”£9... ‘11; , ‘ . ‘ ‘ . it.“ .. . . . . ‘ .. . . ‘ . flwxsnwvfl guauui.... . ‘ . r , n u 2 1‘ m... :26... . ‘m\ t7. . tub :fififii 1 nu... a? 23...: .a . 57. ,. . , L. .. . T I. i .41 .wa , . “#3.. $33.70 .1 .. hiya LIBRARY Michigan State ,7 L015 University This is to certify that the dissertation entitled BIOCHEMICAL ANALYSIS OF THE CHLOROPLAST DIVISION PROTEINS FTSZ‘I AND FTSZZ presented by BRADLEY JESSE STANFORD CARNAHAN OLSON has been accepted towards fulfillment of the requirements for the Doctoral degree in Biochemistry and Molecular Biology 42:441UL;€/JVDJ/( c/g/zm _ Major Professor’s Si ature mam (gr (fig Date MS U is an affirmative-action, equal-opportunity employer --.-—---r—.-n—-—-.--a-.-—--—-.- o.—.—-—v—.-.—-A—.-—.—-n-u--.—— - 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 5108 KilProi/Accspres/CIRC/DateDue indd BIOCHEMICAL ANALYSIS OF THE CHLOROPLAST DIVISION PROTEINS FTSZl AND FTSZZ Bv Bradley Jesse Stanford Carnahan Olson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry and Molecular Biology 2008 ABSTRACT BIOCHEMICAL ANALYSIS OF THE CHLOROPLAST DIVISION PROTEINS FTSZl AND FI'SZZ By Bradley Jesse Stanford Carnahan Olson Chloroplast division in plants is mediated by Ftle and FtsZZ, which are homologues of the bacterial cell division protein FtsZ. The bacterial FtsZ protein is thought to be the progenitor to tubulin. FtsZ is a GTP-dependent, filament-forming protein that encircles the bacterial division site. Similarly, plant FtsZ proteins encircle the mid-plastid. Plant Ftle and FtSZZ family proteins differ primarily at their C-termini with FtsZZ proteins possessing a motif similar to those found in bacterial FtsZ that is responsible for binding to other bacterial cell division factors. In plants, this motif has been shown to bind to the chloroplast division protein ARC6, which may be involved in FtsZ filament remodeling. The focus of this dissertation is on understanding the composition of the in vivo FtsZ complex and understanding the function and topology of FtsZ protofilaments. To understand the composition of the FtsZ complex in vivo, FtsZ was purified from pea stroma. Pea Ftle and FtsZZ co-fractionate with the chloroplast division proteins ARCS and ARCS when separated by sucrose density gradient, anion-exchange chromatography, Size-exclusion chromatography and native polyacrylamide gel electrophoresis. In addition, a ZipA-antibody cross-reactive protein was found to co- fractionate with Ftle/FtsZZ/ARCB/ARCG-containing complexes. Data from these experiments suggest that a minimal unit of FtsZ assembly is a complex containing Ftle, FtsZZ, ARC3, ARCS and protein that cross-reacts with an antibody to E. coli ZipA. The identify of the ZipA cross-reactive protein was not able to be determined Plant FtsZZ contains a C—terminus similar to the ZipA-binding C-terminus of bacterial FtsZ. However, ZipA has not been identified in plants. A structural search algorithm was created find a chloroplastic protein that is structurally similar to ZipA or to identify the protein that cross-reacts with the E. coli ZipA antibody. This algorithm identified a strong ZipA candidate, $521, which was found to be a stromal protein. A homozygous sszl mutant has slightly fewer and enlarged chloroplasts. However, 5521 did not interact with FtsZZ in the yeast two-hybrid assay. At this time 5521 can not be definitively linked to a role in chloroplast division. To understand the biochemical properties of plant FtsZ, recombinant FtsZ1 and FtSZZ were produced individually in E. coli. Ftle and FtsZZ were both found to be GT Pases, but had approximately 15-fold lower GTPase activity than E. coli FtsZ. When mixed equally, Ftle and FtsZZ co-assemble into bundled ribbon-like structures. This result differs from E. coli FtsZ, which only assembles into single protofilaments when polymerized under similar conditions. Maximal Ftle/FtsZZ co-assembly requires equal concentration of both proteins. In addition, a T7-loop mutant, FtsZZDBZZA , was found to be a sub-stoichiometric competitive inhibitor of Ftle, but not FtsZZ. Together these results support an FtZl/FtsZZ heterofilament model similar to ocB-tubulin. Dedication To my wife Valerie ACKNOWLEDGMENTS Special thanks to fellowship support from the Department of Biochemistry and Molecular Biology with a N. E. Tolbert Memorial fellowship and from the Michigan State University Plant Sciences fellowship program. In addition, special thanks to the MSU- DEO Plant Research Laboratory for supporting travel and professional development. This work was supported by the National Science Foundation. Jonathan Glynn, Aaron Schmitz, Dr. Cecilia Chi-Ham, Dr. Shin-ya Miyagishima, Dr. Stanislav Vitha and Dr.Yue Yang were invaluable resources to discuss ideas and for critical assessment of this work. Thanks to Valerie Olson for editorial assistance Special thanks to Dr. Alicia Pastor for training and support of electron microscopy Dr. Robert Hausinger and Dr. Kenneth Keegstra deserve special thanks for supporting pursuit of external awards. TABLE OF CONTENTS List of Tables ................................................................................................................ x List of Figures ............................................................................................................... xi Key to Abbreviations .................................................................................................. xiii Introduction ................................................................................................................. 1 Specific Aims ..................................................................................................... 2 Bacterial cell division is best understood in E. coli ............................................. 5 FtsZ ............................................................................................................. 5 Tubulin biochemistry ....................................................................................... 6 E. coli FtsZ is structurally and biochemically similar to tubulin ..................... 10 FtsZ polymerization and structure ................................................................. 11 E. coli FtsZ GTP hydrolysis drives depolymerization ...................................... 13 Cooperative assembly of bacterial FtsZ ......................................................... 15 Model Of FtsZ polymerization ........................................................................ 16 Regulation of Z-ring assembly and placement ................................................. 19 Nucleoid occlusion positions the Z-ring in bacteria ...................................... 19 Positioning of the Z-ring by the MinCDE system ........................................... 20 ZipA and FtSA organize and anchor the Z-ring at the membrane ................. 21 ZapA ........................................................................................................... 22 Eer ........................................................................................................... 23 SulA ........................................................................................................... 23 Chloroplast Division ........................................................................................ 24 Ultrastructural Observations of chloroplast division ..................................... 24 The stromal chloroplast division machinery .................................................. 25 The eukaryotic-derived chloroplast division machinery ............................... 30 Perspective ..................................................................................................... 32 In vivo quantitative relationship between plastid division proteins Ftle and FtsZZ and identification of ARCG and ARC3 in a native FtsZ complex. .............................. 33 Synopsis .......................................................................................................... 34 Introduction .................................................................................................... 35 Experimental Procedures ................................................................................ 37 Plant material ................................................................................................ 37 Chloroplast morphology phenotyping ........................................................... 38 vi lmmunofluorescence labeling ....................................................................... 38 Preparation of antibodies .............................................................................. 39 Expression, purification and calibration of recombinant protein standards 40 Chloroplast isolation ...................................................................................... 43 Determination of chloroplast volume ........................................................... 43 Quantitative immunoblotting ........................................................................ 44 lmmunoprecipitation of in vitro FtsZ translation products before and after import into isolated intact chloroplasts ............................................ 45 Isolation of an endogenous FtsZ complex by two-dimensional gel electrophoresis .................................................................................. 47 Isolation and analysis of FtsZ-containing complexes from Arabidopsis and pea ..................................................................................................... 48 Results ............................................................................................................ 50 Specificity of Arabidopsis FtsZ antibodies ..................................................... 50 AtFtsZZ-Z functions in chloroplast division ................................................... 51 Quantitative analysis of FtsZ levels in Arabidopsis chloroplasts ................... 54 AtFtsZ proteins are stably associated following import into pea chloroplasts ........................................................................................................... 64 Ftle and FtsZZ are in a stable complex in vivo ............................................ 66 The native FtsZ complex also contains ARC6 and ARC3 ................................ 76 Discussion ....................................................................................................... 82 Supplementary Results ................................................................................... 87 Preparation of Quantitative lmmunoblotting Standards .............................. 87 Specificity of the ARC6 Antibody ................................................................... 88 Acknowledgments .......................................................................................... 89 Plant Ftle and FtSZZ are GT Pases and Assemble into Heterofilaments that Bundle into Ribbons ........................................................................................................... 90 Introduction .................................................................................................... 91 Experimental Procedures ................................................................................ 98 Expression and purification of recombinant FtsZ proteins ........................... 98 Site-directed mutagenesis ........................................................................... 103 GTP binding and hydrolysis assays .............................................................. 104 Polymerization assays .................................................................................. 105 Results .......................................................................................................... 107 Production and refolding of recombinant Arabidopsis Ftle-l and FtsZZ-l ......................................................................................................... 107 Both FtsZ1 and FtsZZ bind and hydrolyze GTP ............................................ 110 Do Ftle and FtsZZ assemble? ....................................................................... 112 Ftle/FtsZZ co-assembly EM and light-scattering ....................................... 112 Ftle and FtsZ2 homofilaments are unstable ............................................. 120 Ftle/FtsZZ co-polymer length correlates to GTPase activity ..................... 121 The critical concentration for polymerization ............................................. 123 vii Ftle/FtsZZ co-polymers are larger and more stable than E. coli FtsZ polymers .......................................................................................... 124 Ftle/FtsZZ copolymers are not disassembled by GDP .............................. 127 Stoichiometric requirements for GTPase activity and polymerization ....... 128 Adding additional Ftle or FtsZZ to Ftle/FtsZZ polymerization assays decreases the extent of polymerization .......................................... 128 FtsZZ-1 promotes lateral bundling of Ftle-1/FtsZZ-1 ribbons ................... 129 FtsZZD322A competitively inhibits the polymerization of FtsZ1 ................ 135 FtsZZD322A inhibits the GTPase activity of Ftle, but not FtsZZ ................ 136 Plant FtsZ and T7-loop mutants assemble into rings in E. coli .................... 139 Discussion 139 Previous in vitro studies .............................................................................. 139 Ftle and FtsZZ polymerize into heterofilaments ....................................... 142 The in vivo stoichiometry of Ftle and FtsZZ .............................................. 144 Heterofilament bundling ............................................................................. 146 FtsZ1/FtsZZ are different than tubulin and bacterial FtsZ ........................... 147 Why do plants require two types of FtsZ? ................................................... 148 Acknowledgements ..................................................................................... 149 l n D I I l : Eu The search for an analogue to the bacterial cell division protein ZipA by computational and experimental approaches 151 i: Introduction 152 ; Results 154 ; Chloroplasts contain a protein that cross-reacts with a ZipA antibody and co- 1% fractionates with FtsZ during partial purification ............................ 154 4. Identification of a structural analogue to the bacterial cell division protein ZipA f, 153 52‘ Phenotype of a T-DNA insertional mutant for $521 ..................................... 163 ‘2" Sszl is a Stromal Chloroplast protein 163 E Removal of the $521 homologue in Synecococcus elongatus does not disrupt I: FtsZ filament morphology ................................................................ 168 i 5521 does not interact with FtsZZ in the yeast-two hybrid assay ................ 169 3+ Discussion 176 i Structure Similarity Search Algorithm Development .................................. 176 1} Materials and Methods 181 i Generation of a SszlA strain of Synechococcus PCC 7942 .......................... 181 .3 Database of Putative Chloroplast Proteins for Threading ........................... 182 ; Threading of the Unknown Chloroplast Proteins ........................................ 182 ‘5‘.- Growth Conditions of Plants ........................................................................ 182 Microscopy of Fixed Arabidopsis Tissue ...................................................... 183 Genotype of SALK insertions in 5521 ............................................................ 183 Yeast-two Hybrid Assay ............................................................................... 184 viii Summary and Future Directions ............................................................................... 189 Summary ...................................................................................................... 190 Future Directions .......................................................................................... 191 Structural Search Algorithms ....................................................................... 191 The protein that cross-reacts with the E. coli ZipA antibody ...................... 193 Chloroplast division protein purification from inclusion bodies ................. 194 FtsZ biochemistry ......................................................................................... 196 Probing heterofilament formation by Ftle and FtsZZ ............................... 196 Ftle and FtsZZ lateral interactions ............................................................. 200 The biochemical effects of ARC6 and ARC3 ................................................. 201 Concluding remarks ...................................................................................... 203 Appendix A .............................................................................................................. 204 References ............................................................................................................... 222 LIST OF TABLES Table 2.1: Effects of whole plant age on FtsZ in Arabidopsis chloroplasts .......... 60 Table 2.2: Comparison of FtsZ levels in actively dividing E. coli cells with those in chloroplasts of 3-week-Old Arabidopsis ................................................... 62 Table 3.1: GTP binding and hydrolysis properties of Ftle and FtsZZ compared to E. coli FtsZ .......................................................................................................... 106 Table 4.1: Candidate proteins structurally similar to E. coli ZipA ...................... 161 Table 4.2: Residue contact comparison between ZipA bound to E. coli FtsZ and $521 bound to the plant FtsZZ C-terminus ...................................................... 167 Table 4.3: Summary of a yeast two-hybrid interaction assay between $521 and Ftle ................................................................................................................. 178 LIST OF FIGURES Figure 1.1: Cartoon representation sequence similarity between plant Ftle, FtsZZ, bacterial FtsZ and a- and B-tubulin ............................................................. 4 Figure 1.2: Model of the assembly dynamics Of E. coli FtsZ ................................ 18 Figure 2.1: Expression and purification of recombinant protein standards for quantitative analysis ................................................................................. 42 Figure 2.2: Antibody specificity for AtFtsZ1-1, AtFtsZZ-l, and AtFtsZZ-Z ............ 52 Figure 2.3: Chloroplast morphology and AtFtSZZ-l and AtFtsZZ-Z localization in WT Arabidopsis and atftsZZ-l and atftsZZ-Z T-DNA insertional mutants ...... 56 Figure 2.4: Linear ranges of detection of FtsZ protein standards and FtsZ in chloroplast lysates for quantitative immunoblotting .................................................. 59 Figure 2.5: Quantitative analysis of molecular levels of Ftle and FtsZZ in wild-type Arabidopsis ................................................................................................ 63 Figure 2.6: lmmunoprecipitation of in vitro [3SS]-Iabeled AtFtsZ translation and import products by FtsZ-specific antibodies ................................................. 68 &70 Figure 2.7: Ftle and FtsZZ cofractionate in isolated pea chloroplasts .............. 72 Figure 2.8: 2-D native SDS-PAGE separation of an endogenous protein complex from pea chloroplasts containing Ftle and FtsZZ ............................................ 74 Figure 2.9: Specificity of ARC6 Antibody .............................................................. 78 Figure 2.10: ARC6 and ARC3 are associated with a protein complex containing FtsZ1 and FtsZZ in both pea and Arabidopsis ............................................................ 81 Figure 3.1: Sequence alignment of AtFtle-l, AtFtsZZ-Z and E. coli FtsZ ...94 8: 95 Figure 3.2: GTPase kinetics of FtsZ1, FtsZZ, equally mixed FtsZ1/FtsZZ and E. coli FtsZ ................................................................................................................. 114 Figure 3.3: 90° Light-scattering polymerization assays of Ftle and FtsZZ ....... 117 Figure 3.4: Negative stain electron micrographs of Ftle and FtsZZ ................ 119 xi Figure 3.5: Critical concentration for FtsZ1/FtsZZ polymerization monitored by 90° light- scattering ................................................................................................ 125 Figure 3.6: Light-scattering comparison Of polymerization of E. coli FtsZ and Ftle with FtsZZ both at 5 [1M in HMK buffer .......................................................... 126 Figure 3.7: Equally mixed Ftle and FtsZZ are not depolymerized by GDP ...... 130 Figure 3.8: The effect of varying the ratio between Ftle and FtsZZ, while keeping the total FtsZ concentration at 5 uM monitored by light-scattering ............ 132 Figure 3.9: Inhibition of Ftle polymerization by FtsZZDSZZA .......................... 134 Figure 3.10: Inhibition of the GTPase activity of Ftle and FtsZZ by FtsZZD322A in 2 mM GTP in a coupled GTPase assay ............................................................... 138 Figure 3.11: Fluorescence microscopy of plant FtsZs fused to GFP expressed in E. coli ................................................................................................................. 142 Figure 3.12: Model of plant FtsZ protofilaments ............................................... 150 Figure 4.1: Ftle, FtsZZ, ARC6 and a protein that cross-reacts with an antibody to E. coli ZipA cofractionate in sucrose density gradients, native PAGE and by hydrophobic interaction chromatography ................................................................... 156 Figure 4.2: Overview of the structural threading search algorithm used to identify a putative Arabidopsis chloroplastic protein with structural similarity to ZipA, $521 ................................................................................................................. 159 Figure 4.3: Alignment of E. coli ZipA with its potential structural analogue in Arabidopsis ................................................................................................................. 166 Figure 4.4: PCR analysis to identify a homozygous SALK_039451 mutant of in $521 ................................................................................................................. 171 Figure 4.5: Comparison of WT and Salk 039451 cells ........................................ 173 Figure 4.6: Chloroplast import assay of $521 ..................................................... 174 Figure 4.7: Immunolocalization of FtsZ in wild-type and SszlA Synechoccocus PCC 7942 ................................................................................................................ 175 Figure 4.8: ClustalW alignment of plant and cyanobacterial Sszl homologues ............................................................................................... 186, 187 & 188 xii ATP ABRC Cc C-termi nus DDM ALS EM EMS E-Site GDP GTP'yS KEY TO ABBREVIATIONS adenosine 5'-triphosphate Arabidopsis Biological Resource Center critical concentration for assembly ca rboxyl terminus of a protein n-dodecyl-B-D-maltoside change in light scattering transmission electron microscopy ethyl methanesulfonate exchangeable GTP binding site of tubulin guanosine 5'-diphosphate guanosine 5'-diphosphate with the hydrolyzed y—P; still bound to the protien y-inorganic phosphate liberated from ATP or GTP guanosine 5'-triphosphate guanosine-5’-O-(3-thiotriphosphate) xiii y—TURC His IEM Min MORN MTOC N-site N-terminus OEM PAGE PD-ring PIP-5K QAAA RuBisCo SDS y—tubulin ring complex histidine inner envelope membrane minute(s) membrane occupation nexus motif Microtubule organization complex non-exchangeable GTP binding stie of tubulin amino terminus of a protein outer envelope membrane polyacrylamide gel electrophoresis plastid dividing ring inorganic phosphate phosphatidyl inositoyl-S-phosphate kinase quantitative amino acid analysis of a protien ribulose 1-5-bisphosphate carboxylase/oxygenase sodium dodecyl sulfate xiv sec second(s) sRuBP small subunit of RuBisCO CHAPTER ONE INTRODUCTION Specific Aims Chloroplast division in plants is mediated by Ftle and FtsZZ, which are homologues of the bacterial cell division protein FtsZ. The bacterial FtsZ protein is thought to be the progenitor to tubulin, and is a GTP-dependent, filament-forming protein that encircles the bacterial division site. Similarly, plant FtsZ proteins encircle the mid-plastid. Plant Ftle and FtsZZ family proteins differ primarily at their C-termini with FtsZZ proteins possessing a motif similar to those found in bacterial FtsZ that is responsible for binding to other bacterial cell division factors. In plants, this motif has been shown to bind to the chloroplast division protein ARC6, which contains a DnaJ-like domain and may be involved in FtsZ filament remodeling, but it is unclear whether ARC6 forms a stable complex with FtsZ in vivo. The focus of this dissertation is on understanding the composition of the in viva FtsZ complex and understanding the function and topology of FtsZ protofilaments. Thg specific aims of this dissertation are: Aim 1: Define the proteins that interact with Ftle and FtsZZ in viva. FtsZ was purified from pea stroma and the composition of proteins that interact with FtsZ were defined. In Chapter 2, FtsZ levels were quantified in isolated Arabidopsis chloroplasts and Ftle and FtsZZ were found at a constant ratio Of"'1:2. FtsZ was partially co- purified in a complex with the chloroplast division proteins ARC3 and ARC6 (chapter 2). The complex also contained a protein recognized by an antibody against Escherichia coli (E. coli) ZipA. A ZipA cross-reactive protein has been tightly linked to FtsZ-containing complexes and is discussed in Chapter 4. Plant FtsZZ proteins contain a C-terminal motif conserved with bacterial FtsZ (Figure 1.1). In bacteria, this motif specifically interacts with the bacterial cell division protein ZipA, but ZipA homologues have not been identified in plants. In an attempt to identify a putative ZipA-like protein, a structural search algorithm was created to derive structural similarity by threading a database of proteins of unknown function into the solved crystal structure of the FtsZ-interacting domain of ZipA. As discussed in Chapter 4, a strong candidate identified from this analysis, 5521, could not be definitively linked to a role in chloroplast division. Aim 2: To determine the topology of the Ftle/FtsZZ pratafilament and define the biadtemical properties of Fts21 and FtsZZ. Why do plants have two types of FtsZ, while bacteria only use one type of FtsZ? In Chapter 3, Ftle and FtsZZ are produced individually as recombinant proteins in E. coli, 3 significant technical advance. Because FtsZ proteins are evolutionary progenitors of tubulin, one possible hypothesis is that Ftle and FtsZ2 have evolved in parallel into heterofilaments like tubulin. Despite both FtsZ1 and FtsZZ being GTPases, maximal Ftle/FtsZZ co-assembly requires equal concentration of both proteins, arguing for a heterofila ment polymer topology. Moreover, the longitudinal, T7-Ioop mutant FtsZZDBZZA was found to be a competitive inhibitor of Ftle, but not FtsZZ, further supporting an FtZl/FtsZZ heterofilament topology. Figure 1.1: Cartoon representation of sequence similarity between plant Ftle, FtsZZ, bacterial FtsZ and a— and B-tubulin. All proteins contain a glycine-rich N-terminal domain that binds GTP. All proteins, except B-tubulin have all residues required for GTP hydrolysis in their T7-loops. B-tubulin has a lysine substituted in its T7-Ioop that prevents hydrolysis in the a-subunit. (xB-tubulin have C-terminal protease-sensitive extensions that bind regulatory proteins such as MAP-Tau. Bacterial FtsZ and plant FtsZZ have similar C-terminal extensions that bind the cell division protein ZipA (Hale and de Boer 1997) and ARC6 (Maple et al. 2005), respectively. Ftle has a plant-specific C-terminal extension that has no known interacting partner, but the entire Ftle protein has been shown to interact with ARC3 in the yeast two-hybrid assay (Maple et al. 2005). a-Tubulin GTP Binding Hydrolysis MAP-Tau GGGTGS G GxxNxDxE Binding GTP Binding l-l-ydielysis MAP—Tau ‘3 Tubun" GGGTGS G W ‘ Binding . GTP Binding Hydrolysis ZipA Baden“ FtsZ GGGTGS G GxxNxDxD Binding Plant FtsZZ CTP l I GTP Binding Hydrolysis (GGGTGS/T G) GxxNxDxD Plant Ftle CTP GTP Binding Hydrolysis GGGTGS G GxxNxDxD ' Bacterial cell division is best understood in E. coli The stromal chloroplast division machinery is derived from the cell division machinery of the ancient cyanobacterial endosymbiont. Bacterial cell division is best understood in the Gram-negative bacterium E. coli. Cell division mutant screening in E. call has been powerful for understanding the cell division mechanism of other bacteria including the Gram-positive bacteria Bacillus subtills (B. subtilis). While the division mechanisms of E. coli and B. subti/is are Similar, many functional details differ. Much of the difference between E. coli and B. subtilis division probably is because B. subtilis can undergo asymmetric cell division, termed sporulation. The division machinery of cyanobacteria is not well understood, but contains components found in both Gram- positive and Gram-negative bacteria and could represent a unique class of division machinery (Miyagishima et al. 2005). Because chloroplast division is derived from bacterial cell division, an understanding of bacterial cell division is essential for understanding the stromal chloroplast division machinery. FtsZ Bacteria contain tubulin-like cell-division proteins, called FtsZ (Erickson 1995, 1997) that assemble into a ring at mid-cell just after DNA replication and immediately before cell division. FtsZ contracts with the midcell division furrow as the cell-cycle progresses (Addinall et al. 1996; Bi and Lutkenhaus 19903, 1991; den Blaauwen et al. 2003; Harry et al. 1999; Lin et al. 1997; Regamey et al. 2000). FtsZ has been proposed to be a scaffold for binding additional cell division proteins (Margolin 2000; Rothfield et al. 1999), but it also have been demonstrated to generate contractile force (Osawa et al. 2008) Tubulin biochemistry FtsZ is structurally and biochemically similar to tubulin despite sharing less than 10% sequence similarity. Due to a high degree Of structural similarity (LOwe 1998; Lowe and Amos 1998) both proteins are believed to share a common ancestor (Erickson 1995, 1997; Nogales et al. 1998a). Tubulin biochemistry is well understood and the biochemistry of tubulin will be discussed as a foundation for understanding FtsZ biochemistry. The basic subunit of tubulin is a 043 dimer that are tightly bound and do not dissociate. Because a— and B-tubulin monomers are unstable alone, chaperone complexes assemble OLB-tubulin dimers. Imbalanced expression of either a- or B-tubulin causes blocks in dimer assembly (Geissler et al. 1998; Hirata et al. 1998; Melki et al. 1996; Tian et al. 1999; Tian et al. 1996; Tian et al. 1997). Binding of GTP to aB-tubulin dimers causes assembly into microtubules. Microtubules are composed of longitudinally arranged, head-to-tail, repeating dimers of ocB-tubulin that form heterofilaments (Tian et al. 1996). Tubulin heterofilaments laterally associate into tubes (Downing and Nogales 1998a, b, c; Nogales 2000). Structures of OLB-tubulin dimers have been generated to investigate how GTP binding and hydrolysis regulate tubulin assembly into microtubules (Nogales 1999, 2000; Nogales and Wang 2006; Nogales et al. 1999; Nogales et al. 1998b; Sackett 1995). 