzuki, et al. 2009, Reddy, Dinkins, et al. 2002). Because MORN domains
have been shown to be important for membrane association (Ma, Lou, et al. 2006), an
alternative possibility for the MORN-mediated PARC6-ARC3 interaction is that it only
occurs near the surface of a membrane. Removal of the MORN sequence from ARC3

could prevent its association with the IEM, thereby preventing interaction with PARC6.

PDVl and PDV2 localization at the division site, though independent of one
another (Figure 5.12), both require ARC6 (Glynn, Froehlich, et al. 2008). PDV2
localization is established by direct interaction with ARC6, but it was previously unclear
how ARC6 directs PDVl localization, as we did not detect interaction between ARC6
and PDVl using two hybrid assays (Glynn, Froehlich, et al. 2008). Our current results
show that PARC6 acts as an intermediary between ARC6 and PDVl, functioning to
organize PDVl at the division site (Figure 5.9 and Figure 5.10). PDVl positioning by
PARC6 may involve direct interaction with PDVl within the cytosol, based on our two
hybrid experiments (Figure 5.11). Beyond this question, the mid-plastid localization of
either PDVl or PDV2 is clearly sufficient for ARC5 recruitment, but our collective data
indicate that localization of both PDV] and PDV2 at the division site is required for full
ARC5 contractile activity (Glynn, Froehlich, et al. 2008, Miyagishima, Froehlich, et al.

2006). This aspect of chloroplast division remains to be explored.

During division, PDVl initially localizes to a medial punctate ring that becomes

reduced to a single spot in deeply constricted chloroplasts. The PDVl spot persists at the

pole of one of the two daughter organelles following their separation. ARC5 behaves

170

similarly (Miyagishima, F roehlich, et al. 2006). Intriguingly, the localization patterns we
observed for PARC6 (Figure 5.4) suggest that it also follows a PDVl-like progression
during division, possibly being retained at the pole of a single daughter chloroplast. We
speculate that the lingering PARC6 spot could play a role in establishing organelle
polarity within the stroma, which might be important for defining the new division site
prior to the next round of division. In E. coli, cell polarity and division-site positioning
are established by the dynamic behavior of the Min system, which restricts FtsZ ring
assembly to the midcell (Lutkenhaus 2007). The fact that ARC3 interacts with PARC6
and that ARC3 and PARC6 exhibit partial polar localization raises the possibility that
PARC6, perhaps via interaction with ARC3 and/or other components of the plastidic Min
system (Fujiwara, Hashimoto, et al. 2008, Fujiwara, Nakamura, et al. 2004, Itoh,
Fujiwara, et al. 2001, Maple, Chua, et al. 2002, Maple and Moller 2007, Maple, Vojta, et
al. 2007, Nakanishi, Suzuki, et al. 2009), could play a role in directing the orientation of
the Min system within the stroma and hence in placing the new division site. Analysis of
parc6 arc3 double mutants, ARC3 localization in parc6, and PARC6 localization in arc3
mutants will be critical to ordering these factors in a pathway and understanding how

each of them impact division site selection.

The emergence of PARC6 in vascular plants follows the emergence of ARC3 in
green algae (Yang, Glynn, et al. 2008) and PDVl/PDV2 in mosses (Miyagishima,
F roehlich, et al. 2006), perhaps allowing for greater control of FtsZ dynamics and
chloroplast morphology. Presumably, the dissimilar regions of ARC6 and PARC6 confer

protein-specific functions and will provide clues to the individual domains responsible

171

for their unique functions. For example, the N-terminal stromal domain of PARC6 does
not interact with Arabidopsis FtsZ2 (Figure 5.7), in contrast to that of ARC6 (Maple,
Aldridge, et al. 2005), but has apparently evolved to interact with ARC3, perhaps via the
ARC3 MORN domain. Similarly, while the C-termini of PARC6 and ARC6 share a high
degree of sequence similarity (Figure 5.1) (Glynn, Yang, et al. 2009), subtle differences
within this conserved region might mediate specific interactions with other proteins,

though the precise location of the C-terminus of PARC6 remains to be determined.

The activities of the F tsZ and dynamin rings are presumably coordinated with
those of the inner and outer plastid-dividing (PD) rings, electron-dense components of the
plastid division machinery whose compositions are not yet known (Kuroiwa, Kuroiwa, et
al. 1998, Miyagishima, Nishida, et al. 2003b, Yoshida, Kuroiwa, et al. 2006). It will be
interesting to learn whether PARC6 and ARC6 interact with or otherwise influence the
operation of the PD rings. Further studies to elucidate the functional and evolutionary
relationship between PARC6 and other plastid division components should deepen our
understanding of the mechanisms contributing to coordination of the internal and external

division complexes.

172

Materials and Methods

Sequence and Phylogenetic Analysis.

These sequences were used for alignments and phylogeny: Arabidopsis thaliana
(AtPARC6, NP_188549); Medicago truncatula (MtPARC6, IMGAG-annotated
pseudomolecule AC157350_28.4); Oryza sativa cv. japonica (OsPARC6,
NP_001054252); Populus trichocarpa (PtPARC6, a GENSCAN-based prediction (Burge
and Karlin 1997) of Populus trichocarpa genome Scaffold_122 204390-215000); Vitis
vinifera (VvPARC6, CAO48483); Arabidopsis thaliana (AtARC6, NP_199063); Oryza
sativa cv. japonica (OsARC6, NP_001045726); Ostreococcus lucimarinus (OlARC6,
XP_001421185); Physcomitrella patens (PpARC6, XP_001778770); Zea mays
(ZmARC6, ACG29776); Thermosynechococcus elongatus BP-l (TeFth, BAC08309);
Synechococcus elongatus PCC 7942 (SeFtn2, ABB57973); Nostoc sp. PCC 7120
(N sFtn2, BAB74406); Crocosphaera watsonii WH 8501 (CwFtn2, EAM48783); and
Cyanothece sp. ATCC 51142 (CsFtn2, ACB49642). Evolutionary history was inferred
by Neighbor-Joining (Saitou and Nei 1987) following multiple sequence alignment with
MEGA4 (Tamura, Dudley, et al. 2007) using an identity matrix. The bootstrap
consensus-tree inferred from 1000 replicates represents the evolutionary history of the
taxa analyzed (Felsenstein 1985). Branches corresponding to partitions reproduced in
less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in
which the associated taxa clustered together in the bootstrap test (1000 replicates) are
shown next to the branches (Felsenstein 1985). Tree is drawn to scale, with branch

lengths in the same units as evolutionary distances used to infer the phylogenetic tree.

173

Evolutionary distances were computed using the Poisson correction method (Zuckerkandl
and Pauling 1965) and are given as the number of amino acid substitutions per site. All
positions containing gaps and missing data were eliminated from the dataset (Complete

deletion option). There were 441 positions in the final dataset.

Analysis of Chloroplast Morphology.

Tissue preparation was carried out as described previously (Miyagishima,
Froehlich, et al. 2006, Pyke and Leech 1991, Vitha, Froehlich, et al. 2003). Quantitative
analysis of Col-0 and parc6-1 phenotypes was performed using the distal 3-4 mm of the

largest leaf present on plants 30 days after sowing.

Analysis of Subcellular Localization of PARC 6

35Spro-PARC6 Tp— YF P and 35Spm-PARC6-YF P constructs were made using the

Gateway System (Invitrogen, http://www.invitrogen.com/). A PARC6 cDNA was used
as template for PCR. Analysis of PARC6 localization in tobacco and in Arabidopsis was

performed as described (Glynn, Yang, et al. 2009).

Production of PARC 6 antibody.

The coding sequence for the N-terminus of mature PARC6 (PARC6 A A 77-357)
was cloned into pHIS9 (Yang, Xu, et al. 2008) and His-tagged protein was expressed in
E. coli Rosetta DE3 (N ovagen). The insoluble recombinant protein was solubilized with

6M urea and purified using Ni-agarose. Subsequent SDS-PAGE removed urea and

purified the antigen. Afier brief coomassie-blue staining, antigen was excised from the

174

gel and rabbit antibodies for PARC6 were generated by Covance Research Products, Inc.

Exsaguination serum was used at 1:5,000 for immunoblotting.

Analysis of PARC 6 Fractionation and Topology.
Pea chloroplast isolation and fiactionation into membrane and soluble fractions
was performed as described previously (Bruce, Perry, et al. 1994); additional details can

be found elsewhere (Glynn, Yang, et al. 2009).

Analysis of F tsZ Localization.
Tissue preparation, fixation, and immunofluorescence analysis were carried out as
described previously (Miyagishima, Froehlich, et al. 2006, Vitha, Froehlich, et al. 2003,

Vitha, McAndrew, et al. 2001).

Two-hybrid Analysis.
Clones for two hybrid analysis were made using methods as described (Glynn,
Yang, et al. 2009). HIS3 reporter assays were performed as described previously (Glynn,

Froehlich, et al. 2008, Maple, Aldridge, et al. 2005).

Tests of Epistasis.

arc6-I was crossed to parc6-1 and arco parc6 double mutants were identified by
genotype. SALK genotyping primers were used as described (Alonso, Stepanova, et al.
2003) for identifying homozygous parc6-I plants. An NlaIII-containing dCAPS marker

(N eff, Neff, et al. 1998) was used to identify the arc6-1 mutation; the product generated

175

from these two primers harbors an additional NlaIII site when arc6-1 genomic DNA is
amplified. Lines homozygous for both mutations were used for double mutant analysis.

All primers used for genotyping are described elsewhere (Glynn, Yang, et al. 2009).

Analysis of ARC 6, PDV], PDV2, and ARC5 Localization.

Transgenes for ARC6pm-ARC6-GF P, PD VIpm-GFP-PD VI, PD V2pm- YFP-
PD V2, and ARC5 pro-CF P-A RC 5 were introduced by Agrobacterium-mediated

transformation (Clough and Bent 1998). Following selection of T1 individuals, young

leaves of transgenic plants were examined by fluorescence microscopy using a Leica

DMRA2 microscope outfitted with a Retiga EXi camera.

Extract Preparation and Immunoblotting.

Whole-cell extracts for anti-Ftle-l and anti-FtsZ2-1 immunoblots were prepared
by grinding tissue from 11 day-old plate-grown plants in liquid nitrogen. The resulting
powder was homogenized in 5 volumes of Extraction Buffer (25mM Tris, pH 7.4
containing 4mM dithiothreitol, 0.1% Triton X-100, and protease inhibitors). An equal
amount of loading buffer was added to this homogenate before boiling. Four microliters
of each sample was loaded for SDS-PAGE and transferred to nitrocellulose. The blot

was probed with Ftle -1 or FtsZ2-1 antibodies at 1:15000 in 5% blocking buffer.

176

Acknowledgements

I thank Yue Yang for developing the PARC6 antibodies, determining preliminary
topology of PARC6, providing localization of PARC6 in tobacco, and sharing data on
ARC6-PARC6 interaction. 1 thank Stan Vitha for examining FtsZ localization in parc6
mutants and determining localization of PARC6-GFP in Arabidopsis. I thank Aaron
Schmitz (ARC3) and Mia Hemmes (PARC6) for generating the two hybrid clones used
here. Finally, I thank Shin-ya Miyagishima for providing the parc6-2 and parc6-3

alleles.

Most of this chapter and the figures therein are reproduced from (Glynn, Yang. et
al. 2009) (httn://www.wilev.com/bw/iournflrsp?ref=0960-74l2/) and is Copyright
Blackwell Publishing. Blackwell Publishing allows the author to re-use your own article
in another publication providing you are editor or co-editor (author or co-author) of the

new publication.

177

Chapter 6

A Conserved Aspartate within ARC6 Influences Plastid Size and Z-ring Position.

178

Abstract

The systems controlling positioning and assembly of the complexes that mediate
plastid division are likely to be integrated with each other to allow for efficient regulation
of the division process. The site of bulk assembly of the FtsZ polymers within the stroma
is restricted to the mid-plastid by the Min system, composed of ARC3, MinD, MinE,
MCDl , and PARC6. However, ARC6 is absolutely required for Z-ring formation in
vivo. It is unclear if or how the activities of ARC6 and the Min system are coordinated to

ensure efficient assembly and operation of the plastidic Z-ring. Here, we perform

preliminary characterization of a novel hypomorphic allele of ARC6, arc6 D20 5 N-

arc6 D 20 5N mutants possess subtle alterations in chloroplast size and morphology, relative

to wild type, perhaps due to decreased amounts of ARC6 protein. Most intriguingly,

arc6 D 20 SN mutants harbor misplaced and miniaturized Z-rings within the stroma,

implicating ARC6 in the operation of the plastidic Min system. The conservation of

ARC6D205 amongst land plants, algae, and cyanobacteria suggests that this mechanism

of crosstalk between factors that position and stabilize the Z-ring probably occurs in all

ARC6- and F tn2-encoding organisms.

179

Introduction

The FtsZ ring (Z-ring) is a central player in bacterial cell division and plastid
fission, providing some small amount of contractile force and serving as a scaffold for
other division proteins (Erickson 2009, Osawa, Anderson, et al. 2008, Rothfield,
Taghbalout, et al. 2005). In most bacteria, a single FtsZ gene encodes a polymer-forming
GTPase with structural similarity to eukaryotic tubulins (Erickson, Taylor, et al. 1996,

Mukherjee, Dai, et al. 1993).

In bacteria, positioning of the Z-ring is critical to ensuring an equal distribution of
cytoplasm and genetic material between new daughter cells. At least two systems
provide input into the positioning of the Z-ring in E. coli: the nucleoid occlusion (Noe)

system and the minicell (Min) system (Rothfield, Taghbalout, et al. 2005).

The Noc system prevents chromosomal scission during division by inhibiting Z-
ring formation around the bacterial chromosome (Bernhardt and de Boer 2005).
Uniquely, both plastids and cyanobacteria carry multiple chromosomal copies (Birky and
Walsh 1992, Doolittle 1979, Falkow, Dworkin, et al. 2006, Schneider, Fuhrmann, et al.
2007) and do not utilize proteins with sequences similar to Noc proteins (Glynn,
Miyagishima, et al. 2007). Moreover, it has been shown that Z-rings can form around the
chromosome in the cyanobacterium Synechococcus elongatus (Miyagishima, Wolk, et al.
2005). Based on these and other observations, nucleoid occlusion is probably not a

mechanism employed in these lineages; rather it is likely that chromosomal segregation

180

in the cyanobacteria is largely managed by maintaining multiple chromosomal copies
(Bin, Guohua, et al. 2007). Because of its apparent lack of relevance to plastid division,
we will limit our review of the nucleoid occlusion system and focus on the other Z-ring

positioning input, the Min system.

The Min system in gram-negative bacteria is composed of three primary factors:
MinC, MinD, and MinE (Lutkenhaus 2007). MinC binds FtsZ and inhibits Z-ring
assembly by preventing lateral associations (Hu, Mukherjee, et al. 1999, Scheffers 2008,
Shen and Lutkenhaus 2009). However, the FtsZ assembly-inhibiting activity of MinC is
regulated by MinD, which is tethered to the membrane and promotes MinC activity only
in the polar zones of the cell (Hu and Lutkenhaus 1999, Raskin and de Boer 1999a,
Szeto, Rowland, et al. 2002). MinD is. regulated by MinE through a mechanism in which
MinE binds to MinD and causes it to be released from the membrane (Hu and
Lutkenhaus 2001). The maximum concentration of MinE occurs near the midcell
adjacent to the membrane, thereby causing the concentration of active MinC and MinD to
be highest at the poles (Hale, Meinhardt, et al. 2001). While the determinants for MinE
localization are unknown, the midcell zone created by MinE activity allows for FtsZ
protofilament assembly along the inner leaflet of the plasma membrane at the midcell
(Rothfield, Taghbalout, et al. 2005). In some bacteria that lack MinE, DivIVA tethers
MinD at the cell poles and inhibits FtsZ polymerization within the polar zone by
maintaining a higher concentration of active MinCD at the poles (Marston and Errington

1999).

181

The components of the plastidic Min system are the only known factors that are
responsible for positioning the Z-ring within chloroplasts. The Min system is now known
to be made up of at least five components in Arabidopsis: AtMinD (Colletti, Tattersall, et
al. 2000), AtMinE (Maple, Chua, et al. 2002), ARC3 (Glynn, Miyagishima, et al. 2007,
Maple, Vojta, et al. 2007), MCDl (Nakanishi, Suzuki, et al. 2009), and PARC6 (Glynn,
Yang, et al. 2009, Zhang, Hu, et al. 2009a). There is no MinC ortholog in any of the
vascular plants sequenced to date (Yang, Glynn, et al. 2008), but ARC3 seems to fulfill a
similar functional role in vivo (Glynn, Miyagishima, et al. 2007, Maple, Vojta, et al.
2007). On the basis of their sequence and phenotypic similarity, it was hypothesized that
AtMinD and AtMinE might work similarly to E. coli MinD and MinE (Colletti,
Tattersall, et al. 2000, Fujiwara, Nakamura, et al. 2004, Maple, Chua, et al. 2002).
AtMinD exhibits both polar and equatorial localization in chloroplasts (Fujiwara,
Nakamura, et al. 2004, Nakanishi, Suzuki, et al. 2009), suggesting that it may relocate or
oscillate, similar to E. coli MinD (Raskin and de Boer 1999b), to regulate Z-ring position.
AtMinE localizes to poles in tobacco chloroplasts (Maple, Chua, et al. 2002) in contrast
to its mostly mid-zone localization in bacteria (Hale, Meinhardt, et al. 2001, Sun and
Margolin 2001), though the localization pattern of functional AtMinE under native
expression conditions in Arabidopsis has not yet been demonstrated. MCDl is required
for AtMinD function and binds AtMinD, but the inputs into this plant-specific division
protein are not yet clear (Nakanishi, Suzuki, et al. 2009). Similarly, PARC6 appears to
inhibit Z-ring assembly through its interaction with ARC3, but further work is required in
order to firlly determine how this protein is integrated into the plastidic Min system

(Glynn, Yang, et al. 2009, Zhang, Hu, et al. 2009b).

182

ARC6 is a bitopic inner envelope protein of cyanobacterial origin and was shown
to aid assembly and/or stabilization of the Z-ring (Vitha, Froehlich, et al. 2003).

However, it was not clear if ARC6/F m2 has any role in positioning the Z-ring within

chloroplasts or cyanobacteria. Here we characterize a novel allele of ARC6, arc6Dzo5N,

that exhibits min-like defects in chloroplast morphology and Z-ring positioning.

Consistent with its profound effect upon Z-ring placement, we show that ARC69205 is a

conserved residue in all ARC6-like proteins. The effect of arc6 020 5N upon Z-ring

placement probably does not occur through AtMinE, but we show here ARC6 ring
formation requires both F tsZ and AtMinE. We conclude that ARC6 plays a role in
influencing Z-ring placement and propose that ARC6 might affect division-site selection

through PARC6 or another Min-system component.

Results

arc6 D20 5N is a Hypomorphic Allele of ARC6 that is Associated with Defects in

Chloroplast Morphology and Number.