6 The N-terminus of a- and B-tubulin have a glycine-rich motif (GGGTGS/T G) called the tubulin signature motif found in a series of antiparallel B-sheets called a Rossman fold (Erickson 1995). The GTP binding characteristics of a- and B-tubulin differ in the ability to exchange fresh nucleotide. a-tubulin has a non-exchangeable GTP binding site called the N-site. The N-site is unable to exchange GTP because it is buried within the tight a/B dimer interface that does not dissociate (Menendez et al. 1998). In contrast, in unassembled tubulin the nucleotide binding site in the B-subunit is solvent- exchangeable and termed the E—site. However, the E-site becomes non-exchangeable when tubulin is assembled into filaments because of nucleotide burial in the interface between the a- and B-subunits in the heterofilaments (Menendez et al. 1998; Nogales et al. 1998a). FtsZ differs from tubulin because bound GTP is solvent-exchangeable within assembled filaments (Nogales 2000). GTP is only hydrolyzed in the E-site Of B-tubulin (Nogales 2000; Nogales and Wang 2006; Nogales et al. 1999; Nogales et al. 1998b). The GTP binding-site of B- tubulin is only GTPase competent when assembled with a second OtB-tubulin dimer in a heterofilament. In the B-subunit, GTP is hydrolyzed by the T7-Ioop of a-tubulin from the second OLB-tubulin dimer. The T7-loop contacts the y-phosphate of the bound GTP, completing the active site in trans with an a-subunit. The B-tubulin T7-loop and FtsZ T7- loop have the consensus sequence GxxNxDxx(E/D) (LOwe 1998; Lowe and Amos 1998; Nogales et al. 1998a). In a similar manner, the N-site GTP bound to a-tubulin is contacted by the 17-loop of the B-subunit within and individual aB-tubulin dimer, but the T7-loop of the B-subunit cannot promote GTP hydrolysis because its consensus is 7 GxxNxDxxK, having a lysine substituted for an acidic residue (Nogales et al. 1998a). The inability of a-tubulin to hydrolyze GTP stabilizes the a/B dimer interface, preventing dissociation, and maintaining tubulin as a strict a/B dimer. Because tubulin GTPase active Sites are formed in trans, tubulin filament ends are biochemically different. Microtubules grow quickly at their plus-ends and more slowly on their minus ends. The plus end of ocB-tubulin is always a B-subunit with its exposed E-site bound to GTP. The plus end grows by binding to the T7-IOOp-containing C-terminal interface of an incoming a-subunit dimerized with a B-subunit, thus re- creating a new GTP-bound B-subunit on the plus side. In contrast, the minus-end contains the T7—loops of a-tubulin, which are not GTP bound. Incoming aB-tubulin dimers cannot assemble at the minus end; this process is kinetically unfavorable. The minus end of a tubulin filament is composed primarily of a-tubulin. The current model is that strong lateral contacts between a-subunits stabilize the minus end (Nogales et al. 1999). In contrast, lateral interactions between B-subunits are strong when GTP is bound, but weak when GTP is hydrolyzed to GDP (Derry et al. 1998; Tanaka-Takiguchi et al. 1998). aB-tubulin filaments are thus stable when the plus end is bound to GTP (Derry et al. 1998; Tanaka-Takiguchi et al. 1998). GTP hydrolysis is not required for tubulin assembly and the rates of GTP hydrolysis vary considerably in different types Of tubulin (Daugherty et al. 1998; Vandecandelaere et al. 1999), but generally the GTPase activity of tubulin is slower than the rate of polymerization. Thus, aB-tubulin plus-ends contain primarily GTP, but as the GTPase activity catches up with polymerization, the polymer becomes prone to disassembly due to destabilized lateral interaction between B-subunits at the plus-end in a process termed “dynamic instability”. When examined by EM, destabilized GDP-bound B-tubulin subunits show individual filaments curling away from the microtubule (Nogales 2000). Dynamic instability of aB-tubulin leads to an interesting assembly phenomenon called treadmilling (Margolis and Wilson 1998; Waterman-Storer and Salmon 1998). If OLB- tubulin lacks new incoming GTP-bound dimers to continue assembly, growth stops and when all the GTP is hydrolyzes within the B-subunits, the microtubule disassembles. However, in the presence of sufficient GTP, aB-tubulin dimers released from the minus end can exchange their hydrolyzed GDP for a new GTP and re-assemble at the plus end, thus dynamically maintaining the tubulin polymer, 3 process called treadmilling. Experimentally, fluorescently labeled tubulin added to treadmilling filaments will rapidly traverse from the plus to minus end (Waterman-Storer and Salmon 1998). Tubulin assembly is rate-limited at nucleation. In vivo, tubulin is nucleated by specialized complexes called “microtubule organizing complexes” (MTOCs). In vitro, 3 high critical concentration (abbreviated as Cc) of tubulin is typically required to grow microtubules (“'7 uM). Stabilization agents such as DEAE-dextran or the microtubule inhibitor taxol are often required to stabilize and promote tubulin assembly in vitro due to the high tubulin Cc. The best-described MTOC is the y-tubulin ring complex (y-TURC). y-TURC is a large, 2.2 MDa complex that can directly nucleate microtubules (Moritz et al. 1995; Zheng et al. 1995). Recently, y-TU RC was shown to accomplish this by directly stabilizing a-tubulin at microtubule minus ends (Kollman et al. 2008) and is a key regulator of tubulin polymerization in viva. Tubulin assembly is regulated by several types of microtubule-associated proteins (MAPS). The best-characterized MAPS are MAP-Tau proteins, which promote tubulin assembly. MAP-Tau proteins are unstructured, negatively charged proteins that bind the positively charged C-termini Of ocB-tubulin (Chau et al. 1998) that are solvent accessible in assembled microtubulies (Sackett et al. 1985). MAP-Tau likely inhibits GTP hydrolysis by stabilizing the interaction between OLB-tubulin dimers and possibly allosterically slowing the hydrolysis of GTP by an unknown mechanism (Chau et al. 1998). Finally, many microtubule-destabilizing proteins have been described. Katanin is a microtubule severing factor essential for releasing microtubules from centrosomes (Ahmad et al. 1999; Hartman et al. 1998; Lohret et al. 1998; McNally et al. 1996; McNally and Thomas 1998). Katanin severs microtubules, releasing GDP-capped microtubules from within assembled polymers that are unstable and rapidly depolymerize. Stathmin is a tubulin depolymerization factor that promotes tubuli n disassembly by directly binding ocB-tubulin dimers to promote GTP hydrolysis by an unknown mechanism (Belmont and Mitchison 1996 ; Howell et al. 1999 ). E. coli FtsZ is structurally and biochemically similar to tubulin FtsZ is key to initiating bacterial cell division. FtsZ binds and hydrolyzes GTP in a glycine-rich motif near the N-terminus similar to the GTP-binding motif in tubulin (de Boer et al. 19923; Mukherjee et al. 1993; RayChaudhuri and Park 1992). The presence 10 of a tubulin-like GTP-binding site in the N-terminus of E. coli FtsZ suggested FtsZ might be a bacterial version of tubulin. FtsZ polymerization and structure FtsZ and tubulin only share ~10% sequence Similarity, but evidence was growing that FtsZ might be similar to tubulin. Supporting this hypothesis was the finding that recombinant FtsZ forms multimers when separated by native PAGE (Mukherjee and Lutkenhaus 1994). Electron microscopy of FtsZ in the presence of GTP shows long thin filaments reminiscent of tubulin (Mukherjee and Lutkenhaus 1994). Moreover, use of the non-hydrolysable GTP analogue GTPyS also results in FtsZ assembly, further suggesting that, as in tubulin, GTP binding, but not GTP hydrolysis, is required for assembly (Mukherjee and Lutkenhaus 1994). FtsZ assembly is protein concentration dependent like tubulin assembly with a Cc of "0.5-1 uM (Mukherjee and Lutkenhaus 1994) The first structural link between FtsZ and tubulin came from 3D-reconstruction of electron micrographs of tubulin and FtsZ. From this study, the structure of assembled tubulin and FtsZ were found to have nearly identical structures and arrangement of monomers within filaments (Erickson et al. 1996). However, the structural similarity between bacterial FtsZ and tubulin was fully realized when Methanocaccus jannaschii FtsZ was crystallized and found to have a nearly identical structure to tubulin, despite less than 10% sequence similarity (L6we 1998; Lowe and Amos 1998). Interestingly, the primary difference between tubulin filaments and the proposed FtsZ filament model is 11 that the FtsZ nucleotide-binding site is partially solvent-accessible while tubulin GTP binding sites are non-exchangeable when polymerized (L6we 1998; Lowe and Amos 1999; Nogales et al. 1998a; Small and Addinall 2003). Mechanistically, this means that tubulin is only able to exchange GTP at its plus-ends, whereas FtsZ can maintain polymer stability by exchanging GDP for fresh GTP within assembled filaments (Lu et al. 2000). GTP-assembled FtsZ protofilaments are unstable in vitro and rapidly depolymerize within 10-20 minutes (Mukherjee and Lutkenhaus 1999). The inclusion of CaCIz and DEAE-dextran is typically required to stabilize protofilaments and promote lateral bundling between protofilaments (M ukherjee and Lutkenhaus 1994, 1999; Scheffers et al. 2000; Yu and Margolin 1997). CaClz is used to stabilize FtsZ polymers because it mimics MgClz, which is required for GTP hydrolysis. Thus, CaClz stabilizes FtsZ filaments because it is a competitive inhibitor of GTP hydrolysis. The advantage of CaClz stabilization of FtsZ polymers is that the effect is reversible by chelating CaClz with EGTA. The mechanism by which DEAE-dextran promotes FtsZ polymerization and bundling is unknown. The polymerization dynamics of E. coli FtsZ can be measured in real time by 90° light scattering (M ukherjee and Lutkenhaus 1999). This technique utilizes a spectrofluorimeter with excitation and emission monochromators set at 350 nm, and detects the diffraction of light at a 90° angle using a narrow slit width. 90° light scattering has the advantage that it tolerates high protein concentration, but due to the wavelength of light used for diffraction it can only detect filaments larger than short 12 oligomers. FtsZ nucleation into dimers is not detectable with this technique (Mukherjee and Lutkenhaus 1999). Despite this limitation, light scattering has proven to be a powerful technique for monitoring FtsZ polymerization in real-time (Mukherjee and Lutkenhaus 1999). EM has shown that GTP-FtsZ forms single, long, straight filaments that can laterally associate to form sheets in the presence Of the stabilizing DEAE-dextran (Erickson et al. 1996; Lu et al. 2000; Mukherjee and Lutkenhaus 1994). Assembly of FtsZ stimulates GTP hydrolysis by the completion of the active site in trans similar to tubulin (Scheffers and Driessen 2001; Scheffers et al. 2002; Wang and Lutkenhaus 1993). Following GTP hydrolysis and y-P,- dissociation, GDP-FtsZ polymers adopt a curved conformation that is believed to be less stable than straight filaments, leading to rapid disassembly (Erickson and Stoffler 1996; Lu et al. 2000; Romberg et al. 2001). E. coli F tsZ G TP hydrolysis drives depolymerization GTP binding promotes FtsZ assembly into filaments, but GTP hydrolysis is not required for polymerization since GTPase inhibitors such as GTPyS and CaClz promote polymerization. This suggested that GTP hydrolysis promotes depolymerization of FtsZ filaments similar to the regulation of tubulin stability. The structural similarity between FtsZ and tubulin predicts that the association of two monomers would complete FtsZ active sites. This hypothesis is supported by the conserved GxxNxDxx(D/ E) motif in the T7-loop of bacterial FtsZ. To test if an FtsZ GTPase active site is formed in trans, Scheffers et al. created a series of T7-loop mutations in E. coli FtsZ that were tested for 13 GTPase activity with the wild-type protein. The T7-loop mutants were found to be sub- stoichiometric competitive inhibitors of GTP hydrolysis and polymerization but did not affect GTP binding (Scheffers et al. 2002). Moreover, these mutants did not interfere with the ability of FtsZ to assemble into cell division rings when expressed at low levels in viva, but when expressed at high levels block cell division (Redick et al. 2005). These results confirmed that FtsZ has an active site composed of two monomers and is mechanistically similar to tubulin. The GTPase activity of FtsZ is ”SO-times faster (Huecas et al. 2007b; Romberg and Mitchison 2004) than that of MAP-free tubulin (David-Pfeuty et al. 1977). Within "'10 min. of polymerization in vitro, E. coli FtsZ filaments begin to rapidly disassemble (Mukherjee and Lutkenhaus 1999) as the ratio of available GTP/GDP is reduced (Small and Addinall 2003). This leads to the question, how are FtsZ polymers maintained with a fast GTPase activity? A key difference between tubulin and FtsZ is nucleotide binding. Once tubulin polymerizes, the bound nucleotide can no longer be exchanged for free GTP in solution. Because tubulin has a slow GTPase activity, polymerization can continue for a significant period until the GTPase activity catches up with the polymer in a process called “dynamic instability”. However, FtsZ maintains polymer stability by exchanging nucleotide within assembled protofilaments (Mingorance et al. 2001; Small and Addinall 2003). Thus, instead of re-assembling FtsZ plus-ends, FtsZ regenerates straight, stable polymers by directly exchanging nucleotide within the polymer (Figure 1.2). Because FtsZ polymers can exchange nucleotide, FtsZ is sensitive to the ratio of 14 GTP/GDP. High GDP concentration favors FtsZ depolymerization (Small and Addinall 2003). FtsZ filaments contain mostly GDP and the hydrolyzed y-P; (Scheffers and Driessen 2002). Polymerized FtsZ is able to exchange hydrolyzed GDP+P,- for GTP unlike polymerized tubulin, where the nucleotide is non-exchangeable (Huecas and And reu 2004; Huecas et al. 2007a; Huecas et al. 2007b; Romberg and Levin 2003; Scheffers and Driessen 2002; Small and Addinall 2003). Although FtsZ and tubulin crystal structures are similar, there are considerable differences in the active site. Notably, in the interface between monomers adjacent to the bound GTP in FtsZ is the T3-loop that makes contact with the GTP y-phosphate. In tubulin, the T3-loop prevents nucleotide exchange with solvent, but in FtsZ, the T3-loop is Shorter and allows solvent access to the nucleotide (Diaz et al. 2001). In FtsZ, loss of the y-P,- after GTP hydrolysis causes displacement of the T3-loop away from the nucleotide binding site and likely induces protofilament curvature (Diaz et al. 2001). This conformation is believed to be less stable than straight protofilaments (Diaz et al. 2001; Erickson and Stoffler 1996; Erickson et al. 1996; Huecas and Andreu 2004; Mukherjee and Lutkenhaus 1994, 1998; Oliva et al. 2004) and ultimately leads to depolymerization (Figure 1.2). Cooperative assembly of bacterial FtsZ FtsZ was initially thought to assemble isodesmically because it assembles into single filaments even though FtsZ has an apparent Cc for assembly, which does not support isodesmic assembly (Gonzalez et al. 2005; Romberg et al. 2001). Dimerization 15 Of GDP-FtsZ requires GTP (Rivas et al. 2001; Rivas et al. 2000), while GDP-FtsZ dimerization and assemble is kinetically disfavored (Huecas and Andreu 2004). Like tubulin, the emerging model of FtsZ polymerization suggests the minimal subunit for assembly is a dimer (Chen and Erickson 2005). Despite FtsZ only forming single- stranded filaments (which predicts isodesmic assembly)(Gonzalez et al. 2005 ; Romberg et al. 2001) FtsZ assembles cooperatively (Caplan and Erickson 2003; Chen et al. 2005 ; Huecas and And reu 2003; Huecas et al. 2007a). The precise mechanism of FtsZ cooperative assembly is still unclear, but recent work suggests that conformational switches within FtsZ upon polymerization may self-stabilize protofilaments (H uecas et al. 2007a; Huecas et al. 2007b). Model of FtsZ polymerization In vitro, unassembled FtsZ is GDP-bound. When GTP is available, FtsZ rapidly exchanges nucleotide, forming GTP-FtsZ (Figure 1.2A). GTP-FtsZ cooperatively assembles into protofilaments (Figure 1.28) with a rate Of "0.5 uM sec'1 (and an off rate of 1.2 sec‘l) (Huecas et al. 2007b). However, there is evidence that FtsZ is assembled from dimers of FtsZ and not monomers (Chen et al. 2005; Huecas et al. 2007a). If FtsZ is indeed assembled from dimers, this explains why single-filament FtsZ assembly is cooperative (Chen et al. 2005 ; Huecas et al. 2007a; Rivas et al. 2001; Rivas et al. 2000; Romberg et al. 2001; Sassong et al. 1999)(Figure 1.2). Within protofilaments GTP hydrolysis is "0.13 sec'1 (Figure 1.2C) and is the rate limiting step of the polymerization cycle (Romberg and Mitchison 2004). 16 Figure 1.2: Model of the assembly dynamics of E. coli FtsZ. This model is adapted from (Huecas et al. 2007b; Romberg and Levin 2003; Romberg and Mitchison 2004). FtsZ molecules are indicated by circles and the nucleotide-bound state Of bacterial FtsZ is indicated by ”T” for GTP, ”DPi” for GDP-Pi, and ”D” for GDP. (A) GDP/GTP exchange in monomers is fast in solution. (B) GTP—FtsZ dimers rapidly assemble into straight protofilaments. (C) After GTP hydrolysis, FtsZ filaments are GDP-Pi-FtsZ bound and still straight and (D) Pi release is rapid. (E) Protofilaments can be maintained as straight polymers by a relatively slow nucleotide exchange within the polymer. (G) FtsZ protofilaments are believed to laterally associate, but it is unknown how this affects FtsZ kinetics since laterally associated FtsZ is not observed in vitro. (F) Curved GDP-FtsZ is unstable and rapidly disassembles. (H) GTP- FtsZ can also presumably laterally associate, but this has not been demonstrated under physiological conditions. 17 . A (Fast) (9' " A F (2.615) 18 Following hydrolysis, FtsZ is in the GDP-P,—FtsZ state and P,- release has a rate of 0.1 sec"1 (Figure 1.2D)(Huecas et al. 2007b; Romberg and Mitchison 2004); P,- release results in FtsZ curved filaments, which rapidly disassemble (Romberg et al. 2001). Polymerized filaments disassemble when the y-P,- dissociates, leaving FtsZ GDP-bound, with a rate of "2.6 sec'1 (Figure 1.2F)(Huecas et al. 2007b). However, GTP can exchange within the assembled filament to maintain the assembly (Figure 1.2E)(Lu et al. 2000; Mingorance et al. 2001; Romberg and Mitchison 2004). Bundling of bacterial FtsZ is believed to occur, but has not been observed under physiological conditions and it is unclear how lateral association between FtsZ filaments would affect polymerization kinetics (Figure 1.2G). Regulation of Z-ring assembly and placement Nucleoid occlusion positions the Z-ring in bacteria Z-ring assembly is negatively regulated by the presence of the bacterial nucleoid. In order for the cell division apparatus to assemble at mid-cell the nucleoid must be duplicated and separated toward the poles, which leaves a small nucleoid-devoid furrow in the middle of the cell where FtsZ can assemble (Rothfield et al. 2005; Wold ringh et al. 1990). The nucleoid occlusion model predicts the presence of a Z-ring inhibitory factor that uniformly binds the nucleoid. Recently, SlmA and Noc were found to be DNA binding proteins required for cell-division (Bernhardt and de Boer 2005; Marston et al. 1998). Double SlmA/Noc mutants are lethal, suggesting both have critical roles in cell division. Both SlmA and Noc are located to the polar side of the nucleoid. SlmA and Noc 19 are predicted to be FtsZ inhibitors, but biochemical roles in Z-ring assembly have not been described. Positioning of the Z—ring by the MinCDE system The position Of the Z-ring at mid-cell is controlled not only by nucleoid occlusion, but also by a system of three proteins, MinC, MinD and MinE, collectively termed the Min system. Mutations in components of the Min system result in ”mini” cells from improper placement of the Z-ring (de Boer et al. 1989a, 1992b). Because the cells divide before the nucleoid is properly segregated, they are typically anucleate (de Boer et al. 1989a, 1992b). Min proteins are found in Gram-negative (e.g. E. coli) and Gram-positive bacteria (e.g. B. subtilis) as well as cyanobacteria (Mazouni et al. 2004) and homologues of MinD and MinE, but not MinC, are found in plants (Aldridge et al. 2005). MinC is a direct inhibitor of FtsZ assembly (Hu et al. 1999; Levin et al. 2001; Marston and Errington 1999; Pichoff and Lutkenhaus 2001). However, MinC lacks spatial site selection. MinC heterodimerizes with the membrane recruitment factor MinD (Hu and Lutkenhaus 1999; Raskin and de Boer 1999a, b). The balance between MinCD heterodimers and FtsZ is essential for proper cell division. For example, hyper- accumulation of MinC leads to blocked cell division, which can be overcome by an overproduction of FtsZ (Bi and Lutkenhaus 1990b; Justice at al. 2000; Levin et al. 2001). MinC is composed of two domains; the N-terminal domain directly inhibits FtsZ assembly (Hu and Lutkenhaus 2001). The N-terminus Of MinC is connected by a linker to 20 the C-terminal domain that heterodimerizes with MinD (Cordell et al. 2001; Hu and Lutkenhaus 2000; Szeto et al. 2001). MinE is a topological specificity factor for MinCD. MinE restricts the FtsZ- inhibitory activity of MinCD to the poles in E. coli (Fu et al. 2001; Hu and Lutkenhaus 1999; Raskin and de Boer 1999a, b; Rowland et al. 2000) by forming a MinE-ring that caps MinCD to the polar region (Fu et al. 2001; Hale et al. 2001; Raskin and de Boer 1997; Shih et al. 2002). MinE rapidly cycles between the poles of the E. coli cell, rapidly assembling and disassembling with a periodicity of approximately 1-2 min (Raskin and de Boer 1999b). Assuming a simple oscillator model for pole-to-pole movement of MinCDE would result in a time averaged accumulation of MinCDE at the poles (Rothfield et al. 2005), suggesting that the determinant for polar localization is dependent upon intrinsic assembly properties of MinE, although direct evidence of this does not exist. Gram-positive bacteria such as B. subtilis lack MinE and instead use DileA to recruit MinCD to the poles. MinE dynamically oscillates between the cell poles, whereas B. subtilis DileA is statically localized to the cell pole (Edwards et al. 2000). DileA is believed to be targeted specifically to a lipid or protein domain at the cell pole (Rothfield et al. 2005), thus explaining its static polar localization. ZipA and FtsA organize and anchor the Z-ring at the membrane In bacteria, FtsZ proteins are anchored to the membrane by two proteins, ZipA and FtsA; the latter is in the same structural super-family as actin/HSP70 proteins (Bork et al. 1992; van den Ent and Lowe 2000). Both ZipA and FtsA interact with the extreme 21 C-terminus of E. coli FtsZ (Addinall and Lutkenhaus 1996; Hale and de Boer 1999; Hale and de Boer 1997; Haney et al. 2001; Liu et al. 1999; Wang et al. 1997; Yan et al. 2000) and anchor the FtsZ ring to the inside of the cell membrane (Hale and de Boer 1997). Both FtsA and ZipA are required for interaction with downstream cell division proteins (Pichoff and Lutkenhaus 2002). The crystal structure of ZipA complexed with the carboxyl terminus Of FtsZ has been solved (Mosyak et al. 2000). Based on the quaternary structure of tubulin (Li et al. 2002) the ZipA/FtsZ interacting domain is believed to stick out of FtsZ polymers (LOwe 1998; Lowe and Amos 1999). Despite the conservation of the ZipA/FtsZ binding motif in FtsZZ (Figure 1.1), neither ZipA nor FtsA have been identified in plants, but both functions are essential to bacterial cell division. ZapA ZapA promotes the assembly of FtsZ and is found in many bacteria, but is not essential for proper cell division (Gueiros-Filho and Losick 2002). ZapA was identified in B. subtilis by a novel genetic screen. Constitutive overexpression of MinD results in loss of topological Specificity of MinC, an FtsZ assembly inhibitor, inhibiting Z-ring assembly throughout the cell. This MinD overexpressor was transformed with an overexpression library and screened for cell lines that were able to overcome the constitutive block in FtsZ assembly and would thus be FtsZ assembly factors (Gueiros-Filho and Losick 2002). This screen resulted in the identification of ZapA, (yshA) which is a relatively short protein (85 amino acids) and was found to co-localize to Z-rings. ZapA is required for proper cell division only in the absence Eer and DivIVA, both of which regulate Z-ring 22 dynamics (Gueiros-Filho and Losick 2002). ZapA has a functionally equivalent ortholog in E. coli called Yng (Small et al. 2007). Eer Eer is found only in Gram-positive bacteria such as B. subtilis and is a negative regulator of Z-ring assembly. Loss of Eer function results in Z-rings that mis-Iocalize to the cell pole (Levin et al. 1999). Eer contains a transmembrane domain that localizes it to the cell membrane, but Eer is also associated with the Z-ring. The GTPase activity of FtsZ is slightly enhanced by Eer, even though GTP binding is reduced (Chung et al. 2007). Eer destabilizes Z-rings by interacting with the C-terminus of FtsZ near the conserved C-terminal motif, suggesting Eer is a negative regulator of FtsZ. Thus a working model is that FtsA, a positive regulator of assembly, and Eer compete to regulate Z-ring formation (Singh et al. 2007)(Gram-positive bacteria do not have ZipA). SuIA SulA is a negative regulator of FtsZ assembly that is not essential for proper cell division. SulA prevents cell division in response to various cell damage signals (Higashitani et al. 1997; Huang et al. 1996; Justice et al. 2000; Mukherjee et al. 1998). SulA is interesting because it blocks FtsZ assembly by binding the T7-loop, or the hydrolysis loop, effectively acting as a filament cap by preventing the addition of new GTP-bound subunits. The effect of SulA can be mimicked by mutating the T7 loop, which blocks GTP hydrolysis but not GTP binding (Redick et al. 2005 ; Scheffers et al. 2002). 23 Chloroplast Division Ultrastructural observations of chloroplast division Electron microscopy of fixed plant and algal chloroplasts has been a powerful tool for observing the ultrastructure of dividing plastids. These experiments have identified electron dense rings, termed PD rings. An inner PD-ring is found adjacent to the stromal side of the inner-envelop membrane (IEM). A outer PD-ring is found on the cytosolic side of the outer-envelope membrane (OEM)(Leech et al. 1981). An inter- membrane space localized middle PD-ring has also been observed, but only in the alga Cyanidioschyzon merolae (Miyagishima et al. 20013). It is unknown if plants also have a middle PD-ring. After the discovery of stromal FtsZ rings in chloroplasts (discussed below) (Vitha et al. 2001) a logical hypothesis was that the inner PD-ring is composed of FtsZ, but it was later found that the FtsZ ring assembles much earlier. Additionally, the inner PD- ring is probably dependent on FtsZ assembly at mid-plastid (Kuroiwa et al. 2002; Miyagishima et al. 2001a). FtsZ probably assembles before of the inner PD-ring, which in turn promotes assembly of the middle PD ring (in C. merolae) and finally the outer PD-ring assembles at the site of chloroplast division. The inner and middle PD-rings disassemble after constriction and prior to completion of fission (Miyagishima et al. 20013), but the outer ring stays intact through completion Of division. The composition of the PD-rings is still unknown. Preliminary biochemical studies of the outer PD-ring 24 found it to be composed of a "56 kDa filament forming protein with an unknown sequence or identity (Miyagishima et al. 2001b). The stromal chloroplast division machinery FtsZ Supporting the endosymbiotic origin of chloroplasts (Gray 1993) was the finding that plants contain chloroplast-targeted homologues of the bacterial cell division protein FtsZ (Osteryoung et al. 1998; Osteryoung and Vierling 1995). Bacteria typically have one type of FtsZ, but plants have two phylogenetically distinct families of FtsZ, Ftle and FtsZ2. Arabidopsis contains three nuclear-encoded FtsZ genes: one Ftle gene, AtFtsZI-l, and two FtsZZ genes, AtFtsZZ-I and AtFtsZZ-Z. In Arabidopsis, AtFtsZZ- 1 and AtFtsZZ-Z are genetically redundant (Aaron Schmitz, unpublished observation). All three Arabidopsis FtsZs are targeted to the chloroplast stroma by cleavable transit peptides (McAnd rew et al. 2001; McAndrew et al. 2008). FtsZ1 may associate with thylakoids early in development (El-Kafafi et al. 2008). Plant FtsZZ proteins are more similar to bacterial FtsZ proteins and have the conserved C-terminal motif (Figure 1.1) that in bacteria binds to ZipA and FtsA. In plants, the FtsZZ C—terminal motif binds to the chloroplast division protein ARC6 (Maple et al. 2005) through a discrete portion Of ARC6 that may be structurally similar to the FtsZ binding domain of ZipA (Jonathan Glynn, unpublished). Ftle, which lacks this conserved domain, does not bind ARC6 (Maple et al. 2005). In contrast to FtsZZ, Ftle proteins have shorter C-termini and there are no known proteins that bind Ftle C-termini. However, Ftle is distinguished from FtsZZ by 25 specifically binding the chloroplast division protein ARC3, but not ARC6 (Maple at al. 2005). In E. coli, blocked cell division, such as caused by mutations in FtsZ, results in long, filamentous, multinucleate cells with evenly spaced Z-rings (Addinall et al. 1996). Blocked chloroplast division results in fewer enlarged chloroplasts (Osteryoung et al. 1998; Osteryoung and Vierling 1995 ; Stokes et al. 2000). Loss of function or overexpression of any of the three Arabidopsis FtsZ genes leads to blocked chloroplast division, which suggests stoichiometric interaction between the FtsZ proteins may be important for function (Osteryoung et al. 1998; Stokes et al. 2000). Similar to bacterial FtsZ, plant FtsZ proteins are localized to rings at mid-plastid that encircle the site of division on the stromal side of the IEM of the chloroplast (Vitha et al. 2001). Moreover, using double-stain immunofluorescence micrOSCOpy, FtsZ1 and FtsZZ are found precisely co-localized (within the resolution limit of immunofluorescence microscopy) to the same division rings at mid-plastid (Vitha et al. 2001). FtsZ rings have been found to be remodeled in less than 30 sec (Vitha et al. 2005) similar to the dynamics Of E. coli FtsZ in viva (Stricker et al. 2002). Disruption of FtsZ levels leads to defects in FtsZ-ring morphology (Vitha et al. 2001). Loss of Ftle function results in long disorganized FtsZZ filaments throughout the chloroplasts, whereas the loss Of FtsZZ results in short disorganized FtsZ1 filaments in the plastid (Stokes et al. 2000; Vitha et al. 2001; Yoder et al. 2007). Because the FtsZZ-interacting 26 protein ARC6 (discussed below) may promote FtsZZ organization, it is unclear if the long FtsZZ-filaments are a property of FtsZZ, or a result Of ARC6 promoting FtsZZ assembly. ARC6 is a homologue of Fth ARC6 was identified as having a role in chloroplast division in a series of T-DNA insertion lines with enlarged chloroplasts. In arc6 mutants, plants typically have only one enlarged chloroplast that takes up much Of the available cytosolic volume (Pyke and Leech 1994). The gene encoding ARC6 was identified as a homologue of the cyanobacterial cell division gene Ftn2 (Koksharova and Walk 2002) found in the arc6 mapping interval (Vitha et al. 2003). ARC6 is a transmembrane protein that contains an N-terminal DnaJ-like co-chaperone motif, which is in the chloroplast stroma, while the C-terminus extends into the intermembrane space. ARC6 has been found to interact specifically with the conserved C-terminal motif found in plant FtsZZ proteins, but not Ftle family proteins (Maple at al. 2005). Immunolocalization of FtsZ in the arc6 mutant shows disorganized Z-filaments throughout the entire plastid, but the lines overexpressing ARC6 have hyper-stabilized Z-filaments that form spirals at mid-plastid, suggesting ARC6 organizes and stabilizes Z- rings (Vitha et al. 2003) by binding the C-terminus of FtsZZ (Maple et al. 2005; McAndrew et al. 2008). AtMinD and AtMinE Chloroplasts have homologues of MinD and MinE, but lack an obvious homologue of MinC. Because not all bacteria use the Min system, the functional 27 consequences of missing MinC are unknown (Barak and Wilkinson 2007; Lutkenhaus 2007; Margolin 2002; Rothfield et al. 2005). MinD and MinE were first found in plants by BLAST searches (Colletti et al. 2000; Itoh et al. 2001; Maple et al. 2002). In the plastid, MinD and MinE specify proper symmetric division of the plastid and loss of MinD or overexpression of MinE leads to asymmetric plastid division, resulting in heterogeneous plastid sizes. As in bacteria, the opposite experiment of overexpression of MinD or loss of MinE results in blocked plastid division (Colletti et al. 2000; Fujiwara et al. 2004; Maple et al. 2002). AtMinD has been reported to bind and hydrolyze ATP like its bacterial counterpart (Hu and Lutkenhaus 2001), and is stimulated by AtMinE (Aldridge et al. 2005). In bacteria, MinD and MinE homodimerize and heterodimerize together and the same is true for AtMinD and AtMinE, suggesting functional conservation (Maple et al. 2005; Maple and Moller 2007b). In vivo, the mutant arc11 has a mutation in the AtMinD dimerization motif resulting in misplacement of the division furrow and heterogeneoust-sized and multiply constricted chloroplasts (Fujiwara et al. 2004) again suggesting the Min system is functionally conserved in chloroplast division. However, many functional details of the Arabidopsis Min system remain unknown. ARC3 ARC3 was identified in a mutant screen for plants with fewer enlarged chloroplasts (Pyke and Leech 1994) and the arc3 mutation was cloned by map-based cloning (Shimada et al. 2004). The N-terminus of ARC3 contains an FtsZ-like, glycine- 28 rich, GTP binding domain and a eukaryotically derived C-terminal repeat MORN (membrane occupation and nexus) motif followed by a domain similar to a phosphatidyl inositoyl-S-phosphate kinase (PIP-5K) domain. However, biochemical activity of these domains has not been described. Moreover, the PIP-5K domain lacks key residues required for catalytic activity (Maple et al. 2007). ARC3 localizes to a mid-plastid ring (Shimada et al. 2004) and specifically interacts with Ftle, but not FtsZZ proteins (Maple at al. 2007) and can be partially purified in a salt-stable complex with Ftle, FtsZZ and ARC6 (McAndrew et al. 2008). Curiously, immunofluorescence labeling of FtsZ in arc3 mutants shows multiple parallel Z-rings (Glynn et al. 2007) and multiple parallel division furrows (Maple and Moller 2007a, b). This suggests a role for ARC3 in Z-ring positioning similar to that of MinC, although its mid-plastid localization by immunofluorescence suggest that ARC3 positioning of Z-rings is mechanistically different than that of MinC (which is polar localized in bacteria). However, the biochemical effect of ARC3 on plant FtsZ filaments is unknown. AtSulA/6C1 In response to DNA damage in bacteria, FtsZ filaments are capped and depolymerized by the cell division inhibitor SulA. A weak homologue of SulA, AtSulA (also called GC1) has been identified in plants (Maple at al. 2004; Raynaud et al. 2004a). AtSulA is a transmembrane protein that does not assemble into a ring at mid-plastid (Maple et al. 2004). In contrast to other chloroplast division proteins, AtSulA does not form rings or other sub-plastidic localized structures, but instead is found diffusely throughout the IEM. Loss Of AtSulA leads to blocked chloroplast division, but it is 29 unclear if overexpression leads to blocked chloroplast division (Maple et al. 2004; Raynaud et al. 2004a). Bacterial SulA proteins interact with FtsZ to block the GTPase activity by capping Z-filaments blocking incoming subunits to complete a trans—GTPase active site(Bi and Lutkenhaus 1993). However, AtSulA does not interact with either Ftle or FtsZZ in yeast-two-hybrid assays and no other functional information exists for AtSulA/6C1. Recently it has been argued that AtSulA may be only weakly similar to bacterial SuIA and may have other roles in chloroplast division (Maple and Moller 2007b). Immunolocalization of FtsZ in AtSulA mutants will be essential to assignment of AtSulA function. The eukaryotic-derived chloroplast division machinery ARCS/CmDan ARCS and its homologue in the alga C. merolae, CmDan, were identified simultaneously and are plant-specific dynamin proteins (Gao et al. 2003; Miyagishima et al. 2003b). CmDan was identified by sequence similarity to CmDnml which is involved in C. merolae mitochondrial division. In contrast to mitochondrial localized CmDnml, CmDan was found to localize to the cytosolic side of the chloroplast division site in C. merolae (Miyagishima et al. 2003b). Interestingly, CmDng is found to only localize to the chloroplast division site during the process of fission and is not a part of cytosolic PD-rings (Kuroiwa et al. 2002; Miyagishima et al. 2001a). Furthermore, CmDan stays associated with the division furrow late into division long after the FtsZ ring has disassembled (Miyagishima et al. 2003b) 30 ARCS was identified in an ethyl methanesulfonate (EMS) mutant screen for Arabidopsis plants with defects in chloroplast division having altered chloroplast morphology (Pyke and Leech 1994). In the arc5 mutant, chloroplasts are dumbbell- shaped suggesting ARCS may have a role in completing chloroplast division (Pyke and Leech 1994; Robertson et al. 1996). ARCS is a dynamin-like protein that localizes to a ring outside the mid-plastid division furrow (Gao et al. 2003). Dynamin proteins are known to be involved in cellular membrane fission and fusion and most notably are involved in mitochondria fission (Bleazard et al. 1999; lngerman et al. 2005; Tieu and Nunnari 2000; Tieu et al. 2002). CmDan rings have been isolated and found to exert force on optical tweezers supporting the idea that ARCS/CmDan provide the external force of chloroplast division (Yoshida et al. 2006). PDV1/PDV2 PDV1 was identified as an ARCS phenocopy mutant in a population of EMS mutagenized Arabidopsis plants. Plants contain two types of PDV proteins, PDV1 and PDV2. Mutation in either PDV1 or PDVZ gene show defects in late-stages of chloroplast division, often having dumbbell shaped chloroplasts in mutant lines. pdv1/pdv2 double mutants show an additive, more severe defect in chloroplast division (Miyagishima et al. 2006). ARCS-GFP localizes to mid-plastid rings in pdv1 and pdv2 mutants, but not in the pdv1/pdv2 double mutant. These results suggestthat PDV1 and PDV2, though partially redundant, may work together to recruit ARCS to the division site (Miyagishima et al. 2006). However, the biochemical roles of PDV1 and PDV2 in chloroplast division are unknown. 31 Perspective Bacterial cell division has been a useful model for the discovery of several chloroplast division proteins such as Ftle, FtsZZ, MinDE and others. However, the biochemical properties of plant homologues of bacterial cell division proteins are not well understood. In the case of plant FtsZs, there is reason to believe their fundamental properties may vary from those of bacterial FtsZ, primarily because there are two-types of FtsZ in plants. Moreover, many of the key FtsZ regulatory proteins such as ZipA, FtsA and MinC have no Obvious homologues in chloroplasts, further complicating our understanding of chloroplast division. The work described below seeks to understand the composition of the stromal FtsZ complex and to define the functional properties of plant Ftle and FtsZZ and determine if the Ftle/FtsZZ filament topology is a heterofilament or homofilament. 32 CHAPTER TWO IN VIVO QUANTITATIVE RELATIONSHIP BETWEEN PLASTID DIVISION PROTEINS ”521 AND FI'SZZ AND IDENTIFICATION OF ARC6 AND ARC3 IN A NATIVE FI'SZ COMPLEX. McAndrew, R. S*., B. J. S. C. Olson*, D. Kadirjan-Kalbach, C. L. Chi-Ham, S. Vitha, J. E. Froehlich and K. W. Osteryoung Biochemical Journal (2008). 412: 367-368 *These authors contributed equally to this work 8. J. S. C. Olson contributed the following data: Figures 2.7, 2.10 and 2.11 Confirmation of the data reported in Figure 2.8 B. J. S. C. 0 also co-wrote the discussion, contributed other major portions of the manuscript, and submitted the manuscript for publication in consultation with K. W. Osteryoung. 33 Synopsis Ftle and FtsZZ are phylogenetically distinct homologues of the tubulin-like bacterial cell division protein FtsZ that play major roles in the initiation and progression of plastid division in plant cells. Both proteins are components of a mid-plastid ring, the Z-ring, which functions as a contractile ring on the stromal surface of the chloroplast inner envelope membrane. Ftle and FtsZZ have been shown to interact, but their in viva biochemical properties are largely unknown. To gain insight into the in viva biochemical relationship between Ftle and FtsZZ, we investigated their molecular levels in wild-type Arabidopsis thaliana plants and endogenous interactions in Arabidopsis and pea. Quantitative immunoblotting and morphometric analysis showed that the average total FtsZ concentration in chloroplasts of 3-week-old Arabidopsis plants is comparable to that in E. coli. FtsZ levels declined as plants matured, but the molar ratio between Ftle and FtsZZ remained constant at approximately 1:2, suggesting this stoichiometry is regulated and functionally important. Density gradient centrifugation, native gel electrophoresis, gel filtration and co-immunoprecipitation experiments show that a portion of the FtsZi and FtsZZ in Arabidopsis and pea chloroplasts is stably associated in a complex of "200-245 kDa. This complex also contains the FtsZZ-interacting protein ARC6, an inner envelope membrane (IEM) protein, and analysis of density gradient fractions suggests the presence of the Ftle- interacting protein ARC3. Based on the mid-plastid localization of ARC6 and ARC3 and their postulated roles in promoting and inhibiting chloroplast FtsZ polymer formation, respectively, we hypothesize that the Ftle/FtsZZ/ARCS/ARCG complex represents an 34 unpolymerized, IEM-associated pool of FtsZ that contributes to the dynamic regulation of Z-ring assembly and remodeling at the plastid division site in viva. Introduction The essential cell division protein FtsZ is a polymer-forming, tubulin-like GTPase found in most prokaryotes (reviewed in (Margolin 2005a; Michie and Lowe 2006; Romberg and Levin 2003)). Prior to cytokinesis, FtsZ assembles at the mid-cell division site, just inside the cytoplasmic membrane, to form a contractile ring termed the Z-ring. The in viva molecular structure of the Z-ring is unknown, but in vitro studies suggest it is built from overlapping segments of short protofilaments composed of FtsZ monomers assembled end-to-end and stabilized at the division site through interactions with accessory factors (Anderson et al. 2004; Chen et al. 2005; Chen and Erickson 2005; Redick et al. 2005). Mutations in F132 disrupt cell division, resulting in the formation of long, multi-nucleate bacterial filaments. The Z-ring in Escherichia coli and Bacillus subtilis contains 30-35% of the total FtsZ, but is constantly remodeled by exchange of subunits with a cytoplasmic FtsZ pool in a dynamic process that requires the GTPase activity of FtsZ (Anderson et al. 2004; Redick et al. 2005 ; Stricker et al. 2002). FtsZ concentration is critical for its cell division activity (Sassong et al. 1999; Wang and Lutkenhaus 1993). Alterations in FtsZ levels or in the stoichiometry between FtsZ and other division proteins cause lethal blocks in cell division in vivo, and FtsZ polymerization and GTPase activity are concentration-dependent in vitro (Dai and Lutkenhaus 1992; Dewar et al. 1992; Hale and de Boer 1997; Takada et al. 2005). FtsZ is known to interact 35 with several other proteins that are recruited to the ring in a defined order, but the physiologically relevant sub-complexes deployed to the division site in vivo and the mechanisms regulating Z-ring dynamics in bacteria remain unclear (Margolin 2005b; Michie and Lowe 2006; Pradel et al. 2006; Romberg and Levin 2003). Consistent with the endosymbiotic origin of chloroplasts, plants possess nuclea r- encoded, plastid-targeted homologues of bacterial FtsZ (Osteryoung et al. 1998; Osteryoung and Vierling 1995). Most prokaryotes, including the cyanobacterial relatives of chloroplasts, have a single form of FtsZ; however, plants contain two distinct FtsZ protein families, Ftle and FtsZZ, both of which are required for the proper division of plastids (McAndrew et al. 2001; Osteryoung et al. 1998; Stokes et al. 2000). Similar to their bacterial counterparts, Ftle and FtsZZ colocalize to a mid-plastid Z-ring in the chloroplast stroma adjacent to the inner envelope membrane. The Z-ring assembles prior to the ordered recruitment of other sub-assemblies of the division complex and constricts throughout plastid division (McAndrew et al. 2001; Miyagishima et al. 2001c; Vitha et al. 2001). Recently, Ftle and FtsZZ have been shown to interact separately with the chloroplast division proteins ARC3 and ARCG, respectively (Maple et al. 2005 ; Maple et al. 2007), and may interact with other novel components of the chloroplast division machinery (Glynn et al. 2007; Maple et al. 2007). Thus, Ftle and FtsZZ are functionally distinguished in part by distinct protein-protein interactions. Several lines of evidence suggest that Ftle and FtsZZ function in a complex. They are tightly colocalized to rings in immunofluorescence labeling experiments in both 36 wild type plants and in chloroplast division mutants in which FtsZ filament morphology is perturbed (Fujiwara and Yoshida 2001; McAndrew et al. 2001; Vitha et al. 2003; Vitha et al. 2001). Ftle and FtsZZ interact directly, both individually and with each other, in yeast two-hybrid and bimolecular fluorescence complementation assays (Maple et al. 2005). Ftle is not absolutely required for plastid division since ftle null mutants in Arabidopsis are viable (El-Kafafi et al. 2008; Yoder et al. 2007), but a change in the level of either Ftle or FtsZZ perturbs plastid division (Osteryoung et al. 1998; Raynaud et al. 2004b; Stokes et al. 2000; Strepp et al. 1998), suggesting their stoichiometry relative to one another and/or to other division factors such as ARC3 and ARC6 is important for normal plastid division in viva. As an important foundation for understanding the roles of Ftle and FtsZZ in chloroplast Z-ring dynamics, we are investigating their in viva biochemical properties. In the studies described here, we report endogenous FtsZ1 and FtsZZ protein levels and molar ratios in chloroplasts of the model plant Arabidopsis thaliana. In addition, we show that Ftle and FtsZZ in pea and Arabidopsis chloroplasts are stably associated in a native complex that also contains ARC6; analysis in pea indicates that the complex contains ARC3 as well. These studies are the first in which the in viva quantitative relationship between chloroplast FtsZ proteins and their interactions with accessory factors have been investigated in wild-type plants. Experimental Procedures Plant material 37 Arabidopsis thaliana ecotype Columbia (Col-0) plants were grown as in (Stokes et al. 2000). Pea plants (Pisum sativum var. Little Marvel) were grown in vermiculite as in (Bruce et al. 1994). The mutant lines SALK_134970 and SALK_050397, carrying T-DNA insertions in the AtFtsZZ-l and AtFtsZZ-Z genes were identified in the Salk Institute Genomic Analysis Laboratory database (http://signal.sa|k.ed u/cgi-bin/tdnaexpress) (Alonso et al. 2003) and obtained from the Arabidopsis Biological Resource Center (http://www.arabidopsis.org/abrc/). The positions of the T-DNA inserts were confirmed by sequencing of PCR products amplified from the mutants using the T-DNA left border primer LBb1 (5’-GCGTGGACCGCTTGCTGCAACT-3’) and either an AtFtsZZ-l-specific primer (5’-AGGGGG'l‘l'CGTGGGATATCTG-B’) or AtFtsZZ-Z-specific primer (5’- TATI'GTGTGAATTI'GCTGCC-B’). Individuals homozygous for the T-DNA insertions were identified by segregation analysis. Chloroplast morphology phenotyping Leaf tissue was prepared for microscopic analysis and viewed with a BH-2 (Olympus) microscope as described (Osteryoung et al. 1998; Pyke et al. 1991). lmmunofluorescence labeling Fixation, embedding and immunofluorescence labeling of leaf and floral bud tissue with antibodies specific for AtFtsZ1-1, AtFtsZZ-1 and AtFtsZZ-Z was performed as described (Vitha et al. 2001). Specimens were viewed with a Leica DMR A2 microscope (Leica Microsystems, Wetzlar, Germany) and images were processed as described (Vitha et al. 2003). 38 Preparation of antibodies Polyclonal antipeptide antibodies specific for AtFtle-l (1-1 antibodies) and AtFtsZZ-l (2-1A antibodies) were generated previously (Stokes et al. 2000). The same procedure was used to generate and affinity-purify antipeptide antibodies against AtFtsZZ-l (2-1B antibodies) and AtFtsZZ-Z (2-2 antibodies) using synthesized peptides corresponding to amino acid residues 367-380 (TRRRSSSFRESGSVEI) of AtFtsZZ-l and 244-261 (EGRRRAVQAQEGLAALRD) of AtFtSZZ-2, respectively. Affinity-purified 1-1, 2-18 and 2-2 antibodies were concentrated to 1.7, 1.2, and 2.8 mg/ml, respectively. The specificity of the 2-1A antibody was maximized by pre-absorbing IgG-enriched 2-1A serum on an AtFtsZZ-Z peptide column prior to affinity purification on a 2-1A peptide column (Stokes et al. 2000). 2-1A antibodies were concentrated to 1.3 mg/ml in P85 (140 mM NaCl, 3 mM KCI, 10 mMNazHPO4, 2 mM KHZPO4). ARC6 antibodies were raised against amino acids 216-482 created from a Xhol- Dral fragment of the ARC6 cDNA (At5g42480) ligated into the Xhol-Smal site of the expression vector pJC40 (Clos and Brandau 1994) resulting in a C-terminally His-tagged ARC6. This ARC6 fragment does not include the conserved .I-like domain (Vitha et al. 2003). Recombinant proteins were expressed in E. coli Rosetta (0E3) pLysS cells (Novagen) with 1 mM IPTG and purified by Ni-affinity chromatography (Novagen). Antiserum produced in New Zealand White rabbits (Covance Research Products, Denver, PA) was affinity-purified over a 1 ml protein-A column (Pierce); the column was washed with PBS, and bound antibodies eluted in 0.2M glycine, pH 1.85. After buffering the 39 eluted antibody in Tris-HCI with a final pH of 7.4, the eluate contained 0.72 mg/ml protein. Affinity-purified antibodies were diluted 1:4900 for immunoblotting. An antibody to ARC3 (Shimada et al. 2004) was obtained from Dr. Hiroshi Shimada and verified for specificity by immunoblotting using extracts from wild type Arabidopsis (Col-0) and arc3 mutant plants, extracts from E. coli expressing recombinant ARC3-His, and purified ARC3-HIS (not shown). Expression, purification and calibration of recombinant protein standards Quantitative i mmunoblotting standards were created from fragments of the AtFtle-l, AtFtsZZ-l, and AtFtsZZ-Z cDNAs encoding partial or full-length proteins as shown (Figure 2.1A) expressed in E. coli as N-terminal 10-histidine-tagged fusion proteins. Recombinant proteins were produced as described (Stokes et al. 2000) except that AtFtsZ2-2 was produced in C43(DE3) cells (Miroux and Walker 1996) (Avidis) and the mature AtFtsZZ-l construct (AtFtsZ2-1(m)) was expressed from pDB328 in C41(DE3) cells (de Boer et al. 1989b; Hale and de Boer 1997). Soluble recombinant proteins were purified from cell lysates by Ni-affinity chromatography and their purity was evaluated by SDS-PAGE (Figure 2.18). The purified proteins (>95% pure) were lyophilized, resuspended in storage buffer (50 mM Tris-HCI, pH 8, 1 mM EDTA, 1 mM 4-(2- aminoethyl) benzenesulfonyl fluoride (AEBSF), 10% glycerol), and stored at —20°C. FtsZ protein protein concentration was determined by the BCA and Bradford assays (Olson and Markwell 2007) calibrated against purified E. coli FtsZ (Supplementary Figure 2.28, lane 5) quantified as described (Lu et al. 1998). Protein standards were 40 Figure 2.1: Expression and purification of recombinant protein standards for quantitative analysis (A) Schematic diagram of recombinant protein sequences expressed in E. coli for use as quantification standards. Numbers on right indicate lengths in amino acid residues of pre-proteins encoded by full-length cDNAs (McAndrew et al. 2001). Arrow above indicates approximate length of transit peptides (TP). Residues encoded by recombinant protein constructs (black bars) were 41-269 for truncated AtFtle-1(t), 173-363 for AtFtsZZ-1(t) (25), 81-478 for mature AtFtsZZ-1(m) (Stokes et al. 2000), and 1-473 for AtFtsZZ-Z. (B) Coomassie-stained gel showing recombinant AtFtle-1(t), AtFtsZZ-1(t), AtFtsZZ-1(m), and AtFtsZZ-Z proteins (lanes 1—4, respectively, "'1 mg protein/lane) resolved by SDS-PAGE. Recombinant proteins were analyzed and calibrated for protein concentration and purity relative to recombinant E. coli FtsZ (lane 5). Approximate molecular masses are indicated on the left. 41 A TP H 433 AtFtsZ1-1lt) E—Zl 41 269 478 AthsZZ-m) I 173 363 AtFtsZ2-1(m) :— 81 478 AtFtsZZ-Z 42 adjusted to 1 mg/ml, diluted serially (500 to 2.5 ng in 10 pi), and processed on immunoblots probed with FtsZ-specific primary antibodies and [lzslldonkey-anti-rabbit secondary IgG (8.8 uCi/ug, GE Healthcare) to generate standard curves (Supplementary Figure 2.3) for quantitative analysis of plant FtsZ levels (described below). Standard curves were constructed based on the linear range of detection of radioactive signals. Data generated from the fusion proteins labeled AtFtsZZ-1(m) and AtFtsZZ-1(t) in Figure 2.1 yielded nearly identical standard curves (not shown); therefore, AtFtsZZ-1(t) was used subsequently as the quantification standard for AtFtsZZ-1. Chloroplast isolation Arabidopsis chloroplasts were isolated from protoplasts (Fitzpatrick and Keegstra 2001) prepared from whole shoots of 3-7week-old Arabidopsis plants and all steps after isolation included a protease inhibitor cocktail (5 mM EDTA, 5 mM EGTA, 0.05 mg ml'1 1- chloro-3-tosylamido-7-aminO-2-heptanone, 1 mM benzamidine HCI, 5 mM B-amino-N— caproic acid, 1 LIM leupeptin, 1 uM pepstatin A, and 1 mM AEBSF). Pea chloroplasts were isolated from leaves of 7-9 day-old plants as described previously (Bruce et al. 1994). All steps included a protease inhibitor cocktail (Sigma P2714). Chlorophyll content of chloroplast suspensions was determined as in (Arnon 1949) and chloroplast concentration (organelles/ml) was determined using a heamocytometer. Determination of chloroplast volume Chloroplasts isolated from 3-week Old Arabidopsis plants were immobilized by mixing chloroplasts in import buffer 1:1 with warm low-melting-point agarose, 0.2% 43 (w/v) in import buffer, and cooling the suspension on a slide under a cover glass. 3-D images of chlorophyll autofluorescence (excitation 488 nm, emission 600-700 nm) were acquired using a Zeiss LSM 5 PASCAL confocal microscope (Carl Zeiss) equipped with a 60x/1.4 oil immersion objective. After correction for refractive index mismatch between the immersion oil and the specimen, the effective voxel size was 0.089 pm in the XY, and 0.43 um in the 2 direction, yielding a voxel volume of 3.4 x 10'3 Lima. Image stacks were opened using lmageJ ver. 1.34 software (http://rsb.info.nih.gov/ij), filtered (Median, radius = 4), set to auto-threshold, and voxels for each chloroplast were counted using the VoxelCounter plug-in (http://rsb.info.nih.gov/ij/plugins/voxel- counter.html). Volume was calculated as the number of voxels multiplied by voxel volume. Plastid circumference was measured from image stacks that were rotated using the VolumeViewer plug-in (http://rsb.info.nih.gov/ij/plugins/volume-viewer.html) to show the plastid projection along the longest axis of the plastid, which corresponds to the plastid cross-section at mid-length. The circumference of this projection was then measured in ImageJ. A total of 66 chloroplasts were measured to determine average volume. Quantitative immunoblotting Isolated chloroplasts in import buffer were solubilized by addition of 2X Laemmli sample buffer (1:1) (Laemmli 1970), and diluted with 1X Laemmli sample buffer to various chlorophyll concnetrations. lmmunoblotting on PVDF was performed as previously described (McAndrew et al. 2001) except that blots were incubated overnight 44 with 1-1(1:6000), 2-1A (1:5000), or 2-2 (1:6000) antibodies in TBSTG which is composed of T85 (50 mM Tris-HCI, pH 7.4, 200 mM NaCl) with 0.2% [v/v] Tween 20, 0.25% (v/v) fish gelatin (Sigma-Aid rich). Following washing, blots were incubated for 4 h in TBSTG containing 125l-donkey-anti-rabbit IgG (8.8 uCi/ug, GE Healthcare), at 2-3 uCi/42.5 cmz. Blots were washed, dried, and placed on phosphorimager screens for periods ranging from 1-16 h, then digitized with a phosphorimager (Personal FX, Bio-Rad) and QUANTITY ONE software (Bio-Rad). Three identical blots containing a duplicate set of serially diluted standards and chloroplast lysate proteins were developed concurrently with signal intensities showing less than 10% (.t 5.0.) variation. FtsZ concentrations in chloroplast lysates were determined within the linear range of the standards Figure 2.4 and showed <10% (:I: S.D.) variation between replicate blots. The results of these quantitative immunoblots were combined with microscopic measurements to estimate the number of FtsZ molecules per chloroplast. Immunoprecipitation of in vitro FtsZ translation products before and after import into isolated intact chloroplasts Affinity-purified 1-1 (8 mg), 2-1A (10 mg), 2-18 (10 mg), and 2-2 (10 mg) kTM Gel (Pierce) agarose beads, as antibodies were separately coupled to Ca rboLin recommended by the manufacturer, with a coupling efficiency of "85-90%. Coupled antibodies (A-beads; "8.5 ng antibodies/LII beads), pre-conditioned with 0.1 M glycine, pH 2.5, and equilibrated in TBS plus 0.02% (w/v) thimerosol, were used in immunoprecipitation assays as a 1:1 (bead:buffer) slurry. 2-1A and 2-18 antibodies 45 were coupled to beads and mixed together (1:1) prior to use for immunoprecipitation of AtFtsZZ-l. 5 AtFtle-l, AtFtsZZ-l, or AtFtsZZ-Z were translated in the presence of [3 Sj-Met and imported into pea chloroplasts as previously described (McAnd rew et al. 2001). Import reactions contained 2.1 X 106, 2.0 X 106, and/or 3.3 X 106 dpm of radiolabeled AtFtsZ1-1, AtFtsZZ-l, and AtFtsZZ-Z, respectively, and intact pea chloroplasts (150 pg chlorophyll) in a final volume of 900 pl. Control assays were performed using the nuclear-encoded, small subunit of RuBisCO (SSU). Following import, chloroplasts were treated with thermolysin to degrade non-imported proteins (Cline et al. 1984), re- isolated through 40 % (v/v) Percoll, and washed in import buffer. Intact chloroplasts (75-100 pg chlorophyll) were either solubilized in 2X Laemmli sample buffer for SDS- PAGE, or resuspended in 200 pl of buffer containing 25 mM HEPES-KOH, pH 8.0, 4 mM MgCl2 and protease inhibitors (5 mM EDTA, 5 mM EGTA, 1 mM benzamidine HCI, 5 mM e-amino-N-caproic acid, 1 uM leupeptin, 1 mM AEBSF) and placed on ice for 30 min before initiating immunoprecipitation reactions. Radiolabeled translation products were immunoprecipitated directly from in vitro translation reactions (48 pl containing "120,000 dpm/LII), as described (Anderson and Blobel 1983), using FtsZ-specific A-beads (100 Ill of1:1 slurry, "5 ng antibody ul‘l). Imported proteins (“1.0 x 105 dpm/reaction) were immunoprecipitated from chloroplast lysates, as described (Nielsen et al. 1997), with the following modifications. Briefly, 1 ml of lP-DDM buffer (25 mM HEPES, pH 7.5, 50 mM NaCl, 2 mM EDTA, 2 mM EGTA, 0.2 mM 46 n-dodecyl-fl-D-maltoside (DDM, Anatrace) and protease inhibitors) was added to the chloroplast lysate (200 pl), followed by addition of FtsZ-specific A-beads (100 pl of 1:1 slurry). Reactions were incubated ("12 h, 4 °C) on a rocking table. Beads were collected by centrifugation (10,000 g, 15 s), washed three times with IP-DDM, and once in buffer with no detergent (50 mM Tris, pH 7.5, 150 mM NaCI, 2 mM EDTA, 2 mM EGTA). Bound proteins were eluted by addition of 2X Laemmli sample buffer (no dye) and incubation at 80 °C (5 min) and collected as above. Beads were washed once in water and combined eluants were analyzed by SDS-PAGE and fluorogra phy. Isolation of an endogenous FtsZ complex by two-dimensional gel electrophoresis Intact pea chloroplasts ("2.3 mg chlorophyll/ ml) were lysed in buffer containing 50 mM Tris, pH 7.5, 50 mM MgCl2, 50 mM KCI, 50 mM NaCl, 0.2 mM DDM, supplemented with protease inhibitors. The lysate was incubated on ice (30 min) and passed several times through a syringe fitted with a 25-gauge needle. Stromal protein complexes (400 pg total protein) were recovered in the supernatant following centrifugation (17,000 g, 45 min, 4 0C), mixed 1:1 with native-PAGE sample buffer (40% (w/v) sucrose, 1 M Tris, pH 8.0, 5% (w/v) bromophenol blue, 0.6 mM PMSF) and immediately resolved on native gradient gels (4-20% resolving, 4% stacking, 4°C) (Weigel and Glazebrook 2002). Molecular mass marker complexes ranging from 66-669 kDa (GE Healthcare) were run in parallel for estimating masses of unknown complexes. FtsZ migration in native gels was determined from immunoblots of adjacent duplicate lanes 47 probed with FtsZ-specific antibodies. Putative complexes containing FtsZ proteins were excised from native gels stained with 0.05% aqueous Coomassie Brilliant Blue G Colloidal (Sigma-Aldrich), solubilized by maceration of the gel slices in Laemmli sample buffer, and separated by 10% SDS-PAGE in the second dimension. Pea chloroplast stromal proteins (7 pg) were loaded onto adjacent lanes as controls. Ftle and FtsZZ proteins were detected by immunoblotting as described (McAndrew et al. 2001). Isolation and analysis of FtsZ-containing complexes from Arabidopsis and pea Chloroplasts isolated from pea were lysed hypotonically in lysis buffer (50 mM HEPES-KOH pH 8.0, 4 mM MgCl2, 1 mM DDM) and passed through a 25-guage needle several times. Insoluble material was removed by centrifugation (16,000 g, 10 min, 4 0C). All buffers used for lysis and subsequent analysis contained protease inhibitors. Soluble proteins in isolated pea chloroplasts (50-100 mg chlorophyll) were resolved by sedimentation through a sucrose density gradient (5-20% (w/v), in lysis buffer, 130,000 g, 12 h, 4 °C) and collected in 24 fractions. Gradient fraction proteins were precipitated with 80% ice cold acetone, dried, resuspended (1:20) in Laemmli buffer, and analyzed by SDS-PAGE and immunoblotting using antibodies raised against AtFtle-l, AtFtsZZ-l (McAndrew et al. 2001), and ARC6 (Figure 2.9). In a separate series of experiments, a similar gradient was fractionated into 13 fractions and examined for FtsZZ and ARC3 (Shimada et al. 2004) . Molecular mass standards (66-669 kDa, GE Healthcare) dissolved in lysis buffer were applied to a separate sucrose gradient, and run in parallel. Protein 48 peaks in gradient fractions were detected by UV absorbance at 280 nm and the standard curve is shown in Supplementary Figure 2.6A. To investigate the co-fractionation of radiolabeled, imported Arabidopsis AtFtle-l and ARC6, both AtFtle-l and ARC6 were translated in the presence Of [355]- Met as described above and imported independently into isolated pea chloroplasts. Chloroplasts were treated with thermolysin (Cline et al. 1984), reisolated through 40% Percoll in import buffer, and disrupted in lysis buffer. Proteins were resolved in parallel through identical sucrose density gradients, collected in fractions, precipitated with 80% acetone, dried, and resuspended 1:100 (v/v) in Laemmli buffer. 10 pl from each of two sequentially collected fractions were combined and resolved by SDS-PAGE. Following electrophoresis the gel was soaked for 1 h in 20% (w/w) 2,5-diphenyloxazole (PPO) in dimethysulfoxide (DMSO), washed with water and dried for autoradiography. To investigate the stability of FtsZ-containing complexes, sucrose gradient fractions containing FtsZ were pooled in sets of two, dialyzed in lysis buffer with protease inhibitors, and concentrated by ultrafiltration (Amicon Ultra 15, 30,000 MWCO; Millipore). Proteins from the concentrated pool (“'50 pg total protein) were applied to a Q-sepharose (GE Healthcare) anion exchange column (1 ml bed volume) equilibrated with 50 mM HEPES-KOH pH 8.0, 4 mM MgCl2. Proteins were eluted at 1 ml/min in a linear gradient (0-1M NaCl) applied over 100 min. Fractionated proteins were acetone-precipitated and Ftle, FtsZZ and ARC6 were detected by immunoblotting. 