To identify important functional domains within ARC6, we used TILLING
(McCallum, Comai, et al. 2000) to identify novel alleles of ARC6 that are associated with
defects in chloroplast division. A major advantage of TILLING is that in vivo relevance
for any new allele can be quickly determined by the same established assay(s) used to
identify mutants in forward genetic screens (Henikoff, Till, et al. 2004). We targeted the

two large conserved regions within ARC6 (see Figure 1.4) for TILLING (V itha,

183

Froehlich, et al. 2003) and identified 14 novel mutations within the ARC6 locus that lead
to amino acid changes within the ARC6 protein: P858, P878, D205N, SZSON, G293D,
L309F, A328V, A592T, A621V, R677K, E705K, D708N, S763F, and T7731. We were
able to isolate homozygous lines for each of these polymorphisms, with one exception

(A328V), by screening with molecular markers (Neff, Neff, et al. 1998). Of these, only

arc6 D 20 5N exhibited defects in chloroplast morphology and number. arc6 D 205N mutants

have slightly enlarged chloroplasts and possess slightly fewer chloroplasts than the

corresponding wild type TILLING parental line, Col-er105 (Figure 6.1); the phenotype

associated with arc6 D 20 5 N is recessive to the wild type allele. In addition to the enlarged

chloroplasts present within arc6 D 20 5 N, we occasionally observed mini-chloroplasts

within this line (see later sections). These mini-chloroplasts could arise from asymmetric
positioning of the Z-ring, in a process similar to the asymmetric division events that

generate chloroplasts of varying sizes in Arabidopsis min mutants (Colletti, Tattersall, et

al. 2000); however, the number and incidence of mini-chloroplasts in arc6Dzo5N was

highly variable. From these results, we conclude that ARC6DZO5 is a functionally-

irnportant residue in vivo that could be involved in the operation of the plastidic Min

system.

184

 

 

O

Chloroplasts per cell

 

Col-er105 arc6Dzo5N

 

 

 

Figure 6.1. arc6D205N is associated with a defect in chloroplast size and number.
Chloroplast phenotypes are shown from expanded leaf cells of Col-er105 (A) and

arc6 D20 5N (B). Scale bars = 10 um. Quantitative analysis of chloroplast number is
shown in panel (C). Error bars represent standard deviation from the mean; 25 cells were
counted for each line; the average cell plan area was 3279 :t 467 umz.

185

ARC 6 D 20 5 is conserved in land plants, algae, and cyanobacteria.

To examine the evolutionary conservation of ARC6DZO5 and gain insight into its
function, we generated multiple sequence alignments between several ARC6, PARC6,
and Ftn2 proteins (Figure 6.2). The alignment revealed that ARC6DZO5 is a conserved
residue (Figure 6.2, arrowhead) that resides within a 9 amino acid motif (Figure 6.2,
line). From this analysis, we conclude that ARC6Dzo5 is likely to be a functionally-

important amino acid that impacts the operation of the divisomes of chloroplasts and

cyanobacteria by a similar mechanism.

arc6 D 20 5 N is Associated with Miniaturized and Misplaced Plastidic Z-rings, as well as a

Decrease in ARC 6 Protein Levels .

To gain insight into the mechanism by which arc6 D 20 5N might impact chloroplast

division, we examined F tsZ localization in arc6Dzo5N mutants. In wild type chloroplasts,

FtsZ proteins colocalize to a central ring (Figure 6.3A) (V itha, McAndrew, et al. 2001).

In contrast, arc6 D 20 5 N mutants possess a miniature FtsZ ring that is localized near the

chloroplast periphery (Figure 6.3B). This F tsZ localization pattern suggests that ARC6
not only regulates FtsZ assembly within plastids (V itha, Froehlich, et al. 2003), but may
also influence the size and position of the Z-ring. Moreover, this peripheral Z-ring

localization might be connected to the unusual mini-chloroplast phenotype observed in

arc6Dzo 5N plants (Figure 6.3C-D).

186

 

 

  
 

YLE
YLE

VVPARCG PAAVAD I T L
PtPARCG .....

 

AtARC6 ..FKQDVVLV FLDVS
OsARC6 ..FKQDVVLA YVDLS
PpARC6 ..LHRDVVFA YMELS
OlARC6 RRHRRDVALT LCEYG
ZmARC6 ..FKQDVVLA YVDIS
NsFtn2 KPYIHDIFLS ECAIA
SeFtn2 RPYVHDVLLA ECSIA
CwFtn2 KPYADDLVLS ECTVA
TeFtn2 KPYIHDLILS ECAIA
CsFtn2 KPYVHDVLLS ECAIA
AtPARcs ..... T SLELG
OsPARCG ..... A QQSLG
MtPARCG ..... s YLEIS

T LG

s ;s

 

 

 

 

 

 

 

 

Figure 6.2. Multiple sequence alignment showing conservation of ARC6D205.
Boxed regions indicate sequence similarity and shaded regions indicate sequence
identity. This portion of the alignment has been cropped from an alignment of the full-
length proteins and is published elsewhere (Glynn, Yang, et al. 2009). Black line (top)
indicates 9-mer amino acid motif and aspartate 205 of Arabidopsis ARC6 is indicated
with a red arrowhead. Arabidopsis thaliana (At); Medicago truncatula (Mt); Oryza
sativa cv. japonica (Os); Populus trichocarpa (Pt); Vitis vinifera (Vv); Ostreococcus
lucimarinus (OI); Physcomitrella patens (Pp); Zea mays (Zm); T hermosynechococcus
elongatus BP-l (Te); Synechococcus elongatus PCC 7942 (Se); Nostoc sp. PCC 7120
(N s); Crocosphaera watsonii WH 8501 (Cw); and Cyanothece sp. ATCC 51142 (Cs).

187

 

Figure 6.3. FtsZ localization and mini-chloroplasts in arc6Dzo5N.

Immunolocalization of FtsZZ-l in young leaves of Col-er105 (A) and arc6 D 2 501V (B).
Chlorophyll autofluorescence is indicated in red and FtsZ protein is shown in green;

arrowheads point to Z-rings. Mini-chloroplast phenotype observed in arc6 D 20 SN is
shown in (C), boxed region is magnified and shown in (D). Arrowheads point to mini-
chloroplasts in (D). Scale bars in (A-C) are 10 um. Scale bar in (D) is 2 pm.

188

To determine if ARC6 levels are affected in the arc6 D 20 SN background, we

examined the relative amounts of ARC6 and FtsZ protein in Col-er105 and arc6 D 20 5N

flower buds using ARC6- and FtsZ-specific antibodies (McAndrew, Olson, et al. 2008,
Vitha, McAndrew, et al. 2001). Consistent with a defect in ARC6 function, we observed
less ARC6 protein in arc6Dzo5N lines than in wild type (Figure 6.4). While no ARC6
protein was detected in this background by immunoblotting, there must be a small
amount of ARC6 protein present, as arc6 null mutants have only one or two chloroplasts
per cell (Glynn, Froehlich, et al. 2008, Vitha, Froehlich, et al. 2003); arc6Dzo5N likely
has a reduced amount of ARC6 protein that is below the threshold of detection for our
ARC6 antibody (McAndrew, Olson, et al. 2008). However, it is unclear if this decrease

in ARC6 protein level is the basis for the defect observed in Z-ring morphology and

position (Figure 6.3B). Unlike the arc6-1 mutant (V itha, Froehlich, et al. 2003), the

levels of Ftle and F tsZ2 in arc6 D20 5 N were mostly unchanged relative to wild type

plants (Figure 6.4).

AtMinE- YF P has a unique localization pattern in Arabidopsis and ARC6 D 20 5 mutants do

not interact with AtMinE.

It was previously noted that both arc6-1 mutants and arc12 (atminE) mutants
have very similar phenotypes with respect to chloroplast morphology, as the leaf cells
within both of these mutants typically contain only one or two large chloroplasts (Glynn,

Miyagishima, et al. 2007, Pyke, Rutherford, et al. 1994, Rutherford 1996). Interestingly,

189

 

l 2

Félhflfl‘m-fiflfhfl a' L- l' I -‘ .‘n'-“ .H I “HMR'E'IJDIZI ”HIT-r .9
. 1
v . p

i;

 

 

 

 

a-ARC6

a-FtsZI

a-FtsZZ

RuBisCo

 

Figure 6.4. Preliminary analysis of ARC6, Ftle, and FtsZZ levels in arc6pzo5N.

ARC6 protein levels (top panel) in arc6 020 5N (lane 2) are diminished relative to wild
type (lane 1), while Ftle and FtsZ2-1 levels (center panels) are mostly unchanged. An
estimate of relative loading is indicated by the Coomassie stain showing RuBisCo levels
in each lane; based on this, there is slightly more protein in the lane carrying extract from

an arc6 020 5N plant, probably explaining the higher Ftle and FtsZ2-l signals in the

mutant.

190

 

not only are their chloroplast morphologies similar, but F tsZ immunolocalization patterns
observed within arc6 and arc12 are indistinguishable from one another (Glynn,
Miyagishima, et al. 2007), suggesting that the functions of the two proteins might be
connected. Because previous studies have shown that ARC6 and AtMinE have similar
phenotypes, we aimed to determine if AtMinE localizes to equatorial rings in
Arabidopsis, similar to ARC6 (V itha, Froehlich, et al. 2003). Curiously, previous studies
have shown that AtMinE localizes to polar spots within plastids of tobacco leaf cells
(Maple, Chua, et al. 2002), in contrast to its largely mid-zone localization in bacteria
(Hale, Meinhardt, et al. 2001, Sun and Margolin 2001). However, these localization
results for AtMinE are disputed, as they are observed in tenninally-differentiated leaf
cells with very high levels of 35S-AtMinE expression — AtMinE expression in
Arabidopsis is highest in regions surrounding and including the apical meristem in
vegetative tissues (Itoh, Fujiwara, et al. 2001). Other studies have shown that polar
localization of Arabidopsis Min proteins within tobacco may not be representative of
their native localization in Arabidopsis, as AtMinD localizes to a polar spot when
transiently overexpressed in tobacco (Fujiwara, Nakamura, et al. 2004), but also localizes
to a ring-like structure in Arabidopsis chloroplasts (N akanishi, Suzuki, et al. 2009). To

clarify the precise localization pattern for AtMinE in Arabidopsis, we introduced an

AtMinEpro-MinE-YFP transgene into C 01-0 and arc12 mutants; arc12 is an AtMinE loss-

of—function allele (Glynn, Miyagishima, et al. 2007). We examined emerging leaves for
a YFP signal using epifluorescence microscopy. To our surprise, we saw neither
AtMinE-YFP spots nor rings in these samples; AtMinE-YFP was localized to a diffuse

pattern within very small plastids (Figure 6.5D), suggesting that AtMinE may only have a

191

 

 

 

 

 

Figure 6.5. Localization of AtMinE-YFP in Arabidopsis. Micrographs from a young
emerging leaf expressing AtMinE-YFP in the arc12 background is shown in the upper
panels: bightfield (A); chlorophyll autofluorescence (B); AtMinE-YFP fluorescence (C);
and a merged image of chlorophyll and YFP fluorescence (D). In the lower panels,
verification of functional complementation the plastid division defect in the arc12
(atminE) mutant using an AtMinEpro-AtMinE- YFP transgene is shown: Col-0 (E); arc12
(F); arc12 rescued with the AtMinEpro-AtMinE- YFP transgene. The boxed region in (D)
is magnified and shown in (H). Scale bars in (A-G) = 10 um. Scale bar in (H) = 1 pm.

192

significant impact upon proplastid division rather than chloroplast division, consistent
with its high level of expression within the apical meristem (Itoh, Fujiwara, et al. 2001).

We did not observe any AtMinE-YFP signal within plastids of older, expanded leaves

(not shown). Our AtMinEpm-MinE- YFP transgene was fully ftmctional, as most arc12

lines carrying this transgene possessed wild type chloroplasts in fully expanded leaf cells

(Figure 6.5G). Further work is required to determine if AtMinE-YFP exhibits aberrant

localization in the arc6 020 5N background.

Because of their similar phenotypes and expression patterns, we aimed to
determine if a stromal portion of ARC6 (AA154-340) interacts with processed AtMinE
(AA34-229) using yeast two hybrid assays; this region of ARC6 (AA154-340) has no
defined firnction, but is conserved in all ARC6-like proteins. Previous work showed no

interaction between these two proteins, but because of the defect in ARC6 function

observed in arc6 D 20 5 N mutants and the prospect of aspartate phosphorylation in

cyanobacteria and chloroplast response regulators (Jacobs, Connell, et al. 1999, Li and
Kehoe 2005, Maeda, Sugita, et al. 2006, Ruiz, Salinas, et al. 2008), we hypothesized that

perhaps ARC6 might conditionally interact with AtMinE as a result of posttranslational

modification. We generated site-directed mutants of ARC6 A A] 54-340, coding for D205A
and D205E amino acid changes. In no instance did ARC6 interact with AtMinEAA34-

229 using two hybrid assays (not shown), suggesting that these proteins probably do not

interact in Arabidopsis, despite their similar mutant chloroplast and FtsZ morphology

193

phenotypes (Glynn, Miyagishima, et al. 2007, Itoh, Fujiwara, et al. 2001, Pyke,

Rutherford, et al. 1994, Rutherford 1996, Vitha, Froehlich, et al. 2003). Taken together,

we conclude that the mispositioned Z-rings observed in the arc6 D 20 5 N background occur

through a MinE-independent mechanism, but the localization of all of the Min system

components needs to be examined in the arc6 D 20 5 N background.

ARC 6 probably fimctions downstream of F tsZI/F tsZZ assembly and downstream of
AtMinE.

Despite its lack of ARC6 interaction with AtMinE, our FtsZ immunolocalization

data for arc6 D 20 5 N still suggest that ARC6 might be somehow connected to the plastidic

Min system, as arc6 D 20 SN mutants clearly possess mispositioned Z-rings (Figure 6.3 B).

To confirm where ARC6 functions within the pathway of plastidic Z-ring formation, we
introduced an ARC6-GFP transgene (Glynn, F roehlich, et al. 2008) into ftsZZ mutants
(Schmitz, Glynn, et al. 2009) and arc12/atminE (Glynn, Miyagishima, et al. 2007) to
examine its localization in these backgrounds; we cannot use traditional tests of epistasis
in this case because the arc6, arc12, and ftsZ2 phenotypes are not easily distinguished
from one another. In both cases, we were unable to observe ARC6-GP P localization to
organized structures (Figure 6.6), though we were able to confirm expression of the
fusion protein in all cases by immunoblotting (data not shown). The absence of ARC6-

GFP rings was probably not a consequence of plastid size, as the enlarged plastids

194

 

 

 

 

 

Figure 6.6. ARC6-GFP does not localize to rings in ftsZZ or arc12 mutants. Shown is
preliminary ARC6-GFP localization in Col-0 (A); an fisZZ mutant; and (C) an arc12
(atminE) mutant. The background of the mutants shown above is Col-0. The ftsZ2
mutant carries homozygous T-DNA insertions in the loci for F tsZZ-I and F tsZZ-Z
(Schmitz, Glynn, et al. 2009). Scale bar = 5 pm.

195

of pdv2 mutants still form ARC6-GF P rings (Glynn, Froehlich, et al. 2008). These
results indicate that at least some degree of FtsZ assembly is required for localization of
ARC6 to mid-plastid rings in vivo; when FtsZ is either depleted (as in ftsZZ mutants) or
filament assembly is constituitively inhibited (as in atminE/arcIZ mutants), ARC6 does
not form a ring. From these results, we conclude that the plastidic Min system, or at least
AtMinE, functions upstream of ARC6 and that some degree of F tsZ assembly is required

for ARC6 localization to a ring in vivo.

Discussion

Here we perform preliminary characterization of a novel allele of ARC6,

arc6 D 20 5N: that possesses defects in chloroplast morphology and Z-ring placement that

appear somewhat similar to Arabidopsis min mutants. The precise connection to the Min
system was initially unclear, so we looked for interaction with a Min system component,
AtMinE, whose mutant phenotype is strinkingly similar to arc6 mutants. While we were

unable to detect any interaction between ARC6 and AtMinE, nor provide further

indication of the basis for the arc6 D 20 5N phenotype, our preliminary results do shed light

on a frequently-encountered question relating to plastid division: does FtsZ ring
formation precede ARC6 localization or does ARC6 localize independently of the Z-
ring? Our observations of ARC6-GFP in an fisZZ depletion mutant show that FtsZ is
absolutely required for ARC6 localization to the division site. Further, our examination
of ARC6-GFP localization in arc12 (an AtMinE loss-of-fimction allele) shows that the

presence of FtsZ protein within the plastid is insufficient for ARC6 localization;

196

chloroplast polarity and probably some degree of FtsZ polymerization are required for

ARC6 localization to the mid-plastid.

While the connection of ARC6 to the plastidic Min system does not appear to
occur through AtMinE, it is possible that another Min system component interfaces
ARC6 with the Min system. In three very recent studies, novel components or novel
localization patterns of known components have come to light. Similar to ARC6, native
AtMinD protein was shown to localize to mid-plastid rings in Arabidopsis (N akanishi,
Suzuki, et al. 2009), in contrast to the polar localization of AtMinD when overexpressed
in mature tobacco leaf cells (Maple, Chua, et al. 2002). It is possible that ARC6
conditionally interacts with AtMinD, but previous work suggests that it does not (Maple,
Aldridge, et al. 2005) and the plastid phenotypes of arc6 and atminD loss-of—function
mutants are quite different (Fujiwara, Hashimoto, et al. 2008, Fujiwara, Nakamura, et al.
2004, Vitha, Froehlich, et al. 2003). A novel division protein, MCDl, was recently
shown to be a modifier of the plastidic Min-system and possesses partial equatorial
localization within the plastid (Nakanishi, Suzuki, et al. 2009), similar to ARC6 (V itha,
Froehlich, et al. 2003). However, the precise topology and function of MCDl is still
unclear, making its involvement with the Min system (and possibly ARC6) difficult to
address. Most recently, PARC6 (a Paralog of ARC6) was shown to play a role in
inhibiting Z-ring assembly and positioning, possibly through another Min system
component, ARC3 (Glynn, Yang, et al. 2009, Zhang, Hu, et al. 2009b). Like ARC6,

PARC6 localizes to a ring during plastid division (Glynn, Yang, et al. 2009) and

PARC6AA77-355 interacts with ARC6 A A1 54-340, possibly through the conserved motifs

197

(Figure 6.2) harboring PARC6D211/ARC6DZO5 (Y. Yang and K. Osteryoung,

unpublished). Consistent with this, both ARC6 (Maple, Aldridge, et al. 2005) and
PARC6 (Zhang, Hu, et al. 2009b) have been shown to be capable of self-interaction.
Furthermore, a recent study has shown that Cdv3, a cyanobacterial protein with some
sequence similarity to a portion of the stromal domain of PARC6, binds the cytosolic

region of Ftn2/ARC6 in bacterial two hybrid assays (Marbouty, Saguez, et al. 2009).
These observations indicate that the arc6 D 20 5 N mutant may lead to a change in affinity
between ARC6 and PARC6 (a component of the plastidic Min system), possibly

explaining the mislocalization of the Z-ring observed in arc6 D 20 5N mutants (Figure 6.3).

Alternatively, the irregular position of the Z-ring in arc6 D 20 5N mutants may somehow be

a side-effect of low ARC6 protein levels (Figure 6.4). Regardless, further work that

analyzes the ARC6-PARC6 interaction and the impact of site-directed mutations at or

near ARC6Dzo5 might provide insights into the intriguing division defects observed in

the arc6 D 20 SN mutant.