49 Because it was difficult to estimate accurately the molecular mass of the FtsZ containing complex in sucrose density gradients, chloroplast stromal proteins prepared in lysis buffer were applied to a Superdex 200 10/300 GL column (GE Healthcare) equilibrated in lysis buffer (containing 1 mM DDM) and eluted at a flow rate of 0.5 ml/min; 178, 0.2 ml fractions were collected. Fractions were acetone-precipitated and examined by immunoblotting with Ftle and FtsZZ antibodies. The column was standardized with HMW calibration markers (GE Healthcare) prepared in lysis buffer and the molecular mass of the complex was calculated based on a standard curve (Figure 2.11). The retentions of the standards differed slightly in the presence and absence of DDM (not shown). Results Specificity of Arabidopsis FtsZ antibodies Arabidopsis thaliana contains one Ftle gene, AtFtle -1 (At5g55280), and two FtsZZ genes, AtFtsZZ-l (At2g36250) and AtFtsZZ-Z (At3g52750). Previously, we produced antipeptide antibodies against AtFtle-l (1-1 antibodies) and AtFtsZZ-l (2-1 antibodies) that specifically target Ftle and FtsZZ in plant extracts (Vitha et al. 2001). For the studies described here, to ensure discrimination between AtFtsZZ-l and AtFtsZ2- 2, which share "82% amino acid identity, we generated and affinity-purified new antibodies against a peptide from AtFtsZZ-Z (2-2 antibodies), and further purified our previously generated 2-1 antibodies (Stokes et al. 2000; Vitha et al. 2001). Antibody specificity was established by immunoblotting of leaf extracts (Figure 2.2) prepared 50 from WT Arabidopsis plants, transgenic plants expressing antisense constructs for either AtFtle-l or AtFtsZZ-l (Osteryoung et al. 1998), and a homozygous atftsZZ-Z knockout mutant (described below). Consistent with previous results (Stokes et al. 2000; Vitha et al. 2001), the 1-1 antibodies recognized a single protein of “'40 kDa in WT plants, AtFtsZZ-l antisense plants and atftsZZ-Z knockout mutants that was not detected in AtFtle-I antisense plants (Figure 2.2, lanes 1-4) or in a mutant null for AtFtle-l (Yoder et al. 2007). Likewise, the 2-1 antibodies detected two closely migrating proteins of about 45 and 46 kDa in all extracts except those of the AtFtsZZ-l antisense line (Figure 2.2, lanes 5-8). The 2-2 antibodies recognized a single protein in WT and AtFtle-l antisense plants (lanes 9 and 10) that was not detected in the AtFtsZZ-Z knockout mutant (lane 12). These results demonstrate the specificity of the 1-1, 2-1 and 2-2 antibodies for AtFtle- 1, AtFtsZ2-1 and AtFtsZZ-Z, respectively. In addition to establishing antibody specificity, the results shown in Figure 2.2 revealed that the AtFtsZZ-I antisense transgene (Osteryoung et al. 1998) silenced both AtFtsZZ-I (lane 7) and AtFtsZZ-Z (lane 11) and that the two closely migrating proteins recognized by the 2-1 antibodies (Stokes et al. 2000; Vitha et al. 2001) (lanes 5, 6 and 8) are both products of the AtFtsZZ-l gene. AtFtsZZ-Z functions in chloroplast division Prior studies (Osteryoung et al. 1998; Vitha et al. 2001; Yoder et al. 2007) have established unequivocal roles for AtFtle-l and AtFtsZZ-l in plastid division in Arabidopsis. To determine whether AtFtsZZ-Z is also a functional plastid division gene, 51 Figure 2.2: Antibody specificity for AtFts21-1, AtFtsZ2-1, and AtFtsZZ-Z protein (A) lmmunoblot analysis Of proteins (1 mg fresh leaf tissue/lane) from WT Arabidopsis plants (WT, lanes 1, 5, and 9), transgenic plants carrying antisense constructs for AtFtle-l (lanes 2, 6, and 10) or AtFtsZZ-l (lanes 3, 7, and 11), and knockout mutants with a T-DNA insertion in the AtFtsZZ-Z gene (lanes 4, 8, and 12) were probed with affinity-purified antibodies raised against peptide sequences from AtFtle- 1 (lanes 1-4), AtFtsZZ-l (lanes 5-8), and AtFtsZZ-Z (lanes 9-12). The two proteins detected by the AtFtsZZ-l antibodies are consistently detected in Arabidopsis extracts (McAndrew et al. 2001; Stokes et al. 2000). Approximate molecular masses are shown. Antibody Probe 2-1 ...a I 'i" N 1-1 antisense 2-1 antisense -‘ _ 2-2 knockout 1-1 antisense 2-1 antisense !. 2-2 knockout I g 1-1 antisense 2-1 antisense 2-2 knockout I. 0 WT WT WT kDa -46 9101112 C0 1234 567 52 we characterized a T-DNA insertional mutant Of AtFtsZZ-Z (SALK_050397). We also characterized a T-DNA insertion allele of AtFtsZZ-l (SALK_134970). We sequenced the annotated T-DNA insertion sites (httpz/lsignal.sa|k.edu/cgi-bin/tdnaexpress) (Alonso et al. 2003) and confirmed their positions in intron 4 after nucleotide 1363 for the atftsZZ- 1 mutant, and in exon 4 after nucleotide 1417 for the atftsZZ-Z mutant (Figure 2.3A). AtFtsZZ-Z protein was not detected by immunoblotting at any stage of development in the atftsZZ-Z mutant (Figure 2.2, lane 12), indicating that the atftsZZ-Z T-DNA insertion allele is null. AtFtsZ2-1 and AtFtle-l levels were unaffected in the atftszZ-Z mutant (Figure 2.2, lanes 4 and 8). In homozygous atftsZZ-l individuals, AtFtsZZ-l protein was not detected in plants grown from the original seed stock obtained from the ABRC, but AtFtsZZ-l proteins were often detected in progeny of these plants and in later generations (not shown). We conclude that atftsZZ-l is a knockdown allele of AtftsZZ-l with variable expression, probably due to the location of the T-DNA insertion in an intron (Figure 2.3A). Chloroplast morphology in fully expanded leaf mesophyll cells of four-week-old atftsZZ-l and atftsZZ-Z mutants was examined by light microscopy (Osteryoung et al. 1998) and compared with that in WT plants (Figure 2.38). In the atftsZZ-l mutant, chloroplast morphology was highly variable. Large and small chloroplasts were frequently observed in the same cells (Figure 2.3C) and chloroplast number was reduced, often to one per cell. Related phenotypes have been reported in AtFtsZZ-l antisense plants (Osteryoung et al. 1998; Raynaud et al. 2004b). Chloroplast division defects were less pronounced in the atftsZZ-Z mutant (Figure 2.3D), but quantitative 53 analysis showed that these chloroplasts were visibly larger and fewer in number than those in cells of comparable size from WT plants (Figure 2.38) (A. Schmitz and K. W. Osteryoung, unpublished), indicating impaired chloroplast division. The difference in severity of division defects in the afttsZZ-I and atftsZZ-Z mutants may reflect their respective contributions to the total FtsZZ pool, as described below. To further define the role of AtFtsZ2-2 in chloroplast division, we investigated its localization by immunofluorescence labeling. In green chloroplasts of fully expanded leaves, AtFtsZ2-2 was detected in mid-plastid ring structures in WT plants, but not in the atftszZ-Z knockout mutant (Figure 2.3, G and H, respectively). AtFtsZZ-Z also localized to rings in small non-green plastids of young floral buds (Figure 2.3G, inset). The absence of AtFtsZZ-Z in the enlarged chloroplasts of atftszZ-Z mutants did not interfere with ring formation by either AtFtsZZ-l (Figure 2.3F) or AtFtle-l (not shown), which is consistent with the mild division defect observed in the atftszZ-Z plants. Taken together, the localization of AtFtsZZ-Z to a mid-plastid ring in WT plants and the reduced chloroplast division capacity of the atftszZ-Z knockout mutants indicate that AtFtsZZ-Z, like AtFtsZZ-l, is a functional chloroplast division protein. These results also suggest that AtFtsZZ-l and AtFtsZZ-Z are at least partially redundant Quantitative analysis of FtsZ levels in Arabidopsis chloroplasts AtFtsZ levels were measured by quantitative immunoblotting in chloroplasts isolated from whole rosettes of plants ranging in age from 3-7 weeks. 54 Figure 2.3: Chloroplast morphology and AtFtsZZ-1 and AtFtsZ2-2 localization in WT Arabidopsis and atftsZZ-l and atftsZZ-Z T-DNA insertional mutants (A) Positions of T-DNA insertions in mutant alleles of AtFtsZZ-l (At2g36250) and AtFtsZZ-Z (At3g52750). (B-D) Chloroplast morphology in leaf mesophyll cells of WT Col- 0(8), and homozygous atftsZZ-l knockdown (KO; C) and atftsZZ-Z knockout (KO; 0) mutants. (E-G) lmmunofluorescence labeling of AtFtsZZ-l (E and F) and AtFtsZZ-Z (G and H) in leaf mesohphyll cells and floral buds (insets) of WT plants (E and G) and atftsZZ-Z knockout mutants (F and H). FtsZ rings are indicated by arrowheads. AtFtsZZ-l and AtFtsZZ-Z colocalize in WT (E and G). In the atftsZZ-Z knockout mutant, AtFtsZZ-l forms rings (F) and AtFtsZZ-Z is not detected (H). In all panels, the bar = 20 pm. 55 A AtFtsZ2-1 r? SALK—1'34”” _ . (At2936250) .'—. —E'E_E;—E;—E: .. SALK_050397 AtFtsZZ-Z p ' (At3952750) *M ,. AtFtsZ2 2 knockout 8'; N :3 9:. 6‘ < o :9 ‘E < “5 RI 2 ..'-'-.. < 56 Preliminary blots established the linear working range for the purified FtsZ recombinant protein standards (Fig 52) and the range over which FtsZ protein levels and chlorophyll content were linearly correlated in leaf extracts ("‘7-15 pg chlorophyll; Figure 2.48). Leaves, leaf cells and chloroplasts in older plants are on average larger than those in younger plants (Pyke and Leech 1994; Pyke et al. 1991), but are variable in size at any given age; thus the population of chloroplasts isolated at each age represents a developmental range and the measurements performed on these populations represent averages. The average number of FtsZ molecules per chloroplast at each plant age was estimated by combining immunoblotting results with measurements of the average number of chloroplasts per unit chlorophyll in isolated chloroplast suspensions (Table 2.1). In 3-week-old plants, the average number of FtsZ molecules per chloroplast was 101,200 i 6000, with a molar distribution of approximately 33% AtFtle-l, 47% AtFtsZZ- 1, and 20% AtFtsZZ-Z. Although total FtsZ levels declined 10-fold between 3 and 7 weeks, the molar ratio between Ftle and FtsZZ remained at approximately 1:2 (Figure2.5), suggesting this stoichiometry in chloroplasts may be important for FtsZ function. The ratio between AtFtsZZ-l and AtFtsZZ-Z was also somewhat stable (Table 2.1). Comparison of signals on immunoblots suggests Ftle-tO-FtsZZ ratios may be similar in pea, tobacco and spinach (Vitha et al. 2001) (not shown), but we lack the protein standards and knowledge of complete FtsZ gene complement required for rigorous quantitative analysis in other species. Previous calculations by Lu et al. (Lu et al. 1998) have shown there are "15,000 FtsZ molecules in an average log-phase E. coli 57 Figure 2.4: Linear ranges of detection of FtsZ protein standards and FtsZ in chloroplast lysates for quantitative immunoblotting (A) Average densities (CNT/mmz) of phosphorimager signals, within 10% (i 5.0.) agreement after 2 h exposures, were recorded from immunoblots of recombinant protein standards AtFtle-1(t) (circles), AtFtsZZ-1(t) (squares), and AtFtsZZ-Z (triangles) probed with FtsZ-specific primary antibodies and 125l-secondary antibodies, and plotted relative to total protein (ng). FtsZ standards demonstrated reproducible linearity over the protein ranges shown: 10-83.5 ng, 10-65 ng, and 10-62.5 ng, for AtFtle-1(t), AtFtsZZ-1(t), and AtFtsZZ-Z, respectively. (B) To determine the limits of quantitative immunoblotting relative to plant sample load, total FtsZ (ng) in chloroplast lysates of known chlorophyll content (pg) were extrapolated from the standard curves shown in (A). The linear range of chlorophyll (pg), relative to that of the FtsZ standards, is 7.5-20 pg/lane for AtFtle—l (circles) or AtFtsZZ-l (squares), and 10-20 pg/lane for AtFtsZZ-2 (triangles). Experiments were repeated four times in duplicate and data was averaged from separate blots containing both standards and plant extracts generating signals within 10% (i 5.0.) agreement. 58 > Density (CNT/mm2 x 104) Total FtsZ (rig) 10 ' o AtFtsZ1-1(t) r 8 . a AtFtsZZ-1(t) 1 A AtFtsZ2-2 6 .. 4 . 2 n o ' ' 80 100 Total protein (ng) 50 ' a AtFtsZ1-1 40 _ - AtFtsZ2-1 A AtFtsZZ-Z 30 ' 20 ' 1O - 4/ o l n ' I 5 10 15 20 25 Total chlorophyll (pg) 59 Table 2.1: Effects of whole plant age on FtsZ in Arabidopsis chloroplasts Plant age FtsZ levels (average malecules/ chloroplast)” Molar ratio (weeks)' AtFtsZ1-1 AtFtsZZ-l AtFtsZZ-Z Total FtsZ (1-1:2-1:2-2) 3 33,100 47,700 20,400 101,200 1:1.4:0.6 3.5 27,100 45,500 19,500 92,100 1:1.6:0.7 4 20,500 25,900 11,100 57,500 1:13:05 5 6,900 12,000 3,600 22,500 1:13:05 7 3,500 5,100 2,100 10,700 1:1.4:0.6 a Time (weeks post-germination) of plant harvest and chloroplast isolation. b FtsZ molecular levels were calculated from ng FtsZ/chloroplast, and based on molecular masses of 40, 45, and 46 kDa, respectively, for AtFtle-l, AtFtsZZ-l, and AtFtsZZ-Z. (Molecular averages did not vary by more than 8% (iS.D.). cell, which if assembled end-to-end, could in theory form a protofilament encircling the cell’s circumference "20 times. We combined our measurements of chloroplast FtsZ levels with morphometric measurements to compare FtsZ levels in chloroplasts isolated from 3-week-old Arabidopsis plants with those in E. coli (Table 2.2). The average circumference Of the short axis of these chloroplasts at midpoint, measured from 3-D image stacks, was 19.1 i 2.3 pm. Assuming FtsZ monomer dimensions equivalent to those of bacterial FtsZ, 4- 4.5 nm long and “'5 nm wide (Gonzalez et al. 2003; LOwe 1998; Lu et al. 2000; Romberg et al. 2001), the average "101,200 total molecules of FtsZ in chloroplasts of 3-week-old plants (Table 2.1) could theoretically encircle the plastid division site ”21-23 times, close to the estimate of 20 times calculated for E. coli FtsZ (Lu et al. 2000). Although only 30% of the total FtsZ in E. coli is in the Z-ring (Stricker et al. 2002) and the in viva molecular structures of the bacterial and chloroplast Z-rings are not yet known, these calculations suggest that FtsZ levels and overall ring structure in the cell and organelle are comparable with respect to their sizes and the dimensions of their division sites. We estimated the average FtsZ concentration in chloroplasts of 3-week-old plants. Confocal microscopy indicated that the average volume of chloroplasts isolated from these plants was 131 i 55 pm3. If the stroma, in which Ftle and FtsZZ are localized (McAndrew et al. 2001), occupied the entire chloroplast volume, then the average total FtsZ concentration would be "1.28 pM. Because the thylakoids occupy a considerable proportion of the chloroplast volume (Musser and Theg 2000), 61 Table 2.2: Comparison of FtsZ levels in actively dividing E. coli cells with those in chloroplasts of 3-week-old Arabidopsis. Species Cell/Organelle Total FtsZ Polymer Division site Refer circumference” (molecules) lengthd encirclede ence (pm) (pm) (# times) E. coli 3.03 15,000 64 "20 (Lu et (log phase cells)a al. 1998) A. thaliana 21.23 101,200 435 “'21 This (chloroplasts)a work a The majority of chloroplasts in 3-week-old seedlings or log phase E. coli cells are actively dividing. Circumference was calculated as 21tr, for an E. coli cell with "0.48 pm radius or A. thaliana chloroplast with a ~3.37 pm radius, respectively. cThe length of FtsZ polymers was calculated as the total molecules of FtsZ multiplied by FtsZ monomer length, "4.3 nm (Gonzalez et al. 2003; Romberg et al. 2001). dThe number of times FtsZ polymers theoretically encircle the division site was calculated by dividing polymer length (pm) (end-to-end assembly) by cell or organelle circumference (pm). 62 80000 60000 40000 20000 Molecules per chloroplast 3 3.5 4 5 7 Whole plant age (weeks) Figure 2.5: Quantitative analysis of molecular levels of Ftle and FtsZZ in wild-type Arabidopsis Total Ftle (black bars) and FtsZZ (gray bars) in Arabidopsis chloroplasts were determined for plants of different ages. Total molecular levels of FtsZ were 101,200 (15050), 92,930 ($4650), 57,310 ($2860), 18,690 (i940), and 10,720 (1540) in chloroplasts of 3-, 3.5-, 4-, 5-, and 7-week-old plants, respectively. Error bars are i SD. 63 this value represents a minimum concentration; if the stroma represents 50% of the total volume, as has been reported, the concentration would be twice this, or 2.56 pM. These values are within the range of concentrations required for both cooperative assembly (0.3-3 pM) and full GTPase activity (0.5-2 pM) of E. coli FtsZ in vitro (Anderson et al. 2004; Caplan and Erickson 2003; Chen et al. 2005; Redick et al. 2005; Rueda et al. 2003). AtFtsZ proteins are stably associated following import into pea chloroplasts Ftle and FtsZZ tightly co-Iocalize to rings in viva (McAndrew et al. 2001; Vitha et al. 2003) and recombinant and GFP-fused forms Of Ftle and FtsZZ interact in yeast and transgenic plants (Maple et al. 2005). To begin investigating in viva interactions between Ftle and FtsZ2, we performed a series of immunoprecipitation assays on radiolabeled precursor (p) AtFtle-l, AtFtsZZ-l, and AtFtsZZ-Z proteins generated by in . . . 35 . . vrtro translation In the presence of[ S]-Met that were subsequently Imported Into isolated pea chloroplasts to yield processed, mature (m) import products (McAndrew et al. 2001). Following import, chloroplasts were treated with protease to remove unimported precursors and reisolated. Radiolabled precursor and imported mature proteins were detected by SDS-PAGE and autoradiography (shown for p-AtFtle-l and m-AtFtle—l in Figure 2.6A). For immunoprecipitation experiments, 1-1, 2-1, and 2-2 antibodies coupled to agarose beads were incubated with either radiolabeled mature proteins present in post- 64 import chloroplast lysates or radiolabeled precursor proteins. Analysis of bound (8) and unbound (U) fractions showed that all radiolabled AtFtsZ proteins, both precursor (Fig 4B, lanes 1, 2, 7 and 9 and not shown) and mature (Fig 4B, lanes 3, 8, 10 and not shown) forms, were immunoprecipitated by their corresponding antibodies. Due to the complexity of the experiments, only a subset of the results is shown in Figure 2.6. In control assays, radiolabeled, imported RuBisCO small subunit (SSU) was not precipitated by any of the AtFtsZ antibodies (shown for 1-1 antibodies in Figure 2.68, lanes 5 and 6), and neither agarose beads alone (not shown) nor pre-immune IgG coupled to agarose beads (PI) pulled down any of the radiolabeled AtFtsZ proteins (shown for m-AtFtsZZ-Z incubated with PI in lane 12). To investigate FtsZ interactions, radiolabeled AtFtsZ precursors were first incubated with pea chloroplasts in combinations of two: AtFtle-l and AtFtsZZ-Z, AtFtsZZ-l and AtFtsZZ-Z, or AtFtle-l and AtFtsZZ-l. In all cases, co-import of both proteins was observed (Figure 2.6C, lanes 1-3, respectively). Following import and recovery of intact chloroplasts, equal amounts (pCi) of rad iolabeled mature proteins in chloroplast lysates were incubated with FtsZ-specific antibodies and analyzed as above. The results indicated that any two coimported AtFtsZ proteins (Figure 2.6C, lanes 1-3) were coprecipitated (~1o—20% yield) by antibodies specific for either protein (Figure 2.6C, lanes 4-9). In contrast, when two precursor proteins were incubated together with any of the FtsZ antibodies, only the radiolabeled protein recognized by the antibody was precipitated ("5% yield) (not shown). In experiments in which all three AtFtsZ proteins were coimported, all three mature proteins could be coprecipitated by any one FtsZ- 65 specific antibody (”12-30% yield) indicating their interaction in a stable complex (Figure 2.60). Moreover, coimported AtFtsZZ-l and AtFtsZZ-Z (Figure 2.6E), as well as singly imported AtFt522-1 (not shown), could be pulled down by the Ftle antibody. These results Show that the imported Arabidopsis FtsZZ proteins associate with the endogenous pea Ftle in organello. FtsZI and FtsZZ are in a stable complex in viva Ultracentrifugation experiments were used to examine whether Ftle and FtsZZ are in a complex in vivo as suggested by the coimport/coimmunoprecipitation experiments. Chloroplasts from 7-9 day-old pea leaves were lysed in the presence of DDM to partially solubilize the membranes and the soluble protein fraction was sedimented through a 5-20% sucrose gradient. Gradient fractions were analyzed by immunoblotting. Replicate experiments showed that Ftle and FtsZZ primarily cosedimented in the upper region Of the gradient in a mass range between ~215-240 kDa (Figure 2.7A). The association between Ftle and FtsZZ was further examined by pooling the peak FtsZ-containing density gradients fractions and subjecting them to anion exchange chromatography, where both proteins co-eluted at "300-400 mM NaCl (Figure 2.78). The DDM-solubilized chloroplast proteins were also separated by size exclusion chromatography and the fractions examined by immunoblotting (Fig 5C). Consistent with the sedimentation analyses, both Ftle and FtsZZ co-eluted in fractions corresponding to a mass range of 213-243 kDa. 66 Figure 2.6: Panels A and B, Immunoprecipitation of in vitro [355]-labeled AtFtsZ translation and import products by FtsZ-specific antibodies 35 AtFtle-l, AtFtsZZ-l and AtFtsZZ-Z cDNAs were translated in vitro with [ S]-Met to produce precursor proteins and imported into isolated pea chloroplasts to yield processed, mature proteins. Following import, chloroplasts were treated with thermolysin and reisolated. lmmunoprecipitation reactions, carried out using 1-1, 2-1, 2-2 or preimmune (PI) antibodies coupled to agarose beads, were performed directly on the in vitro translation mixture for precursor proteins (p-1-1, p-2-1, p-2-2, precursor AtFtsZ1-1, AtFtsZZ-l and AtFtsZ2-2, respectively) and on soluble chloroplast fractions for mature proteins (m-1-1, m-2-1, m-2-2, mature AtFtle-l, AtFtsZZ-l and AtFtsZ2-2, respectively). Translation, import and immunoprecipation products were detected by SDS-PAGE and autoradiography. (A) p-AtFtle—l (lane 1) and m-AtFtle—l (lane 2). (B) Precursor (lanes 1-2, 7, 9) or individually imported (lanes 3-6, 8, 10-12) proteins were immunoprecipitated with the antibodies shown above panel and proteins either bound (8) or unbound (U) to the antibody beads are shown. All precursor and mature proteins were specifically bound by their corresponding antibodies, which did not recognize SSU. AtFtsZ2-1 is a doublet as previously reported (Vitha et al. 2003). In a control experiment, imported SSU was detected only in the unbound fraction after immunoprecipitation with 1-1-antibodies (lanes 5-6). 67 A B In vitro translation and import Antibody used for '\ Nb Coimmunoprecipitation él/ 8 kDa 50 <- p-1-1 _ m-1-1 36 29 21 m-SSU 1 2' Precursor/Mature Protein 68 Figure 2.6 (Continued): (C) Pairs of the precursor AtFtsZ proteins were co-imported into pea chloroplasts and the mature proteins (left panel in C) immunoprecipitated with the FtsZ antibodies indicated. (D) All three AtFtsZ proteins were co-imported into chloroplasts and immunoprecipitated with the different AtFtsZ antibodies (lanes 1-3) but not by uncoupled beads (lane 4). (E) Following co-import of AtFtsZZ-l and AtFtsZZ-Z, the mature import products could be co-precipitated by 1-1 antibodies (lane 1), but not by Pl antibodies (lane 2), indicating their interaction with endogenous pea Ftle. 69 Antibody used for C Coimport Coimmunoprecipitation Products m-AtFtsZZ-1 _> m-AtFtsZ2-2 —> m-AtFtsZI-1 _, Mature Protein x’N Antibody used for Antibody used for Coimmunoprecizaitation Coimmunoprecipitation D E :5. 2'. it E 4— m-AtFtsZZ- 1 4— m-AtFtsZZ- 2 -._....... [:1 1 2 Mature X‘Vq’ q’fi/ Protein q? 5" \x (1’ 70 Figure 2.7: Fts21 and FtsZZ cofractionate in isolated pea chloroplasts A Isolated intact pea chloroplasts were lysed hypotonically in the presence of DDM and soluble proteins were fractionated as indicated. Proteins in the fractions shown were acetone-precipitated and analyzed by SDS-PAGE and immunoblotting using AtFtle-l and AtFtsZZ-l antibodies. (A) Analysis of fractions from a 5-20% sucrose density gradient. Approximate peak locations of molecular mass standards separated in a parallel sucrose gradient and identified in fractions by UV absorbance are indicated. Ftle and FtsZZ cosedimented in a mass range of "215-240 kDa. (B) The peak FtsZ- containing fractions (5-8) from panel A were pooled and subjected to anion-exchange chromatography using a 0-1 M NaCl elution gradient. Ftle and FtsZZ coeluted in 300- 400 mM NaCI. (C) Soluble proteins from chloroplasts lysed in the presence of DDM were fractionated by gel filtration chromatography. lmmunoblots of the subset of fractions (0.2 ml) near the FtsZ complex peak are shown. The standard curve is shown in Figure 2.1. 71 kDa 140 232 9 l I 440‘ 661 N 7' ‘9 00 o v o ‘— " v- n— (\I N “.1 ‘l' “P °? ‘7 .'. (J) “5 .L 0') J. ‘— (O m N a) ‘— ‘_ \— ‘— ‘_ N 2 0M NaCI 1M NaCI J I C (232 kDa (140 kDa Fraction 60) Fraction 98) Bass iii 82 11:31:;ng FtsZ1 .n_-._—-.~. WI. 72 Figure 2.8: 2-D native SDS-PAGE separation of an endogenous protein complex from pea chloroplasts containing Ftle and FtsZZ (A) Isolated intact pea chloroplasts were lysed hypotonically in the presence of DDM and soluble proteins were separated by native PAGE through a 4-20% gradient gel. Replicate samples loaded onto the same gel were Coomassie-stained (lane 1) or analyzed by immunoblotting with AtFtsZ1-1(Iane 2) and AtFtsZZ-Z (lane 3) antibodies. Relative to the molecular mass markers (left), the native RuBisCO complex (550 kDa, right) migrated at an apparent mass of "470 kDa and a complex containing only Ftle migratesjust below this mass (indicated by *). A putative complex containing pea Ftle and FtsZZ (boxed) migrated at an apparent mass of "200 kDa. (B) The putative FtsZ1- and FtsZZ-containing complex (C) was excised from the Coomassie-stained native gel shown in panel A and separated by SDS-PAGE in the second dimension. Pea chloroplast stromal proteins (5) were loaded in adjacent lanes as controls. Replicate samples loaded onto the same gel were Coomassie-stained (lanes 1 and 2) or analyzed by immunoblotting with AtFtle-l (lanes 3 and 4) or AtFtsZZ-Z (lanes 5 and 6) antibodies. Migration of molecular mass markers and FtsZ proteins are indicated on the left and right, respectively. 73 Antibody Probe kDa COO!“ ‘_ Rubisco <- FtsZ containing comp|ex Antibody Pr°be kDa Coom 1'1 2-2 3 C so—is 74 In a separate approach, protein complexes from intact pea chloroplasts were separated by native-PAGE using 4-12% gradient gels (Figure 2.8A, lane 1). A distinct band that resolved at "200 kDa was recognized by both the Ftle and FtsZZ antibodies (Figure 2.8A, lanes 2-3). Notably, the 550 kDa RuBisCO complex [57] migrated with an apparent mass of "470 kDa, indicating the mass of the Ftle—and FtsZZ-containing complex was probably underestimated by this method. Based on the results of native PAGE, density gradient centrifugation and anion exchange chromatography, we conclude that endogenous pea Ftle and FtsZZ associate in a stable, discrete complex in viva with a mass between "200 and 245 kDa. The native immunoblots also showed smearing Of the FtsZ1 and FtsZZ signals, suggesting the presence of larger FtsZ-containing complexes or assembled protofilaments of various lengths (Figure 2.8A, lane 2). In addition, Ftle was detected in a well-resolved complex of ~440 kDa (Figure 2.8A, asterisk) that did not contain detectable FtsZZ; however a similarly sized complex was not evident in the pea density gradient or size-exclusion chromatography fractions (Figure 2.7A, 5C). The composition of this Ftle-containing complex was not further investigated in this study. In a bid to identify other proteins associated with the Ftle- and FtsZZ-containing complex, the "200 kDa band resolved by native PAGE was excised from the native gel and analyzed by SDS-PAGE (Figure 2.88). While Ftle and FtsZZ were readily detectable 75 by immunoblotting, neither was stained by Coomassie, precluding visual detection of other proteins potentially associated with the complex. The native FtsZ complex also contains ARC6 and ARC3 Interaction assays have shown that Ftle and FtsZZ interact with the chloroplast division proteins ARC3 and ARC6, respectively (Maple et al. 2005; Maple et al. 2007). To ask whether FtsZ, ARC6 and/or ARC3 can be detected in an endogenous complex, DDM- solubilized chloroplast proteins from 7-10 day-old pea plants were fractionated on sucrose density gradients and analyzed by immunoblotting using antibodies against Arabidopsis ARC6 (Figure 2.9) and ARC3 (Shimada et al. 2004). The anti-ARC6 antibodies detected two cross-reactive pea proteins of "100 kDa, similar to the mass of ARC6 (Vitha et al. 2003). The doublet could be indicative of posttranslational modification. A significant portion of the smaller protein sedimented at the bottom of the gradient, suggesting it is membrane-associated (Figure 2.10A, fraction M), as is ARC6 (McAndrew et al. 2001; Vitha et al. 2003). Within the gradient, this protein co-sedimented with the peak Ftle- and FtsZZ-containing fractions at ”215-240 kDa (Figure 2.10A, fractions 5-8). In a separate experiment, immunoblots of density gradient fractions were probed with antibodies against Ftle, FtsZZ, ARC6 and ARC3. As in Figure 2.10A, FtsZ1, FtsZ2 and the putative pea ARC6 co-sedimented at 215-240 kDa (not shown). 76 Figure 2.9: Specificity of ARC6 Antibody lmmunoblot analysis shows immunoreactivity of anti-ARC6 with the recombinant ARC6 immunogen (rARC6, lane 1) and proteins (1 mg fresh tissue/ lane) from WT Arabidopsis plants (WT, lane 2), transgenic plants carrying ARC6-GFP (lanes 3) the arc6-1 mutant (lane 4), and overexpressed 35$:ARC6 (lane 5). Approximate molecular masses are indicated at the left. 77 0 (o: 9 B 069 6,0? ’\ (3‘0 O 6’ 6- 53‘ 433 Eg- rzic’ '5‘” 201- . 115__ .. j . <1ARC6-GFP ._.... .. m <1ARC6 54 — * mmwmmm®¥ .. ._M wwv HINNuuhHfl iiiii .HHHHHHH HNH HIH uuuhaa oma HHHHHH. §ZH2E<>HHOHGHIHHOOHHHHSH 09/92 HHHH HEB: H H m. noun HHoou iiiiiiiiiiii omHHHmmmmmmiiiiiiHm<zO HHHHHOHHH QHH Himnmphnfl mHHmHHm-HOHOHHHHZOHZHOO iiiii HmmHZHHHHHHHHHH HHHOHmomH mem HIH unuhefl owe owe owa omH OHH oma H IHHHHHHH HHz mHH>om>mz> oooo oooo_oooo HHHHZZHH HmH H>>H HmHHH H EH H Hm>omeHHHH>HH

B EH mHHH- HmN Hana HHoou a H>> HHHHI owm Hiunmuhed HQHmmi mHm HIH smoked O O O O O O O O O 0 O O O O O O O O O O O O O O O 0 0 O 0 O O O O oma QHH 004 can omm oem mm Hom<> :; H-,H THHH a o HmH uuuh HHoau meHHH H HHmm 2m HHE quur. a Hom Hiuuuuhad H Hmmm H max . HA H-.1 H_HH Hm 0 o Hmm HIH moaned oooo oooo o oooo oooo oooo oooo oooo_o o oooo oooo oooo omm omm cam omm omm OHm l 8222: aboaosfie HHmHHHoH>H a HHHOH H: HmH anon HHooa H>HHmom>H mm m HHH Hiuunuhefl .OHHHOHHHHH . » HH HmH HiH unanna oooo_oooo_oooo_coon—oooo_oooo oooo_oooo oooo oooo oooo N 0mm OPN omm omN 95 Tubulin differs from bacterial FtsZ because the T7-loop of B-tubulin has an amino acid substitution GxxNxDxxE -) GxxNxxDxxK which makes the a-subunit incapable of GTP hydrolysis (Nogales et al. 1998a). In addition, lack of GTP hydrolysis by the a-subunit prevents aB-tubulin dimers from dissociating, resulting in strict heterodimerization of OLB-tubulin. Bacterial FtsZ filaments are formed exactly like tubulin, except there is only one type of FtsZ. Because tubulin and FtsZ have trans GTP binding sites, the ends of FtsZ and tubulin filaments are biochemically distinct. The plus end of a growing microtubule or FtsZ filament is GTP-capped by the glycine-rich, GTP binding motif. Filament growth occurs when the T7-loop of an incoming subunit completes the active site at the plus end. In contrast, filament growth is kinetically disfavored on the T7-loop-containing minus ends of the filament. Thus, FtsZ and microtubules grow primarily at their plus ends. Due to biochemically distinct ends of FtsZ and tubulin, 17-loop mutations block incoming subunit addition, thus preventing protofilament elongation, but T7-loop mutations do not affect lateral interaction with other protofilaments (Redick et al. 2005; Scheffers et al. 2002). ‘For the purposes of this study, 17 loop mutations will be used to probe the difference between head-to-tail protofilament interaction and lateral stabilization of Ftle/FtsZZ co-polymers. Bacterial FtsZ assembly is mechanistically similar to tubulin assembly, differing mainly in kinetics of each step in the process. During cytokinesis, a maximum of 35% of monomeric FtsZ is assembled into division rings (Anderson et al. 2004; Stricker et al. 2002). Thus, most unassembled FtsZ is monomeric and GDP-bound. In vitro, the 96 assembly of FtsZ is modeled as follows: GDP—FtsZ exchanges nucleotide resulting in GTP-FtsZ, which dimerizes (Chen et al. 2005; Huecas et al. 2007a ; Rivas et al. 2001; Rivas et al. 2000; Romberg et al. 2001). Dimeric GTP-FtsZ cooperatively assembles onto protofilament plus-ends. The incoming 17-Ioop from dimeric FtsZ (or OLB-tubulin) completes the active site of the plus-end and GTP hydrolysis begins. GTP hydrolysis is rate-limiting in FtsZ (Romberg and Mitchison 2004) and hydrolysis is not required for assembly (Scheffers et al. 2000). Bacterial FtsZ immediately hydrolyzes GTP upon assembly (Chen and Erickson 2005; Huecas et al. 2007b; Romberg and Mitchison 2004; Scheffers and Driessen 2002), but in tubulin, the GTPase activity lags behind the rate of polymerization (Nogales 1999). In the absence of fresh GTP, FtsZ polymers rapidly hydrolyze GTP and disassemble. FtsZ has a faster GTP hydrolysis rate and FtsZ polymers are less stable than tubulin (Mukherjee and Lutkenhaus 1999). Unlike tubulin, FtsZ is able to maintain its polymerized state by exchanging fresh GTP within the polymer, preventing disassembly. Likewise, polymerized FtsZ (but not tubulin) filaments can be disassembled by addition of GDP which exchanges into polymerized FtsZ (Mingorance et al. 2001). In contrast, tubulin stability is maintained by a slow GTPase activity and the action of many microtubule regulating proteins (Nogales 1999). Tubulin laterally associates into bundled tubes by the lateral interaction primarily between a- and [3- subunits (Nogales 1999). FtsZ filament bundling is believed to occur, but lateral association of FtsZ filaments in the absence of stabilizing agents has not been observed in vitro. It is unclear how lateral association between FtsZ filaments would affect filament stability and GTP hydrolysis and exchange. 97 The aims of this study are to determine if both FtsZI and FtsZZ are GTPases and capable of assembly similar to bacterial FtsZ and tubulin, and to distinguish between assembly into homopolymers or heteropolymers. To answer these questions, Ftle and FtsZZ were produced recombinantly and without their putative chloroplast transit peptides to study their biochemical properties in vitro. In the following series of experiments, we report that Ftle and FtsZZ stably polymerize into heterofilaments that laterally associate into ribbons under dynamic conditions. Experimental Procedures Expression and purification of recombinant FtsZ proteins Previous attempts to characterize the biochemical properties of plant FtsZ (El- Kafafi et al. 2005; Gaikwad et al. 2000) did not remove the chloroplast transit peptides, which interfere with their function. Additionally, FtsZZ examined by El-Kafafi and colleagues (El-Kafafi et al. 2005) had a C-terminal truncation that removed at least "25 amino acids conserved with bacterial FtsZ. Moreover, the ChloroP predicted transit peptide of Ftle would remove part of its glycine rich GTP-binding domain (Figure 3.1), so we sought to produce physiologically relevant forms of Ftle and FtsZZ (design is discussed in the results). FtsZi and FtsZZ were amplified by PCR from cDNA clones of AtFtle-l (ABRC clone U09686) and full length AtFtsZZ-l (McAndrew et al. 2001). Note, previous results from our lab with a cDNA for AtFtle-l (McAndrew et al. 2001; Stokes et al. 2000) (Vitha et al. 2001) contains the mutation 5115F in its GTP binding domain, which interferes with its polymerization properties (not shown), but ABRC clone U09686 98 was sequenced matches the gene sequence annotated in TAIR (http://www.arabidopsis.org). Ftle and FtsZZ were amplified without their transit peptides with the primers 5’-AGTGGTCCATGGCCAGGTCTAAGTCGATGCGATTG-B’ and S’- TGCACCCTCGAGCTAATGATGATGATGATGATGGAAGAAAAGTCTACGGGGA—B’ for Ftle and 5'-AGTGGTCCATGGCCGCCGCTCAGAAATCTGAATC—B’ and 5’- TGCACCCTCGAGTTAATGATGATGATGATGATGGACTCGGGGATAACGAGAGCTG-B’ for FtsZZ. These Ftle and FtsZZ constructs result in the deletion of the putative N-terminal chloroplast transit peptide (57 and 48 amino acids respectively, CT P choice discussed in results) and addition of the amino acids “MA” to complete an Ncol cloning site and insert an ATG start codon for translation. On the C-terminus, a 6X His tag with a stop codon followed by an Xhol site was inserted in the primers. Ftle and FtsZZ PCR fragments were digested with Ncol and Xhol and the digested fragments were sub- cloned into the Ncol and Xhol sites of the expression vector pDB38 (McAnd rew et al. 2008). Note that cloning into the Ncol site removes an HFKT-tag in pDBBZB. Prior to expression, the constructed gene was confirmed by DNA sequencing. Finally, AtFtsZZ-l and AtFtsZZ-Z can substitute for one another in viva, indicating they are biochemically redundant (Aaron Schmitz, in preparation); therefore, we used only AtFtsZZ-l in this study. For expression of recombinant proteins, FtsZ expression plasmids were transformed into Rosetta(DE3) cells (Novagen) harboring pBSSB, which overexpresses the E. coliftsQAZ operon and suppresses filamentation of the cells during expression (Jeong and Lee 2003). Plant FtsZ was optimally expressed as follows: an overnight 99 culture of cells was grown at 37°C in LB supplemented with 100 (Lg/ml carbenicillin, 50 ug/ml spectinomycin and 50 ug/ ml chloramphenicol and then sub-cultured 1:1000 into fresh LB media containing 50 ug/ml carbenicillin, 25 ug/ml spectinomycin and 12.5 ug/ml chloramphenicol, grown to an OD600 of "0.6 and expression was induced with 0.5 mM IPTG for 4h at 37°C. Several different temperatures and conditions were tested for expression, but we consistently found that >80% of the protein was found in inclusion bodies. Because protein purified from the soluble fraction typically had contaminating ATPase activity that was difficult to remove by high-salt washes or other chromatographic techniques we chose to refold plant FtsZ from inclusion bodies. After expression, cells were harvested at 9,000 g for 10 min at 4°C and dry pellets were stored at -80°C until extraction and purification of recombinant protein. Frozen cell pellets were thawed on ice and resuspended in 0.01 volumes the initial culture volume in ice-cold extraction buffer (25 mM TrisCl pH 8.0, 500 mM NaCl, 5 mM imidazole with Roche Complete EDTA-free protease inhibitors). Trition-X 100 was added to 0.1% and the cells were sonicated six times for 30 sec at full tip power using a Bronson microtip sonicator. Insoluble debris and inclusion bodies were collected by centrifugation at 18,000 g for 20 min at 4°C, the supernatant typically contained <20% of the expressed protein and was discarded. Inclusion bodies were resuspended in extraction buffer and sonicated again. Urea was added to 6 M and prior to affinity chromatography, insoluble material was removed by centrifugation at 18,000 g for 20 min at 4°C. 100 Ftle and FtsZZ proteins were purified by affinity chromatography using Ni- Sepharose (GE Healthcare) under denaturing conditions. Crude lysates were applied directly to a 15 ml Ni-Sepharose column (GE Healthcare) at 1 ml/min and washed with 2 column volumes of buffer A (25 mM TrisCl pH 8.0, 500 mM NaCl, 6 M urea). FtsZ was eluted with a linear gradient of 0-1 M imidazole in buffer A at 5 ml/min; FtsZ proteins typically eluted between "100 and 300 mM imidazole and were pooled prior to refolding. Ftle and FtsZZ were refolded by dialysis. Because the high concentration of imidazole from elution interferes with protein refolding, imidazole was first removed by dialysis against 25 mM TrisCl pH 8.0, 500 mM NaCl, 3 M urea. Following imidazole removal, “'1 mM GDP was added directly to the dialysis bag and plant FtsZ was dialyzed against three changes of 25 mM TrisCl pH 8.0, 500 mM NaCl. 500 mM. GDP was required to prevent aggregation during refolding. Some protein aggregation was sometimes observed once the concentration of urea fell below "1 M; this was removed by centrifugation at 14,000 g for 20 min at 4°C. Following dialysis, FtsZ protein was concentrated by ultrafiltration (Amicon Centricon, molecular weight cut-off of 30 kDa) to approximately 15 uM and glycerol was added to 10% before storage at -80°C until use. Prior to assay, FtsZ proteins were thawed on ice and centrifuged for 20 min at 12,000 g at 4°C to remove aggregated protein. Bacterial FtsZ proteins are typically purified with a round of assembly-based purification to select for active protein. However, we find that Ftle and FtsZZ assemble individually with poor efficiency. 101 Typically, "25% of the plant FtsZ is recovered by CaClz, assembly-based purification individually, whereas we typically find >90% of E. coli FtsZ is recovered by assembly- based purification. Initially, poor recovery of Ftle and FtsZZ by assembly-based purification suggested our plant FtsZ preparations were not functional, but the low recovery was found to result from both Ftle and FtsZZ being required for efficient assembly (see Results). Instead, aggregated protein was removed by desalting with a HiPrep 26/10 column at 10 ml/min equilibrated with TMK (25 mM TrisCl pH 7.0, 0.1 mM MgCl2, 100 mM KCI). The yield of plant FtsZ from desalting is 50-80% greater than assembly based purification, but there was little difference in protein activity. Following gel filtration into TMK, FtsZ protein was concentrated to approximately 15-20 uM using an Amicon Centricon (30 kDa molecular weight cut-off) and additional aggregated protein was removed by centrifugation at 18,000 g for 20 min at 4°C (typically less than 2% of the protein was aggregated). To verify proper refolding, Ftle and FtsZZ were subjected to sucrose density gradient fractionation with a linear 0-20% gradient prepared in HMK and both Ftle and FtsZZ fractionated near the 67 kDa marker (not shown). SDS-PAGE and quantitative amino acid analysis (QAAA) were used to assess the purity of recombinant FtsZ. A typical plant FtsZ preparation was >95% pure. QAAA demonstrated Ftle and FtsZZ mature proteins are underestimated by the BCA assay by 20%, similar to E. coli FtsZ. This result is also consistent with our previous standardization for quantification (McAndrew et al. 2008). Occasionally, FtsZZ protein 102 preparations would have a secondary lower MW band, likely due to initiation off of a downstream ATG codon in E. coli. The truncated protein was removed by reverse phase chromatography using a Source RPC 3.1/300 column (GE Healthcare) with a linear gradient of 0.1% trifluoracetic acid (TFA) to 1:10:90, TFA:H20:acetonitrile over 20 CV. The eluted protein was refolded as described above and purity was typically >98%. E. coli FtsZ was prepared by established methods and immediately before assay was subjected to a round of calcium stabilized polymerization and depolymerization (Lu and Erickson 1998; Lu et al. 1998). E. coli FtsZ was typically >90% pure as assessed by Coomassie staining SDS-PAGE gels. Site-directed mutagenesis Expression plasmids for Ftle and FtsZZ were mutagenized using established methods (Fisher and Pei 1997). Briefly, primer pairs 5’- GTCAATGTGGA'lTlTGCAGCTGTGAAGGCAGTCATGAAA-3’ and 5’- ‘lTl’CATGACTGCCTTCACAGCI'GCAAAATCCACATTGAC-3’ or 5’- GTGAATGTGGA'I'TTI'GCTGCTGTGAGAGCTATAATGGCA-S’ and 5’- TGCCA'I'l'ATAGCT CT CACAGCAGCAAAATCCACA‘ITCAC-B’ were used to make D275A or D322A mutations in Ftle or FtsZZ. This results in the T7-loop mutation NVDFAD to NVDFAA. Plasmids were amplified with Ex-Taq (Takara) using the manufacturer’s buffers for 14 cycles. Following amplification, the parental plasmid was digested with Dpnl (New England Biolabs) for 1h at 37°C. 2 ul of the digestion was used to transform competent DHSO. and the mutation was confirmed by sequencing. 103 GTP binding and hydrolysis assays Nucleotide binding to FtsZ proteins was performed as previously described (Redick et al. 2005). The GDP content) of assembled FtsZ was assayed using the same technique developed for E. coli FtsZ (Chen and Erickson 2008; Romberg and Mitchison 2004; Small and Addinall 2003). Briefly, a GTP regeneration system is established with pyruvate kinase that phosphorylates GDP from phosphoenolpyruvate to make GTP. Thus, non-polymer bound GDP is rapidly converted to GTP and the GDP measured is only that bound to polymerized FtsZ. Instead of using radioactive GTP, we detected GDP and GTP by HPLC (Romberg and Mitchison 2004). 5 uM FtsZ was polymerized in the presence of 0.5 mM GTP in HMK, deproteinated with a 5-fold excess of ice-cold 1 M perchloric acid and centrifuged for 20 min at 4°C. The supernatant was injected into a 0.5 ml sample loop on a AKTA Purifier HPLC and loaded onto a TOSOH DEAE SP-SW 0.7 mm x 7 mm anion exchange column equilibrated with 100 mM NH4HC03 at 1 ml/min at 25°C and detected by absorbance at 254 nm. Nucleotide was eluted with a linear gradient to 0.5 M NH4HC03 at 1 ml/min over 3 column volumes. The anion-exchange column was calibrated before and after nucleotide binding assays with GDP and GTP. GDP and GTP elute as two closely spaced peaks at "13.5 and 14.5 ml, respectively, and we estimated the GDP in the polymer by integrating the area underneath the base-line corrected peak corresponding to GDP in Unicorn 5.1 software (GE Healthcare). GDP . . . . . . -1 -1 concentration was calculated usmg the extinction coefficient 13,700 M cm . Control experiments were performed on E. coli FtsZ and our results were consistent with 104 previous estimates of GDP content (Chen and Erickson 2008; Romberg and Mitchison 2004). GTPase activity was initially measured with the malachite green assay (Redick et al. 2005). However, because the activity of Ftle and FtsZZ requires high GTP concentrations, which are incompatible with the malachite green assay, most GTPase assays were performed with a GTP regeneration assay (lngerman and Nunnari 2005). E. coli FtsZ was routinely used as a positive control (Lu and Erickson 1998; Lu et al. 1998). E. coli FtsZ was assayed under conditions optimal for plant FtsZ (lower temperature). Optimal FtsZ1 and FtsZZ GTPase activity was found between 25 and 30°C in 50 mM HEPES-KOH pH 7.0, 100 mM KCl and 5 mM MgSO4 (HMK), similar to optimal conditions for E. coli FtsZ. However, E. coli FtsZ is typically assayed at 37°C (Lu and Erickson 1998; Lu et al. 1998), whereas in this study, E. coli FtsZ is assayed at 30°C. Polymerization assays Electron microscopy of FtsZ polymers was conducted with either 2.5 or 5 uM FtsZ (total Ftle plus FtsZZ). To stabilize FtsZ1, FtsZZ and E. coli FtsZ they were polymerized under non-dynamic conditions in HMKCa (HMK with 5 mM CaClz). Ftle/FtsZZ co-polymers were stable without CaClz conditions and were polymerized in HMK. All buffers were filtered through a 0.22 um filter. Before nucleotide was added, polymerization reactions were centrifuged for 10 min at 14,000 g 105 Table 3.1: GTP binding and hydrolysis properties of Ftle and FtsZZ compared to E. coli FtsZ GTP ATP GTPase GTPase Activity Activity Activity GDP Binding Binding Specific Specific with with with content of Activity‘ Activity, GDPC GTP ATPE polymers and 0 EDTA Ftle 1.05 N/D 0.14 0.028 N/ D N/ D N/ D - 10.03 FtleDZ7SA 1.15 N/D N/D N/D N/D N/D N/D - 10.25 FtsZZ 0.98 N/D 0.14 0.046 N/ D N/ D N/ D - 10.22 FtsZZD322A 0.95 N/D N/D N/D N/D N/D N/D - 10.12 Ftle/FtsZZ - N/D 0.15 - N/D N/D N/D "IO-20% E. coli FtsZ 1.09 1 N/D 2.51 - N/D N/D N/D "40-50% 0.17 Notes A)Maximum activity observed using a coupled GTPase assay at 25°C reported in this column B)Activity determined with a malachite green phosphate assay C)Activity determined with a malachite green phosphate assay D)Determined with both a malachite green phosphate assay and coupled assay 106 at 25°C to remove contaminants. After nucleotide was added, plant FtsZ was polymerized for 20 min at 25°C or 30°C; results were similar at both temperatures. E. coli FtsZ was used as a positive control and results in HMKCa were consistent with previous reports (Erickson et al. 1996; Mukherjee and Lutkenhaus 1994, 1998). 2 ul of a polymerization reaction was pipetted onto a formvar coated ZOO-mesh nickel grid and negative stained with 0.5% or 2% uranyl acetate. FtsZ polymers were examined with a JEOL100 CXII (Japan Electron Optics Laboratories) transmission electron microscope at an accelerating voltage of 100 W at 10,000-450,000X magnification as indicated. 90° light-scattering assays were conducted as previously described for bacterial FtsZ (Mukherjee and Lutkenhaus 1999) using a Photon Technologies Incorporated fluorescence spectrophotometer equipped with a model 814 photomultiplier operated in digital mode at 1000V. Assays were performed with a 0.5 nm excitation and 1 nm emission slit widths and 350 nm excitation and emission wavelengths. Polymerization was conducted in HMK buffer at room temperature unless otherwise specified; all buffers were filtered through a 0.22 pm filter prior to use and all protein preparations were centrifuged for 10 min at 14,000 g at 25°C immediately before assay. Polymerization rates were derived by linear regression fit of the linear increase in light- scattering. Results Production and refolding of recombinant Arabidopsis FtsZI -1 and F tsZZ-1 107 In contrast to previous studies (El-Kafafi et al. 2005 ; Gaikwad et al. 2000), the putative chloroplast transit peptides of Ftle and FtsZZ were removed from recombinant Ftle and FtsZ2 (Figure 3.1). ChloroP (Emanuelsson et al. 1999) predicts a 48 amino acid transit peptide for AtFtsZ2-1. AtFtsZZ-l was aligned with several cyanobacterial FtsZ proteins and the predicted 48 amino acid transit peptide extended beyond the alignments (not shown), suggesting the ChloroP transit peptide prediction for AtFtsZZ-l is correct. Moreover, removal of the AtFtsZZ-l putative transit peptide results in a predicted mass of “'46 kDa, which is close to the “‘45 kDa mass of AtFtsZZ-l determined by immunoblotting of Arabidopsis proteins (McAndrew et al. 2008; Stokes et al. 2000). Arabidopsis has two FtsZZ genes, AtFtsZZ-1 and AtFtsZZ-Z (McAndrew et al. 2008; Osteryoung et al. 1998). The amino acid sequences of these two proteins are 92% similar and 85% identical without their ChloroP predicted transit sequences. In Arabidopsis, AtFtsZZ-l and AtFtsZZ-Z are genetically redundant (Aaron Schmitz et al., in preparation) but AtFtsZZ-l is more abundant than AtFtsZZ-2 (McAnd rew et al. 2008) so AtFtsZZ-l was chosen as a biochemical representative of the FtsZZ family from Ara bidopsis. ChloroP predicts a 90 amino acid transit peptide for AtFtle-l. Removal of 90 amino acids from the N-terminus of AtFtsZ1-1 would remove half of the glycine-rich, GTP binding motif (Figure 3.1) (de Boer et al. 1992a; Erickson 1995; Mukherjee et al. 1993; Mukherjee and Lutkenhaus 1994; RayChaudhuri and Park 1992). Moreover, Ftle without 90 N-terminal amino acids has a predicted mass of "36 kDa, significantly smaller than "40 kDa determined for AtFtle-l by immunoblotting Arabidopsis proteins (Stokes 108 et al. 2000). A better approximation of the AtFtle-l chloroplast transit peptide length was made with the following information: the native AtFtle-l mass is “'40 kDa, AtFtle-l alignments with other Ftle proteins (Stokes and Osteryoung 2003) and chloroplast transit peptides are typically cleaved near charged/arginine residues (Archer and Keegstra 1993; Bruce 2000). These factors together resulted in a new AtFtle-l transit peptide estimate of 57 amino acids. Finally, a AtFtsZ1-1 57 amino acid transit peptide does not remove important structural elements (Nogales et al. 1998a) from a homology based structural model for AtFtle-l (Yoder et al. 2007). Plant FtsZ proteins are cytotoxic when recombinantly expressed in E. coli, resulting in long filamented cells (not shown) and low protein expression. Co—expression of the E. coliftsQAZ operon during plant FtsZ expression suppressed cell filamentation and resulted in significantly increased expression of plant FtsZ (Jeong and Lee 2003). Recombinant FtsZ proteins were found predominantly in inclusion bodies (>80%). The small amounts of soluble recombinant Ftle or FtsZZ were difficult to efficiently separate from contaminating proteins, so plant FtsZ proteins were purified from inclusion bodies. Both FtsZ1 and FtsZZ were purified by Ni-Sepharose (GE Healthcare) affinity chromatography in 6 M urea (denaturing conditions). Both Ftle and FtsZZ typically elute from Ni—Sepharose between "100-300 mM imidazole (not shown). Bacterial FtsZ refolds to an active form when removed from denaturing conditions (Andreu et al. 2002; Santra and Panda 2003) and a similar approach was used to refold plant Ftle and FtsZZ. In some preparations, FtsZZ—1 was contaminated with a small amount of protein initiated from a downstream ATG codon. The truncated, 109 contaminating FtsZZ-1 protein was removed by reverse phase chromatography prior to refolding. Truncated FtsZZ-1 eluted first from a reverse phase column in "25% acetonitrile, while FtsZZ-1 eluted later in m40-45% acetonitrile (not shown). FtsZ1 and FtsZZ purity was assessed by QAAA and was typically >95% pure after refolding. These preparations did not have significant secondary bands when examined by SDS-PAGE and silver-staining (not shown). FtsZZ prepared by reverse phase chromatography was typically >98% pure. Both FtsZI and FtsZZ bind and hydrolyze GTP Plant Ftle and FtsZZ contain FtsZ-consensus, N-terminal GTP-binding motifs (Osteryoung and McAndrew 2001). All FtsZ proteins, including plant Ftle and FtsZZ, contain the GTP hydrolysis motif GxxNxDxxD/E in their T7-loops (Mukherjee et al. 1993; Nogales et al. 1998a; RayChaudhuri and Park 1992; Scheffers et al. 2002) (Figure 3.1). To test for GTP binding, Ftle and FtsZZ were mixed with fresh GTP and unbound nucleotide was removed by gel filtration (Red ick et al. 2005). FtsZ1, FtsZZ and E. coli FtsZ bound 1.05 10.03, 0.98 10.22 and 1.09 i 0.17 mol of GTP per FtsZ respectively (Table 3.1). To test for nucleotide binding specificity, the experiment was repeated with ATP. Ftle, FtsZZ or E. coli FtsZ did not bind detectable levels of ATP (Table 3.1) consistent with previous findings (de Boer et al. 1992a; RayChaudhuri and Park 1992). Together, these data confirm that Ftle and FtsZZ are properly folded and both are capable of specifically binding GTP like other FtsZ proteins. 110 Ftle and FtsZZ were tested for GTP hydrolysis with two assays commonly used with E. coli FtsZ. First, the malachite green assay for inorganic phosphate (P,-) was used to measure free and unbound Pi, but this assay cannot be used with high GTP concentrations (Redick et al. 2005). Second, GTP hydrolysis was measured with a regenerative GTPase assay that measures GDP release (lngerman and Nunnari 2005) presumably from subunit dissociation (Chen and Erickson 2008; Small and Addinall 2003). At 0.5 mM GTP, Ftle and FtsZ2 have low specific activities of 0.028 10.002 and - -1 0.046 10.002 GTP min 1 FtsZ (Table 3.1). The GTPase activities of Ftle and FtsZZ were confirmed in a coupled GTPase assay (lngerman and Nunnari 2005). At higher concentrations of GTP, both Ftle and FtsZZ have specific GTPase activities of 0.14 GTP min.1 FtsZ.1 (Figure 3.2A, Table 3.1). As a control for GTPase activity, T7-loop mutant proteins were created, FtleD275A and FtsZZDBZZA. These proteins have mutated the T7 loops changing GxxNVDFAD to GxxNVDFAA. This mutation in E. coli FtsZ abolishes GTP hydrolysis, but GTP binding is unaffected (Redick et al. 2005; Scheffers et al. 2002). FtleD275A and FtsZZD322A mutant proteins also were able to bind 1.15 10.25 and 0.95 $0.12 molecules of GTP per FtsZ, but did not hydrolyze GTP above background levels. In additional control experiments, hydrolysis was not detected above background when Ftle, FtsZZ and E. coli FtsZ were assayed with GDP, ATP or with GTP and 1 mM EDTA (Table 3.1). The GTP hydrolysis kinetics of Ftle and FtsZZ differ from each other and from E. coli FtsZ. GTP hydrolysis for Ftle and FtsZZ was examined at 2.5 uM total protein, 111 which is within a linear correlation between activity and protein concentration (not shown) and this FtsZ concentration is similar to the predicted FtsZ concentration in chloroplasts (McAnd rew et al. 2008). The maximum GTPase activity of Ftle (0.14 GTP min-1 FtsZ-1) was found between "1-3 mM GTP (Figure 3.2A, I) while the maximum GTPase activity of FtsZZ (0.14 GTP min-1 FtsZ—1) was between "'3-4 mM GTP (Figure 3.2A, A). In contrast, the maximum specific GTPase activity of equally mixed Ftle and FtsZZ was slightly increased to 0.15 GTP min'1 FtsZ-1 and found near "2 mM GTP (Figure 3.2A, O). GTP hydrolysis kinetics were distinctly different for Ftle and FtsZZ (Figure 3.2A). In control experiments, E. coli FtsZ GTP hydrolysis was faster as the concentration of GTP increased (Figure 3.23) from ~2 GTP min‘1 FtsZ-1 to ~25 GTP min‘1 FtsZ-1. At >4 mM GTP, E. coli FtsZ GT P hydrolysis was abolished (Figure 3.2B). E. coli FtsZ did not demonstrate the GTP concentration-dependent increase in activity (Figure 3.23) observed for FtsZ1 and FtsZZ (Figure 3.2A). These results demonstrate that plant have slower rates of GTP hydrolysis compared to E. coli FtsZ. Moreover, the GTPase activity of plant FtsZ is stimulated at high GTP concentrations, which could result from different GTP binding or catalytic properties compared to E. coli FtsZ. Do Ftle and FtsZZ assemble? Ftle/FtsZZ co-assembly EM and light-scattering 112 Figure 3.2: GT Pase kinetics of Ftle, FtsZZ, equally mixed Ftle/FtsZZ and E. coli FtsZ (A) Kinetic GTPase activity of 5 LLM Ftle (I), FtsZZ (A) and 5 uM of equally mixed Ftle and FtsZZ (9) were examined at various GTP concentrations. (b) 5 uM E. coli FtsZ was assayed in parallel as a control. The line connecting the points in both panels is interpolated and does not reflect a curve fit. 113 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 GTP min‘1 FtsZ"1 U) I” N U1 I.i 1 HH I" U'I GTP min'1 FtsZ'1 H .0 U'I O Both Ftle and FtsZZ bind and hydrolyze GTP (Table 3.1), which predicts they assemble like bacterial FtsZ. 1 mM GTP was added to 5 uM of equally mixed Ftle and FtsZZ in HMK, and after a "100 sec lag light-scattering rapidly increased (Figure 3.3B, Cl). The lag before light-scattering rapidly increased is suggestive of cooperative assembly (Caplan and Erickson 2003; Chen et al. 2005; Gonzalez et al. 2005; Huecas and Andreu 2003, 2004; Romberg et al. 2001). As a control, the experiment was repeated with 1 mM GDP instead of GTP and "90% less light-scattering was observed (Figure 3.3B, I). 5 uM of equally mixed Ftle and FtsZZ were polymerized under identical conditions and polymer topology was examined by EM. Ftle and FtsZZ co-polymerized with 1 mM GTP assembled into bundled ribbon-like structures reminiscent of tubulin (Figure 3.4D and E). Ftle/FtsZZ co-polymers are "15-20 filaments in width (Figure 3.4E), or "150 nm wide. Ftle/FtsZZ co-polymers typically range from 1-20 um in length (Figure 3.4D). Ftle/FtsZZ ribbons are significantly longer than those observed with bacterial FtsZ even in the presence of stabilizing CaClz. In contrast, E. coli FtsZ was polymerized in the same conditions as Fts21 and FtsZZ and was found to assemble into single protofilament polymers (not shown) consistent with previous findings (Chen et al. 2005; Huecas et al. 2007a; Romberg et al. 2001). In the absence of nucleotide, or in the presence of 1 mM GDP, assembled filaments were not observed for equally mixed Ftle and FtsZZ or E. coli FtsZ (not shown). 115 Figure 3.3: 90° Light-scattering polymerization assays of Ftle and FtsZZ. Symbols on the traces in the figures were added to distinguish between . 3 experiments. In panels A and B, the vertical axis light scattering data are x10 . All experiments were performed in HMK unless noted otherwise. (A) 5 uM of equally mixed Ftle with FtsZZ was monitored for light-scattering polymerization in the presence of various concentrations of GTP. In the presence of GDP (0), polymerization was observed but was lower than observed when GTP was added. At 0.5 mM GTP (El), Ftle/FtsZZ polymerization steadily increased and saturated as the time approached 2h (not shown). At 1 mM (0), 2 mM (A) and 5 mM (I) GTP, polymerization saturated more quickly. (8) Ftle with FtsZZ polymerizes most efficiently when both are present compared to similar concentrations of Ftle or FtsZZ alone. 2.5 uM Ftle with 2.5 uM FtsZZ was polymerized with 1 mM GTP (Cl), 5 uM Ftle was polymerized with 1 mM GTP (O), 5 uM FtsZZ was polymerized with 1 mM GTP (A) and as a control, 2.5 uM FtsZ1 with 2.5 uM FtsZZ was mixed with 1 mM GDP (I). (C) The rate of polymerization at various GTP concentrations was determined from panel A. The rate of polymerization was maximal with 0.5 mM GTP (when GTP hydrolysis is slow, Figure 3.2) and as the GTP concentration increased (when GTP hydrolysis is faster, Figure 3.2), the rate of polymerization steadily decreased. (D) the rate of polymerization was calculated for the rapid linear increase in polymerization observed for the traces in panel C. Polymerization is maximal only when both Ftle and FtsZZ are present with GTP. ALS = change in light scattering. 116 1500 1500 I» A; g? 1000 1000 Time (sec) D EEE< Time (sec) 500 500 new 5:: H Eb 5...: m 95 ES N Eb 5...: H 0 Eb _>_E md 0 Eb EE 0 100 ~ 80 60 - 40 - 20 - 0 _ B C 120 c=:\m4< 117 Figure 3.4: Negative stain electron micrographs of Ftle and FtsZZ (A) Ftle in HMKCa with 1 mM GTP. Single arrow indicates single protofilament example and double arrow indicates double protofilament example. Bar=200 nm. (8) FtsZZ in HMKCa with 1 mM GTP. Bar=200 nm. (C) FtsZZ in HMKCa with 1 mM GDP as an example of unassembled plant FtsZs. Bar=200 nm. (D) Wide-field view of Ftle/FtsZZ in equal amounts with 0.5 mM GTP in HMK. Bar: 200 nm. (E) Close-up view of Ftle/FtsZZ from paned D. Bar=1 um. (F) Ftle/FtsZZ in equal amounts in 2 mM GTP in HMK. Bar=1 um. (G) 2.5 uM each of Ftle with FtsZZD322A in HMK with 1 mM GTP. Bar=200 nm. (H) FtsZZD322A in HMKCa with 1 mM GTP. Bar=200 nm. (I) FtleD275A in HMKCa with 1 mM GTP. Bar=200 nm. (J) Ftle/FtsZZ mixed at 1:5 with 1 mM GTP in HMK. Bar=1 um. (K) Ftle/FtsZZ mixed at 5:1 with 1 mM GTP in HMK. Bar: 1 pm. (L) Polymers in K stabilized in HMKCa. Bar=1 um. (M) 2:1, Ftle/FtsZZ in HMKCa with 1 mM GTP. Bar=1 um. (N) 1:2, Ftle/FtsZZ in HMKCa with 1 mM GTP. Bar=1 um. 118 Ftle and FtsZZ homofilaments are unstable 2.5 (1M each of mixed Ftle/FtsZZ in HMK co-polymerize into ribbon-like structures (Figure 3.4D and E, Figure 3.3B, Cl), but do Ftle and FtsZZ polymerize individually? 1 mM GTP was added to either Ftle or FtsZZ individually in HMK and light-scattering was monitored. 