In the course of this work, we also analyzed the localization of AtMinE-YFP in

Arabidopsis. The AtMinEpm-AtMinE- YFP transgene was able to complement arc12, an
AtMinE loss-of-function mutant (Glynn, Miyagishima, et al. 2007), suggesting that the
transgene is functional and the localization pattern observed is relevant. Unlike the polar

spot localization of AtMinE-YFP observed in mature tobacco leaf cells (Fujiwara,

Nakamura, et al. 2004), we observed no YFP signals that colocalized with chlorophyll

198

autofluorescence and no obvious expression of AtMinE-YFP in expanded Arabidopsis
leaf tissue. Instead, we observed a diffuse AtMinE-YFP signal associated with very
small plastids within juvenile leaves of Arabidopsis, consistent with the expression of
AtMinE within and proximal to the shoot apical meristem (Itoh, Fujiwara, et al. 2001).
From our analysis (Figure 6.5), it was not possible to determine if this localization pattern
was associated with the inner enve10pe membrane or if it was randomly distributed
throughout the stroma. From these results, we conclude that the native localization
pattern of AtMinE is diffuse within the plastid, and not polar, being more similar to the
localization of E. coli MinE to a broad zone within the cell (Hale, Meinhardt, et al. 2001,
Sun and Margolin 2001) rather than to polar spots observed during transient
overexpression assays in tobacco leaf cells (Itoh, Fujiwara, et al. 2001). Regardless,

analysis of the localization of Min system components (ARC3, MinD, MinE, MCDl ,

PARC6, and ARC6) in the arc6 D20 5N background is still required. In a related line of

investigation, we showed that ARC6-GFP fails to localize to a ring in ftsZZ or arc12
(atminE) mutants, suggesting some basal level of FtsZ assembly must occur for ARC6 to
localize to a ring; this presumably means that ARC6 acts downstream of the plastidic Min
system. Further analysis of the interactions and localization patterns of proteins that
make up the plastidic Min system will be critical to understanding the how the site for Z-
ring assembly is selected and what conditions trigger bulk FtsZ assembly at any

particular site.

199

Materials and Methods

Identification of EMS-induced polymorphisms in ARC6 and genotypic analysis.

Single nucleotide polymorphisms (SNPs) that cause missense mutations were
identified as described previously through the Arabidopsis TILLING Project:
hgp://tilling.fl1crc.org[ (Till, Colbert, et al. 2006). Progeny of mutagenized lines that
harbored the mutation indicated were sown on Linsmaier-Skoog medium and screened by
CAPS or dCAPS-based genotyping (Neff, Neff, et al. 1998). Primers used for screening
the mutant pool for mutations in the ARC6 amino terminus were: CCTCCGATTCCTCC
TCCTCCTCCT (left) and GACAACCCCTGCCAACCAGGTTTC (right). Primers used
for screening the mutant pool for mutations in the ARC6 amino terminus were:
GGCAGGGGTTGTCTTTCCTAGGTTCAG (left) and GATCAAGGAAAAGGGTG
TGCCAAGAAC (right). The polymorphisms in each line are as follows (nucleotide
positions are indicated relative to the primer start position for each TILLING screen:
P858 (C96T); P87L (C103T); D205N (G538R); SZSON (G761A); G293D (G890R);
L309F (C937T); A328V (C995Y); A592T (G377A); A621V (C465T); R677K (G922A);
E705K (G1005A); D708N (G1105A); S763F (C1271T); and T7731 (C1301T). None of
the seeds from the line harboring the A328V mutation could be germinated, but this may

be due to a background mutation.
Phenotypic analysis of plants carrying TILLING-derived missense alleles of ARC6.

Individuals that were shown to be homozygous for the indicated mutation were

examined for chloroplast morphology and compared to the parental TILLING line (C01-

200

er105), using established fixation and analysis protocols (Miyagishima, Froehlich, et al.

2006, Pyke and Leech 1991, Vitha, Froehlich, et al. 2003).

Multiple sequence alignment.

These sequences were used for multiple sequence alignment: Arabidopsis
thaliana (AtPARC6, NP_188549); Medicago truncatula (MtPARC6, IMGAG-annotated
pseudomolecule AC157350_28.4); Oryza sativa cv. japonica (OsPARC6,
NP_001054252); Populus trichocarpa (PtPARC6, a GENSCAN-based prediction (Burge
and Karlin 1997) of Populus trichocarpa genome Scaffold_122 204390-215000); Vitis
vinifera (V vPARC6, CAO48483); Arabidopsis thaliana (AtARC6, NP_199063); Oryza
sativa cv. japonica (OsARC6, NP_001045726); Ostreococcus lucimarinus (OlARC6,
XP_001421185); Physcomitrella patens (PpARC6, XP_001778770); Zea mays
(ZmARC6, ACG29776); Thermosynechococcus elongatus BP-l (TeFtn2, BAC08309);
Synechococcus elongatus PCC 7942 (SeFtn2, ABB57973); Nostoc sp. PCC 7120
(N sFtn2, BAB74406); Crocosphaera watsonii WH 8501 (CwFtn2, EAM48783); and
Cyanothece sp. ATCC 51142 (CsFtn2, ACB49642). Multiple sequence alignment was

performed with MEGA4 (Tamura, Dudley, et al. 2007) using an identity matrix.

F tsZ immunolocalization.
Tissue preparation, fixation, and immunofluorescence analysis were carried out as
described previously (Miyagishima, Froehlich, et al. 2006, Vitha, Froehlich, et al. 2003,

Vitha, McAndrew, et al. 2001).

201

Immunoblotting.

Plant extracts for determination of F tsZ and ARC6 protein levels were taken from
floral bud tissue. Flower buds were ground in liquid nitrogen and prepared for 10% SDS-
PAGE using 6X sample buffer according to previously-established protocols (Wiegel and
Glazebrook 2002). Immunoblotting for Ftle and FtsZZ-l protein was performed as
described previously (Stokes, McAndrew, et al. 2000). Immunoblotting of plant extracts
was performed using an ARC6-specific antibody at 1:2500 in TBS-T containing 5%
nonfat dry milk. The blot was washed several times in TBS-T before applying the
secondary antibody. The anti-rabbit HRP-conjugated secondary was used at 1:5000 in
TBS-T containing 5% nonfat dry milk and the blot was washed several times in TBS-T
before applying the HRP chemilluminescent substrate (Therrno Scientific, Inc.) and

exposing to film.

Yeast two hybrid analysis.

ARC6 and AtMinE clones for two hybrid analysis were made by PCR and cloned
into pGADT7 and pGBKT7, respectively. Mutations causing the missense mutations
D205A and D205E in ARC6 were generated by SOE-PCR and cloned into pGADT7.
Growth assays (HIS3 reporter assays) were performed according to methods as described

previously (Glynn, Froehlich, et al. 2008, Maple, Aldridge, et al. 2005).

202

Generation of AtMinEpm-AtMinE—YFP, transformation, selection, and microscopy.

Analysis of ARC 6—GFP localization.

The AtMinEpm-A tMinE- YFP transgene was generated by amplifying the promoter

and coding regions of the AtMinE locus with primers TTTT'TTTCTGCAGACTTGTTT
CAAAACGACTGTGTTTTTTG and CCAGAGAGATCTCTCTGGAACATAAAAATC
GAACCTGACATC by PCR. This fragment was cloned into a derivative of pCAMBIA-
1302 (Hajdukiewicz, Svab, et al. 1994) carrying C-terrninal EYFP coding sequence by
digesting the vector and PCR products with Pstl and BglII; digestion with Pstl and BglII
removes the 35S promoter from the plant transformation vector. Clone was sequenced
and shown to be free of coding errors prior to transformation of Agrobacterium and
Arabidopsis. Plants were transformed as described previously (Clough and Bent 1998).
Selection of transgenic individuals was performed on Linsmaier-Skoog media containing
hygromycin (25 ug/mL). Hygromycin-resistant lines were transplanted to soil and
allowed to recover for several days prior to analysis. Analysis of ARC6-GFP localization
in arc12 and ftsZZ-I ftsZZ-Z double mutants was conducted using GFP fusion vectors as
previously described (Glynn, Froehlich, et al. 2008, Vitha, Froehlich, et al. 2003) and
selecting on Linsmaier-Skoog media containing hygromycin (25 ug/mL). Hygromycin-
resistant lines were transplanted to soil and allowed to recover for several days prior to

microscopic analysis.

203

Acknowledgements

I thank Mia Hemmes for backcrossing the arc6 020 5N TILLING mutant, Joyce

Bower for assisting with quantitative analysis of the arc6 D 20 5 N phenotype, and Deena

Kadirjan—Kalbach for continuing technical assistance with this project.

204

Chapter 7

Conclusions and Future Directions

205

Summary of ARC6 and PARC6 Functional Analysis and Future Directions.

In Figure 7.1, I show a schematic detailing the functional domains we have

characterized within ARC6 and PARC6, in addition to previously annotated features.

ARC6-FtsZZ Interaction
Our body of work here identifies a prospective domain within ARC6 that binds
the C-terminus of FtsZ2 family members (Figure 7.1, ZBD) and might have structural

similarity to E. coli ZipA. While our pulldown assays and in vivo analysis of
ARC6F44ZD site-directed mutants negates the criticality of the core hydrophobic
phenylalanine (F442) of ARC6ZBD we had hypothesized based on two-hybrid results,

rigorous structural analysis is still required to fully refiite the proposed ZipA-like

structure we show in Chapter 2. While an X-ray crystal structure or NMR-based solution

structure of ARC6ZBD is highly preferred, a comparison between circular dichroism

(CD) spectra from ARC6ZBD and ZipAZBD may be sufficient for a preliminary test of

the model we show in Figure 2.4 (Greenfield 2006, Whitrnore and Wallace 2008). If

ARC6ZBD and ZipAZBD have similar CD profiles, it is reasonable to assume that their

structures are probably similar. However, if their CD profiles are different, it would

suggest that ARC6ZBD has a structure distinct from that of ZipAZBD; this result would
be consistent with secondary structure predictions of ARC6ZBD (Figure 2.5) and in vivo

analysis of ARC6ZBD in Arabidopsis (Figure 2.9 and Figure 2.13). However, we still

206

 

ARC6 801 AA

1.... .1 m:

 

 

 

 

\TM
llimcrimtion‘.’
/ M1 {TM2
1 I a I I
— a a; aw; .2:- , in
ARC3 Interaction l‘JEVI Interaction

PARC6 819 AA

 

 

 

Figure 7.1. Domain Architecture of ARC6 and PARC6 Proteins. ARC6 (top) has a
transit peptide (TP, AA 1-67); a predicted J-domain (J, AA 89-153); a hypothesized
dimerization domain (AA 154-340) containing a conserved aspartate at position 205 (red
dot); an FtsZZ-binding domain (ZBD, AA 351-503); a transmembrane domain (TM, AA
615-635); and a PDV2-binding domain (PBD, AA 636—759) containing a serine residue
at position 744 that influences ARC6-PDV2 interaction and is predicted to be a
phosphoacceptor site (asterisk, *). ARC6AA68_614 resides in the chloroplast stroma and

ARC A A636-801 resides within the intermembrane space. PARC6 (bottom) has a transit
peptide (TP, AA 1-76); an ARC3-interacting region (AA 77-357); a hypothesized
dimerization domain (AA 169—335) containing a conserved aspartate at position 211 (red
dot); two predicted transmembrane domains (TMl, AA 357-377; TM2, AA 574-596);
and a PDVl-interacting region (AA 575-819). PARC6 A A77-3 56 is stromal, but the C-
terminus of the protein has yet-undefined topology within the chloroplast.

207

emphasize that outside of its affinity for the C-terminus of FtsZ2 proteins, the precise

function of ARC6ZBD remains undefined, but hypothesize that this domain is somehow

involved in assembly or stabilization of FtsZ polymers within the chloroplast. To

determine if ARC6ZBD has a direct effect upon the polymerization of Ftle and FtsZ2,

analysis of FtsZ copolymer assembly in the presence of ARC6ZBD could be conducted.

The effect of ARC6ZBD upon F tsZ assembly could be analde using light scattering

assays and electron microscopy following treatment of FtsZ assembly reactions

(containing Ftle and FtsZZ) with recombinant ARC6ZBD protein according to

previously established protocols (Lu and Erickson 1998, Olson 2008).

ARC 6-PD V2 Interaction

We have shown that ARC6 is responsible for positioning PDV2 at the division
site within dividing chloroplasts and that PDV2-positioning activity occurs within the
intermembrane space of the chloroplast (Glynn, F roehlich, et al. 2008). In Chapters 3
and 4 we showed that the ARC6-PDV2 interaction only requires a small domain
contained within the IMS-localized portion of the ARC6 protein (Figure 7.1, PBD).
While we were unable to generate a contiguous structural model of this domain within
ARC6, the structural similarity of the IMS-localized region of PDV2 to the Rst
phosphoprotein phosphatase (Figure 4.3) (Delumeau, Dutta et al. 2004, Dutta and Lewis
2003, Hardwick, Pane-Farre, et al. 2007) led us to hypothesize that interaction between
ARC6 and PDV2 might be controlled by differential phosphorylation of ARC6. Using

multiple sequence comparison of ARC6/F m2 proteins and phosphorylation prediction

208

algorithms, we identified one high-confidence phosphoacceptor site (S744) and two

lower-confidence phosphoacceptor sites (S740 and T742) within ARC6pBD.

Phosphomimetic mutations at one of these sites (ARC6S744E) disrupts the ARC6-PDV2

interaction (Figure 4.5) and impedes chloroplast division (Figure 4.6 and Figure 4.7), but
does not grossly alter PDV2 localization to the division site (Figure 4.8). The presence of

dumbbell-shaped plastids and multiple Z-rings at the division site within chloroplasts of

ARC6S744E-expressing lines are strikingly similar to those reported for both pdv2 and

arc5 mutants (Gao, Kadirjan-Kalbach, et al. 2003, Miyagishima, Froehlich. et al. 2006).

In other words, the ARC6S744 E phosphomimetic mutation generates two effects: (1)

decreased dynamin activity at the outer envelope membrane and (2) an increase in Z-ring

assembly within the stroma. While it is unclear if the ARC6S744 E mutation simply

impedes dynamin activity by altering ARC6-PDV2 interaction affinity or by upregulating
Z-ring assembly within the stroma (thus creating a rigid physical barrier that blocks
constriction of the plastid membranes), the change of a single residue within the IMS-
localized portion of ARC6 certainly leads to upregulation of Z-ring assembly within the
stroma. However, it remains to be seen whether ARC6 residues S740, T742, or S744 are
actually subject to posttranslational modification in vivo. There is no evidence of
phosphorylation of any of these residues based on current phosphoproteomic data
provided in the PhosPhAt database (hfip:flphosphatmpimp—golm.mpg.de[) (Heazlewood,
Durek, et al. 2008), but this might reflect the transient nature of phosphorylation of the
ARC6 protein in vivo — perhaps only a small fraction of ARC6 molecules are

phosphorylated at these sites at any one time or are only phosphorylated under certain

209

environmental or developmental conditions. To address this potential mechanism further,

it is suggested that in vitro assays are conducted in which radiolabeled phosphate is used

to detect phosphorylation and/or dephosphorylation of ARC6S744 following treatment of

recombinant ARC6pBD with plant cell or chloroplast extracts. If ARC6pBD is a substrate

for a phosphatase and/or kinase, a change in the amount of radioactivity should be
observable following treatment of recombinant proteins with plant cell extracts; the
amount of radiolabel carried by S740A, T742A, and/or S744A site-directed mutants
could be compared to that of the wild type recombinant protein to demonstrate that these
residues are actually the target of a phosphatase and/or kinase enzyme contained within
cell extracts. Similar methods have been utilized for other proteins (Ben-Nissan, Cui, et

al. 2008, Xu, Wong, et al. 2008) and may provide insight into the validity of the

hypothesis that ARC6 activity is controlled through phosphomodification of ARC6pBD.

PARC 6-ARC3 Interaction.

In Chapter 5, we introduced PARC6, a novel chloroplast division gene that
probably arose as a result of ARC6 duplication in the tracheophyte lineage. While ARC6
and PARC6 do share some degree of sequence similarity, they vary significantly in
function. Unlike ARC6, which promotes Z-ring assembly (V itha, Froehlich, et al. 2003),
PARC6 appears to inhibit F tsZ assembly within the stroma (Figure 5.6) and does not
directly interact with Ftle or F tsZZ (Figure 5.7). However, we showed that PARC6
interacts with ARC3, a protein with functional similarity to bacterial MinC and that this

interaction requires the MORN region of ARC3 (Figure 5.8) — the MORN region of

210

ARC3 inhibits its interaction with F tle proteins (Maple, Vojta, et al. 2007). Based on
these data, I hypothesize that PARC6 inhibits FtsZ assembly through ARC3: when
PARC6 is present, it sequesters the MORN region of ARC3, allowing ARC3 to bind
Ftle and disrupt the Ftle/FtsZ2 copolymer that makes up the Z-ring; in the absence of
PARC6, the MORN region of ARC3 is exposed and inhibits ARC3 interaction with

F tle, allowing for assembly of the Ftle/FtsZ2 copolymer. Furthermore, because of the
. high PARC6-YFP signal intensity at the isthmus connecting the dividing plastid during
the late stages of division (Figure 5.4) and because parc6 mutants commonly have
chloroplasts that are dumbbell-shaped (Figure 5.3 and Figure 5.9), I hypothesize that
PARC6 functions to aid disassembly of the Z-ring during the late stages of division. The
plastid morphology phenotypes in parc6 mutants could result from disorganized or
inefficient Z-ring disassembly in the absence of PARC6, causing hyperstabilized Z-
ring(s) to block the final stages of plastid division. To determine whether this model is
correct, in vitro assays could be conducted using purified recombinant proteins to
monitor the assembly of the Ftle/FtsZ2 copolymer (Lu and Erickson 1998, Olson 2008)
in the presence of various domains of ARC3 and PARC6. For this example, I
hypothesize that recombinant ARC3 lacking its transit peptide and MORN region would
be capable of inhibiting assembly of the Ftle/FtsZZ copolymer. In contrast, a
recombinant form of ARC3 protein lacking only its transit peptide would have no effect
on Ftle/FtsZ2 copolymer assembly due to the presence of the MORN region of ARC3,
as this region of ARC3 inhibits interaction between ARC3 and Ftle (Maple, Vojta et
al. 2007). The addition of the N-terminal stromal region of PARC6 to assembly assays

containing Ftle/FtsZ2 and mature recombinant ARC3 protein would lead to

211

sequestration of the MORN region of ARC3, thereby allowing ARC3 to bind F tle and
inhibit assembly (or perhaps promote disassembly) of the Ftle/FtsZ2 copolymer, as
bulk FtsZ polymer assembly only occurs in the presence of both FtsZ] and FtsZZ protein
(Olson 2008). The use of complex, multiprotein mixtures in assembly assays will be

critical for determining how plastidic FtsZ assembly is regulated in vivo.

PARC6-PDV] Interaction. .

Another possible explanation for the common observation of dumbbell-shaped
plastids in parc6 mutants might be a defect in dynamin (ARC5) activity at the outer
envelope. Previously, PDVl was shown to be critical for full ARC5 pinchase activity at
the outer envelope (Miyagishima, Froehlich, et al. 2006). In Chapter 5, we demonstrated
that PARC6 is required for PDVl localization to sites of constriction (Figure 5.10) and
that PARC6 interacts with PDVl in two-hybrid assays (Figure 5.11), suggesting that
PARC6 might directly position PDVl during plastid division, thereby influencing ARC5
pinchase activity by directly recruiting PDVl to the division site. However, this
hypothesis is somewhat cursory due to the lack of knowledge regarding the topology of
PARC6. Our preliminary analysis indicated that PARC6 may only traverse the inner
envelope membrane (Glynn, Yang, et al. 2009), but this experiment was preliminary and
may not possess the sensitivity to detect multiple or transient topological orientations of
PARC6 within the envelope membrane(s) in dividing chloroplasts; a complete diagnostic
of PARC6 topology needs to be examined using protease protection assays to determine

the validity and/or plausibility of an interaction between PARC6 and PDVl.

212

Furthermore, we also observed the persistence of a PARC6 spot at one pole of the
chloroplast (Figure 5.4), presumably representing its localization following separation of
the two daughter plastids. Similar localization patterns have been observed for PDVl
and ARC5, but not for PDV2 or ARC6. We suspect that the similarity of PARC6, PDVl ,
and ARC5 at this stage represents a multiprotein complex that is retained on one pole of
the chloroplast, possibly serving as a marker or reference point to aid division-site
selection during the next round of division, perhaps analogous to the way some yeast
species maintain cell polarity following cell division (Amberg, Zahner, et al. 1997,
Drubin 1991, Glynn, Lustig, et al. 2001, Mine, Bratrnan, et al. 2009). Isolation of this
putative polarity-determining complex from chloroplasts using affinity chromatography
of affinity-tag-labeled PARC6 folowed by LC-MS/MS might allow for the identification
of the factors composing this complex and provide insights into how polarity might be

established within daughter plastids following division.

A RC 6 is a Component of the Plastidic Min System.

Historically, I had thought of the plastidic Min system and ARC6 as two separate
entities with little or no connection between their apparent functional roles. However, as
a result of two major observations, 1 now believe ARC6 function is integrated into the

plastidic Min system that firnctions to position the Z-ring at the division site: (1)

expression of ARC65744 E in vivo results in the production of elongated chloroplasts

containing multiple sites of constriction within petiole cells (Figure 4.9), similar to
phenotypes observed in petiole cells of Arabidopsis atminD/arc] I loss-of-function

mutants (Fujiwara, Hashimoto, et al. 2008, Fujiwara, Nakamura, et al. 2004) and 35S-

213

AtMinE overexpressing plants (Reddy, Dinkins, et al. 2002); and (2) the minicelling-like
phenotype and aberrant Z-ring position in arc6 D 20 5 N mutants (Figure 6.3) is consistent

with a role for ARC6 in promoting equatorial positioning of the Z-ring prior to
chloroplast division. To further examine this phenomenon, it is suggested that the
localization patterns of other Min system proteins (AtMinD, AtMinE, ARC3, and
PARC6) are examined in these mutant backgrounds to determine if the D205N or S744E
mutations somehow impact function of the plastidic Min system. Additionally,
preliminary data shows that a conserved segment of the stromal regions of ARC6 and
PARC6 family members (Figure 5.1, grey underline) is important for homo- or hetero-

dimerization between these proteins (Y. Yang and K. Osteryoung, unpublished). The

conserved aspartate residues within these segments of ARC6 (ARC6D205) and PARC6

(PARC69211) might be critical for their ability to dimerize within the chloroplast stroma

(Maple, Aldridge, et al. 2005, Zhang, Hu, et al. 2009a) and/or mediate their
communication with other components of the plastidic Min system. Interestingly, the
cyanobacterial Cdv3/DivIVA division protein (Marbouty, Saguez, et al. 2009,
Miyagishima, Wolk, et al. 2005) also shares some degree of sequence similarity with

PARC6 (J .M. Glynn and K. Osteryoung, unpublished) and like both parc6 and

arc6 D 20 5 N mutants, Cdv3 -deficient cyanobacteria possess ectopic Z-rings (Marbouty,

Saguez, et al. 2009), suggesting that dimerization of ARC6 with PARC6 (or Ftn2 with
Cdv3 in the case of cyanobacteria) serves to interface ARC6/Ftn2 with the Min system.
Consistent with this idea, the loss of Cdv3/DivIVA from the plastid divisome during the

evolution of land plants (J .M. Glynn and K. Osteryoung, unpublished) loosely

214

corresponds with the acquisition of PARC6 in vascular plants (Glynn, Yang. et al. 2009).

Further analysis of the interaction properties and localization of site-directed mutants of

ARC6D205 and PARC69211 may provide key insights into the coordination of Z-ring

placement with organized Z-ring assembly and disassembly in chloroplasts.

Unanswered Questions, Additional Observations, and New Hypotheses.

Does ARC 6 function as a DnaJ-like cochaperone?

While shown to have sequence and probable structural similarity to DnaJ -1ike
cochaperones (Vitha, Froehlich, et al. 2003), it remains to be seen whether the predicted
J-domain of ARC6 actually stimulates the chaperone activity of a DnaK-like protein.
Moreover, it is unclear as to what function this chaperone activity serves within the

chloroplast.

One possibility is that ARC6 is required for efficient import and/or refolding of
FtsZ following import into the chloroplast stroma. Two points support this hypothesis:
DnaJ/DnaK type chaperone systems have been shown to be important for protein import
into eukaryotic organelles (Ivey and Bruce 2000, Zhang, Elofsson, et al. 1999); and
intraplastidic FtsZ protein levels are significantly diminished in arc6 mutants (Figure 1.3)
(V itha, F roehlich, et al. 2003). It is reasonable to suggest that comparative import
experiments might highlight such an activity; perhaps higher efficiency FtsZ protein
import would be observed in the presence of increasing amounts of ARC6 protein if

ARC6 contributes to FtsZ import and/or folding.

215

A second possibility is that the J -domain activity is important for chaperoning the
assembly or disassembly of FtsZ proteins into ordered structures within the stroma.
Bacterial dnaK and dnaJ mutations affect cell division (Clarke, Jacq, et al. 1996,
McCarty and Walker 1994), FtsZ dynamics (Uehara, Matsuzawa, et al. 2001), and
overexpression of FtsZ has been shown to complement the cell division defects observed
in a dnaK mutant (Bukau and Walker 1989). Furthermore, a recent study has suggested
that DnaK chaperone activity is important for proper localization of the Z-ring in E. coli
(Sugimoto, Saruwatari, et al. 2008). Presumably, the role of chaperones in Z-ring
assembly and cell division in bacteria is conserved in plastids. Unfortunately, the
analysis of plastidic hsp 70 T-DNA mutants has revealed partial redundancy between the
two Arabidopsis chloroplast Hsp70 proteins and implicated them in chloroplast
biogenesis in vivo; mutants that lack both plastidic Hsp70 proteins are non-viable (Su and
Li 2008) and as a result are probably uninformative with regard to the role of these
proteins in FtsZ import or assembly in vivo. However, the analysis of J -domain mutants
of ARC6 is currently underway (Y. Yang and K. Osteryoung, unpublished) and may

provide some exciting insights into this aspect of ARC6 function.

Is ARC 6 localized to thylakoid membranes?

Interestingly, analysis of proteomic data from plastid membrane fractions
suggests that the localization and topology of ARC6 might be more complex or more
dynamic than import and fractionation assays suggest. The initial analysis of ARC6

fractionation and t0pology was performed using an in vitro approach with isolated pea

216

chloroplasts; this work placed ARC6 within the inner envelope membrane (V itha,
Froehlich, et al. 2003). However, more recent work indicates that ARC6 is present
within the inner envelope and also the thylakoid membrane system, as multiple ARC6
peptide fragments are found in isolated thylakoid membrane fractions
(http://ppdb.tc.comell.edu/) (Sun, Zybailov, et al. 2009). While it is possible that these
ARC6-derived peptides simply represent contamination of the thylakoid membrane
fraction, this experimentally-determined thylakoid fractionation of ARC6 might represent
a novel function for ARC6 during the early stages of division, or perhaps occurs as a
result of the process of membrane breakage and resealing during the final stage of
daughter plastid separation. It is unclear as to whether the thylakoid membrane is
extricated from the division furrow prior to the final pinch, or if portions of the thylakoid
membrane fuse with the inner envelope. Such a fusion event could lead to temporary
relocation of ARC6 from the inner envelope to the thylakoid membrane, but the fate

and/or purpose of these thylakoid-localized ARC6 molecules is unclear.

How is aynamin (ARC5) recruited to the division site?

While the roles of ARC6 and PARC6 in recruiting the PDV proteins to the
division site are becoming more transparent, it is still unclear how the PDV proteins
might recruit dynamin (ARC5) to the division site. Interestingly, pdvl pdv2 double
mutants are blocked in their ability to recruit ARC5 from cytosolic patches, but pdvl and
pdv2 mutants can localize ARC5 at the division site (Miyagishima, Froehlich, et al.
2006). Similarly, arc6 mutants are defective in their ability to recruit ARC5 patches from

the cytosol (Glynn, Froehlich, et al. 2008) and phenocopy pdvl pdv2 mutants in this

217

regard (Miyagishima, Froehlich, et al. 2006). While PDVl and PDV2 localize to the
chloroplast periphery in arc6 mutants, neither of the PDV proteins localize to a central
ring in the arc6 background (Glynn, Froehlich, et al. 2008); this suggests that PDVl
and/or PDV2 must be concentrated at a single site within the outer envelope to facilitate
dynamin recruitment to the dividing chloroplast. Curiously, neither PDVl (Miyagishima,
F roehlich, et al. 2006) nor PDV2 (M. Hemmes and J .M. Glynn, unpublished) interact
with ARC5 in yeast two-hybrid assays. I have also constructed novel N-terminal BiF C
vectors to test for identify interaction between PDVl-ARCS, PDV2-ARC5, PDVl-
PDV2, PDVl-PDVl, PDV2-PDV2, and ARC5-ARC5 — but these experiments have yet
to be performed. Collectively, these data hint at a factor (or factors) that bridge PDV

proteins to ARC5, and aid ARC5 patch recruitment from the cytosol.

The N-terminal domains of PDVl and PDV2 are both localized in the cytosol,
exhibit some low-level sequence similarity with each other, and have coiled-coil regions,
indicating that they might have related functions (Glynn, Froehlich, et al. 2008,
Miyagishima, Froehlich, et al. 2006). Unfortunately, the PDV proteins of land plants
share little sequence similarity to any other proteins (Miyagishima, Froehlich, et al.
2006). To gain further insight into the proteins or entities that might bridge the PDV
proteins with ARC5 recruitment, I performed comparative structural analysis of the N-
terrninal domains of PDVl (AA 1-206) and PDV2 (AA 1-213) to all extant PDB

structures using the MetaServer at http://bioinfo.pl (Ginalski, Elofsson, et al. 2003).

218

The best match for PDVl A A1-206 was a crystal structure of the C-terminus of

human E31 (1 YIB), a protein domain with homodimeric fold comprised of a coiled-coil
and four-helix bundle (Slep, Rogers, et al. 2005). E31 belongs to a conserved protein
family that localizes to microtubule plus ends and recruits cell polarity and signaling
molecules to growing microtubule tips (Askham, Moncur, et al. 2000, Bu and Su 2003,
Nakamura, Zhou, et al. 2001, Slep and Vale 2007, Wen, Eng et al. 2004). The N-
terrninus of EB] is required for binding microtubles (Barth, Siemers, et al. 2002, Hayashi

and Ikura 2003), while the C-terminus (i.e. the portion of EB] with structural similarity

to PDVl A A1-206) has been shown to be a cargo-recognition domain. The C-terminus of

EB] family proteins bind several cargo proteins that regulate microtubule dynamics,
including APC in mammals (Su, Burrell, et al. 1995), Kar9p in yeast (Miller, Cheng, et
al. 2000), and may be involved in crosslinking the microtubule and actin cytoskeletons

through spectraplakin-like proteins (Slep, Rogers, et al. 2005).

The best match for PDV2 A A1-213 was a crystal structure of the repetitive

elements of fruit fly Spectrin (2SPC). Spectrin is a cytoskeletal protein descended from
an a-actinin-like ancestor (Dixson, Forstner, et al. 2003). Spectrins enhance the
mechanical stability of membranes and promote the assembly of specialized membrane
domains (Broderick and Winder 2005, Kordeli 2000). Spectrin functions as a tetramer
(Liu and Palek 1980) that cross-links transmembrane proteins, membrane lipids, and

actin; this cross-linking activity occurs directly (Brenner and Korn 1979) or through

219

adaptor proteins such as ankyrin and 4.1 (Bennett and Baines 2001, Goodman, Krebs, et
al. 1988).

Dynamins do not generally interact directly with actin, but are typically associated
with membrane structures (Wiejak and Wyroba 2002). My preliminary structural
analysis of the PDV proteins point toward a mechanism where PDVl (and possibly
PDV2) might reorganize the actin and/or microtubule cytoskeleton to facilitate trafficking
of dynamin (ARC5) patches to the cytosolic side of the chloroplast outer envelope
through microtubule and/or actin-anchored motor proteins. Consistent with this
hypothesis, at least two kinesin-like proteins, At5g10470 and At4g38950, have PDVl-
like mRN A expression profiles (J .M. Glynn, unpublished) based on analysis using
Expression Angler (http://bbc.botany.utoronto.ca/) (Toufighi, Brady, et al. 2005).
Furthermore, the membrane remodeling properties of Spectrin-like proteins, perhaps
including PDV2, could create lipid microdomains that may be necessary (Lemmon 2004)
for attracting the pleckstIin homology domain of ARC5 (Gao, Kadirjan-Kalbach, et al.
2003, Hong, Bednarek, et al. 2003), distributing ARC5 around the organelle surface,
and/or facilitating stimulation of ARC5 pinchase activity at the division site (Yoshida,

Kuroiwa, et al. 2006).

220

Working Model of the Coordination of the Chloroplast Divisome in Tracheophytes.

In Figure 7.2, I show a rudimentary model that reflects our current understanding
of the roles of ARC6 and PARC6 in coordinating the assembly and operation of the
plastid divisome. The plastid PD rings (Kuroiwa, Kuroiwa, et al. 1998) are omitted from
this model, as their constituents are yet unknown and uncharacterized. This schematic
reiterates the complexity of the process of plastid division and hints at several new
hypotheses regarding the operation and regulation of some of the components that make
up the plastid divisome. Firstly, while ARC6 certainly plays a role in the assembly
and/or stabilization of the Z-ring, the mechanism by which it does so remains enigmatic.
Our attempt to identify the discrete region of ARC6 that binds FtsZ2 family members
within plant chloroplasts was a necessary step, but indicates that this issue may be more
complicated than previously thought. Beyond its role in regulating FtsZ assembly within
the stroma, we showed that ARC6 also mediates the activity of division factors that act
upon the outer envelope. ARC6 is required to position PDVl and PDV2 within the outer
envelope membrane. ARC6 positions PDV2 by direct interaction, but only influences
PDVl localization indirectly through its paralog, PARC6. Our genetic and cytological
analyses suggest that ARC6 is probably required for PARC6 localization and/or activity
during division. It remains to be seen if ARC6 interacts directly with PARC6 in vivo or if
other proteins bridge these two division factors. Furthermore, the precise topology of
PARC6 is still rather murky, making it difficult to determine the validity of PARC6-

PDVl interaction data. The interaction between PARC6-ARC3 and the similarity of

221

 

®®

/

Cfioskeletal reorganization, Motor Proteins,
Membrane Reorganization?

 

 
  
  
    

 
 
  

 

 

 

 

PDV2 ‘l" "it...“ u C
4; OEM
C C IMS
i IEM
ARC6 , PARC 6
Stroma

Z-ring Assembly Z-ring Disassembly

\/

Z-ring Dynamics

 

 

 

Figure 7.2. Coordination of Z-ring Dynamics and ARC5 Recruitment through
ARC6 and PARC6. PD rings are omitted from this model for simplicity. Division is
initiated by FtsZ assembly at the division site, which is determined by the Min system
(not shown). ARC6 may influence Z-ring positioning and stabilize the Z-ring at the
division site through its interaction with FtsZ2. PARC6 is a Min system component
hypothesized to inhibit Z-ring assembly through ARC3. ARC6 and PARC6 form
homodimers; ARC6 and PARC6 may (?) form heterodimers within the stroma. PDV2 is
positioned in the outer envelope by interacting with ARC6 in the IMS. ARC6 indirectly
positions PDVl in the outer envelope, but does so through PARC6. Preliminary data
suggests that PARC6 binds PDVl in the cytosol. It is unknown (?) if PDVl and PDV2
form hetero- or homodimers in vivo. The recruitment of ARC5 may occur through
cytoskeletal reorganization, kinesin-driven directed movement, and/or membrane lipid
reorganization by the PDV proteins. Not all detail is shown. See text for details.

222

their respective mutant phenotypes indicate that these two proteins might work together
to inhibit FtsZ assembly, perhaps contributing to Z-ring disassembly during the final
stages of division and/or aiding division-site selection along with other components of the

plastidic Min system. While previously thought to be distinct from the Min system,

ARC6 may also contribute to positioning of the Z-ring, based on analysis of ARC6S744 E

and arc6 020 5N alleles. Though the exact connection is unclear, ARC6 may interface

with the Min system through heterodimerization with its paralog, PARC6.

The recruitment and activity of dynamin (ARC5), both regulated by the PDV
proteins, remains a complicated matter. While both PDVl and PDV2 are presumed to be
required for full ARC5 contractile activity, either PDV protein is sufficient for ARC5
recruitment to the division site. ARC5 recruitment may occur through reorganization of
the cytoskeleton, directing the dynamin-like protein to the division site through a
vesicular transport mechanism that utilizes microtubule and/or actin-tracking motor
proteins like kinesins (microtubules) or myosins (actin). Further work on this aspect of
the division process may not only prove useful for improving our understanding of
plastid replication, but may provide novel insights into general aspects of dynamin

recruitment and activity within other types of eukaryotes as well.

223

APPENDIX

224

Appendix A

NEXT and CIENA: Pipelines for Identification of Plastid Division Genes.

225

Summary

Reverse genetic screens for plastid division factors have relied heavily upon gene
identification and characterization of division factors in other systems; when a new
division component is identified in E. coli or another prokaryote, a researcher typically
performs a BLAST query against finished plant genomes to determine if similar factors
might contribute to organelle division. Several plastid division genes have been
identified using this approach, including: FtsZ (Osteryoung, Stokes, et al. 1998,
Osteryoung and Vierling 1995), GCl (Maple, Fujiwara, et al. 2004), AtMinD (Colletti,
Tattersall, et al. 2000), and AtMinE (Itoh, Fujiwara, et al. 2001).' However, this mode of
gene identification is limiting and novel approaches are required to determine the
complete inventory of genes involved in organelle division. Here we highlight two
platforms that we tested in an attempt to identify new plastid division factors. These
platforms utilize genomic context and/or transcriptional profiling to identify new

candidate genes and assign them priority.

NEXT: A Platform for Identification of Candidate Genes by Genome Neighborhood
Analysis and Priority Ranking Based on Gene Expression and Intracellular

Targeting.

To start our search for new plastid division genes, we utilized The Seed gene

neighborhood analysis tool (http://www.theseed.org/mltiMain_Pag§) (Overbeek,

Begley, et al. 2005). In prokaryotes, conserved gene proximity can be indicative of

226

related firnctions, as genes linked to a similar cellular process are often clustered together
on the bacterial chromosome. Prokaryotic gene operons are a popular example of gene
clustering (Overbeek, Fonstein, et al. 1999), and gene-gene association approaches have
been used successfully to define the function of several genes of unknown function (de
Crecy-Lagard 2007, de Crecy-Lagard and Hanson 2007). We pinned known prokaryotic
cell division genes (Miyagishima, Wolk, et al. 2005) and then examined the surrounding
genomic sequence amongst several bacterial species for coding loci that frequently reside
within proximity (< 5000 bases) to a query gene. We identified approximately 80
candidates that commonly co-occur with our set of query genes. Within this set, 55 of
these had orthologous sequences in Arabidopsis using BLAST-based queries (Altschul,

Madden, et al. 1997).