5 uM Ftle (Figure 3.38, O) or FtsZZ (Figure 3.38, A) polymerized individually in HMK with 1 mM GTP exhibits approximately 70% less total light-scattering than 2.5 uM each of FtsZ1/FtsZZ in the same conditions (Figure 3.38, [1). Additionally, the rates of light-scattering increase were compared for each of the light- scattering experiments in Figure 3.3B (Figure 3.3D). Ftle/FtsZZ with 1 mM GTP has a greater rate of polymerization than either Ftle or FtZZ individually with 1 mM GTP. Moreover, Ftle/FtsZZ has a faster rate of light scattering in the presence of GDP than do Ftle and FtsZZ individually with GTP. Thus, Ftle/FtsZZ co-polymerizes in greater quantity and faster than either Ftle or FtsZZ alone. In summary, Ftle and FtsZZ are both required for maximum, cooperative polymerization. FtsZ1 and FtsZZ were polymerized individually in HMK with 1 mM GTP and examined by EM. Consistent with light-scattering experiments, very few individual, 10 nm thick filaments were observed (not shown, but identical to those shown in 3.4A and B). To stabilize individual Ftle and FtsZZ filaments, 5 mM CaClz was added to HMK (HMKCa). Ftle and FtsZZ were polymerized individually with 1 mM GTP and monitored by light-scattering. The rate of FtsZ2 light-scattering did not increase in the presence of CaClz (Figure 3.3D). Very few FtsZZ filaments were observed by EM in HMKCa (Figure 120 3.48). FtsZZ filaments in HMKCa were typically "1 um long and "10 nm in width; filaments greater than 1 pm in length were not observed. In contrast, Ftle light- scattering increased in HMKCa compared to HMK (Figure 3.3D). Ftle polymer abundance correlated to light-scattering when examined by EM. Ftle polymerized into long thin filaments or pairs offilaments in both HMKCa and HMK, but the abundance of filaments was greater in HMKCa (Figure 3.4A, single filament indicated with one arrow and a double filament indicated with two arrows). FtsZ1 filaments were significantly longer in HMKCa and were typically "1—2 um long and "'10 nm wide consistent with single filaments. Ftle sometimes polymerized into a loose polymer network (not shown) similar to Ftle—GFP over-expression in Physcomitrella patens (Reski 2002). In the absence of nucleotide, or in the presence of GDP, both Ftle and FtsZZ appear monomeric (Figure 3.4C, FtsZZ with 1 mM GDP as an example), similar to results with E. coli FtsZ (Mukherjee and Lutkenhaus 1994). Ftle and FtsZZ have low GTPase activity individually, which should correlate to more stable polymers. However, Ftle and FtsZZ do not polymerize by themselves unless CaClz is present. In contrast, when Ftle and FtsZZ are mixed and polymerized with GTP, ribbon-like structures are observed (Figure 3.4D and E). Together, these results further support the hypothesis of Ftle and FtsZZ co-polymerization. FtsZI/FtsZZ co-polymer length correlates to GT Pase activity At 0.5 mM GTP, the GTPase activity of Ftle/FtsZZ is slow, but at > 1 mM the GTPase activity is faster (Figure 3.2A). This result predicts that maximum assembly 121 should occur at low GTP concentrations, but less polymerization should occur at higher concentrations of GTP. This question was addressed by using Iight-scatting and EM to examine Ftle and FtZ2 co-polymerized at various concentrations of GTP. 2.5 uM of Ftszl and FtsZZ in HMK were polymerized with increasing concentrations of GTP. Ftle/FtsZZ have the highest light-scattering signal in the presence of 0.5 mM GTP, (Figure 3.3A, El) and the fastest rate of increase in signal (Figure 3.3C). In addition, Ftle/FtsZZ light-scattering plateaus between "1.5-2h (not shown). However, as the concentration of GTP is increased above 1 mM, the lag of FtsZ1/FtsZZ polymerization increases (Figure 3.3A; 1 mM: 0; 2 mM, A and 5 mM, I) and the rate of polymerization decreases (Figure 3.3C) but the amount of light-scattering plateaus at the same point between 1-5 mM GTP. Significantly, 1-5 mM correlates to the maximum Ftle/FtsZZ GTPase activity (Figure 3.2A). Controls were incubated with 1 mM GDP (Figure 3.3A, 9) or no nucleotide and had low levels of light-scattering (not shown). Does increased GTP concentration correlate to shorter Ftle/FtsZZ ribbons, or less bundling between individual filaments? Light-scattering experiments demonstrate that the amount of Ftle/FtsZZ co-polymerization plateaus at high GTP concentrations (Figure 3.3A). Also, at these concentrations of GTP, Ftle/FtsZZ hydrolyze GTP faster. This predicts that Ftle/FtsZZ ribbons would be shorter, which was tested by examining Ftle/FtsZZ co-polymers by EM in 2 and 5 mM GTP. The width of Ftle/FtsZZ co- polymers was "150 nm and the same as the ribbons polymerized in 0.5 mM GTP (Figure 3.4E). However, Ftle/FtsZZ polymerized in 2 or 5 mM GTP were shorter, never longer than 1 pm, than those polymerized at 0.5 mM (compare Figure 3.4F to Figure 3.4D). 122 Because the width is unaffected we hypothesize that these polymers may be treadmilling (Cleveland 1982; Larsen et al. 2007; Margolin 2007; Margolis and Wilson 1998). The critical concentration for polymerization When mixed equally, Ftle and FtsZZ display a kinetic lag in polymerization that lengthens as the concentration of GTP increases. This suggests Ftle/FtsZZ co-assemble cooperatively. Bacterial FtsZ cooperatively assembles with a critical concentration of “'1 (1M (Anderson et al. 2004; Redick et al. 2005; Stricker and Erickson 2003). The critical concentration for Fts21/FtsZ2 mixtures was determined by light-scattering (Mukherjee and Lutkenhaus 1999). The concentration of Ftle/FtsZZ was varied between 0.125 and 2.5 uM (total Ftle/FtsZ) and plotted versus the rate of polymerization. Polymerization was not observed when the concentration of FtsZ1/FtsZZ was below 0.5 uM in HMK. To verify that assembled protein could be detected at this FtsZ concentration, the experiment was repeated in HMKCa with 1 mM GTP. In HMKCa, 0.5 1.1M Ftle/FtsZZ light-scattering increased (not shown), demonstrating the lack of light-scattering in HMK correlates to a lack of significant assembly. A linear regression was used to fit data points above 0.5 LLM Ftle/FtsZZ (Figure 3.5). The x-intercept of the regression predicts a critical concentration of "0.7 1.1M for co-assembly of Ftle/FtsZZ. The plant Ftle/FtsZZ critical concentration is consistent with calculations of "0.5-1 UM for E. coli FtsZ (Anderson et al. 2004; Redick et al. 2005; Stricker and Erickson 2003). 123 FtsZI/FtsZZ co-polymers are larger and more stable than E. coli FtsZ polymers Polymerization of equally mixed Ftle and FtsZZ (Figure 3.6, top trace) was compared with that of bacterial FtsZ by light-scattering assays in HMK. 5 uM E. coli FtsZ was assayed for light-scattering with 1 mM GTP (Figure 3.6, bottom trace, arrow indicates GTP addition). The signal increased when GTP was added and then returned to baseline within "10 minutes. Moreover, E. coli FtsZ light-scattering did not increase when GDP or no nucleotide was added (not shown). All E. coli FtsZ experiments were consistent with previous results (Mukherjee and Lutkenhaus 1994). Moreover, when 5 mM GTP was added to 5 uM E. coli FtsZ, the light-scattering signal increased in proportion to the GTP concentration. The light-scattering signal persisted longer than with 1 mM GTP (Figure 3.6, middle trace). In contrast, plant Ftle/FtsZZ mixtures have a higher light-scattering signal (Figure 3.6, top trace) in HMK with 1 mM GTP. When examined by EM under these same conditions, E. coli FtsZ was found to assemble into single filaments as previously reported (Caplan and Erickson 2003; Huecas et al. 2007a). Thus, plant Ftle/FtsZZ polymerizes significantly more, and is more stable than E. coli FtsZ. This result is consistent with the difference in GTP hydrolysis rates between plant Ftle/FtsZZ and E. coli FtsZ. 124 Figure 3.5: Critical concentration for Ftle/FtsZZ polymerization monitored by 90° light-scattering. Ftle and FtsZZ were mixed equally and 1 mM GTP was added to the cuvette. Rates of polymerization were derived from light scattering data and plotted versus the FtsZ concentration. All data points except those corresponding to 0.5 and 0.25 uM FtsZ, where polymerization was essentially not observed, were fit with the linear regression y = 5.0328x - 3.7318 (R2 = 0.9413). These results predict a critical concentration for FtsZ1/FtsZZ polymerization of ”0.7 pM. 1o 8 /§ 26 / E 4 */ < 2 %/ /x 0 " " 1‘ . 1 o 1 2 3 Concentration of equally mixed FtsZ1 + FtsZZ (uM) 125 Figure 3.6: Light-scattering comparison of polymerization of E. coli FtsZ and Ftle with FtsZZ both at 5 uM in HMK buffer. (A) Ftle with FtsZZ (upper curve) polymerizes extensively and does not disassemble over the assay monitoring time of 1 h. In contrast, E. coli FtsZ rapidly polymerizes, but disassembles rapidly, likely correlating to the concentration of GTP since a second polymerization of EthsZ with 5 mM GTP (middle curve) was performed and the extent of polymerization was higher than with 1 mM GTP (lower curve). Arrows indicate GTP addition to E. coli FtsZ. 13 Ftle/FtsZZ 16 1 mM GTP 14 12 Z] 10 <1 8 6 4 2 0 r 0 10 200 300 400 Time (sec) E. coli E. coli FtsZ FtsZ 1 mM 5 mM GTP GTP 126 FtsZI/FtsZZ copolymers are not disassembled by GDP Polymerized E. coli FtsZ exchanges nucleotide within the polymer, unlike ocB- tubulin polymers. As a result, E. coli FtsZ can readily be depolymerized by the addition of GDP (Huecas and Andreu 2004). In contrast, plant Ftle/FtsZZ co-polymers are significantly more stable and have higher signals in light-scattering assays than E. coli FtsZ (Figure 3.6). Additionally, FtsZ1/FtsZZ laterally bundle in the absence of stabilizing agents (Figure 3.4D and E). One possible difference between plant FtsZ and E. coli FtsZ could be nucleotide exchange within assembled polymers. Ftle and FtsZZ were co- polymerized with 0.5 mM GTP for 30 min at 25°C and then light-scattering was monitored for '“5 min. 5 mM GDP was added to the cuvette (Figure 3.7, marked) and the light-scattering signal decreased slightly, but recovered in <100 sec. As a control, light-scattering returned to baseline when 2 mM EDTA was added to Fts21/FtsZZ co- polymers (Figure 3.7, marked). To investigate the GDP content of Ftle/FtsZZ co-polymers, E. coli FtsZ and Ftle/FtsZZ co-polymers were maintained in 3 GT P regeneration system previously used to estimate the GDP content of E. coli FtsZ and TubZ (Chen and Erickson 2008; Romberg and Mitchison 2004). 5 LLM EthsZ was found to contain 2-2.5 uM GDP after incubation for 20 minutes with 0.5 uM GTP, while 5 uM Ftle/FtsZZ contained 0.2-1 uM GDP in the same conditions. These results suggest that the consequence of a low GTPase activity of plant FtsZ is a reduced content of GDP within the polymer and provide evidence of why plant Ftle/FtsZZ co-polymers are significantly more stable in HMK than E. coli FtsZ. 127 Stoichiometric requirements for GTPase activity and polymerization Adding additional FtsZI or FtsZZ to FtsZI/Ftszz polymerization assays decreases the extent of polymerization If Ftle and FtsZZ form a head-to-tail heterofilament, varying the ratio of Ftle to FtsZZ should result in reduced polymerization. This possibility was tested by varying the ratio between Ftle and FtsZZ, keeping the total FtsZ concentration at 5 0M and monitoring light-scattering. In the first series of experiments, excess Ftle was added to FtsZZ. As the concentration of Ftle increased, the light-scattering signal decreased (Figure 3.8A: 1:1, El; 10:1, 0; 3:1, A; 2210). Moreover, the rate of light-scattering decreased proportionally to the increase in Ftle levels (Figure 3.8C). In a series of related experiments, excess FtsZZ was added to Ftle while keeping the total concentration of protein at 5 uM. Adding excess FtsZZ decreased light-scattering (Figure 3.88: 1:1, Cl; 1:10, 0; 1:3, A; 1:2, 0) and decreased the initial rate of light- scattering (Figure 3.8C). Light-scattering predicts that excess FtsZ1 or FtsZZ would result in shorter filaments when examined by EM. Ftle and FtsZZ were mixed at 1:5 or 5:1 in HMK and examined by EM. Ftle mixed 1:5 with FtsZZ, shortened and reduced lateral association of filaments (Figure 3.4J) and the reduction in polymerization correlated to the light-scattering signals. In contrast, Ftle mixed 5:1 with FtsZZ, lateral association was almost entirely abolished, and filament length was slightly shortened, but filament quantity was not reduced (Figure 3.4K). These results demonstrate that FtsZZ may 128 mediate bundling with Ftle, but Ftle does not mediate significant bundling between Ftle and FtsZZ. FtsZZ-1 promotes lateral bundling of FtsZI -1/FtsZZ-1 ribbons Ftle/FtsZZ were co-polymerized at varied ratios and examined by EM in the presence of CaCIz to stabilize the polymers. When the ratio of Ftle to FtsZZ is 2:1 (and as high as 5:1, not shown) and polymerized in HMKCa with 1 mM GTP, thin filaments are typically observed (Figure 3.4L) similar to polymerizing Ftle only (Figure 3.4A), but filament pairs were observed more often. In contrast, when the ratio between FtZ1 and FtsZZ is 1:2, Ftle/FtsZZ ribbons increasingly have individual filaments dissociating from the larger cable itself, sometimes crossing from ribbon-to-ribbon (Figure 3.4M). Ftle and FtsZZ were also mixed at a 1:5 ratio. When examined by EM, these polymers adopt a fan-like structure, and bundling is loosened between cable filaments (Figure 3.4N). Moreover, many short ribbon-fragments with fewer bundled filaments are laterally dissociated from the cable. In summary, by stabilizing FtsZ1/FtsZZ polymerized at different ratios two different polymer morphologies were observed. Excess Ftle results in thinner filaments, and sometimes filament pairs. In contrast, excess FtsZZ results in cables with disrupted lateral interaction. In addition, the cable stays somewhat intact, but the lateral disruption is proportional to the increase in ratio. We interpret this result to suggest lateral interactions are predominantly between FtsZZ and FtsZ1. 129 Figure 3.7: Equally mixed Ftle and FtsZZ are not depolymerized by GDP. 5 uM total Ftle and FtsZZ were equally mixed in HMK and polymerized in the presence of 2 mM GTP for 30 min at 25C. Ftle and FtsZZ were co-assembled for 30 min in the presence of 0.5 mM GTP and then placed into a spectrofluorimeter to measure light-scattering. After "' 6 min of monitoring light-scattering in the presence of GTP, 5 mM GDP (in 25 mM HEPES-KOH pH 7.0) was added to the cuvette and monitored for an additional "6 min. To verify that FtsZ1 and FtsZZ co-polymers were able to be disassembled, 2 mM EDTA was then added to the cuvette and light-scattering was monitored for "45 min longer at 25°C. In the presence of 2 mM EDTA, Ftle/FtsZZ co- polymers rapidly disassembled, whereas GDP only decreased the light scattering slightly and then rapidly recovered 80 70 +5 mM GDP +2 mM EDTA 60 l l 50 W .0 k1...“ 3 < 30 20 10 0 I l l T l 0 200 400 600 800 1000 Time (sec) 130 Figure 3.8: The effect of varying the ratio between Ftle and FtsZZ, while keeping the total FtsZ concentration at 5 uM monitored by light-scattering. (A) Adding excess Ftle to the polymerization reaction decreases the total polymerization. The total concentration of FtsZ1 and FtsZZ was maintained at 5 uM while varying the ratio from 1:1(El), 10:1 (0), 3:1 (A), 2:1(0). (8) Excess FtsZZ also decreases the amount of polymerization proportional to the amount of excess FtsZZ added to the cuvette, although the decrease in polymerization is greater than the effect of adding FtsZ1. The total concentration of Ftle and FtsZZ was maintained at 5 uM while varying the ratio from 1:1(l:l), 1:10 (0), 1:3 (A), 1:2 (0). (C) In addition to decreasing the total polymerization of Ftle and FtsZZ, the rate of assembly is decreased by adding excess Ftle or FtsZZ. Rates of polymerization were measured for the light-scattering traces in A and B and compared to each other. Again, adding more FtsZZ decreases the rate of polymerization more than similar experiments adding excess FtsZ1 suggesting a possible capping role for FtsZZ. 131 0 500 1000 1500 0 500 1000 1500 Time(sec) 25 20- : 15— E 10- \ 3 5- <1 0‘ (9° ”.92 a? a? 31.39.? .99 '9- 's; '3‘ '9 y .9/ HeHe~e ..a/ $0.6” 5“?" o" n? '9- r» x "I 132 Figure 3.9: Inhibition of Fts21 polymerization by FtsZZD322A. To distinguish between lateral and longitudinal inhibition of polymerization, a T7 loop capping mutation, FtsZZDBZZA, was tested for competition to FtsZ1 polymerization. (A) Adding FtsZZD322A steadily decreases the extent of Ftle polymerization. 5 uM Ftle/FtsZZ in HMK with 1 mM GTP (D), 2.5 uM FtsZi in HMKCa with 1.25 uM FtsZZD322A (O), 2.5 uM Ftle in HMKCa with 2.5 uM FtsZZD322A (A), the arrow indicates when FtsZZD322A was added to the cuvette for this trace only to demonstrate specific dynamic inhibition. 5 uM FtsZ1 in HMKCa with 1 mM GTP (O). (B) Moreover, the rate of polymerization of 5 uM total Ftle/FtsZ2D322A is 85% less than if 5 uM total Ftle/FtsZZ is polymerized in the presence of 1 mM GTP. Again, the strong competitive effect of FtsZZD322A on Ftle polymerization suggests FtsZZ may cap Ftle/FtsZZ heterofilaments. (C) Linear relationship between the fraction of FtsZZD322A in the reaction and the fraction of the initial rate of polymerization observed in panel A. The data were fit by the equation y = -0.76x + 1.0136, R2 = 0.9962. 133 ALS/min Fraction FtsZZD322A 50 40 30 20 10 0 500 1000 1500 Time (sec) L 5 uM 5 uM 2-5 PM 2.5 uM Ftle+FtsZZ Ftle Ftle Ftle HMK (HMKCa) 1.25 MM 2.5 uM Ft5220322A FtszzogzzA (HMKCa) (HMKCa) 0 0.2 0.4 0.6 0.8 1 1.2 Fraction of Uninhibited Polymerization 134 FtsZZD322A competitively inhibits the polymerization of FtsZI Polymerization of plant FtsZ with excess Ftle or FtsZZ results in shorter polymers. These results suggest Ftle/FtsZZ have a heterofilament polymer topology. Mutations in the T7-loop of Ftle and FtsZZ were used to probe the heterofilament interface. In E. coli, these mutations were previously used to show that the E. coli FtsZ filament interface is similar to that in tubulin (Redick et al. 2005; Scheffers et al. 2002), but lateral incorporation into Z-rings is maintained in vivo suggesting lateral interaction between filaments is not disrupted (Redick et al. 2005). Thus, our FtsZ1D275A and FtsZZSBZZA T7-loop mutants should act as sub-stoichiometric competitive inhibitors (Redick et al. 2005) of GTPase activity and filament formation if FtsZ1 and FtsZZ are heterofilaments. Ftle was polymerized in HMKCa to stabilize the filaments in the presence of increasing concentrations of FtsZZD322A. Ftle polymerization is inhibited at increased concentrations of FtsZZD322A (Figure 3.9A). Furthermore, the reduction in polymerization is proportional to sub-stoichiometric amounts of FtsZZD322A added to the polymerization assay (Figure 3.9C) and plotting the fraction of FtsZZD322A (inhibitor) versus the fraction of the initial rate of polymerization for the uninhibited results in a slope of 0.8. Thus, there is a negative linear correlation between the presence of FtsZZD322A and the initial rate of polymerization. Taken together these data support an Ftle/FtsZZ heterofilament morphology. Moreover, the extent of inhibition of Ftle by FtsZZD322A is strongly inhibited near a 1:1 ratio, suggestive of a strict heterofilament topology. In a parallel line of experiments, the addition of less than a 1:5 ratio of FtleD275 to FtsZZ reduced FtsZZ polymerization in HMKCa below 135 detectable levels suggesting the converse of this experiment is true, but the results are difficult to interpret, because FtsZZ does not polymerize efficiently by itself. FtsZZDBZZA inhibits the GTPase activity of FtsZI, but not FtsZZ In a parallel line of experimentation, the effects of adding increased amounts of FtsZZD322A to Ftle and FtsZZ were examined in coupled GTPase assays (Figure 3.10). FtsZZDBZZA inhibited the GTPase activity of FtsZ1 in 2 mM GTP in HMK as the concentration of FtsZ2D322A increased (Figure 3.10, O). The negative correlation between the fraction of inhibitor and activity is plotted in Figure 3.108 and has a slope of "0.4, suggesting half the active sites are inhibited consistent with co-polymerization. When examined by EM, filaments were significantly shortened (Figure 3.46). Activity was at background levels when the concentration of FtsZ2D322A was greater than 1 uM. In contrast, the GTPase activity of FtsZZ is stimulated in the presence of FtsZZD322A. This may result from lateral allosteric activation of the non-mutated protein. At high concentrations of FtsZZD322A with FtsZZ, visible aggregates were present and were not plotted. Moreover, the high error in the activity estimates for FtsZZ with FtsZZD322A is due to signal drift from accumulation of mesh-like structures (seen when examined by EM in parallel). This is suggestive of laterally associated FtsZZ/FtsZZD322A. Thus the increased GTPase activity of FtsZZ/FtsZZD322A probably results lateral interactions, not head-to-tail protofilament interactions. 136 Figure 3.10: Inhibition of the GT Pase activity of Ftle and FtsZZ by FtsZZDBZZA in 2 mM GT P in a coupled GTPase assay. (A) Ftle (0) and FtsZZ (Cl) were assayed for GTPase activity in the presence of increasing concentrations of FtsZZD322A. Above 1 uM FtsZZD322A, FtsZi activity was at background levels while FtsZZ activity was slightly enhanced. (8) The fraction of FtsZZD322A assayed with Ftle is plotted against the fraction of activity observed. The first three points correlating to detectable levels of activity were fit with the line y = - 0.4041x + 0.3899, R2 = 0.892. The large error observed for FtsZZD322A with FtsZZ is due to the formation of aggregates/mesh-like structures in the presence of high FtsZZD322A concentrations with FtsZZ when examined by EM (Figure 3.4H) 137 GTP min'1 FtsZ'1 Fraction FtsZZD322A 0.5 0_ 4 7' 13$": 1 ; ire“ .l. ‘8er" 0.3 . .x l l 0.2 A... ........ ...Ell 0.1"“: 0 \8“ ~~~~~~~ ;m§.e.«~¢ $ 1 0 1 2 3 11M FtsZZD322A added to 2.5 11M WT fish or FtsZZ 1.2 1 ———e 0.8 ° 0.6 ° 0.4 e 0.2 k 0 \ 1 0 0.5 1 1.5 Fraction Uninhibited Activity 138 Plant FtsZ and T7-loop mutants assemble into rings in E. coli To verify that FtleDZ75A and FtsZZDBZZA mutations do not disrupt protein structure, identical mutations were made in Ftle-GFP and FtsZZ-GFP expression vectors in E. coli. At an 00500 of 1.0, cells harbouring FtsZ-GFP expression plasmids were induced with 0.1 uM IPTG for 2 h. Ftle—GFP, FtsZZ-GFP, FtleDZ75A and FtsZZDBZZA were independantly visualized by fluorescence microscopy. All four proteins assembled into rings equally spaced along filamented E. coli cells (Figure 3.11A and B) reminiscent of the native E. coli FtsZ-GFP (Ma et al. 1996). Because all four proteins localize to identical ring-like structures, the same result as performing a similar experiment with E. coli FtsZ-GFP (Redick et al. 2005), we conclude the T7-loop mutations do not disrupt the plant FtsZ proteins. GFP color variants of E. coli FtsZ are not available , and so plant FtsZ could not be co-localized to native E. coli FtsZ rings. Preliminary results also suggest FtleDZ75A and FtsZZBZZA interact with Ftle and FtsZZ in the yeast two-hybrid assay. Discussion Previous in vitro studies Two previous studies examined the polymerization of Ftle and FtsZZ with their transit peptides (FtsZZ was also C-terminally truncated removing 80 C-terminal amino acids) (El-Kafafi et al. 2005; Gaikwad et al. 2000). In contrast, we constructed AtFtle-l and AtFtsZZ-l expression constructs lacking their putative chloroplast transit peptides to be more relevant to physiological forms of Ftle and FtsZZ. For this study, we chose to 139 use C-terminally His-tagged proteins. In parallel, we examined N-terminally His-tagged proteins that are thrombin cleavable. Although protein recovery rates from this approach were less, we found that the GTPase activity and assembly of untagged plant FtsZ proteins to be indistinguishable from those of C-terminally His-tagged plant FtsZ (not shown). Moreover, C-terminal His-tags have previously been reported not to affect the polymerization or GTPase activities of various bacterial FtsZs (Huecas et al. 2007a ; Oliva et al. 2003). It was previously suggested that full-length Ftle and FtsZZ do not efficiently polymerize without glutaraldehyde cross-linking (El-Kafafi et al. ; Gaikwad et al. 2000) and our results are consistent with this finding. Here we report that both Ftle and FtsZZ are required for efficient polymerization (Figure 3.3). Ftle and FtsZZ possess all residues known to be important for GTP binding and hydrolysis (Osteryoung and McAndrew 2001) (Figure 3.1), and GTP hydrolysis and GTP binding assays confirmed that both are GTPases (Figure 3.2, Table 3.1). However, plant Ftle and FtsZZ have significantly slower rates of GTP hydrolysis than E. coli FtsZ (Figure 3.2) and Ftle/FtsZZ polymers are far more stable than E. coli FtsZ protofilaments. This is not unprecedented; M. tuberculosis FtsZ also has low GTPase activity and slow polymer turnover kinetics compared to those of E. coli FtsZ (Anand et al. 2004; Borhani and White 2004; Chen et al. 2007; White et al. 2000). Consistent with the stability of the Ftle/FtsZZ filaments in vitro, we have observed that Z-rings are visible by immunofluorescence at the mid-plastid in most chloroplasts even in cells in which chloroplasts are not actively dividing (Vitha et al. 2001). Thus, a consequence of a lower GTPase activity may be stabilized Z-rings in vivo. 140 Figure 3.11: Fluorescence microscopy of plant FtsZs fused to GFP expressed in E. coli Ftle-GFP, FtsZZ—GFP, FtleDZ75A-GFP and FtsZZDBZZA-GFP are localized to regularly spaced ring-like structure in E. coli. This is similar to previous results with T7- loop mutants of E. coli FtsZ (Redick et al.). Bar=10 um. FtsZ1-GFP Ftle D322A-G F P FtSZZ'GFP _ FtsZZD275A-GFP 141 This contrasts with E. coli Z-rings, which are short-lived during the cell cycle, appearing onlyjust before fission and then disassembling just before the completion of fission (Addinall et al. 1996; Bi and Lutkenhaus 1990a, 1991; den Blaauwen et al. 2003; Harry et al. 1999; Lin et al. 1997; Regamey et al. 2000). The low GTPase activity and stability of Ftle/FtsZZ polymers suggests that plant Ftle and FtsZZ polymer dynamics may be subject to regulation in vivo. Ftle and FtsZZ polymerize into heterofilaments Ftle and FtsZZ have been shown by immunofluorescence to tightly co-localize to rings at mid-plastid (Vitha et al. 2001), but it was not known if Ftle and FtsZZ assemble into homofilaments that laterally associate, or if they assemble into heterofilamnts (Osteryoung and McAndrew 2001). Ftle and FtsZZ assemble inefficiently individually, but when mixed equally they assemble faster and more extensively (Figure 3.3) suggesting that Ftle and FtsZZ are both required for optimum filament formation. In previous work, our lab has shown that manipulation of FtsZ levels disrupts chloroplast division and FtsZ filament morphology in vivo (McAndrew et al. 2001; McAndrew et al. 2008; Stokes et al. 2000; Yoder et al. 2007) and performing similar experiments in vitro led to the same conclusion: changing the ratio between Ftle and FtsZZ interferes with polymerization. Varying the ratio of Ftle and FtsZZ disrupts the Ftle/FtsZZ co-polymerization longitudinally. l n addition, T7-loop mutations were created in Ftle and FtsZZ that disrupt GTP hydrolysis but not binding (Table 3.1)(Redick et al. 2005; Scheffers et al. 142 2002). T7-loop mutations in E. coli FtsZ specifically disrupt the interface between head- to-tail subunits, impairing protofilament elongation (Redick et al. 2005; Scheffers et al. 2002). Moreover, there is genetic evidence that T7-loop mutations behave analogously in Arabidopsis. Because T7-loop mutants inhibit polymerization, we expect similar mutations to be dominant-negative alleles in vivo. As indicated in Appendix A (Yoder et al. 2007), an Ftle mutant, ft5216267R, was found to have a mutated T7-loop, GxxNxDxxD to RxxNxDxD, and to be dominant-negative. In addition, immunofluorescence of AtFtle-l and AtFtsZZ-l showed significantly shortened Z- filaments in homozygous mutants (in young tissue, where the level of the mutant protein is similar to AtFtle-l levels in wild type plants). We conclude that ftszl 6267R exerts a dominant-negative effect by inhibiting polymerization of Ftle/FtsZZ heteropolymer elongation. The T7-loop mutation FtsZZDSZZA acts as a competitive inhibitor of Ftle, but slightly enhances GTP hydrolysis by FtsZZ (Figure 3.10), suggesting FtsZZD322A is competing for the head-to-tail GT Pase active site of Ftle, but does not interfere with the head-to-tail active site formation between FtsZZ monomers. Moreover, FtsZZD322A at sub-stoichiometric levels proportionately reduces the polymerization of Ftle (Figure 3.9). FtsZZ does not efficiently polymerize by itself and FtsZZD322A did not affect FtsZZ polymerization (not shown). But interpretation of these results is difficult because FtsZZ does not efficiently assemble by itself. Taken together these data support a heterofilament model of Ftle/FtsZZ co-polymers (Figure 3.12). Additionally, and while 143 a ratio of 1:1 Ftle/FtsZZ in bundled heterofilaments is the more stable form of FtsZ, there is no absolute specificity for a strict heterofilament. If Ftle and FtsZZ form heterolfiaments, the minimum subunit for assembly would be predicted to be an Ftle/FtsZZ dimer, similar to aB-tubulin (Caplow and Fee 2002). Indeed, we previously have examined the composition and molecular mass of the FtsZ containing complex (Chapter 2)(McAndrew et al. 2008). We found that Ftle, FtsZZ, ARC3 and ARC6 and possibly a protein that cross-reacts with ZipA (Chapter 4) co- fractionate in a salt-stable complex with a mass of "240 kDa. Using the molecular masses of the individual components, this predicts a minimal stoichiometry of the unassembly complex of 1:1:1:1:1, Ftle/FtsZZ/ARCB/ARCS/"ZipA”. This suggests that the minimal assembly unit found in vivo is a dimer of Ftle and FtsZZ (also with the additional FtsZ binding factors). These facts further support our hypothesis that Ftle and FtsZZ form heterofilament polymers. The in vivo stoichiometry of FtsZI and FtsZZ In Chapter 2 (McAndrew et al. 2008) Ftle and FtsZZ are found to be in a ratio of "1:2. Moreover the total Ftle and FtsZZ concentration in the chloroplast stroma is estimated to average about 1.3 11M to 2.6 11M (depnding on the volume occupied by thylakoids) in three-week old Arabidopsis plants. Ftle/FtsZZ co-polymers assemble rapidly with a critical concentration of "0.7 pM. Moreover, maximal assembly of FtsZ1/FtsZZ ribbons was found at 0.5 mM GTP, which is close to the 0.2 mM concentration of GTP found in the chloroplast (Krause and Heber 1976). Additionally, 144 0.2 mM GTP in the chloroplast stroma would correlate to very low Ftle/FtsZZ GTP hydrolysis (Figure 3.2A). Thus, under physiological conditions the assembly of Ftle/FtsZZ co-polymers is strongly favored, except that the ratio between Ftle and FtsZZ is approximately 1:2 (McAndrew et al. 2008). In vitro, this ratio wouldinhibit polymerization of FtsZ1/FtsZZ co-polymers. So how do Ftle and FtsZ2 attain maximum polymerization into cables? One answer may be that ARC6, an FtsZZ-interacting protein, may stabilize FtsZ filaments and Z-rings in vivo (Vitha et al. 2003) by regulating FtsZZ availability or activity. In chloroplasts, ARC3 is essential for proper Z-ring positioning to mid-plastid (Glynn et al. ; Maple and Moller 2007a) and is a chimera of an FtsZ GTP-binding domain and a PlP-S-kinase (Shimada et al. 2004). ARC3 specifically interacts with Ftle. How might ARC3 function in plastids? In the absence of FtsZ1 in vitro, FtsZZ filament formation is disfavored and stabilization factors are required to visualize polymers by EM (Figure 3.48). Thus, ARC3 may be an Ftle GTP-exchange factor that stimulates GTP exchange, and hence stabilizes FtsZ rings at mid-plastid. Alternatively ARC3 may sequester Ftle from FtsZZ, thus promoting Z-ring disassembly at sites other than the mid-plastid. However, the finding that ARC3 is localized to mid-plastid rings strongly favors a role of ARC3 being integral to the Z-ring and acting as a GTP exchange factor. The precise localization of ARC3 in the plastid will need to be further clarified to understand its biochemical effects on Z-rings. 