We then used Genevestigator (hgpszflwwwgenevestigatorcomD (Zimmerrnann,
Hirsch-Hoffmann, et al. 2004) to prioritize our panel of ~55 candidates. Genevestigator
is a gene expression analysis tool that allows the user to compare the degree of
correlative expression between any two genes in Arabidopsis; we used ARC 6
(At5g42480) as a reference gene, as it is likely to be a critical regulator of plastid division
in vivo (Marrison, Rutherford, et al. 1999, Pyke, Rutherford, et al. 1994, Vitha,

Froehlich, et al. 2003), and arranged our candidates in order, based on the degree of

expression profile similarity to ARC6.

To further prioritize candidates coming out of the pipeline described above, we

used the TargetP (httgflwwwcbs.dtu.dl_</services/TargetP/) intracellular prediction

227

targeting algorithm (Emanuelsson, Nielsen, et al. 2000) to identify genes encoding
proteins bearing chloroplast transit peptides, as we guessed that chloroplast division
genes might be more likely to be inside the chloroplast stroma. We referred to this
complete pipeline as the NEXT (Neighborhood Analysis, Expression Analysis, and
Targeting Analysis) Functional Identification Platform. For our pilot screen, we elected
to examine chloroplast phenotypes in the 12 loci that had the highest-priority ranking.
We obtained T-DNA insertion lines for this set of 12 Arabidopsis candidate genes and
analyzed chloroplast morphology in each of these lines: At5g19850 (SALK_026840);
At5g40500 (SALK_029848); At3g25470 (SALK_043556); At5g10720 (SALK_051629);
At3g07430 (SALK_056049); At5g53920 (SALK_070621); At3g04870 (SALK_O79674);
At4g01900 (SALK_095650); At1g53120 (SALK_099429); At1g08530 (SALK_101302);
At5g14800 (SALK_127043); and At1g32440 (SALK_142845). Unfortunately, we were
unable to identify homozygous insertions in a few of these lines (SALK_026840,
SALK_029848, SALK_101302, and SALK_142845) and we did not observe any lines in
this candidate set with noticeable defects in chloroplast morphology. From these results,
we conclude that these genes probably do not encode plastid division factors and that our
NEXT-based approach will need further refinement if it is to be a successful platform for

identifying gene function.

CIENA: Identification of Candidate Division Genes by Gene Network Analysis.

Because we were unable to identify any new division genes using NExT, we

aimed to identify new candidate division genes based on their siliarity of expression to

228

known plastid division genes using publicly available Arabidopsis microarray data. This
screen for new candidates is arguably less biased than NeXT, as it relies solely upon
expression data and may more easily allow for the identification of factors that were not
inherited from the cyanobacterial endosymbiont. Recently, more advanced expression-
clustering algorithms have become widely accessible and their power will continue to
improve as more genome-wide expression data becomes available (Kinoshita and
Obayashi 2009). We used the ATTED-II database (Impzl/attedjfip) (Obayashi, Hayashi, et
al. 2009) to identify Arabidopsis genes that reside in gene expression networks with
known plastid division genes. This approach was termed CIENA (Candidate
Identification by Expression Network Analysis). As proof of concept, we examined the
gene expression network surrounding ARC6 and found two of its three nearest-neighbors
to be F tsZI -1 and F tsZZ-I (Table A.l). We performed this same query for 21 genes
known to affect plastid morphology (Table A.l). We compared network members arising
from this analysis and found 3 candidates that appear in more than one network:
At3g28760, At4g19710, and At4g28210 (Table A.l). At4g28210 is annotated as
embryo-defective (Tzafrir, Pena-Muralla, et al. 2004). At4g19710 has already been
characterized as a bifunctional aspartate kinase/homoserine dehydrogenase (Paris,
Wessel, et al. 2002) and is already present within the Plastid 2010
(http://plastid.msu.edu/) (Lu, Savage, et al. 2008) chloroplast morphology pipeline. As
such, we chose not to analyze T-DNA alleles within At4g28210 or At4g19710.
However, At3 g28760 is a gene of unknown function and was not present within the

Plastid 2010 pipeline, so we analyzed homozygous T-DNA insertions within an exon of

229

Table A.1. Known Plastid Division Genes and their Network Members.

 

Division Gene

ATTED-II Network Members

 

ARC] (At4g23940)

At5g49030, At3g04340,At3g28760, At4g19710, At5g50110

 

 

 

 

 

 

 

ARCZ (At2g28000) At5g20720, At3gl3470, At2g04030, At4g24280

ARC3 (At1g75010) At1g80480, At2g39930, At3g57430

ARC5 (At3g19720) At5g06060, At4g19710, At5g58550, At2g28100

ARC6 (At5g42480) At2g36250, At2g39670, At5g55280

BR04 (At1g55490) At5g63310, At2g44650, At1g56050

CRF2 (At4g23 750) At1g74890, At1g69040, At1g68400

CRL (At5g51020) At3g52230, At1g68660, At5g44000, At1g21065, At4g36530,

Atl g70580, At2g34310

 

C YOI (At3g19220)

At4g28210, At2g39080, At3g04550

 

FtsZI-I (At5g55280)

At1g69200, At5g42480, At3g54090, At2g29690

 

FtsZZ-I (At2g36250)

At5g42480, At1g09795, At1g13270

 

FtsZ2-2 (At3g52750)

Atl g63610, At2g32500, At2g47590, At3g10940, At3g28760,
At4g20760, At3g01510

 

GCI (At2g21280)

At4g34l90, At3g63140, At1g31800, At1g80030, At4g13670,
At3g05350

 

 

 

 

 

 

 

 

 

 

AtMinD (At5g24020) At2g33430, At4g28210, At3g46740

AtMinE (Atlg69390) Atlg65260, At1g12410, At2g04700

MSL2 (At5g10490) At1g56560, Atlg77800, At4g37320

MSL3 (Atl g58200) At5g53050, At3g62010, At5g54170, At4g01026, At3g06500,
At1g75960

0R (At5g61670) At4g34350, At5g64840, At5g43850, At5g44l90

PARC 6 (At3g19180) Affymetrix data not available.

PDVI (At5g53280) At1g78580, At4g04870, At5g09250

PD V2 (At2g16070) At4g37460, At4g32770, At2g21320, At3g45900

 

Network members that appear in more than one network and have undefined roles in
plastid division are indicated in bold text.

230

 

the At3g28760 locus (SALK_083209). We observed no difference in chloroplast
morphology between SALK_083209 homozygotes and Col-0. It remains to be seen if
At4g19710 has a role in plastid fission, but based on our analysis thus far we must
conclude that At3 g28760 is not involved in chloroplast division. It would be worth the
effort to examine lines carrying T-DNA insertions (if available) within the other loci that
appear Table A. 1 ., even if they only appear in one network, but it is certainly worth
establishing priority amongst this large number of candidate genes to make analysis more
efficient. Alternatively, TILLIN G or RNAi-based approaches may be used to gain
perspective on gene function if analysis of T-DNA insertion mutants proves unfruitful. If
the remaining candidates turn out to be involved in plastid division, firrther refinement of

this approach will be required.
Acknowledgements

1 thank Dean Dellapenna for introducing me to gene neighborhood analysis and
Ross Overbeek and Sveta Gerdes for their instruction on using the SEED database. I
thank the horde of rotation students and undergraduate helpers that assisted with this
project: Wei-hing Huang, Christine Shyu, Mia Hemmes, James Morgan, and John
Sherbeck genotyped and phenotyped most of the T-DNA insertion lines analyzed in the

NEXT and CIENA pipelines.

231

Appendix B

Mapping Novel Alleles of Plastid Division Genes

232

Summary

Forward and reverse genetic approaches have identified slightly more than 20
genes in Arabidopsis that affect plastid division. About half of these can be directly
traced to a cyanobacterial gene and the others are inventions of the host organism, arising
after stable endosymbiosis was established. It is presumed that several more factors are
involved in plastid division in land plants, because: (1) a number of cyanobacterial
division genes have been lost, and presumably were replaced by other factors, during
chloroplast evolution; (2) the components of the plastid dividing (PD) rings of higher
plants remain unknown; (3) some molecules must exist to link certain division
components to one another; (4) the scope of control mechanisms, including metabolites
and transcription factors, that act as inputs and outputs of plastid division have yet to be

determined. Despite the predicted near-saturation of forward genetic screens for division

mutants, we attempted to identify novel factors by examining the chloroplasts in M2

individuals derived from EMS-mutagenized Col-0 lines. This analysis has yielded 13
new division mutants. Some of these have been shown to be novel alleles of known loci
and several are yet unidentified. Additionally, we provide rough mapping data for the
ARC9 mutation, which was previously obtained from a T-DNA mutant screen for

chloroplast division mutants.

New Plastid Division Alleles Discovered within an EMS-mutagenized Population.

To identify new plastid division genes we grew pools of M2 progeny of EMS-

mutagenized Col-0 individuals (Lehle Seeds) on Linsmaier-Skoog medium. This M2

233

population is part of the same batch that yielded two mutant alleles of PD V1
(Miyagishima, F roehlich. et al. 2006). At 15-25 days post-germination, we examined the

largest leaves for defects in chloroplast morphology by light microscopy. We identified

22 plants with aberrations in chloroplast morphology from a pool of ~14,000 M2

individuals. We confirmed the chloroplast morphology in fresh tissue samples and then
transplanted plants to soil. Out of the original 22 plants, 13 survived transplantation and
were fertile. We termed these lines chloroplast altered number (can) mutants. We
prioritized the mutants based on their division phenotype and plant vigor; the mutants
with more interesting phenotypes were given higher p1iority and amongst these, the
healthier plants were given the highest priority to facilitate quick identification of the

causative mutation (Table 8.1).

We crossed each of our high-priority candidates (can2, CAN3, can4, can5, and
can8) to known division mutants with similar chloroplast phenotypes to test for allelism.

In parallel, we outcrossed each of our high-priority candidates to Ler-O to generate a
mapping population for each of these mutants. We verified F1 hybrids by genotypic
analysis, examined their chloroplast morphology phenotypes, and allowed these
individuals to self-fertilize to generate an F2 mapping population. We quickly identified

can2 and can4 as recessive alleles of arc5 and ftsZI, respectively, based on crosses to
previously characterized mutants (Gao, Kadirjan-Kalbach, et a1. 2003, Yoder, Kadirjan-
Kalbach, et al. 2007). The relevant loci were then sequenced to identify the causative

mutation. The remaining three alleles (CAN3, can5, and can8) were rough-mapped to a

234

Table B.1. A Summary of the Arabidopsis can Mutants.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

produced less than 20 seeds. Unknown
inheritance pattern.

 

Mutant Phenotypes and inheritance pattern Status

can] arcl-like chloroplasts. Plant is severely Causitive mutation unknown.
stunted and pale, requiIing more than 40 Check for allelism with arc] .
days to produce flowers. Unknown Mapping population needed
inheritance pattern.

can2 arc5-like chloroplasts. Recessive ARC5 codon 564 GAG to
inheritance pattern. AAG (E564K).

CAN3 Unique chloroplast phenotype. Dominant F T SZZ-I codon 319 GAT to
inheritance pattern. AAT (D319N).

can4 ftsZI-like chloroplasts. Recessive F T SZI codon 275 GAT to
inheritance pattern. AAT (D275N).

can5 Unique chloroplast phenotype. Recessive ARC3 codon 106 TGG to
inheritance pattern. TGA (W 106*). Mutation

occurs at splice acceptor site.

can6 Unique chloroplast phenotype. Unknown Causitive mutation unknown.
inheritance pattern. Mapping population needed.

can7 Unique chloroplast phenotype. Unknown Causitive mutation unknown.
inheritance pattern. Mapping population needed.

can8 Unique chloroplast phenotype. Plant Linked to right arm of chr 5
flowers earlier than wild type. Recessive near F T SZI locus; non-allelic
inheritance. with [621 and arc6 mutants.

can9 Unique chloroplast phenotype. Unknown Causitive mutation unknown.
inheritance pattern. Mapping population needed.

can] 0 Unique chloroplast phenotype. Unknown Causitive mutation unknown.
inheritance pattern. Mppping population needed.

can12 crl-like plant and chloroplast phenotypes. Causitive mutation unknown.
Unknown inheritance pattern. Mapping population needed.

can13 ftsZI -like chloroplast phenotype. Causitive mutation unknown.
Unknown inheritance pattern. Mapping population needed.

can] 6 Unique chloroplast phenotype. Plant Causitive mutation unknown.

Mapping population needed.

 

235

 

specific Arabidopsis chromosome using molecular markers on a small F2 population of

20-30 plants and then identified based on their proximity to known plastid division genes.

ARC9 is an allele of PDV2.

The original arc9 mutant was originally isolated from a population of T-DNA-
mutagenized Ws individuals and previously annotated as a recessive mutation (Aldridge,
Maple, et al. 2005, Rutherford 1996). We received seeds from the arc9 line (CS283)
from the Arabidopsis Biological Resource Center (ABRC), with the hope of identifying
the mutation which causes the chloroplast morphology defect within this line. We
quickly noted two populations of plants emerging from this single seed stock, one with
wild-type chloroplasts (~90%) and the others with aberrant chloroplast morphology that
we presumed to be arc9 mutants (~10%) (Figure B. 1), indicating that the seed stock may
have been contaminated or segregating during propagation at ABRC. We selected one of
the individuals with abnormal chloroplasts for propagation and again noticed two
populations occurring within ~50 offspring from this single plant. One population
(~40%) possessed wild-type chloroplasts and the other (~60%) harbored chloroplasts
similar to the parental plant. We hypothesized that the mutation in C8283 was dominant
and that a background mutation might be contributing to the skewed phenotypic ratio
observed within the progeny, as the expected phenotypic ratios observed in the progeny
from an individual carrying a heterozygous dominant mutation should be 3: 1. Consistent
with this hypothesis, we examined embryos within developing siliques of a heterozygous

plant and observed that ~3 0% of the embryos were stunted and/or undeveloped.

236

 

 

 

 

 

Figure B.1. Chloroplast phenotype observed in ARC9/arc9 (C8283) plants. DIC
images showing mesophyll cells from from Ws-2 (A) and and ARC9/arc9 heterozygote.
Scale bar = 10 pm.

237

To identify the causative mutation in ARC9, we crossed an ARC9/arc9 plant to

Col-0 and allowed the mutant F1 individuals to self-fertilize to produce an F2 generation

for rough mapping. From a population of 80 F 2 individuals, we identified 30 with wild

type chloroplast morphology and collected genomic DNA from these plants for genotypic
analysis. Using 10 polymorphic PCR markers that allow for the differentiation of Ws and
Col-0 backgrounds by agarose gel electrophoresis, we observed the highest degree of
linkage between the wild type phenotype and two molecular markers, both located on
chromosome 2; of the 56 chromosomes analyzed from 28 wild type individuals, 79%
carried the Col-0 RGA polymorphism and 71% carried the Col-0 NGA168
polymorphism (Bell and Ecker 1994); both of these polymorphisms reside on
Arabidopsis chromosome 2, suggesting that the ARC9 locus also resides on this
chromosome. At least four plastid division genes reside on this chromosome: PD V2

(At2gl6070), GC I (At2g21280), ARC2 (At2g28000), and FtsZZ-I (At2g36250).

Notably, the PD V2 locus resides in the center of the interval bounded by the RGA
and NGA168 markers; and the ARC9/arc9 chloroplast morphology phenotype is
remarkably similar to PD V2/pdv2 (Glynn, F roehlich, et al. 2008, Miyagishima,
Froehlich, et al. 2006). Furthermore, like ARC9, the two characterized mutations in
PD V2 are dominant-negative (Glynn, Froehlich, et al. 2008, Miyagishima, F roehlich, et
al. 2006) due to gene dosage effects (J .M. Glynn and K. Osteryoung, unpublished)
(Okazaki, Kabeya, et al. 2009). Therefore, we wanted to determine if ARC9 might be an
allele of PD V2. To do this we crossed a PDV2-deficient heterozygous T-DNA line

(SAIL_875E10) to an ARC9/arc9 heterozygote, using the ARC9/arc9 plant as a pollen

238

donor. We genotyped F1 individuals from this cross to verify that they were Col/Ws

hybrids using the ATPase molecular genetic marker, which easily distinguishes Ws and

Col accessions. Out of 11 marker-verified Col/Ws F1 hybrids resulting from the

ARC9/arc9 x SAIL_875E10 (heterozygous) cross, 3 (27%) had severe chloroplast
morphology phenotypes similar to lines lacking any PDV2 protein (Glynn, Froehlich, et
al. 2008, Miyagishima, Froehlich, et al. 2006). These results suggest that either ARC9 is
an allele of PD V2 or that the aggregated effect of these two dominant mutations (ARC9
and PD V2) results in a severe chloroplast morphology phenotype. To determine if the
severe phenotype observed in the cross progeny were due to a lack of PDV2 protein, we
examined PDV2 protein levels in cell extracts from a wild-type plant, three cross progeny
with PD V2/pdv2 heterzoygote-like phenotypes, and the three cross progeny with severe
phenotypes (Figure B.2). Consistent with ARC9 and PD V2 allelism, we observed that all
three progeny with severe phenotypes contained no detectable PDV2 protein (Figure B.2,
lanes 6-8). From this we conclude that the dominant mutation that leads to the
chloroplast morphology phenotype in ARC9/arc9 plants is likely due to PDV2
haploinsufficiency (Okazaki, Kabeya, et al. 2009) resulting from a T—DNA insertion in
the PD V2 locus; the cause of the embryo lethality observed in this line is still unknown,
but is probably due to a background mutation since neither pdv2-1 nor pdv2-2 alleles
exhibit embryo lethality defects (Miyagishima, Froehlich, et al. 2006). Further work
would be required to identify the precise borders of the T-DNA insertion in ARC9/arc9

plants, if necessary.

239

 

--.1- 2 3 _4__5_,_§_. 7 3

 

 

 

 

 

Figure B.2. ARC9 is associated with a decrease in PDV2 protein level. Molecular
weight marker (lane 1); extract from the eighth leaf of a wild-type plant (lane 2);
ARC 9/arc9 x PD V2/pdv2 cross progeny (F1) with PD V2/pdv2 heterozygote-like
phenotypes (lanes 3-5); and ARC9/arc9 x PDV2/pdv2 cross progeny (F1) with severe
chloroplast morphology phenotypes (lanes 6-8). Approximately 2 mg of whole cell
extract is loaded in each lane; extracts were made from 30-day old plants. kD =
kiloDaltons.

240

Acknowledgments

Afiqah Ahmadhisham and John Sherbeck screened the bulk of the EMS-
mutagenized Col-0 population. Afiqah Ahmadhisham performed rough mapping of can8
and did so with great care and efficiency. 1 am very much appreciative of John and
Afiqah’s efforts and for being such hard-working, flexible, and cheerful members of the
Osteryoung lab over the past few years. I thank Joyce Bower for providing seeds from

the ABRC ARC9 mutant stock and Kathy Osteryoung for allowing me to identify ARC9.

241

REFERENCES

242

REFERENCES

Aldridge, C., Maple, J. and Moller, S.G. (2005) The molecular biology of plastid
division in higher plants. .1. Exp. Bot, 56, 1061-1077.

Aldridge, C. and Moller, S.G. (2005) The Plastid Division Protein AtMinD] Is a Ca2+-
ATPase Stimulated by AtMinE]. J. Biol. Chem, 280, 31673-31678.