145 Curiously, despite the reduced rate of GTP hydrolysis (13-53 fold less that that of E. coli FtsZ) by Ftle/FtsZZ in vitro, the in vivo dynamics measured by FRAP of Ftle-GFP are about 3-times slower (a preliminary measurement) (Vitha et al. 2005) than the in vivo dynamics of bacterial FtsZ-GFP (Stricker et al. 2002). This suggests that Ftle/FtsZZ co-polymerization may not be regulated by their internal GTP hydrolysis and assembly rates but possibly by modulation of their biochemical properties by accessory factors such as ARC3 and ARC6. Future in vitro work will be essential to defining the biochemical roles of ARC3 and ARC6 and their effect on Ftle/FtsZZ filaments. Heterofilament bundling Ftle/FtsZZ heterofilaments bundle into ribbons when examined by EM. When excess FtsZZ is co-polymerized with Ftle (and stabilized with CaClz), Ftle/FtsZ2 ribbons appear to spread into fan-like structures with many intersecting individual unbundled filaments apparent (Figure 3.4M and N). The results were similar in the absence of CaClz; mixing Ftle and FtsZZ at 1:5 causes Ftle/FtsZZ filaments within ribbons to ”fray” (Figure 3.41), whereas excess Ftle completely disrupts lateral interactions when mixed at 5:1 in HMK (Figure 3.4K). This suggests a lack of availability for FtsZZ to laterally interact with Ftle, since the amount of Ftle is reduced. This result is different than tubulin, where lateral interactions are (H: and B-B (Nogales 2000). However, we interpret lateral stabilization within Ftle/FtsZZ ribbons to be primarily between FtsZZ and Ftle (Figure 3.12). Further site-directed mutagenesis 146 studies will be important to deciphering lateral bundling interactions between Ftle and FtsZZ. FtsZI/FtsZZ are different than tubulin and bacterial FtsZ Plant Ftle and E. coli FtsZ share 41% identity and 61% similarity and FtsZZ and E. coli FtsZ share 37% identity and 61% similarity. This suggests plant and bacterial FtsZ should be biochemically similar. Indeed, plant and bacterial FtsZ both bind and hydrolyze GTP and assemble into protofilaments with similar critical concentrations. However, the biochemical details differ between Ftle/FtsZZ and E. coli FtsZ. The most obvious difference is that plant Ftle/FtsZZ assemble more stably than E. coli FtsZ and form ribbon-like structures in the presence of GTP without stabilizing agents. In contrast, E. coli FtsZ has a high GTPase activity (Figure 3.28, Table 3.2) and E. coli FtsZ filaments disassemble rapidly (Mukherjee and Lutkenhaus 1999) and do not bundle in the absence of stabilizing agents. Ftle/FtsZZ bundling may block nucleotide exchange, perhaps explaining why plant Ftle/FtsZZ polymers cannot be disassembled by GDP. FtsZ1/FtsZZ share some similarities with bacterial FtsZ, but in other ways also resemble tubulin. Tubulin forms strict head-to-tail heterofilaments that laterally associate into tubes (Nogales 1999). Laterally associated Ftle/FtsZZ heterofilament ribbons look similar to tubulin when examined by EM (Figure 3.4D and E). However, because Ftle and FtsZZ are both GTPases and can assemble independently, they probably are not strict heterodimers. However, co-assembled Ftle/FtsZZ is clearly more stable. ocB-tubulin heterofilament formation is a direct result of a mutation in the 147 T7-loop of the B-subunit, disrupting GTP hydrolysis in a-tubulin (Nogales 1999). Thus, the most kinetically stable form of tubulin is a strict heterodimer that subsequently assembles into heterofilaments. Likewise, FtsZ1 and FtsZZ polymerize into a heterofilament (Figure 3.12). In conclusion, plant FtsZ has retained many similarities to bacterial FtsZ, but has evolved to polymerize more stably like eukaryotic tubulin. Likewise, aB-tubulin has evolved an extremely complex regulatory network of accessory proteins, nucleation complexes and depolymerization factors to strictly regulate microtubule assembly spatially and temporally (Nogales 1999). For plant Ftle/FtsZZ heterofilaments the regulation point seems to be at remodeling and disassembly. ARC3 and ARC6 are strong candidates to perform this function. Why do plants require two types of FtsZ? Ta ken together, this study allows speculation as to the evolutionary driving force that resulted in two types of FtsZ in plants. There is precedent for gene expansion of cytoskeletal proteins, notably as FtsZ evolved to tubulin, the array of tubulin types has expanded significantly (Dutcher 2001). Thus, two hypotheses could explain the divergence of two FtsZ types in plants. First, the simplest hypothesis is that there are two types of FtsZ in plants to help overcome dimerization/ nucleation for assembly like OLB-tubulin. In this work, we demonstrate maximum cooperative assembly of Ftle and FtsZZ requires equal amounts of both. A second hypothesis is that two types of FtsZ are required for Z-ring regulation. Indeed, Ftle and FtsZZ differ in their specific binding to 148 ARC3 and ARC6, respectivley. Moreover, ARC3 is a eukaryotically created chimera of an FtsZ, GTP binding domain and a eukaryotic C-terminus. ARC3 binds Ftle and specifies mid-cell positioning of the Z-ring (Glynn et al. ; Maple and Moller 2007a). In contrast, ARC6 specifically binds FtsZZ to stabilize and organize Z-rings at mid-plastid. Finally, despite two possible explanations for the evolution of two FtsZ types, it is most likely that both nucleation and assembled Z-ring control by the eukaryotic host drove the divergence of two FtsZ types. Future experiments clarifying plant FtsZ nucleation and the biochemical roles of ARC3 and ARC6 will be crucial to further understanding why two types of FtsZ are used for chloroplast division. Acknowledgements Special thanks to Dr. Alicia Pastor for training and support of electron microscopy. Dr. Joseph Leykam was an invaluable resource to discuss protein purification and Dr. William Wedemeyer consulted on protein refolding confirmation. Jonathan Glynn consulted in many useful discussions and helpful reading of the manuscript and thanks to Aaron Schmitz for helpful discussions and sharing preliminary findings. This work was supported by fellowships from the MSU Plant Sciences Fellowship program and an N. E. Tolbert memorial fellowship to B..l.S.C. Olson. This work was supported by grant numbers 0092448 and 0544676 from the National Science Foundafion. 149 Figure 3.12: Model of plant FtsZ heterofilament interactions. (A) Lateral association between (xB-tubulin is a—a and B-B. aB-tubulin is capped by B-tubulin bound to GTP. (B) Bacterial FtsZ protofilaments are believed to laterally interact, but the mechanism is unknown. (C) Model of the Ftle/FtsZZ heterofilament. Ftle/FtsZZ lateral interaction is between Ftle and FtsZZ, differing from tubulin. Ftle/FtsZZ filaments are also GTP capped, but the cap could be either Ftle or FtsZZ. A: Tubulin B: Bacterial C: Plant FtsZ Model FtsZ a—Tubulin B-Tubulin »" GTP Cap 150 CHAPTER FOUR THE SEARCH FOR AN ANALOGUE TO THE BACTERIAL CELL DIVISION PROTEIN ZIPA BY COMPUTATIONAL AND EXPERIMENTAL APPROACHES B. J. S. C. Olson performed all experiments except those outlined Figures 4.6 and 4.7 and wrote the chapter. 151 Introduction The division of chloroplasts is mediated by two families of FtsZ proteins, Ftle and FtsZZ, which are homologues of the bacterial cell division protein FtsZ (Osteryoung 2000; Osteryoung and McAndrew 2001; Osteryoung and Pyke 1998; Stokes et al. 2000). Plant and bacterial FtsZ share significant sequence similarity (Osteryoung 2001), supporting the endosymbiotic origin of chloroplasts (Margulis 1971, 1975), and FtsZZ family proteins share a conserved C-terminal motif with bacterial FtsZ (Osteryoung and McAndrew 2001). Despite the conservation of FtsZ in the division of chloroplasts, several bacterial cell division proteins have no obvious homologues in plants, such as ZipA, FtsA and MinC (discussed in chapter 1). A key question is: have new proteins evolved to replace the functions of these “missing” proteins, or have the progenitors of these proteins retained their core structures, but diverged enough in their sequences to be no longer detectable by sequence similarity searches? The work described here attempted to address this question and possibly identify a plant analogue to the bacterial cell division protein ZipA. In bacteria, FtsZ proteins are anchored to the membrane by two proteins, ZipA and FtsA; the later is in the same structural super-family as actin/HSP70 proteins (Bork et al. 1992; van den Ent and Lowe 2000). Both ZipA and FtsA interact with the extreme carboxyl terminus of E. coli FtsZ (Addinall and Lutkenhaus 1996; Hale and de Boer 1999; Hale and de Boer 1997; Haney et al. 2001; Liu et al. 1999; Wang et al. 1997; Yan et al. 2000) and anchor the FtsZ ring to the inside of the cell membrane by an integral 152 membrane domain (Hale and de Boer 1997). In contrast, FtsA is a peripheral membrane protein (Addinall and Lutkenhaus 1996). Both FtsA and ZipA are required for interaction with downstream cell division proteins (Pichoff and Lutkenhaus 2002). The crystal structure of ZipA complexed with the carboxyl terminus of FtsZ has been solved (Mosyak et al. 2000). Based on the quarternary structure of tubulin (Li et al. 2002) the ZipA/FtsZ interacting domain is believed to stick out of assembled FtsZ filaments (Nogales 2000). Despite the conservation of the ZipA/FtsA binding motif in FtsZZ (chapter 1, Figure 1.1)(Osteryoung 2001), neither ZipA nor FtsA have been identified in plants, though both functions together are essential to bacterial cell division. The presence of the ZipA binding domain in FtsZZ suggested that a ZipA-like protein may be involved in chloroplast division. In an attempt to identify a ZipA protein in chloroplasts, two parallel approaches were utilized. First, a partially purified FtsZ-containing complex from pea chloroplasts (McAndrew et al. 2008) was probed by immunoblotting with an antibody to E. coli ZipA. The antibody detected a protein of "35 kDa that co-fractionates with FtsZ through various purification steps. Although the identity of the protein that cross-reacts with the ZipA antibody was not determined, future efforts may elucidate its identity now that the work described below tightly links this protein to the pea FtsZ complex. Second, a novel structural search algorithm was developed and utilized to identify a putative Arabidopsis protein, $521, with structural similarity to ZipA. However functional analysis of this gene did not unequivocally link it to a role in chloroplast division. 153 Results Chloroplasts contain a protein that cross-reacts with a ZipA antibody and co-fractianates with FtsZ during partial purification To find if a ZipA-like protein is involved in chloroplast division, an antibody to E. coli ZipA was obtained (Hale and de Boer 1999) and tested for cross-reactivity with chloroplastic proteins. The ZipA-specific antibody did not recognize a protein in isolated intact Arabidopsis chloroplasts, but recognized a "35 kDa protein in isolated pea chloroplasts (R. S. McAnd rew, unpublished observation). To test if the ZipA cross- reactive protein may have a role in chloroplast division, n-dodecyl-B-D-maltoside (DDM) solubilized pea chloroplasts were separated by sucrose density gradient fractionation as described above (chapter 2)(McAndrew et al. 2008). Sucrose density gradient fractions were probed by immunoblot with antibodies to Ftle, FtsZZ, ARC6 and ZipA. Proteins recognized by these antibodies co-fractionated near the top of the sucrose density gradient between 210-240 kDa (Figure 4.1A). In separate experiments these proteins were also found to co-fractionate with ARC3 (McAndrew et al. 2008). To further investigate interaction between the protein recognized by the ZipA antibody and Ftle and FtsZZ, a series of additional) purification steps were performed after sucrose density gradient fractionation. First, fractions 3-10 from the sucrose 154 Figure 4.1: Ftle, FtsZZ, ARC6 and a protein that cross-reacts with an antibody to E. coli ZipA cofractionate in sucrose density gradients, native PAGE and by hydrophobic interaction chromatography. Isolated pea chloroplasts were lysed in the presence of DDM and soluble proteins were fractionated on a sucrose density gradient. Proteins in the fractions shown were acetone-precipitated and analyzed by SDS-PAGE and immunoblotting using AtFtle-l, AtFtsZZ-l and AtFtsZZ-2 specific antibodies as described (McAndrew et al. 2008). (A) Analysis of fractions from a 5-20% sucrose density gradient. Ftle, FtsZZ, ARC6 and a protein that cross-reacts with an E. coli ZipA antibody co-sedimented in a mass range of "210-240 kDa, indicated with a bar above the immunoblots. (B) The peak FtsZ-containing fractions (indicated by the bar in panel A) were pooled and subjected to native PAGE and analyzed by immunoblot. Ftle, FtsZZ and the protein that cross- reacts with the E. coli ZipA antibody again co-separate at "220 kDa. (C) The peak FtsZ- containing fractions indicated by the bar in panel A were pooled and subjected to hydrophobic interaction chromatography using a 1 M to 0 M NaCI elution gradient. FtsZ1, FtsZZ, and a protein that cross-reacts with the E. coli ZipA antibody co-eluted in the same fractions, between 1 M and 0.8 M NaCI, indicated by the bar above the immunoblots 155 (U D x G <1- ,_ 232 kDa 440 kDa 669 kDa Ftle FtsZ2-1 FtsZZ-2 ARC6 EcZipA Antibody Probe B .— .. <5. Antibody ID], w H Probe E I ”‘3 C [NaCI] 5 5 Column Fraction 1 2 3 4 5 6 7 FtsZ1 FtsZ2 EcZipA 156 density gradient corresponding to "210-240 kDa were pooled (McAndrew et al. 2008)(Chapter 2, Figure 2.7), concentrated by ultrafiltration, separated by native PAGE and analyzed by immunoblot using antibodies to FtsZ1, FtsZZ and ZipA (Figure 4.18). The mass of this complex is approximately 220 kDa, significantly higher than the 35 kDa mass of the protein recognized by the ZipA antibody under denaturing conditions, suggesting they are in a complex together. The stability of the complex containing Ftle, FtsZ2 and the ZipA cross-reactive protein was investigated by performing hydrophobic interaction chromatography on sucrose density gradient fractions containing these proteins (Figure 4.13 indicated with a bar). lmmunoblot analysis of hydrophobic interaction chromatography fractions eluted with a 1 M to 0 M NaCI gradient shows that Ftle, FtsZZ and the ZipA cross-reactive protein co-elute (Figure 4.1C). In order to identify the protein that cross-reacts with the ZipA-specific antibody, a series of proteomic experiments were employed on various fractions containing FtsZ and ZipA. First, bands from the native PAGE gel, at a molecular mass of "220 kDa where Ftle, FtsZ2 and the ZipA cross-reactive protein migrated, were cut out of the gel, digested with trypsin and subjected to tandem mass-spectrometry. Proteins were identified by molecular ion fingerprinting using the predicted mass fragments of proteins in the Arabidopsis genome (even though the protein was isolated from pea). The positive control for the experiment is identification of either Ftle or FtsZZ in the complex; however, neither protein was identified and none of the other identified proteins were strong candidates to be ZipA (not shown). Moreover, most of the strong 157 fingerprints identified were that of the chloroplast metabolic pathways and probably contaminants. A series of additional experiments were attempted by immunoprecipitation of the FtsZ complex from sucrose density gradient fractions (Figure 4.1A, indicated by the bar) and then performing proteomic analysis. However, no strong candidates for a putative ZipA were identified (not shown). The current identity of the ZipA cross- reactive protein is unknown, but the recent 454 sequencing of a pea cDNA library recently created at Michigan State University (Andreas Weber and Andrea Braeutigam, personal communication) could be pursued to identify the ZipA cross-reactive protein in pea. Identification of a structural analogue to the bacterial cell division protein ZipA In bacteria, the FtsZ ring is believed to be organized near the membrane at the division site by the action of two essential division proteins, ZipA and FtsA. FtsZZ family proteins contain a conserved C-terminal domain that in bacteria is responsible for binding to these proteins. This suggests that ZipA and/or FtsA homologues are involved in chloroplast division. Attempts to identify a homologue of bacterial ZipA in plants based on sequence similarity has been unsuccessful due to the general lack of conservation of primary amino acid sequence among known ZipA proteins (Du and Arvidson 2003). 158 Figure 4.2: Overview of the structural threading search algorithm used to identify a putative Arabidopsis chloroplastic protein with structural similarity to ZipA, Sszl. sic-«H. .cf~l9.‘r. - 1' .1. '24.» ram .; ‘ J.- k (4‘ amt-- .511 ;’..“4 63.11.517.19 MM ‘13:. 15%: $.31 C ‘5 4 _{n‘l ~ ' v ' I ‘TJJI'E L {‘1}? L1? ‘4... 'J y “‘2' ”‘W‘V—y' 4K" ~’~. " H", iri'ffih‘tui,‘ _‘vn‘, (g: i q ._1 ‘155'CH‘? ij‘uir'_ ”673591;.‘3‘ mw~ .‘t—‘A—J'“ ’51" 732.15... 2.. J-‘gszvVfi “'r. ' Jail? i." :1 'Ed:‘ .""‘ "‘2; 1 j 1 - . :1 Ii 1" 159 Table 4.1: Candidate proteins structurally similar to E. coli ZipA. List of high-ranking proteins from a database of putative chloroplastic Arabidopsis proteins in the filtered database that had a high likelihood of being the same structure as the C-terminus of ZipA (1f46). Z-score is a measure of likelihood that the structure is correct and a number greater than 6 is significant. Asterisk indicates the gene for 5521. Genes where an associated T-DNA insertion was available and examined are listed. 160 Gene (AGI Z-score Annotation in TAIR Chloroplast Division Number) Phenotype? INitrilase activity, nitrogen compound 'No* AT4GO8790 5.35 metabolic process Chloroplast protein, molecular 'No* Ifunction unknown, biological process AT3G19680 6.45 unknown Chloroplast protein, molecular 'Not Available Ifunction unknown, biological process AT1650040 4.85 unknown Transd ucin family protein / WD-40 'Not Available ATBGS 1930 4.43 repeat family protein DNA repair protein recA; Identical to 'Not Examined DNA repair protein recA homolog 1, ATlG79050 9.65 chloroplast precursor (RECA) IM etallo-beta-Iactamase family Slightly fewer enlarged protein; similar to hypothetical chloroplasts AT4G33540" 12.9 protein OsJ_008663 Transposable element gene; Not available transposase-related, weak similarity AT1637063 757 to Tam3-transposase *Genotype was only confirmed once, both homozygous and heterozygous plants were examined. 161 Since there is a crystal structure for E. coli ZipA (EcZipA, Mosyak et al. 2000; May et al. 2000) I attempted to detect a structural analogue of ZipA in a database of proteins predicted to be targeted to chloroplasts (overview of the approach, Figure 4.2). The amino acid sequence for E. coli ZipA was used to produce a structural profile using the threading algorithm in the PROSPECT 2 software package (Xu and Xu 2000). E. coli ZipA has 3 MW of 36.4 kD so a database of approximately 4,000 predicted chloroplast proteins was filtered for those that had 3 MW range of 35-55 kD. The logic was that a chloroplastic ZipA homologue would have to acquire a chloroplast transit peptide, thus lengthening the protein. Moreover, an E. coli ZipA-specific antibody recognizes a protein of about 35 kDa in pea. A filtered database was used for this analysis because of the large amount of computational time required for the analysis. Proteins in the database were threaded into the crystal structure for E. coli ZipA. The threaded structures were then sorted and analyzed by their z-scores. Seven strong hits were identified as putative candidates (Table 4.1). Analysis of the structural alignments between known ZipA proteins and these candidates revealed that the locus At4g33540 (called Sszl, structurally similar to gipA, $521) was a strong candidate to contain the C- terminal FtsZ binding domain of ZipA. Examination of the threaded Sszl structures showed that structural contacts between E. coli ZipA and E. coli FtsZ would be conserved if 5521 were to bind AtFtsZZ, thus further supporting this locus as a candidate ZipA analogue (Figure 4.3, Table 4.2). Of note is the fact that only the carboxyl terminal region of the $521 is structurally similar to ZipA and that $521 lacks an amino terminal transmembrane domain believed to be involved in anchoring ZipA to the membrane. 162 $521 on the other hand has weak similarity to B-Iactamases and glyoxylatses. It is unclear if 5521 has B-Iactamase or glyoxylase activity. However, B-lactamases are found in the cell wall of bacteria and there is some evidence that chloroplasts may have bacterial cell wall material (murein) at the division site (Katayama et al. 2003; Machida et al. 2006). Phenotype of a T-DNA insertional mutant for 5521 The SALK T-DNA insertional line 039451 contains a T-DNA insertion annotated as being in exon 4 of S521 (Figure 4.4). This insertion is expected to abolish expression or produce a truncated Sszl protein. Seeds were obtained from the ABRC at Ohio State University, grown in soil and analyzed for their genotype and phenotype. Out of 37 plants initially grown, line number 17 was found to be homozygous for the T-DNA insertion based on PCR analysis of the At4g33540 locus (Figure 4.43). Samples from the leaves of these plants were examined by microscopy (Figure 4.5A). Compared to three- week-old wild-type plants there are slightly fewer, enlarged chloroplasts in three-week- old SALK_039451 plants (Figure 4.53), suggesting that $521 may be involved in chloroplast division. Furthermore, analysis of chloroplast number per cell plan area showed fewer chloroplasts per cell area. In the future the phenotype needs to be unequivocally linked to the genotype by backcrossing the SALK_039451-171 into a wild- type background. 5521 is a Stromal Chloroplast protein 163 In order to interact with FtsZZ in the chloroplast, Sszl would have to be stromal- localized like FtsZ2. To test if 5521 is a chloroplastic protein, it was translated in vitro in 35 . . the presence of [ S]-methionine and imported Into chloroplasts as prevnously done to determine stromal localization of AtFtle-l, AtFtsZZ-1 and AtFtsZZ-Z (McAndrew et al. 2001) and inner envelope membrane localization of ARC6 (Vitha et al. 2003). Following import, chloroplasts were treated with thermolysin and trypsin, which digests proteins outside the chloroplast and in the inner membrane space (Cline et al. 1984). Chloroplasts were fractioned into membrane and stromal fractions and analyzed by SDS-PAGE and autoradiography. The $521 translation product (Figure 4.6A, TP) migrates near its predicted molecular mass of “'50 kDa. A band corresponding to the translation product (indicated as ”pSszl”) was observed in chloroplasts that were untreated by thermolysisn as well as a lower mass band (indicated by “is21”, Figure 4.6A). The upper mass band was not present in chloroplasts treated with thermolysin or trypsin, but the lower band was present and corresponds to the "40 kDa mass predicted for $521 without its putative transit peptide (53 amino acids)(Figure 4.6A, lanes 4 and 6). Lysing the chloroplasts with Triton X-100 resulted in the digestion of both pSszl and iszl (not shown), demonstrating that the proteases were active. In control experiments, the small- subunit of RuBisCo (Figure 4.6B) was also found in the stroma, demonstrating S521 is an imported protein. 164 Figure 4.3: Alignment of E. coli ZipA with its potential structural analogue in Arabidopsis. Identities are highlighted in black, similarities determined by a PAM250 matrix are highlighted in gray. ZipA is structurally similar to AT4G33540 from residue 190-328 in ZipA. Columns marked with "+" indicate residues in E. coli ZipA that are involved in interacting with FtsZ (Mosyak et al., 2000) and are chemically similar in AT4G33540 and predicted to maintain favorable contact with the C-terminal region of AtFtsZ2-1. Columns marked with a ""‘" are not chemically identical, but in AT4GB3540 they would make favorable contacts to AtFtsZZ-l. Columns marked with a "it" would not be favorable contacts. A full alignment of a sub-set of proteins with sequence similarity to $521 is in Figure 4.8. 165 éfififii emu A4 mogOWAHonmmazugn—e swag—HA Hm omeannmzmmnmaomm>=znmam mAHmam magnum mammaAmzwupwe MA--I >m4mm¢ om panda co «mom Emom0mm>mo - Hm gnmgfimawofiou. mm 0.530 mAioflamwmmofigézzfimAnomm%on§Hmg: m .-omnmmoo>m «madam mm.- womomm -meemm----m dz- -mzpm ->uo>om -pn ....... mm ; ... . .: .: -. .. A 58 So Hnowpzo my»... mumumfiamm ummmowmummzmmmemez _ --- --wmn --mxmna---m HHK-HAG>HH -n--Aao=s ........... H Humm «AAA Aumm «AAA Aumm AmAu Anmm Adan Axum «man 166 Table 4.2: Residue contact comparison between ZipA bound to E. coli FtsZ and $521 bound to the plant FtsZZ C-terminus. Residues in the structures of the E. coli FtsZ C-terminus that contact residues in E. coli ZipA (from structure 1f46) compared to how the threaded structure of S521 (Structural based sequence alignment in Figure 4.3) would contact the AtFtsZ2-1 C- terminus. Amina Acid in the EcZipA interacts with this crystal structure or amino acid in Ecl-‘tsZ or equivalent residue in $521 AtFtsZZ-I EcZipA S__s_z_; EthsZ AtFtsZZ-l Notes R305/ R306 K348/ N349 Y371 S460 M248 V294 L372 V461 K250 M296 0373 E462 Contact is between the non-polar portion of K250, this would be a similar in the AtFtsZZ-l and AT4633540 interaction T267 P313 None None Side chain is solvent exposed I196 I250 I374 I463/P464 P375 P464 Solvent exposed, locks a side chain turn in place A376 E465 A376 of EthsZ interacts with the hydrophobic region of R379 in EthsZ, in AtFtsZ2-1 E465 would interact favorably with K468. Also see F269 in ZipA M226 V272 F377 F466 F377 in EthsZ interacts with L378 in EthsZ, in AtFtsZZ-l this would be F466 interacting with L467 V194 V248 L378 L467 See the above note F269 0315 A376/F377 E465/F466 See the note for A376 in EthsZ R379 K468 Solvent Exposed K380 K469 Solvent Exposed M226 V278 0381 K470 In ZipA, EthsZ interaction the interaction is between the hydrophobic part of 0381 and M226, The charged part of the side chain is solvent exposed; a similar interaction would be expected between V278 of AT4G33940 and AtFtsZZ-l with the charged head group solvent exposed. 167 Following import, protease protection and protease removal, chloroplasts were hypotonically lysed and fractionated into soluble and membrane fractions. The membrane fractions were washed four times with lysis buffer to remove residual soluble protein and analyzed by SDS-PAGE and autoradiography. $521 was found to be a soluble stromal protein and in the soluble fraction (Figure 4.6A, lane 4 and 6). These results confirm the prediction that $521 is indeed a stromal chloroplast protein as predicted from bioinformatic analysis opening up the potential to be involved in chloroplast division. Removal of the $521 homologue in Synecococcus elongatus does not disrupt FtsZ filament morphology The phenotype of an Arabidopsis T-DNA insertion in $521 is slightly fewer and enlarged chloroplasts. $521 was identified on the basis of being a candidate FtsZZ interacting protein with structural similarity to bacterial ZipA. This predicts that $521 mutants would exhibit perturbed Z-ring morphology (McAndrew et al. 2001) as has been observed for disruptions in the FtsZ2 binding protein ARC6 (Vitha et al. 2003). 5521 has an ortholog in the cyanobacterium Synechococcus elongatus (locus syc1416) and disruption of the gene should lead to a perturbed Z-ring morphology. A knockout allele, SszlA, was created by homologous recombination mutation using the cloned 5521 allele with a resistance marker to spectinomycin inserted into the middle of the gene. After segregation of the knockout line by selection on spectinomycin Sszl transcript was no 168 longer observed (not shown) and the wild-type 5521 could not be detected by PCR (not shown). 5521A mutants and wild-type cells were analyzed by immunofluorescence microscopy with an antibody to FtsZ (Miyagishima et al. 2005) and the Z-ring morphology was the same in both wild-type and 552121 strains of Synechococcus PCC 7942 (Figure 4.7). 5521 does not interact with FtsZZ in the yeast-two hybrid assay To test if Sle interacts with plant FtsZZ, an initial yeast two-hybrid assay was performed between $521 and FtsZZ in the Matchmaker |Go system (Clontech). Initial results suggested the proteins interact (not shown). However, when 5521 was re-tested against an empty vector it was found to autoactivate the His reporter (not shown). Problems with autoactivation are well known in this yeast two-hybrid system and so this interaction was re-tested in an alternate system with fewer technical problems. 169 Figure 4.4: PCR analysis to identify a homozygous SALK_039451 mutant of in $521. (A) PCR analysis of SALK_039451-17 (homozygous), SALK_039451-18 (heterozygous) and WT Col-0. For each line, two PCR reactions were performed. The first reaction probed for the presence of an intact WT allele (primers gAt4g33540-L and gAt4g33540-R, expected product size 1685 bp) and the second reaction confirmed the presence of the T-DNA in the gene (primers LBb1 and gAt4g33540-R). Arrows indicate the PCR product for the WT allele and T-DNA PCR products. (B) Diagram of the gene structure for $521 (At4g33540) and the position of PCR primers used to analyze the T- DNA insertion of SALK_039451 and the expected bands for WT and insertional alleles. In addition, the location of the T-DNA is indicated in exon 4. 170 33093932.: 8-388163% GnomfioEorv n T — mvmmolv3>_ m 1 A .8 2 mz. 00 A0. 3.o.. .mmF m.mwmAAz 0A---.oAx. monmm 23 0A.. o.msz mHuxmz.omumm n<_ mmwmoAA1zgmww---.EAmm2m.Aom.2mwzu mm --m .AAJmAAmQx . AAAmAm:;mwnAm1mmAA--- AAAmzmoAOAoAmo.m. m4. HH.AAmmAwm A mA<<4xmuwmmmuAmmczomAmzmmo ---OAmmmA2m>oum>mwzzw.m nvmnfixo. A m 2M. moo memmzmmommeAz _mQA---zHAz¢> mAOA. m 2w.m _ ._. ._. 4. ._. .4. 4 ..._.... mm omm owm omm omm OHM «oxHmQOthmUMBm©mEEHdWthOHngmUZOdeum mA memmm--mx<>mmwemAmx-.mxwmAu>mouHmwmemHAmcoawAmzmwzoEAx m m _AA-.z_--meamiowemAmwmemwmemAAmmomgAmzmouwmAm m AHmA--.mm--azHAmowemAmwmmmwmemAAm>oamezmwmwmAx m _...._...._...._...._...._...._1...—...._...._...._...._. com 0mm 0mm Ohm 0mm 0mm ma m Aw m.mo>>>oumA.moHB. xmw GommerAz.>. cmoA.¢mzom_sEm AA 4m- .m<.oo Q A muqu.o> .omo<>o .mmHA ..m wwzm. m..Am.msmmkmA . mHmm _AAmou. m:mo. HA 2A.H oz.mAm.> a. mmmmo> QAP omm>owmmAAmmuxAmmazxmmo<>a .. .BAmzwmmowAmmAzn.Azeewmmmo> maem.m>mwmmHAmmuzAmmo-.mmHAmzwm 921m. mEAAAmgomAmmaszmmofi/Q-QmmmAmzwmawwAmmA.m AVE. mamas, om mmm>oommAAm m.mmgdzxxmod>m-.mmHAmzwm>ww.ZAH.w.Axm .mmmo> _...._1..._...._...._....—...._....—...._...._...._...._.... O¢N 0mm CNN OHN OON omH mflumhoonuuqhm mflouooonoodkm ooumoz m oofim 4 coax monumm mAmmooAnmue mfiumhuonuudhm maooooonoonhm uoumoz m woflm 4 ooAm aosumm mAmmoeAnmue mfiumhoonuunhm mnuoouonuonhm ooumoz m woAm d uufim monumm mAmmocAnmue 187 Figure 4.8 (continued) mflmwmm mum mum/h mflumhoonoodhm msooooonoodhm UOumoz m moflm 4 coax oUSHQm mAmmouAnmue 188 CHAPTER FIVE SUMMARY AND FUTURE DIRECTIONS 189 Summary This dissertation aims to advance the knowledge of the biochemical properties of plant Ftle and FtsZZ. When this work was initiated there was a limited amount of function information for Ftle and FtsZZ. Moreover, the putative plant FtsZ regulatory proteins ARC3 and ARC6 were not yet discovered. Thus the aim of these studies was to use a biochemical approach to understand why plants use two types of FtsZ proteins. To understand the composition of the FtsZ complex in viva, FtsZ was purified from pea stroma. Pea Ftle and FtsZZ co-fractionate with the chloroplast division proteins ARC3 and ARC6 when separated by sucrose density gradient, anion-exchange chromatography, size-exclusion chromatography and native polyacrylamide gel electrophoresis. In addition, a ZipA-antibody cross-reactive protein was found to co- fractionate with FtsZ1/FtsZZ/ARC3/ARC6-containing complexes. Finally, FtsZ levels were quantified in isolated Arabidopsis thaliana chloroplasts and Ftle and FtsZZ were found at a constant ratio of "1:2. Ftle and FtsZZ were also found to be co- immunoprecipitated from pea stroma. Plant FtsZZ contains a C-terminus similar to the ZipA-binding C-terminus of bacterial FtsZ. However, ZipA had not been identified in plants. A structural search algorithm was created find a chloroplastic protein that is structurally similar to ZipA. This algorithm identified a strong ZipA candidate, 5521, but 5521 could not be definitively linked to a role in chloroplast division. Moreover, exon 4 of ARC6 has recently been found to be structurally similar to ZipA and binds the C-terminus of FtsZZ. 