Alonso, J.M., Stepanova, A.N., Leisse, T.J., Kim, C.J., Chen, H., Shinn, P.,
Stevenson, D.K., Zimmerman, J., Barajas, P., Cheuk, R., Gadrinab, C.,
Heller, C., Jeske, A., Koesema, E., Meyers, C.C., Parker, H., Prednis, L.,
Ansari, Y., Choy, N., Deen, H., Geralt, M., Hazari, N., Hom, E., Karma, M.,
Mulholland, C., Ndubaku, R., Schmidt, 1., Guzman, P., Aguilar-Henonin, L.,
Schmid, M., Weigel, D., Carter, D.E., Marchand, T., Risseeuw, E., Brogden,
D., Zeko, A., Crosby, W.L., Berry, C.C. and Ecker, J.R. (2003) Genome-wide
insertional mutagenesis of Arabidopsis thaliana. Science, 301, 653-657.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, L, Miller, W. and
Lipman, DJ. (1997) Gapped BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids Res, 25, 3389-3402.

Amberg, D.C., Zahner, J.E., Mulholland, J.W., Pringle, J.R. and Botstein, D. (1997)
Aip3p/Bud6p, a yeast actin-interacting protein that is involved in morphogenesis
and the selection of bipolar budding sites. Mol Biol Cell, 8, 729-753.

Askham, J.M., Moncur, P., Markham, A.F. and Morrison, E.E. (2000) Regulation
and function of the interaction between the APC tumour suppressor protein and
EB]. Oncogene, 19, 1950-1958.

Austin II, J. and Webber, A.N. (2005) Photosynthesis in Arabidopsis thaliana mutants
with reduced chloroplast number. Photosynth Res, 85, 373-3 84.

Barth, A.I., Siemers, K.A. and Nelson, W.J. (2002) Dissecting interactions between
EBl, microtubules and APC in cortical clusters at the plasma membrane. J Cell
Sci, 115, 1583-1590.

Baskin, T.I., Cork, A., Williamson, RE. and Gorst, J.R. (1995) STUNTED PLANT l,
A Gene Required for Expansion in Rapidly Elongating but Not in Dividing Cells
and Mediating Root Growth Responses to Applied Cytokinin. Plant Physiol, 107,
233-243.

Bearer, EL. (2003) Overview of image analysis, image importing, and image processing
using freeware. Curr Protoc Mol Biol, Chapter, Unit 14.15.

243

Beemster, GT. and Baskin, T.I. (2000) Stunted plant 1 mediates effects of cytokinin,

but not of auxin, on cell division and expansion in the root of Arabidopsis. Plant
Physiol, 124, 1718-1727.

Bell, C.J. and Ecker, J.R. (1994) Assignment of 30 microsatellite loci to the linkage
map of Arabidopsis. Genomics, 19, 137-144.

Ben-Nissan, G., Cui, W., Kim, D.J., Yang, Y., Yoo, B.C. and Lee, J.Y. (2008)
Arabidopsis casein kinase l-like 6 contains a microtubule-binding domain and
affects the organization of cortical microtubules. Plant Physiol, 148, 1897-1907.

Bennett, V. and Baines, A.J. (2001) Spectrin and ankyrin-based pathways: metazoan
inventions for integrating cells into tissues. Physiol Rev, 81, 1353-1392.

Bernhardt, T.G. and de Boer, RA. (2005) SlmA, a nucleoid-associated, FtsZ binding
protein required for blocking septal ring assembly over Chromosomes in E. coli.
Mol Cell, 18, 555-564.

Bi, E.F. and Lutkenhaus, J. (1991) FtsZ ring structure associated with division in
Escherichia coli. Nature, 354, 161-164.

Bin, H., Guohua, Y., Weixing, Z., Yingjiao, Z. and Jindong, Z. (2007) MreB is
important for cell shape but not for chromosome segregation of the filamentous

cyanobacterium Anabaena sp. PCC 7120. Molecular Microbiology, 63, 1640-
1652.

Birky, C.W., Jr. and Walsh, J.B. (1992) Biased gene conversion, copy number, and
apparent mutation rate differences within chloroplast and bacterial genomes.
Genetics, 130, 677-683.

Blom, N., Gammeltoft, S. and Brunak, S. (1999) Sequence and structure-based

prediction of eukaryotic protein phosphorylation sites. J Mol Biol, 294, 1351-
1362.

Boffey, S.A. and Lloyd, D. (1988) Division and Segregation of Organelles. Cambridge,
UK: Cambridge University Press.

Bracha-Drori, K., Keren, 8., Aviva, K., Moran, 0., Ruthie, A., Shaul, Y. and Nir, O.
(2004) Detection of protein-protein interactions in plants using bimolecular

fluorescence complementation. The Plant Journal, 40, 419-427.

Braun, D.M. and Walker, J.C. (1996) Plant transmembrane receptors: new pieces in the
signaling puzzle. Trends Biochem Sci, 21, 70-73.

244

Brenner, S.L. and Korn, ED. (1979) Spectrin-actin interaction. Phosphorylated and
dephosphorylated spectrin tetramer cross-link F-actin. J Biol Chem, 254, 8620-
8627.

Broderick, M.J. and Winder, SJ. (2005) Spectrin, alpha-actinin, and dystrophin. Adv
Protein Chem, 70, 203-246.

Bruce, B.D., Perry, 8., Froehlich, J. and Keegstra, K. (1994) In vitro Import of Protein
into Chloroplasts. Boston: Kluwer Academic Publishers.

Bu, W. and Su, L.K. (2003) Characterization of functional domains of human EBl
family proteins. J Biol Chem, 278, 49721-49731.

Bukau, B. and Walker, G.C. (1989) Cellular defects caused by deletion of the
Escherichia coli dnaK gene indicate roles for heat shock protein in normal
metabolism. J Bacteriol, 171, 2337-2346.

Burge, C. and Karlin, S. (1997) Prediction of complete gene structures in human
genomic DNA. J Mol Biol, 268, 78-94.

Castellano, M.d.M., Boniotti, M.B., Caro, E., Schnittger, A. and Gutierrez, C. (2004)
DNA Replication Licensing Affects Cell Proliferation or Endoreplication in a Cell
Type-Specific Manner. Plant Cell, 16, 2380-2393.

Cavalier-Smith, T. (2000) Membrane heredity and early chloroplast evolution. Trends in
Plant Science, 5, 174.

Cerveny, K.L., Tamura, Y., Zhang, Z., Jensen, RE. and Sesaki, H. (2007) Regulation
of mitochondrial fusion and division. Trends Cell Biol, 17, 563-569.

Chen, J.G., Shimomura, 8., Sitbon, F., Sandberg, G. and Jones, A.M. (2001) The role
of auxin-binding protein 1 in the expansion of tobacco leaf cells. Plant J, 28, 607-
617.

Chen, Y., Asano, T., Fujiwara, M.T., Yoshida, 8., Machida, Y. and Yoshioka, Y.
(2009) Plant cells without detectable plastids are generated in the crumpled leaf
mutant of Arabidopsis thaliana. Plant Cell Physiol, 50, 956-969.

Christie, J.M. (2007) Phototropin blue-light receptors. Annu Rev Plant Biol, 58, 21-45.

Clarke, D.J., Jacq, A. and Holland, LB. (1996) A novel DnaJ-like protein in
Escherichia coli inserts into the cytoplasmic membrane with a type III topology.

Mol Microbiol, 20, 1273-1286.

Clough, SJ. and Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-
mediated transformation of Arabidopsis thaliana. Plant J, 16, 735-743.

245

Colletti, K.S., Tattersall, E.A., Pyke, K.A., Froelich, J.E., Stokes, K.D. and
Osteryoung, K.W. (2000) A homologue of the bacterial cell division site-
deterrnining factor MinD mediates placement of the chloroplast division
apparatus. Current Biology, 10, 507.

Cordell, S.C., Robinson, E.J. and Lowe, J. (2003) Crystal structure of the 808 cell
division inhibitor SulA and in complex with FtsZ. Proc Natl Acad Sci U S A, 100,
7889-7894.

Cramer, W.A., Engelman, D.M., Von Heijne, G. and Rees, D.C. (1992) Forces
involved in the assembly and stabilization of membrane proteins. FASEB J, 6,
3397-3402.

Cui, L., Veeraraghavan, N., Richter, A., Wall, K., Jansen, R.K., Leebens-Mack, J.,
Makalowska, I. and dePamphilis, CW. (2006) ChloroplastDB: the Chloroplast
Genome Database. Nucleic Acids Res, 34, D692-696.

Davila-Aponte, J.A., Inoue, K. and Keegstra, K. (2003) Two chloroplastic protein
translocation components, Tie] 10 and Toc75, are conserved in different plastid
types from multiple plant species. Plant Mol Biol, 51, 175-181.

de Crecy-Lagard, V. (2007) Identification of genes encoding tRNA modification
enzymes by comparative genomics. Methods Enzymol, 425, 153-183.

de Crecy-Lagard, V. and Hanson, A.D. (2007) Finding novel metabolic genes through
plant-prokaryote phylogenomics. Trends Microbiol, 15, 563-570.

de Souza, S.J., Long, M., Schoenbach, L., Roy, SW. and Gilbert, W. (1996) Intron
positions correlate with module boundaries in ancient proteins. Proc Natl Acad
Sci U S A, 93, 14632-14636.

Delumeau, O., Dutta, 8., Brigulla, M., Kuhnke, G., Hardwick, S.W., Volker, U.,
Yudkin, MD. and Lewis, RJ. (2004) Functional and structural characterization
of Rst, a stress signaling protein phosphatase 2C. J Biol Chem, 279, 40927-
40937.

DePamphilis, M.L. (2003) The 'ORC cycle': a novel pathway for regulating eukaryotic
DNA replication. Gene, 310, 1-15.

Dixson, J.D., Forstner, M.J. and Garcia, D.M. (2003) The alpha-actinin gene family: a
revised classification. J Mol Evol, 56, 1-10.

Dobson, GM. (2003) Protein folding and misfolding. Nature, 426, 884.

Doolittle, W.F. (1979) The cyanobacterial genome, its expression, and the control of that
expression. Adv Microb Physiol, 20, 1-102.

246

Dreher, K. and Callis, J. (2007) Ubiquitin, hormones and biotic stress in plants. Ann Bot
(Lond), 99, 787-822.

Drubin, D.G. (1991) Development of cell polarity in budding yeast. Cell, 65, 1093-1096.

Dutta, 8. and Lewis, R.J. (2003) Crystallization and preliminary crystallographic
analysis of the kinase-recruitment domain of the PP2C-type phosphatase Rst.
Acta Crystallogr D Biol Crystallogr, 59, 191-193.

Dyall, S.D., Brown, M.T. and Johnson, P.J. (2004) Ancient Invasions: From
Endosymbionts to Organelles. Science, 304, 253-257.

Emanuelsson, O., Nielsen, H., Brunak, S. and von Heijne, G. (2000) Predicting
subcellular localization of proteins based on their N-terrninal amino acid
sequence. J Mol Biol, 300, 1005-1016.

Emanuelsson, O., Nielsen, H. and von Heijne, G. (1999) ChloroP, a neural network-

based method for predicting chloroplast transit peptides and their cleavage sites.
Protein Sci, 8, 978-984.

Erickson, H.P. (2009) Modeling the physics of F tsZ assembly and force generation.
Proc Natl Acad Sci U S A, 106, 9238-9243.

Erickson, H.P., Taylor, D.W., Taylor, K.A. and Bramhill, D. (1996) Bacterial cell
division protein FtsZ assembles into protofilament sheets and minirings, structural
homologs of tubulin polymers. Proc Natl Acad Sci U S A, 93, 519-523.

Ewald, R., Kolukisaoglu, U., Bauwe, U., Mikkat, S. and Bauwe, H. (2007)
Mitochondrial protein lipoylation does not exclusively depend on the thAS
pathway of de novo fatty acid synthesis in Arabidopsis. Plant Physiol, 145, 41-48.

Falkow, 8., Dworkin, M., Rosenburg, E., Schleifer, K.-H. and Stackebrandt, E.
(2006) The Prokaryotes: F irmicutes, Cyanobacteria 3rd edn. New York: Springer
Science and Business Media.

Felsenstein, J. (1985) Confidence Limits on Phylogenies: An Approach Using the
Bootstrap. Evolution, 39, 783-791.

Fischer, A., Hofmann, 1., Naumann, K. and Renter, G. (2006) Heterochromatin
proteins and the control of heterochromatic gene silencing in Arabidopsis. J Plant
Physiol, 163, 358-368.

Fodor, A.A. and Aldrich, R.W. (2009) Convergent Evolution of Alternative Splices at
Domain Boundaries of the BK Channel. Annual Review of Physiology, 71, 19-36.

247

Forth, D. and Pyke, K.A. (2006) The suffulta mutation in tomato reveals a novel
method of plastid replication during fi'uit ripening. J. Exp. Bot, 57, 1971-1979.

Fujiwara, M.T., Hashimoto, H., Kazama, Y., Abe, T., Yoshida, 8., Sato, N. and Itoh,
RD. (2008) The assembly of the FtsZ ring at the mid-chloroplast division site
depends on a balance between the activities of AtMinEl and ARCl l/AtMinDl.

Plant Cell Physiol, 49, 345-361.

Fujiwara, M.T., Nakamura, A., Itoh, R., Shimada, Y., Yoshida, S. and Moller, S.G.
(2004) Chloroplast division site placement requires dimerization of the
ARC l l/AtMinDl protein in Arabidopsis. J Cell Sci, 117, 2399-2410.

Gao, H., Kadirjan-Kalbach, D., Froehlich, J.E. and Osteryoung, K.W. (2003) ARC5,
a cytosolic dynamin-like protein from plants, is part of the chloroplast division
machinery. PNAS, 100, 4328-4333.

Gao, H., Sage, T.L. and Osteryoung, K.W. (2006) F ZL, an F ZO-like protein in plants,
is a determinant of thylakoid and chloroplast morphology. PNAS, 103, 6759-6764.

Geissler, B., Elraheb, D. and Margolin, W. (2003) A gain-of-function mutation in ftsA
bypasses the requirement for the essential cell division gene zipA in Escherichia
coli. Proc Natl Acad Sci U S A, 100, 4197-4202.

Ghosh, B. and Sain, A. (2008) Origin of contractile force during cell division of
bacteria. Phys Rev Lett, 101, 178.

Ginalski, K., Elofsson, A., Fischer, D. and Rychlewski, L. (2003) 3D-Jury: a simple
approach to improve protein structure predictions. Bioinformatics, 19, 1015-1018.

Glynn, J.M., Froehlich, J.E. and Osteryoung, K.W. (2008) Arabidopsis ARC6
coordinates the division machineries of the inner and outer chloroplast
membranes through interaction with PDV2 in the intermembrane space. Plant

Cell, 20, 2460-2470.

Glynn, J.M., Lustig, R.J., Berlin, A. and Chang, F. (2001) Role of bud6p and tealp in
the interaction between actin and microtubules for the establishment of cell

polarity in fission yeast. Curr Biol, 11, 836-845.

Glynn, J.M., Miyagishima, S.Y., Yoder, D.W., Osteryoung, K.W. and Vitha, S.
(2007) Chloroplast division. Trafiic, 8, 451-461.

Glynn, J.M., Yang, Y., Vitha, 8., Schmitz, A.J., Hemmes, M., Miyagishima, S. and
Osteryoung, K.W. (2009) PARC6, a novel chloroplast division factor, influences
FtsZ assembly and is required for recruitment of PDVl during chloroplast

division in Arabidopsis. Plant J, 59, 700-711.

248

Goehring, N.W. and Beckwith, J. (2005) Diverse paths to midcell: assembly of the
bacterial cell division machinery. Curr Biol, 15, R514-526.

Goodman, S.R., Krebs, K.E., Whitfield, C.F., Riederer, B.M. and Zagon, 1.8. (1988)
Spectrin and related molecules. CRC Crit Rev Biochem, 23, 171-234.

Greenfield, NJ. (2006) Using circular dichroism spectra to estimate protein secondary
structure. Nat Protoc, 1, 2876-2890.

Hajdukiewicz, P., Svab, Z. and Maliga, P. (1994) The small, versatile pPZP family of
Agrobacterium binary vectors for plant transformation. Plant Mol Biol, 25, 989-
994.

Hale, C.A., Meinhardt, H. and de Boer, P.A. (2001) Dynamic localization cycle of the
cell division regulator MinE in Escherichia coli. EMBO J, 20, 1563-1572.

Hale, C.A., Rhee, A.C. and de Boer, P.A. (2000) ZipA-induced bundling of FtsZ
polymers mediated by an interaction between C-terminal domains. J Bacteriol,
182, 5153-5166.

Hales, K.G. and Fuller, M.T. (1997) Developmentally Regulated Mitochondrial Fusion
Mediated by a Conserved, Novel, Predicted GTPase. Cell, 90, 121.

Hardwick, S.W., Pane-Farre, J., Delumeau, 0., Marles-Wright, J., Murray, J.W.,
Hecker, M. and Lewis, R.J. (2007) Structural and functional characterization of

partner switching regulating the environmental stress response in Bacillus subtilis.
J Biol Chem, 282, 11562-11572.

Haswell, ES. and Meyerowitz, EM. (2006) MscS-like Proteins Control Plastid Size
and Shape in Arabidopsis thaliana. Current Biology, 16, 1-11.

Hayashi, I. and Ikura, M. (2003) Crystal structure of the amino-terminal microtubule-
binding domain of end-binding protein 1 (BED. J Biol Chem, 278, 36430-36434.

He, Y. and Amasino, RM. (2005) Role of chromatin modification in flowering-time
control. Trends Plant Sci, 10, 30-35.

Heazlewood, J.L., Durek, P., Hummel, J., Selbig, J., Weckwerth, W., Walther, D.
and Schulze, W.X. (2008) PhosPhAt: a database of phosphorylation sites in

Arabidopsis thaliana and a plant-specific phosphorylation site predictor. Nucleic
Acids Res, 36, D1015-1021.

Heckman, KL. and Pease, L.R. (2007) Gene splicing and mutagenesis by PCR-driven
overlap extension. Nat Protoc, 2, 924-932.

249

 

Henikoff, 8., Till, DJ. and Comai, L. (2004) TILLING. Traditional mutagenesis meets
functional genomics. Plant Physiol, 135, 630-636.

Hinshaw, J.E. (2000) Dynamin and its role in Membrane Fission. Annual Review of Cell
and Developmental Biology, 16, 483-519.

Hinshaw, J.E. and Schmid, S.L. (1995) Dynamin self-assembles into rings suggesting a
mechanism for coated vesicle budding. Nature, 374, 190-192.

Holzinger, A., Kwok, E.Y. and Hanson, MR. (2008) Effects of arc3, arcS and arc6
mutations on plastid morphology and stromule formation in green and nongreen
tissues of Arabidopsis thaliana. Photochem Photobiol, 84, 1324-1335.

Hong, Z., Bednarek, S.Y., Blumwald, E., Hwang, 1., Jurgens, G., Menzel, D.,
Osteryoung, K.W., Raikhel, N.V., Shinozaki, K., Tsutsumi, N. and Verma,
DP. (2003) A unified nomenclature for Arabidopsis dynamin-related large
GTPases based on homology and possible functions. Plant Mol Biol, 53, 261-265.

Hoppins, 8., Lackner, L. and Nunnari, J. (2007) The machines that divide and fuse
mitochondria. Annu Rev Biochem, 76, 751-780.

Hu, Y., Poh, H.M. and Chua, N.H. (2006) The Arabidopsis ARGOS-LIKE gene
regulates cell expansion during organ growth. Plant J, 47, 1-9.

Hu, Z. and Lutkenhaus, J. (1999) Topological regulation of cell division in Escherichia
coli involves rapid pole to pole oscillation of the division inhibitor MinC under
the control of MinD and MinE. Mol Microbiol, 34, 82-90.

Hu, Z. and Lutkenhaus, J. (2001) Topological regulation of cell division in E. coli
spatiotemporal oscillation of MinD requires stimulation of its ATPase by MinE
and phospholipid. Mol Cell, 7, 1337-1343.

Hu, 2., Mukherjee, A., Pichoff, S. and Lutkenhaus, J. (1999) The MinC component of
the division site selection system in Escherichia coli interacts with F tsZ to prevent
polymerization. PNAS, 96, 14819-14824.

Huber, S.C. and Hardin, S.C. (2004) Numerous posttranslational modifications provide
opportunities for the intricate regulation of metabolic enzymes at multiple levels.
Curr Opin Plant Biol, 7, 318-322.

Huisman, O., D'Ari, R. and Gottesman, S. (1984) Cell-division control in Escherichia
coli: specific induction of the 808 function SfiA protein is sufficient to block
septation. Proceedings of the National Academy of Sciences of the United States
of America, 81, 4490-4494.

250

Itoh, R., Fujiwara, M., Nagata, N. and Yoshida, S. (2001) A Chloroplast Protein
Homologous to the Eubacterial Topological Specificity Factor MinE Plays a Role
in Chloroplast Division. Plant Physiol, 127, 1644-1655.

Ivey, R.A., 3rd and Bruce, RD. (2000) In vivo and in vitro interaction of DnaK and a
chloroplast transit peptide. Cell Stress Chaperones, 5, 62-71.

Jackson, D.T., Froehlich, J.E. and Keegstra, K. (1998) The hydrophilic domain of
Tic110, an inner envelope membrane component of the chloroplastic protein

translocation apparatus, faces the stromal compartment. J Biol Chem, 273, 16583-
16588.

Jacobs, M.A., Connell, L. and Cattolico, RA. (1999) A conserved His-Asp signal
response regulator-like gene in Heterosigma akashiwo chloroplasts. Plant Mol
Biol, 41, 645-655.

Jander, G., Norris, S.R., Rounsley, S.D., Bush, D.F., Levin, I.M. and Last, RL.
(2002) Arabidopsis map-based cloning in the post-genome era. Plant Physiol,
129, 440-450.

Jensen, 8.0., Thompson, LS. and Harry, E.J. (2005) Cell division in Bacillus subtilis:
F tsZ and F tsA association is Z-ring independent, and F tsA is required for efficient
midcell Z-Ring assembly. J Bacteriol, 187, 6536-6544.

Jeong, W.J., Park, Y.-I., Suh, K., Raven, J.A., Yoo, OJ. and Liu, J.R. (2002) A
Large Population of Small Chloroplasts in Tobacco Leaf Cells Allows More

Effective Chloroplast Movement Than a Few Enlarged Chloroplasts. Plant
Physiol, 129, 112-121.

Jones, A.M., 1m, K.H., Savka, M.A., Wu, M.J., DeWitt, N.G., Shillito, R. and Binns,
A.N. (1998) Auxin-dependent cell expansion mediated by overexpressed auxin-
binding protein 1. Science, 282, 1114-1117.

Kang, C.M., Vijay, K. and Price, C.W. (1998) Serine kinase activity of a Bacillus
subtilis switch protein is required to transduce environmental stress signals but
not to activate its target PP2C phosphatase. Mol Microbiol, 30, 189-196.

Kasahara, M., Kagawa, T., Oikawa, K., Suetsugu, N., Miyao, M. and Wada, M.
(2002) Chloroplast avoidance movement reduces photodamage in plants. Nature,
420, 829-832.

Kinoshita, K. and Obayashi, T. (2009) Multi-dimensional correlations for gene

coexpression and application to the large-scale data of Arabidopsis.
Bioinformatics, 25, 2677-2684.

251

 

Kohler, R.H., Cao, J., Zipfel, W.R., Webb, W.W. and Hanson, M.R. (1997) Exchange
of protein molecules through connections between higher plant plastids. Science,
276, 2039-2042.

Koksharova, O.A. and Wolk, GP. (2002) A Novel Gene That Bears a DnaJ Motif
Influences Cyanobacterial Cell Division. J. Bacteriol, 184, 5524-5528.

Kordeli, E. (2000) The spectrin-based skeleton at the postsynaptic membrane of the
neuromuscular junction. Microsc Res Tech, 49, 101-107.

Kuroiwa, T., Kuroiwa, H., Sakai, A., Takahashi, H., Toda, K. and Itoh, R. (1998)
The division apparatus of plastids and mitochondria. Int Rev Cytol, 181, 1-41.

Lan, G., Daniels, B.R., Dobrowsky, T.M., Wirtz, D. and Sun, 8.x. (2009)
Condensation of F tsZ filaments can drive bacterial cell division. Proc Natl Acad
Sci USA, 106, 121-126.

Lan, G., Wolgemuth, C.W. and Sun, s.x. (2007) Z-ring force and cell shape during
division in rod-like bacteria. Proc Natl Acad Sci U S A, 104, 16110-16115.

Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A.,
McWilliam, H., Valentin, E, Wallace, I.M., Wilm, A., Lopez, R., Thompson,
J.D., Gibson, T.J. and Higgins, D.G. (2007) Clustal W and Clustal X version
2.0. Bioinformatics, 23, 2947-2948.

Leech, R.M., Thomson, W.W. and Platt-Aloia, K.A. (1981) Observations on the
mechanism of chloroplast division in higher plants. New Phytologist, 87, 1-9.

Lemmon, M.A. (2004) Pleckstrin homology domains: not just for phosphoinositides.
Biochem Soc Trans, 32, 707-711.

Levin, P.A., Shim, J.J. and Grossman, A.D. (1998) Effect of minCD on FtsZ ring
position and polar septation in Bacillus subtilis. J Bacteriol, 180, 6048-6051.

Li, L. and Kehoe, D.M. (2005) In vivo analysis of the roles of conserved aspartate and
histidine residues within a complex response regulator. Mol Microbiol, 55, 1538-
1552.

Li, Z., Trimble, M.J., Brun, Y.V. and Jensen, G.J. (2007) The structure of FtsZ
filaments in vivo suggests a force-generating role in cell division. EMBO J, 26,
4694-4708.

Liu, S.C. and Palek, J. (1980) Spectrin tetramer-dimer equilibrium and the stability of
erythrocyte membrane skeletons. Nature, 285, 586-5 88.

Low, H.H. and Lowe, J. (2006) A bacterial dynamin-like protein. Nature, 444, 766-769.

252

Low, H.H., Moncrieffe, M.C. and Lowe, J. (2004) The crystal structure of ZapA and its
modulation of FtsZ polymerisation. J Mol Biol, 341, 839-852.

Lu, C. and Erickson, H.P. (1998) Purification and assembly of F tsZ. Methods Enzymol,
298, 305-313.

Lu, C., Reedy, M. and Erickson, H.P. (2000) Straight and Curved Conformations of
FtsZ Are Regulated by GTP Hydrolysis. J. Bacteriol, 182, 164-170.

Lu, Y., Savage, L.J., Ajjawi, 1., Imre, K.M., Yoder, D.W., Benning, C., Dellapenna,
D., Ohlrogge, J.B., Osteryoung, K.W., Weber, A.P., Wilkerson, CG. and
Last, R.L. (2008) New connections across pathways and cellular processes:
industrialized mutant screening reveals novel associations between diverse
phenotypes in Arabidopsis. Plant Physiol, 146, 1482-1500.

Lutkenhaus, J. (2007) Assembly dynamics of the bacterial MinCDE system and spatial
regulation of the Z ring. Annu Rev Biochem, 76, 539-562.

Ma, H., Lou, Y., Lin, W.H. and Xue, H.W. (2006) MORN motifs in plant PlPKs are
involved in the regulation of subcellular localization and phospholipid binding.
Cell Res, 16, 466-478.

Ma, X. and Margolin, W. (1999) Genetic and Functional Analyses of the Conserved C-
Terminal Core Domain of Escherichia coli FtsZ. J. Bacteriol, 181, 7531-7544.

Maeda, 8., Sugita, C., Sugita, M. and Omata, T. (2006) A new class of signal
transducer in His-Asp phosphorelay systems. J Biol Chem, 281, 37868-37876.

Maple, J., Aldridge, C. and Moller, S.G. (2005) Plastid division is mediated by
combinatorial assembly of plastid division proteins. The Plant Journal, 43, 811-
823.

Maple, J., Chua, N.H. and Moller, S.G. (2002) The topological specificity factor
AtMinEl is essential for correct plastid division site placement in Arabidopsis.
Plant J, 31, 269-277.

Maple, J., Fujiwara, M.T., Kitahata, N., Lawson, T., Baker, N.R., Yoshida, S. and
Moller, S.G. (2004) GIANT CHLOROPLAST 1 Is Essential for Correct Plastid
Division in Arabidopsis. Current Biology, 14, 776-781.

Maple, J. and Moller, S.G. (2006) Plastid Division: Evolution, Mechanism and
Complexity. Ann Bot (Lond).

Maple, J. and Moller, S.G. (2007) Interdependency of formation and localisation of the
Min complex controls symmetric plastid division. J Cell Sci, 120, 3446-3456.

253

 

Maple, J., Vojta, L., 8011, J. and Moller, S.G. (2007) ARC3 is a stromal Z-ring
accessory protein essential for plastid division. EMBO Rep, 8, 293-299.

Marbouty, M., Saguez, C., Cassier-Chauvat, C. and Chauvat, F. (2009) ZipN, an
FtsA-like orchestrator of divisome assembly in the model cyanobacterium
Synechocystis PCC6803. Molecular Microbiology.

Margolin, W. (2003) Bacterial division: the fellowship of the ring. Curr Biol, 13, R16-
18.

Margolin, W. (2005) FtsZ and the division of prokaryotic cells and organelles. Nat Rev
Mol Cell Biol, 6, 862-871.

Marrison, J.L., Rutherford, S.M., Robertson, E.J., Lister, C., Dean, C. and Leech,
RM. (1999) The distinctive roles of five different ARC genes in the chloroplast
division process in Arabidopsis. Plant J, 18, 651-662.

Marston, A.L. and Errington, J. (1999) Selection of the midcell division site in

Bacillus subtilis through MinD-dependent polar localization and activation of
MinC. Mol Microbiol, 33, 84-96.

Masuda, H.P., Ramos, G.B., de Almeida-Engler, J., Cabral, L.M., Coqueiro, V.M.,
Macrini, C.M., Ferreira, RC. and Hemerly, AS. (2004) Genome based

identification and analysis of the pre-replicative complex of Arabidopsis thaliana.
FEBS Lett, 574, 192-202.

Mazouni, K., Domain, F., Cassier—Chauvat, C. and Chauvat, F. (2004) Molecular
analysis of the key cytokinetic components of cyanobacteria: FtsZ, ZipN and
MinCDE. Molecular Microbiology, 52, 1 145-1 158.

McAndrew, R.S., F roehlich, J.E., Vitha, S., Stokes, K.D. and Osteryoung, K.W.
(2001) Colocalization of plastid division proteins in the chloroplast stromal
compartment establishes a new functional relationship between FtsZ] and FtsZ2
in higher plants. Plant Physiol, 127, 1656-1666.

McAndrew, R.S., Olson, B.J.S.C., Kadirjan-Kalbach, D., Chi-Ham, C.L., Vitha, S.,
Froehlich, J.E. and Osteryoung, K.W. (2008) In vivo quantitative relationship
between plastid division proteins Ftle and FtsZ2 and identification of ARC6 and
ARC3 in a native F tsZ complex. Biochem J, 412, 367—378.

McCallum, C.M., Comai, L., Greene, EA. and Henikoff, S. (2000) Targeted screening
for induced mutations. Nat Biotechnol, 18, 455-457.

McCarty, J.8. and Walker, G.C. (1994) DnaK mutants defective in ATPase activity are

defective in negative regulation of the heat shock response: expression of mutant
DnaK proteins results in filamentation. J Bacteriol, 176, 764-780.

254

 

 

McDonald, MS. (2003) Photobiology of Higher Plants. West Sussex, England: Wiley.

McNiven, M.A. (1998) Dynamin: a molecular motor with pinchase action. Cell, 94, 151-
154.

Michniewicz, M., Zago, M.K., Abas, L., Weijers, D., Schweighofer, A., Meskiene, 1.,
Heisler, M.C., Ohno, C., Zhang, J., Huang, F., Schwab, R., Weigel, D.,
Meyerowitz, E.M., Luschnig, C., Offringa, R. and Friml, J. (2007)
Antagonistic regulation of PIN phosphorylation by PP2A and PINOID directs
auxin flux. Cell, 130, 1044-1056.

Miller, R.K., Cheng, S.C. and Rose, MD. (2000) Bimlp/Yeblp mediates the Kar9p-
dependent cortical attachment of cytoplasmic microtubules. Mol Biol Cell, 11,
2949-2959.

Mine, N., Bratman, S.Y., Basu, R. and Chang, F. (2009) Establishing new sites of
polarization by microtubules. Curr Biol, 19, 83-94.

Miyagishima, S.-y. (2005) Origin and evolution of the chloroplast division machinery.
Journal of Plant Research, 118, 295.

Miyagishima, S.-y., Froehlich, J.E. and Osteryoung, K.W. (2006) PDVl and PDV2
Mediate Recruitment of the Dynarnin-Related Protein ARC5 to the Plastid
Division Site. Plant Cell, 18, 2517-2530.

Miyagishima, S.-y., Kuroiwa, H. and Kuroiwa, T. (2001) The timing and manner of
disassembly of the apparatuses for chloroplast and mitochondrial division in the
red alga C yanidioschyzon merolae. Planta, V212, 517-528.

Miyagishima, S.-y., Nishida, K., Mori, T., Matsuzaki, M., Higashiyama, T.,
Kuroiwa, H. and Kuroiwa, T. (2003a) A Plant-Specific Dynarnin-Related
Protein Forms a Ring at the Chloroplast Division Site. Plant Cell, 15, 655-665.

Miyagishima, S.-y., Takahara, M., Mori, T., Kuroiwa, H., Higashiyama, T. and
Kuroiwa, T. (2001) Plastid Division Is Driven by a Complex Mechanism That

Involves Differential Transition of the Bacterial and Eukaryotic Division Rings.
Plant Cell, 13, 2257-2268.

Miyagishima, S.-y., Wolk, GP. and Osteryoung, K.W. (2005) Identification of
cyanobacterial cell division genes by comparative and mutational analyses.

Molecular Microbiology, 56, 126-143.

Miyagishima, S.Y., Nishida, K. and Kuroiwa, T. (2003b) An evolutionary puzzle:
chloroplast and mitochondrial division rings. Trends Plant Sci, 8, 432-438.

255

Miyagishima, S.Y., Nozaki, H., Nishida, K., Nishida, K., Matsuzaki, M. and
Kuroiwa, T. (2004) Two types of FtsZ proteins in mitochondria and red-lineage

chloroplasts: the duplication of FtsZ is implicated in endosymbiosis. J Mol Evol,
58, 291-303.

Moreira, 1.8., Femandes, P.A. and Ramos, M.J. (2006) Detailed Microscopic Study of
the Full ZipAthsZ Interface. Proteins, 63, 811-821.

Mosyak, L., Zhang, Y., Glasfeld, E., Haney, 8., Stahl, M., Seehra, J. and Somers,
W.S. (2000) The bacterial cell-division protein ZipA and its interaction with an
FtsZ fragment revealed by X-ray crystallography. EMBO J, 19, 3179-3191.

Mukherjee, A., Dai, K. and Lutkenhaus, J. (1993) Escherichia coli cell division
protein FtsZ is a guanine nucleotide binding protein. Proc Natl Acad Sci U SA,
90, 1053-1057.

Nakamura, M., Zhou, X.Z. and Lu, KR (2001) Critical role for the EB] and APC
interaction in the regulation of microtubule polymerization. Curr Biol, 11, 1062-
1067.

Nakanishi, H., Suzuki, K., Kabeya, Y. and Miyagishima, S.Y. (2009) Plant-specific
protein MCDl determines the site of chloroplast division in concert with bacteria-
derived MinD. Curr Biol, 19, 151-156.

Nakaya, M., Tsukaya, H., Murakami, N. and Kato, M. (2002) Brassinosteroids
control the proliferation of leaf cells of Arabidopsis thaliana. Plant Cell Physiol,
43, 239-244.

Neff, M.M., Neff, J.D., Chory, J. and Pepper, A.E. (1998) dCAPS, a simple technique
for the genetic analysis of single nucleotide polymorphisms: experimental
applications in Arabidopsis thaliana genetics. Plant J, 14, 387-392.

Obayashi, T., Hayashi, 8., Saeki, M., Ohta, H. and Kinoshita, K. (2009) ATTED-II
provides coexpressed gene networks for Arabidopsis. Nucleic Acids Res, 37,
D987-991.

Okazaki, K., Kabeya, Y., Suzuki, K., Mori, T., Ichikawa, T., Matsui, M., Nakanishi,
H. and Miyagishima, S.-y. (2009) The PLASTID DIVISIONI and 2
Components of the Chloroplast Division Machinery Determine the Rate of
Chloroplast Division in Land Plant Cell Differentiation. Plant Cell, 21, 1769-
1780.

Olson, B.J.S.C. (2008) Biochemical Analysis of the Chloroplast Division Proteins Ftle

and F tsZ2. In Department of Biochemistry and Molecular Biology. East Lansing:
Michigan State University.

256

Osawa, M., Anderson, DE. and Erickson, H.P. (2008) Reconstitution of contractile
F tsZ rings in liposomes. Science, 320, 792-794.

Osteryoung, K.W. (2001) Organelle Fission in Eukaryotes. Curr Opin Microbiol, 4,
639-646.

Osteryoung, K.W. and Nunnari, J. (2003) The division of endosymbiotic organelles.
Science, 302, 1698-1704.

Osteryoung, K.W., Stokes, K.D., Rutherford, S.M., Percival, A.L. and Lee, W.Y.
(1998) Chloroplast division in higher plants requires members of two functionally
divergent gene families with homology to bacterial ftsZ. Plant Cell, 10, 1991-
2004.

Osteryoung, K.W. and Vierling, E. (1995) Conserved cell and organelle division.
Nature, 376, 473.

Overbeek, R., Begley, T., Butler, R.M., Choudhuri, J.V., Chuang, H.Y., Cohoon, M.,
de Crecy-Lagard, V., Diaz, N., Disz, T., Edwards, R., Fonstein, M., Frank,
E.D., Gerdes, 8., Glass, E.M., Goesmann, A., Hanson, A., Iwata-Reuyl, D.,
Jensen, R., Jamshidi, N., Krause, L., Kubal, M., Larsen, N., Linke, B.,
McHardy, A.C., Meyer, F., Neuweger, H., Olsen, G., Olson, R., Osterman, A.,
Portnoy, V., Pusch, G.D., Rodionov, D.A., Ruckert, C., Steiner, J., Stevens,
R., Thiele, 1., Vassieva, O., Ye, Y., Zagnitko, O. and Vonstein, V. (2005) The
subsystems approach to genome annotation and its use in the project to annotate
1000 genomes. Nucleic Acids Res, 33, 5691-5702.

Overbeek, R., Fonstein, M., D'Souza, M., Pusch, G.D. and Maltsev, N. (1999) Use of
contiguity on the chromosome to predict functional coupling. In Silico Biol, 1, 93-
108.

Pagel, K. and Koksch, B. (2008) Following polypeptide folding and assembly With
conformational switches. Curr Opin Chem Biol, 12, 730-739. Epub 2008 Oct
2007.

Paris, 8., Wessel, P.M. and Dumas, R. (2002) Overproduction, purification, and
characterization of recombinant bifunctional threonine-sensitive aspartate kinase-
homoserine dehydrogenase from Arabidopsis thaliana. Protein Expr Part]; 24,
105-110.

Paves, H. and Truve, E. (2007) Myosin inhibitors block accumulation movement of
chloroplasts in Arabidopsis thaliana leaf cells. Protoplasma, 230, 165-169.

Pichoff, S. and Lutkenhaus, J. (2001) Escherichia coli division inhibitor MinCD blocks
septation by preventing Z-ring formation. J Bacteriol, 183, 6630-6635.

257

Pichoff, S. and Lutkenhaus, J. (2002) Unique and overlapping roles for ZipA and FtsA
in septal ring assembly in Escherichia coli. EMBO J, 21, 685-693.

Pichoff, S. and Lutkenhaus, J. (2005) Tethering the Z ring to the membrane through a
conserved membrane targeting sequence in FtsA. Mol Microbiol, 55, 1722-1734.

Pichoff, S. and Lutkenhaus, J. (2007) Identification of a region of FtsA required for
interaction with FtsZ. Mol Microbiol, 64, 1129-1138.

Possingham, J.V. and Lawrence, ME. (1983) Controls to plastid division.
International Review of Cytology, 84, 1-56.

Praefcke, G.J. and McMahon, H.T. (2004) The dynamin superfamily: universal
membrane tubulation and fission molecules? Nat Rev Mol Cell Biol, 5, 133-147.

Pyke, K. (1998) Plastid Division: The Origin of Replication. Plant Cell, 10, 1971-1972.

Pyke, K. (2006) Plastid Division: The Squeezing Gets Tense. Current Biology, 16, R60.

Pyke, K.A. (1999) Plastid Division and Development. Plant Cell, 11, 549-556.

Pyke, K.A. (2009) Plastid Biology lst edn. Cambridge: Cambridge University Press.

Pyke, K.A. and Leech, RM. (1991) Rapid Image Analysis Screening Procedure for
Identifying Chloroplast Number Mutants in Mesophyll Cells of Arabidopsis
thaliana (L.) Heynh. Plant Physiol, 96, 1193-1195.

Pyke, K.A. and Leech, RM. (1992) Chloroplast Division and Expansion ls Radically
Altered by Nuclear Mutations in Arabidopsis thaliana. Plant Physiol, 99, 1005-
1008.

Pyke, K.A. and Leech, RM. (1994) A Genetic Analysis of Chloroplast Division and
Expansion in Arabidopsis thaliana. Plant Physiol, 104, 201-207.

Pyke, K.A., Rutherford, S.M., Robertson, E.J. and Leech, RM. (1994) arc6, A Fertile
Arabidopsis Mutant with Only Two Mesophyll Cell Chloroplasts. Plant Physiol,,
106, 1169-1177.

Raskin, D.M. and de Boer, P.A. (1999a) MinDE-dependent pole-to-pole oscillation of
division inhibitor MinC in Escherichia coli. J Bacteriol, 181, 6419—6424.

Raskin, D.M. and de Boer, P.A. (1999b) Rapid pole-to-pole oscillation of a protein

required for directing division to the middle of Escherichia coli. Proc Natl Acad
Sci U S A, 96, 4971-4976.

258

RayChaudhuri, D. (1999) ZipA is a MAP-Tau homolog and is essential for structural
integrity of the cytokinetic FtsZ ring during bacterial cell division. EMBO J, 18,
23 72-23 83.

Raynaud, C., Cassier-Chauvat, C., Perennes, C. and Bergounioux, C. (2004) An
Arabidopsis homolog of the bacterial cell division inhibitor SulA is involved in
plastid division. Plant Cell, 16, 1801-1811.

Raynaud, C., Perennes, C., Reuzeau, C., Catrice, 0., Brown, S. and Bergounioux, C.
(2005) Cell and plastid division are coordinated through the prereplication factor
AtCDTl. PNAS, 102, 8216-8221.

Reddy, M.S., Dinkins, R. and Collins, GB. (2002) Overexpression of the Arabidopsis
thaliana MinEl bacterial division inhibitor homologue gene alters chloroplast size

and morphology in transgenic Arabidopsis and tobacco plants. Planta, 215, 167-
176.

Reski, R., Wehe, M., Hadeler, B., Marienfield, J.R. and Abel, W.O. (1991) Cytokinin
and light quality interact at the molecular level in the chloroplast mutant PC22 of
the moss Physcomitrella. J. Plant Physiol, 138, 236-243.

Ricard, M. and Hirota, Y. (1973) Process of cellular division in Escherichia coli:
physiological study on thermosensitive mutants defective in cell division. J
Bacteriol, 116, 314-322.

Robertson, E.J., Pyke, K.A. and Leech, RM. (1995) arc6, an extreme chloroplast
division mutant of Arabidopsis also alters proplastid proliferation and
morphology in shoot and root apices. J Cell Sci, 108, 2937-2944.

Rost, B. and Eyrich, V.A. (2001) EVA: large-scale analysis of secondary structure
prediction. Proteins, Suppl, 192-199.

Rothfield, L., Taghbalout, A. and Shih, Y.L. (2005) Spatial control of bacterial
division-site placement. Nat Rev Microbiol, 3, 959-968.

Ruiz, D., Salinas, P., Lopez-Redondo, M.L., Cayuela, M.L., Marina, A. and
Contreras, A. (2008) Phosphorylation-independent activation of the atypical
response regulator NblR. Microbiology, 154, 3002-3015.

Rutherford, SM. (1996) The genetic and physical analysis of mutants of chloroplast
number and size in Arabidopsis thaliana: University of York, UK.

Saitou, N. and Nei, M. (1987) The neighbor-joining method: a new method for
reconstructing phylogenetic trees. Mol Biol Evol, 4, 406-425.

259

Sakr, S., Jeanjean, R., Zhang, C.-C. and Arcondeguy, T. (2006) Inhibition of Cell
Division Suppresses Heterocyst Development in Anabaena sp. Strain PCC 7120.
J. Bacteriol, 188, 1396-1404.

Scheffers, DJ. (2008) The effect of MinC on FtsZ polymerization is pH dependent and
can be counteracted by ZapA. F EBS Lett, 582, 2601-2608.

Schimper, A. (1883) Uber die Entwicklung der Chlorophyllkdmer und Farbkdrper. Bot
Zeit, 41, 105-112.

Schmidt von Braun, S. and Schleiff, E. (2008) The chloroplast outer membrane protein
CHUPl interacts with actin and profilin. Planta, 227, 1151-1159.

Schmitz, A.J., Glynn, J.M., Olson, B.J.S.C., Stokes, K.D. and Osteryoung, K.W.
(2009) Arabidopsis F tsZZ-l and FtsZZ-2 are Functionally Redundant, but FtsZ-
based Plastid Division is Not Essential for Chloroplast Partitioning or Plant
Growth and Development. Molecular Plant, 2, 1211-1222.

Schneider, D., Fuhrmann, E., Scholz, 1., Hess, W.R. and Graumann, P.L. (2007)
Fluorescence staining of live cyanobacterial cells suggest non-stringent

chromosome segregation and absence of a connection between cytoplasmic and
thylakoid membranes. BMC Cell Biol, 8, 39.

Schwacke, R., Schneider, A., van der Graaff, E., Fischer, K., Catoni, E., Desimone,
M., Frommer, W.B., Flugge, U.I. and Kunze, R. (2003) ARAMEMNON, a
novel database for Arabidopsis integral membrane proteins. Plant Physiol, 131,
16-26.

Scott, NS. and Possingham, J.V. (1980) Chloroplast DNA in Expanding Spinach
Leaves. J. Exp. Bot, 31, 1081-1092.

Shaw, J.M. and Nunnari, J. (2002) Mitochondrial dynamics and division in budding
yeast. Trends Cell Biol, 12, 178-184.

Shen, B. and Lutkenhaus, J. (2009) The conserved C-terminal tail of FtsZ is required
for the septal localization and division inhibitory activity of MinC(C)/MinD. Mol
Microbiol, 72, 410-424.

Shimada, H., Koizumi, M., Kuroki, K., Mochizuki, M., Fujimoto, H., Ohta, H.,
Masuda, T. and Takamiya, K.-i. (2004) ARC3, a Chloroplast Division Factor, is
a Chimera of Prokaryotic FtsZ and Part of Eukaryotic Phosphatidylinositol-4-
phosphate 5-kinase. Plant Cell Physiol, 45, 960-967.

Singh, J.K., Makde, R.D., Kumar, V. and Panda, D. (2007) A membrane protein,

Eer, regulates assembly dynamics of FtsZ by interacting with the C-terrninal tail
of F tsZ. Biochemistry, 46, 11013-11022.

260

Slep, K.C., Rogers, S.L., Elliott, S.L., Ohkura, H., Kolodziej, P.A. and Vale, RD.
(2005) Structural determinants for EBl-mediated recruitment of APC and
spectraplakins to the microtubule plus end. J Cell Biol, 168, 587-598.

Slep, KC. and Vale, RD. (2007) Structural basis of microtubule plus end tracking by
XMAP215, CLIP-170, and EB1. Mol Cell, 27, 976-991.

Sokolovski, S., Hills, A., Gay, R., Garcia-Mata, C., Lamattina, L. and Blatt, M.R.
(2005) Protein phosphorylation is a prerequisite for intracellular Ca2+ release and

ion channel control by nitric oxide and abscisic acid in guard cells. Plant J, 43,
520-529.

Stokes, K.D., McAndrew, R.S., Figueroa, R., Vitha, S. and Osteryoung, K.W. (2000)
Chloroplast division and morphology are differentially affected by overexpression
of Ftle and FtsZ2 genes in Arabidopsis. Plant Physiol, 124, 1668-1677.

Stokes, K.D. and Osteryoung, K.W. (2003) Early divergence of the Ftle and FtsZ2
plastid division gene families in photosynthetic eukaryotes. Gene, 320, 97.

Strepp, R., Scholz, S., Kruse, S., Speth, V. and Reski, R. (1998) Plant nuclear gene
knockout reveals a role in plastid division for the homolog of the bacterial cell
division protein FtsZ, an ancestral tubulin. Proc Natl Acad Sci U S A, 95, 4368-
4373.

Su, L.K., Burrell, M., Hill, D.E., Gyuris, J., Brent, R., Wiltshire, R., Trent, J.,
Vogelstein, B. and Kinzler, K.W. (1995) APC binds to the novel protein EBl.
Cancer Res, 55, 2972-2977.

Su, P.H. and Li, H.M. (2008) Arabidopsis stromal 70-kD heat shock proteins are
essential for plant development and important for therrnotolerance of germinating
seeds. Plant Physiol, 146, 1231-1241.

Suetsugu, N. and Wada, M. (2007) Chloroplast photorelocation movement mediated by
phototropin family proteins in green plants. Biol Chem, 388, 927-935.

Sugimoto, S., Saruwatari, K., Higashi, C. and Sonomoto, K. (2008) The proper ratio
of GrpE to DnaK is important for protein quality control by the DnaK-DnaJ-GrpE
chaperone system and for cell division. Microbiology, 154, 1876-1885.

Sun, Q. and Margolin, W. (2001) Influence of the nucleoid on placement of FtsZ and
MinE rings in Escherichia coli. J Bacteriol, 183,, 1413-1422.

Sun, 0., Zybailov, B., Majeran, W., Friso, G., Olinares, P.D.B. and van Wijk, K.J.

(2009) PPDB, the Plant Proteomics Database at Cornell. Nucl. Acids Res, 37,
D969-974.

261

Szeto, T.H., Rowland, S.L., Rothfield, LI. and King, G.F. (2002) Membrane
localization of MinD is mediated by a C-terrninal motif that is conserved across
eubacteria, archaea, and chloroplasts. PNAS, 99, 15693-15698.

Tamura, K., Dudley, J., Nei, M. and Kumar, S. (2007) MEGA4: Molecular
Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol, 24,
1596-1599.

Thomer, M., May, N.R., Aggarwal, B.D., Kwok, G. and Calvi, BR. (2004)
Drosophila double-parked is sufficient to induce re-replication during
development and is regulated by cyclin E/CDK2. Development, 131, 4807-4818.

Till, B.J., Colbert, T., Codomo, C., Enns, L., Johnson, J., Reynolds, S.H., Henikofl',
J.G., Greene, E.A., Steine, M.N., Comai, L. and Henikofl', S. (2006) High-
throughput TILLING for Arabidopsis. Methods Mol Biol, 323, 127-135.

Toufighi, K., Brady, S.M., Austin, R., Ly, E. and Provart, NJ. (2005) The Botany
Array Resource: e-Northems, Expression Angling, and promoter analyses. Plant
J, 43, 153-163.

Tranel, P.J., Froehlich, J., Goyal, A. and Keegstra, K. (1995) A component of the
chloroplastic protein import apparatus is targeted to the outer enve10pe membrane
via a novel pathway. EMBO J, 14, 2436-2446.

Trusca, D., Scott, S., Thompson, C. and Bramhill, D. (1998) Bacterial SOS
Checkpoint Protein SulA Inhibits Polymerization of Purified FtsZ Cell Division
Protein. J. Bacteriol, 180, 3946-3953.

Tzafrir, 1., Pena-Muralla, R., Dickerman, A., Berg, M., Rogers, R., Hutchens, S.,
Sweeney, T.C., McElver, J., Aux, G., Patton, D. and Meinke, D. (2004)
Identification of genes required for embryo development in Arabidopsis. Plant
Physiol, 135, 1206-1220.

Uehara, T., Matsuzawa, H. and Nishimura, A. (2001) HscA is involved in the
dynamics of FtsZ-ring formation in Escherichia coli K12. Genes Cells, 6, 803-
814.

Ungewickell, E.J. and Hinrichsen, L. (2007) Endocytosis: clathrin-mediated membrane
budding. Curr Opin Cell Biol, 19, 417-425.

van den Ent, F. and Lowe, J. (2000) Crystal structure of the cell division protein FtsA
from Therrnotoga maritima. EMBO J, 19, 5300-5307.

Vitha, S., Froehlich, J .E., Koksharova, 0., Pyke, K.A., van Erp, H. and Osteryoung,
K.W. (2003) ARC6 Is a J-Domain Plastid Division Protein and an Evolutionary

262

Descendant of the Cyanobacterial Cell Division Protein F m2. Plant Cell, 15,
1918-1933.

Vitha, S., Holzenburg, A. and Osteryoung, K.W. (2005) Assembly Dynamics of FtsZ
and ARC6 Plastid-Dividing Proteins. Microsc Microanal, 11, 1154-1155.

Vitha, S., McAndrew, RS. and Osteryoung, K.W. (2001) FtsZ Ring Formation at the
Chloroplast Division Site in Plants. The Journal of Cell Biology, 153, 111-1 19.

Wada, M., Kagawa, T. and Sato, Y. (2003) Chloroplast movement. Annu Rev Plant
Biol, 54, 455-468.

Wang, Q., Canutescu, A.A. and Dunbrack, R.L., Jr. (2008) SCWRL and MolIDE:
computer programs for side-chain conformation prediction and homology
modeling. Nat Protoc, 3, 1832-1847.

Warrens, A.N., Jones, MD. and Lechler, RJ. (1997) Splicing by overlap extension by
PCR using asymmetric amplification: an improved technique for the generation of
hybrid proteins of immunological interest. Gene, 186, 29-35.

Wen, Y., Eng, C.I-I., Schmoranzer, J., Cabrera-Poch, N., Morris, E.J., Chen, M.,
Wallar, B.J., Alberts, A.S. and Gundersen, G.G. (2004) EB] and APC bind to

mDia to stabilize microtubules downstream of Rho and promote cell migration.
Nat Cell Biol, 6, 820-830.

Whitmore, L. and Wallace, B.A. (2008) Protein secondary structure analyses from

circular dichroism spectroscopy: methods and reference databases. Biopolymers,
89, 392-400.

Wiegel, D. and Glazebrook, J. (2002) Arabidopsis: A Laboratory Manual. New York:
Cold Spring Harbor Press.

Wiejak, J. and Wyroba, E. (2002) Dynamin: characteristics, mechanism of action and
function. Cell Mol Biol Lett, 7, 1073-1080.

Xu, C., Fan, J., Cornish, A.J. and Benning, C. (2008) Lipid trafficking between the
endoplasmic reticulum and the plastid in Arabidopsis requires the extraplastidic
TGD4 protein. Plant Cell, 20, 2190-2204.

Xu, S., Wong, C.C., Tong, E.H., Chung, S.S., Yates, J.R., 3rd, Yin, Y. and Ko, B.C.
(2008) Phosphorylation by casein kinase 1 regulates tonicity-induced osmotic
response element-binding protein/tonicity enhancer-binding protein
nucleocytoplasmic trafficking. J Biol Chem, 283, 17624-17634. Epub 12008 Apr
17614.

263

Yang, X., Kang, C.M., Brody, MS. and Price, C.W. (1996) Opposing pairs of serine
protein kinases and phosphatases transmit signals of environmental stress to
activate a bacterial transcription factor. Genes Dev, 10, 2265-2275.

Yang, Y., Glynn, J.M., Olson, B.J.S.C., Schmitz, A.J. and Osteryoung, K.W. (2008)
Plastid division: across time and space. Curr Opin Plant Biol, 11, 577-584.

Yang, Y., Xu, R., Ma, C.J., Vlot, A.C., Klessig, D.F. and Pichersky, E. (2008) Inactive
methyl indole-3-acetic acid ester can be hydrolyzed and activated by several

esterases belonging to the AtMES esterase family of Arabidopsis. Plant Physiol,
147, 1034-1045.

Yoder, D.W., Kadirjan-Kalbach, D., Olson, B.J.S.C., Miyagishima, S.Y., Deblasio,
S.L., Hangarter, RP. and Osteryoung, K.W. (2007) Effects of mutations in
Arabidopsis F tle on plastid division, FtsZ ring formation and positioning, and
FtsZ filament morphology in vivo. Plant Cell Physiol, 48, 775-791.

Yoon, H.S., Hackett, J.D., Ciniglia, C., Pinto, G. and Bhattacharya, D. (2004) A
molecular timeline for the origin of photosynthetic eukaryotes. Mol Biol Evol, 21,
809-818.

Yoshida, Y., Kuroiwa, H., Misumi, O., Nishida, K., Yagisawa, F., Fujiwara, T.,
Nanamiya, H., Kawamura, F. and Kuroiwa, T. (2006) Isolated Chloroplast
Division Machinery Can Actively Constrict After Stretching. Science, 313, 1435-
1438.

Zhang, M., Hu, Y., Jia, J., Gao, H. and He, Y. (2009a) A plant MinD homologue
rescues Escherichia coli HLl mutant (DeltaMinDE) in the absence of MinE. BMC
Microbiol, 9, 101.

Zhang, M., Hu, Y., Jia, J., Li, D., Zhang, R., Gao, H. and He, Y. (2009b) CDPl, a
novel component of chloroplast division site positioning system in Arabidopsis.
Cell Res, 19, 877-886.

Zhang, X.P., Elofsson, A., Andreu, D. and Glaser, E. (1999) Interaction of
mitochondrial presequences with DnaK and mitochondrial hsp70. J Mol Biol,
288, 177-190.

Zimmermann, P., Hirsch-Hoffmann, M., Hennig, L. and Gruissem, W. (2004)
GENEVESTIGATOR. Arabidopsis Microarray Database and Analysis Toolbox.
Plant Physiol. , 136, 2621 -2632.

Zuckerkandl, E. and Pauling, L. (1965) Molecules as documents of evolutionary
history. J Theor Biol, 8, 357-366.

264