190 To understand the biochemical properties of plant FtsZ, recombinant FtsZ1 and FtsZZ were produced individually in E. coli. Ftle and FtsZZ were both found to be GTPases, but approximately 15-fold slower than the GTPase activity of E. coli FtsZ. When mixed equally Ftle and FtsZZ co-assemble into bundled ribbon-like structures. Maximal Ftle/FtsZZ co-assembly requires an equal concentration of both proteins. In (addition, a T7-loop mutant, FtsZZD322A, was found to be a sub-stoichiometric competitive inhibitor of Ftle, but not FtsZZ. Together these results support a FtZl/FtsZZ heterofilament model similar to OLE-tubulin. In summary, work described in this dissertation advances our knowledge of biochemical properties of Ftle and FtsZ2. More importantly, this work lays the groundwork for future study of the stromal chloroplast division machinery. Ftle and FtsZZ co-polymerize into heteropolymers and can be co-purified with ARC3, ARC6 and a ZipA cross-reactive protein. ARC3 and ARC6 are known to regulate FtsZ filament morphology in vivo (Glynn et al. 2007; Vitha et al. 2003). Thus a logical model is that Ftle/FtsZZ heterofilaments are directly bound to proteins that regulate their assembly, organization and dynamics. Future work will be needed to test this new model of FtsZ dynamics and regulation. Future Directions Structural Search Algorithms Sequence alignment search tools such as BLAST (Altschul et al. 1990; Gish and States 1993) infer structural and functional similarity from sequence similarity, and not 191 protein structural similarity. This approach works well for closely related genes and proteins. However, sequence alignment searches generally do not work well when sequences differ substantially. The result of this problem is that approximately 40% (as of 2007) of the Arabidopsis genome remains functionally unannotated because these genes (or deduced proteins) do not have sequence similarity to other proteins (Swarbreck et al. 2007). However, the inability to annotate genes in plant genomes may result from some of these genes being plant-specific. Plants contain stromal targeted homologues to some bacterial cell division proteins such as FtsZ, MinD, MinE and SulA. However, plants lack homologues to several bacterial cell division proteins that have been predicted to exist, such as MinC and ZipA. This prompted the concept of searching for proteins with similar function, not by sequence similarity, but by structural similarity. Searching a database for proteins with structural similarity has not been previously done to my knowledge. Instead of developing a new computational framework to do this analysis, the threading algorithm of the PROSPECT 2 software package was used to find a structural analogue of E. coli ZipA in plants. While one candidate, Sszl, looked promising (Chapter 4), ultimately 5521 could not be definitively linked to chloroplast division. However, the prediction that plants contain a structural analogue of ZipA has been borne out. Recent work in our laboratory by Jonathan Glynn has found that a de novo derived structure of a short portion of ARC6 is structurally similar to ZipA. Furthermore, this portion of ARC6 also interacts with the FtsZZ C—terminal motif similar to the E. coli FtsZ ZipA binding motif, and ARC6 has a role in remodeling and organizing FtsZ filaments (Vitha et al. 2003), 192 indicating ARC6 had functional as well as partial structural similarity to ZipA. recent studies from our laboratory have revealed that ARC6 bears a domain with structural and functional similarity to ZipA (J. Glynn, unpublished). In the future, structural similarity algorithms may become useful, especially as more genomes are sequenced with high throughput mechanisms. In order for this approach to be successful in the future, the major weakness of this approach should be addressed, the threading algorithm itself. This analysis used the threading algorithm from the structure prediction software package PROSPECT 2, which is designed for structural modeling and not database searching. Future structural search algorithms should instead focus on a different approach such as comparing the arrangement and content of a and 8 sheets in a secondary structure prediction or other methods without a sequence similarity bias. Finally, a good lesson from this work is that while computation predictions can be extremely useful for making scientific progress, there is no substitute for the power of the scientific experiment itself. The protein that cross-reacts with the E. coli ZipA antibody An antibody to E. coli ZipA specifically recognizes a "35 kDa protein in pea chloroplasts, similar to the mass of the E. coli ZipA protein. The identification of this putative chloroplastic protein is an exciting prospect. In the research presented here, this protein has been shown to be tightly associated with the FtsZ-containing complex by sucrose density gradient fractionation, native PAGE, anion exchange and 193 hydrophobic interaction chromatography, suggesting this protein might be a chloroplast division protein. Unfortunately, due to a lack of an appropriate molecular ion template database, a strong candidate for the 35 kDa ZipA cross-reactive protein was not identified. The pea orthologues of ARC6 and its homologue ASH1 can be ruled out as candidates for the ZipA cross-reactive protein because of their high molecular masses (80-100 kDa based on the Arabidopsis sequences and detection on immunoblots; McAndrew et al. 2008; Yue Yang, unpublished). Future experiments should focus on identifying the ZipA cross- reactive protein by proteomic identification from a—ZipA immunoprecipitation, instead using a recently generated pea cDNA library to generate molecular ions for identification (Andreas Weber, personal communication). Chloroplast division protein purification from inclusion bodies Chloroplastic proteins are difficult to produce recombinantly in E. coli. Most stromal proteins are encoded in the nucleus, translated in the cytoplasm and imported into the chloroplast. Further, most chloroplast proteins have a chloroplast targeting sequence on their N-terminus that is cleaved as the protein is imported into the chloroplast (Bruce 2000). Moreover, prediction of the presence of the chloroplast transit peptide is generally straightforward (although many proteins may be missed), but the prediction of the cleavage site for the chloroplast targeting sequence is still enigmatic. For example, the predicted cleavage site of the Ftle transit peptide is at amino acid 90 and would result in the removal of half of the Rossman-fold GTP binding 194 motif that is characteristic of FtsZ and tubulin (see Chapter 3 for more detail). Indeed, this cleavage site was chosen for yeast-two-hybrid constructs between plant FtsZs (Maple et al. 2005) and while it does not appear to have affected interaction between the FtsZs themselves, it does call into question the interaction between Ftle and ARC3, since ARC3 contains an FtsZ Rossman-fold GTP binding domain that may be compensating for this missing domain in the Ftle construct. So how should researchers approach this in the future? The expression of plastid division proteins (such as Ftle, FtsZZ, ARC6 and ARC3 and many others) recombinantly in E. call has been an extremely difficult procedure because A) the proteins frequently block E. coli cell division and are thus cytotoxic; 8) due to toxicity, they are frequently found in inclusion bodies and C) stromal plastid division proteins tend to bind tightly to the native E. coli division machinery and are difficult to remove, similar to the difficulty of purifying E. coli FtsZ beyond 90% under native conditions (Lu and Erickson 1998). Typically, accessory factor proteins bind extremely tightly and are not removed even under the high salt conditions of anion and cation exchange chromatography. The work described here makes a significant leap forward in the ability for plant FtsZ proteins to be purified in an active form. The key to this process is the co- expression of the ftsQAZ operon from E. call during expression (Jeong and Lee 2003) to prevent a block in cell division. Further, expression levels are increased in strains 195 overexpressing the ftsQAZ operon suggesting blocked cell division reduces protein expression. The approach to expression and purification of Ftle and FtsZZ recombinantly in E. coli has also been successful for expressing the stromal, soluble domain of ARC6 and fragments of ARC3 (Bradley Olson, unpublished observations). Moreover, refolding protocol derivatives of those used to refold plant Ftle and FtsZZ (Chapter 3) are successful at refolding ARC6 and ARC3 from inclusion bodies and have significantly increased the capacity to study cytotoxic chloroplast division proteins in vitro, by purification of recombinant proteins. FtsZ biochemistry Future biochemical study of FtsZ should be focused in three areas. First, the biochemical relationship between Ftle and FtsZZ in head-to-tail protofilaments should be defined in detail. These experiments can be complemented with in viva assessment of the T7-loop mutations (FtsZZ described in Chapter 3) and others that target the interface between Ftle and FtsZ2. Second, defining lateral interactions between Ftle and FtsZZ will need to be addressed. Third, what are the biochemical effects of FtsZ interacting proteins such as ARC6 and ARC3? Probing heterofilament formation by FtsZI and FtsZZ In Chapter 2, Ftle and FtsZZ were found in a salt-stable complex that represents unassembled FtsZ (McAndrew et al. 2008). Moreover, the molecular mass of the native FtsZ complex suggests the presence of one each of Ftle, FtsZZ, ARC6 and ARC3 196 (McAndrew et al. 2008) and possibly the presence of a protein that cross-reacts with ZipA (Chapter 4). In Chapter 3, Ftle and FtsZZ were found to assemble into heterofilaments. Thus, ignoring ARC3 and ARC6, the basic unit of plant FtsZ assembly is a head-to-tail dimer of Ftle and FtsZZ similar to OLB-tubulin. In contrast to tubulin, both Ftle and FtsZZ are GTPases, thus there is not an obvious explanation for why Ftle/FtsZZ polymerize into heterofilaments. One possibility is that plant FtsZ disassembly is regulated, contrasting with bacterial FtsZ which is assembly regulated. Ftle/FtsZZ head-to-tail polymerization should be explored for a mechanistic explanation. Several structures of afl-tubulin dimers have been solved and described in detail (Nogales 1999, 2000; Nogales et al. 1998a ; Nogales and Wang 2006; Nogales et al. 1999; Nogales et al. 1998b). Ftle and FtsZZ homology based structures (Yoder et al. 2007) should be examined for residues that might mediate head-to-tail interactions. The residues that are found to be in the Ftle/FtsZZ interface should be sequentially mutated and tested for the effect on polymerization and GTPase activity, in a similar approach that was used with 'l7-loop mutations (Chapter 3). These experiments will provide significant insight into the exact specificity for heterofilament formation. Moreover, they could yield insight into the head-to-tail filament specificity for bacterial FtsZ and aB-tubulin. In concert with these experiments, it would be useful to collaborate with a structural lab to attempt to crystallize an Ftle/FtsZZ dimer. 197 There is growing evidence that bacterial FtsZ polymers are assembled by the addition of FtsZ dimers, not monomers (Chen et al. 2005; Chen and Erickson 2005; Huecas et al. 2007a). This is kinetically similar to aB—tubulin, with the exception that the OLB-tubulin dimer is initiated by chaperones (Lewis et al. 1997). In the bacterial FtsZ model, there is growing evidence that E. coli FtsZ exists as a dimer and not monomers (Chen et al. 2005; Chen and Erickson 2005; Huecas et al. 2007a; Rivas et al. 2001). After dimer formation there is a kinetic lag (cooperativity) to add a FtsZ dimers onto a growing protofilament (Caplan and Erickson 2003; Chen et al. 2005; Huecas and Andreu 2003, 2004; Huecas et al. 2007a ; Romberg and Levin 2003). If true for plant FtsZ, this would predict that Ftle/FtsZZ dimerization is is the minimal subunit of filament formation, and addition of Ftle/FtsZZ dimers to heterofilaments is cooperative, as observed by light-scattering in Chapter 3. Ftle/FtsZZ dimerization can be examined with multiple approaches. First, variants of Ftle and FtsZZ recombinant proteins described in Chapter 3 have been created that allow N-terminal fluorescent tags to be attached to Cys residues. These Cys residues correspond to the plus-side of the growing protofila ment and would be GTP- capped. To probe for dimerization, fluorescently labeled Ftle could be tested for quenching upon mixing with unlabelled FtsZZ (or the converse). This approach can be used to measure the dissociation constant between head-to-tail associated FtsZ, since the fluorescent tagging site lies in the FtsZ1/FtsZZ interface. Indeed, preliminary data to this effect have been generated. N-terminal fluorescently labeled Ftle and FtsZZ are biochemically indistinguishable from unlabelled proteins (Bradley Olson, unpublished 198 observation). Preliminary experiments demonstrate the fluorescent tags are quenched when mixed in trans, but a technical hurdle of photobleaching needs to be overcome to measure the dissocation constant in detail (Bradley Olson, unpublished observations). If fluorescent tagging does not work, an approach such as calorimetry (Caplan and Erickson 2003; Huecas et al. 2007b) could be used to confirm the energy of Ftle/FtsZZ heterofilament polymerization. Second, non-hydrolyzable fluorescent GTP analogues could be used to probe the nucleotide dissociation constants for Ftle and FtsZZ individually and mixed (Huecas et al. 2007b).. These experiments will answer the question of which nucleotide binding site is in the interface between Ftle and FtszZ. In Chapter 3, there is evidence that polymerized plant FtsZ does not exchange nucleotide for free nucleotide. In this future line of experiments, the nucleotide binding sites would be probed with non- hyd rolyzable, fluorescent nucleotide analogues similar to previous approaches (Huecas and And reu 2003, 2004; Huecas et al. 2007b). The site that is found within the interface between Ftle and FtsZ2 would be found to have a lower dissociation constant. To distinguish between FtsZ1 and FtsZ2 nucleotide binding sites, Ftle and FtsZZ could be mutated in their glycine-rich N-terminal domain to convert them to ATP binding proteins as has been done previously for E. coli FtsZ (RayChaudhuri and Park 1994). Thus, ATP-binding and GTP-binding variants of FtsZ1 and FtsZZ could be mixed and analyzed for fluorescence nucleotide quenching to determine which nucleotide binding site is in the interface. 199 Finally, an Arabidopsis Ftle-I mutant was found to have the mutation, G267R which is in its T7-loop (Appendix A)(Yoder et al. 2007). In lines or tissues where ft5216267R is expressed at normal levels, this allele is dominant-negative. Moreover, ftszlGZG7R leads to short disorganized filaments, suggesting it is blocking polymerization of FtsZ1/FtsZ2 heterofila ments. An important confirmation of the hypothesis that Ftle and FtsZZ polymerize into heterofilaments would be to attempt to rescue ftszI and ftszZ knockouts (Aaron Schmitz, in preparation) with T7-loop mutants similar to those described in Chapter 3. If Ftle and FtsZZ do polymerize into heterofilaments in viva, it would be expected that Fl' SZ10275A would act dominant- negatively inizan ftszl knockout mutant and exacerbate the knockout phenotype. Moreover, overexpression of FT SZZDZ75A should lead to strong depolymerization of wild-type FtsZZ when observed by immunofluorescence. Important to these experiments will be doing the reciprocal experiment and controls with wild-type alleles. F tsZI and FtsZZ lateral interactions The lateral interactions between bacterial FtsZ are poorly understood, mostly because in the absence of stabilizing agent, bacterial FtsZ only assembles into single protofilaments (Chen et al. 2005; Mukherjee and Lutkenhaus 1994). Moreover, most of the differences in amino acid sequence between Ftle and FtsZZ (Stokes and Osteryoung 2003) lie in regions predicted to be solvent—exposed in putative structures created by homology modeling (Ftle model in Yoder et al. 2007). Residues found to be putatively within the interface between plant FtsZ filaments should be probed with site- 200 directed mutagenesis and analysis of the mutation effect analyzed by light-scattering and electron microscopy, similar to the techniques used in Chapter 3. In this set of experiments, mutations that decrease filament bundling should be identified and characterized further. It should also be pointed out that the inherent stability of plant Ftle/FtsZZ co-polymers may be an experimental advantage over E. coli FtsZ for understanding FtsZ lateral interactions. In a second line of experiments, an approach such as partial proteolysis of polymerized Ftle/FtsZZ could be used to define filament interfaces. The idea behind this experiment is that residues that lie within the interface between Ftle and FtsZZ would be protease-protected, but solvent exposed faces would be protease sensitive. These experiments could be followed by fluorescently labeling residues on various surfaces of Ftle and FtsZZ (similar to experiments proving the interface head-to-tail above) and testing for quenching after polymer bundling. These experiments could use results from a previous study with E. coli FtsZ as a guide (Lu et al. 2001). In conclusion, the experiments described above aim to structurally define the relationship between Ftle and FtsZZ in more detail. These experiments will be extremely useful for furthering our understanding of FtsZ and tubulin polymerization. Finally, these experiments will describe the assembly of a component of the chloroplast fission machinery in detail and be a scaffold for understanding the role of other division factors such as ARC3 and ARC6. The biochemical effects of ARC6 and ARC3 201 A series of biochemical experiments based on light-scattering and EM should be performed to define the biochemical effects of ARC6 and ARC3 on the GTPase activity and polymerization of Ftle/FtsZZ. Based on genetic analysis showing that ARC6 promotes (Vitha et al. 2003) and ARC3 inhibits FtsZ filament formation (Glynn et al. 2007) in chloroplasts, we hypothesize that ARC6 should promote and ARC3 inhibit the polymerization of Ftle/FtsZZ in vitro through their respective interactions with FtsZZ and FtsZ1. Based on the presence of both proteins in the FtsZ-containing complex described in (McAndrew et al. 2008)(Chapter 2), we might further expect ARC6 and ARC3 to act antagonistically on polymerization when mixed. There is evidence that ARC6 may also promote the polymerization of FtsZZ in the absence of FtsZ1. In antisense AtFtle -1 and AtFtsZZ-l lines, the reduction of AtFtsZ1-1 or AtFtsZZ-l have different effects on Z-fila ment topology (Vitha et al. 2001). The loss of AtFtsZZ-l leads to short disorganized AtFtle-l filaments (Vitha et al. 2001). In contrast, loss ofAtFtle-l leads to long spiral AtFtsZZ-l filaments (Vitha et al. 2003). This suggests that FtsZZ is capable of forming long Z-filaments on its own, or that ARC6 is promoting FtsZZ filament stability in the absence of AtFtle-l. In vitro, it was found that FtsZZ does not assemble efficiently on its own (Chapter 3) and this would suggest the long FtsZZ-filaments observed in the absence ofAtFtle-1 are due to the stabilization effects of ARC6. This idea could be tested by creating an ftszl/arc6 double mutant and performing immunofluorescence for AtFtsZZ-l. If ARC6 is in fact promoting FtsZZ stability, short AtFtsZ2-1 Z-filaments would be expected to be observed in the ftszl/arc6 double mutant (Vitha et al. 2001). 202 Finally, ARC3 has a role in the positioning of the Z-ring properly to mid-plastid. Immunolocalization of FtsZ in the arc3 mutant shows evenly spaced, multiple Z-rings (Glynn et al. 2007) suggesting ARC3 may have a MinC-like function (Glynn et al. 2007; Maple and Moller 2007a). Additionally, ARC3 localizes bath to the pole (Maple et al. 2007) and mid-plastid (Maple at al. 2007; Shimada et al. 2004). ARC3 contains an N- terminal, FtsZ-type GTP binding domain and could either modulate the GTP activity or cap FtsZ filaments. This could be tested in vitro by examining the GTPase activity and polymerization of Ftle and Ftle/FtsZZ in the presence of increasing concentrations of ARC3. Concluding remarks The establishment of an in vitro system to study plant FtsZ biochemistry will open a new functional understanding of chloroplast division at a biochemical level. Current efforts in understanding chloroplast division have been focused on identifying the molecular players involved. However as more chloroplast division genes are identified, their functional role in plastid division will need to be understood. This dissertation contributes significantly to this goal. 203 APPENDIX A Yoder, D. W*., D. Kadirjan-Kalbach*, B. J. S. C. Olson, S. Y. Miyagishima, S. L. Deblasio, R. P. Hangarter, K. W. Osteryoung and S. Vitha (2007). "Effects of mutations in Arabidopsis Ftle on plastid division, FtsZ ring formation and positioning, and FtsZ filament morphology in vivo." Plant and Cell Physiology 48(6): 775-791. *These authors contributed equally to this work Note: This reprint is reproduced under license number 1912310855884 from Oxford University Press granted to B. J. S. C. Olson and Michigan State University. B. J. S. C. Olson contributed the following: performed the initial immunoblots for this paper (immunoblots for publication including atFtsZI-l-Al were performed by D. W. Yoder); built the homology-based structure and created the structure-based alignment between plant and bacterial FtsZs; made the discovery that pmi4 is semi-dominant; made the discovery that expression levels of the mutant proteins vary in young versus aId leaf tissues; interpreted the possible effect of the mutations; wrote the early drafts of the discussion and contributed editorial assistance to the final manuscript. 204 Plant Cell thsiai. 48(6): 775—79l (2007! doi.10.ltfl3r‘]:cpftun04‘). available cnhnc at wwwpcpaxlurdiournalserg O The Author 200?. Publish-s... by Oxford University Press on behalf of Japanese Society of Plant Physwlogists. All rights reserved. For permissions, please entail. journals pernussicrs@oxfordjournals.org Effects of Mutations in Arabidopsis FtsZ1 on Plastid Division, FtsZ Ring Formation and Positioning, and F tsZ Filament Morphology in vivo David w. Yoder M, Deena Kadirjan-Kalhach M, Bradley J. s. C. Olson 1'2, Shin-ya h’Iiyagish' ima 1.5, Stacy L. DeBlasio 3'°, Roger P. Hangarter 3 and Katherine W. Osteryoung "* 1 Departing"! of Plant Biology, Michigan State University, East Lansing, MI 48824, USA :Deparrmenl of Biocherriistry and Molecular Biology, Michigan Stare University, East Lansing. MI 48324. USA " Department of Biology, Indium (.I'nii'ersity, Bloommgtari. IN 47405. USA In plants. chloroplast division FtsZ proteins have diverged into two families, FtsZ] and FtsZ2. FtsZ] is more divergent from its bacterial counterparts and lacks a C -terminal motif comerved in most other FtsZs. To begin investigating FtsZ] structure—function relationships, we first identified a T-DNA insertion mutation in the single FtsZ] gene in Arabidopsis thaliana, AtFtsZI-I. Homozygotes null for FtsZ], though impaired in chloroplast division, could be isolated and set seed nonnally, indicating that FtsZ] is not essential for viability. We then mapped five additional atftle-I alleles onto an FtsZ1 structural model ainl diaracterized chloroplast morphologies, FtsZ protun levels and FtsZ filament morphologies in young and mature leaves of the corresponding mutants. atftsZI-l ( 626711), atfrsZI- 1(R298Q) and arftsZI-I ( 11404—433) exhibit reduced FtsZ1 acctnnulation but wild-type FtsZ2 levels. The semi-dominant a1ftsZ]-I(G26.7R) mutation caused the most severe pheno- type, altering a conserved residue in the predicted T7 loop. atftsZ]-l(GZfi7R) protein acumiulates normally in young leaves but is not detected in rings or filaments. atftsZI- 1(R298Q) has midplastid Ftle-containing rings in young leaves, indicating that R298 is not critiml for ring formation or positioning despite its conservation. atfisZI-NDISSlN) and arfisZI-I(G366A) both have overly long, sometimes spiral-like FtsZ filaments, suggesting that FtsZ dynamics are altered in these mutants. However, atfrle-NDIS9N) exlu'bits loss of proper midplastid FtsZ positioning while atftsZI-I (G366A) does not. Finally, truncation of the FtsZ] C-termirns in atfisZI-If'A-104—433) impairs chloroplast division somewhat but does not prevent midplastid Z ring formation. These alleles will facilitate understanding of how the in vitro biochemical properties of Ftle are related to its in vivo function. Keywords: arc/(l ——- FtsZ —- pmi4. Abbrevtations. arc, accumulation and replication of chloro- plasts; CAPS. cleaved amplified polymorphic sequences; CTD, Cetcrminal domain; DIC, ditlcrcntail interference contrast; EMS, ethylmethane sulfonatc; FITC. fluorescein isothiocyanate; pmi, plasud mobility impaired; NTD, N-tcrminal domain; RT PCR, reverse transcription PCR," SSLP, simple sequence length polymorphism. Introduction Plastids arose from an ancestral cyanobacterial endo- symbiont and have retained a division apparatus reminis- cent of that in cyanobacterial cell division (Osteryoung and Vierling 1995, Osteryoung ct al. 1998, Colletti ct al. 2000, Itoh at al. 2001, Maple ct al. 2002, Vitha ct al. 2003, Maple ct al. 2004, reviewed in Osteryoung and Nunnari 2003. Aldridge ct al. 2005). Plastid division requires assembly of F152] and F 1322, plant-Specific homologs of the tubulin- likc bacterial cytoskeletal protein FtsZ (Osteryoung cl 3]. 1998). into a ring (the Z ring) at the midplastid division site (Vitha ct al. 2001). The Z ring is localized to the midplastid through the activity of the FtsZ-positioning proteins MinD and MinE (Colletti ct al. 2000, Itoh ct al. 2001, Maple ct al. 2002) and is thought to be stabilized by the .l-domain-likc protein ARC6 (Vitha ct al. 2003). Midplastid positioning of the Z ring presumably ensures that normal plastid papula- tions are maintained in all plant cells. Bacterial FtsZ has two functional domains, an N-tciminal domain (NTD) and a C-tcrrninal domain (CTD), that fold independently of one another (Oliva ct al. 2004). The NTD contains a Rossman fold for GTP binding (Lowe 1998, dec and Amos 1998), and the CT!) contains a T7 loop for GTP hydrolysis (dc Boer et al. 1092. RayChaudhuri and Park 1992, Lowe 1998. Léwc and Amos 1998, Scheffers ct al. 2002). The CTD also contains at its extreme C—tcrminus a short conserved stretch of amino acids that has been shown to interact with the bacterial cell division proteins FtsA and ZipA (Ma and Margolin 1999, Vaughan at al. 2004‘). Similar to tubulin, longitudinal contacts in F152 polymers are made bctwacn thc rcgion of 4These authors contributed equally to this work. :jPrcscnt address Miyagishima Initiative Research Unit, Frontier Research System. RIK EN. 2—1 Hirosawa, Wake. Saitama, 3510193 Japan. °Prcsc11t address: Departmzmt of Microl'nology. The State Universrty of New York. Stony Brook, NY l I794, USA ‘Corrcsponding author. E-mdll, osteryoufiansu.cdu: Fax, +1-517-353-1936. 205 776 Mutations in Ftle affect chlorOplas: division the. CPD bearing the T7 loop in one monomer and the GTP-binding ate in the NTD of the next monomer. The GTP—hydrolytic site is formed by interaction between two monomers, and GTP hydrolysis is stimulated by FtsZ polymerization (Erickson 1998, Lowe and Amos 1998, Nogales ct al. 1998, Oliva ct al. 2004). Full and FtsZ.? arose early in plant evolution by duplication of a common ancestral FtsZ gm: of cyano- bacterial origin. Ftle is more divergent from its cyano- bacterial counterparts than is FtsZZ, and is 1miqu: to plants and green algae (Stokes and Osteryoung 2003, Wang et al. 2003, Rensing at al. 2004). F132] also lacks the short C-terminal motif found in most bacterial FtsZ proteins; this motif is conserved in FtsZZ proteins (Fig. 2A), where it mediates an FtsZZ-specific interaction with ARC6 (Maple ct al. 2005). Thus FtsZ] and FtsZ2 are functionally as well as phylogenetically divergent Nevertheless, except for lack of the C-tcrminal conserved motif, FtsZ1 has all the hallmarks of a typical FtsZ and is predicted to be a GTPase (Osteryoung and McAndrew 2001). Consistent with this observation, recombinant FtsZ1 has been reported to undergo GTP-dependent multimcrization and assembly into filament-like structures in vitro and to complement a cell division defect in an Escherichia coli fLSZ mutant (Gaikwad ct al. 2000, Bl-Kafafi ct al. 2005). However, the structural requirements for FtsZ1 chloroplast division activity in vivo have not been investigated. To investigate further FtsZ] function in plants and begin defining the features important for its chloroplast division activity, we compared the predicted secondary and tertiary structures of Ftle with those of bacterial FtsZ. We show that FtsZ] contains predicted tit-helices, loops and fi-shects that in bacteria form a Rossman fold for GTP binding and a T7 loop for GTP hydrolysis. We also isolated six mutant alleles of the single FtsZ] gene in Arabidopsis thaliana, ArFrsZI-I (A15g55280), including a null allele. Four alleles alter conserved amino acids, and one tnrucatcs ArFtsZI-I near its C-terminus. Analysis of chloroplast morphologies, Hell and FtsZZ protein levels, and Hal ring and filament morphologies in the mutant backgrounds show that: (1) plants homozygous for the FtsZ] null allele are viable and FtsZ2 protein levels are not affected; (ii) predicted structural features in FtsZ] homologous to those in bacterial FtsZ are important for FtsZ filament (Z filament) and chloroplast morphology in vivo, and some residues may be important for accumulation of Ftle, including the C-tcrminus; (iii) FtsZ2 is capable of forming rings in very small chloroplasts in the absence of a functional FtsZ] protein; (iv) changes in Ftle protein level or function do not alter expression of FtsZ2, but do affect FtsZZ filament morphology; and (v) mutations in key msidues appear to impact FtsZ polymer dynamics in viva. Results aIfLrZI -J-A1 is a null allele ofAr'FrsZ-I A search of the SIGnAL ‘T—DNA Express’ database (httpz/lsignal.salkcdufcgi-binftdnacxpress) revealed a mutant line, SALK_073878 (Alonso at al. 2003), harboring a T—DNA insertion in the first exon of AtFtsZr'J (Fig. 1A). We have designated this gene aIflJZI -1 —A1 . The genotype of the homozygous T-DNA insertion mutant was verified by PCR. atftsZI-I—AI mutant plants exhibit altered chloro- plast morphology and a block in chloroplast division, as indicated by a dramatic reduction in chloroplast numbers compared with wild-type Columbia (Col-0; Fig. 18, C). Lcaf mesophyll cells in atftsZI-I—AI typically contain an average of five chloroplasts per cell, with a range of l 12. chloroplasts (Table 1), compared with a mean of 41 in wild- type Col-0. These chloroplasts are enlarged and exhibit heterogeneity in size and number from cell to cell (Fig. ID). No aplastidic cells were detected in any leaf samples. Homozygous mutants set seed and produced viable plants. In order to confirm that the PCR-detected TDNA insertion caused the mutant phenotypes, atfstl-I-Ar‘ plants were transformed with wild-type AzFrle-I under control of its native promoter. Wild-type chloroplast morphology and number were restored in the transforrnant, confirming that the T-DNA insertion is the cause of the mutant phenotypes (data not shown). A sarrgomn AMI-1 r' V _ . “5955230) M -.—WW 8382 1-1-A 1 D ”a r. fi'c‘J.’ “'1." '1‘ ‘ E m Null a-AiFuzr-r BEE a—AtFteZ'l-t — Coomaasle "" 1 2 Fig. 1 (A) AtFtsZIal gene depicting the position of the T—DNA insertiOn in attiszl-l-A? (Salk_073878) in the 3' region of exon 1. Light micrographs representing chloroplast morphology in wild- type Col-O (B) and atftszl-I-di (C and D). (E) lmmunoblot of protein extracts from mature leaves of Col-0 and the nuil a‘lele (atftszl-I-Al) using a—AtFtle-l cop panel), a-AtFtsZZ-l (middle panel}, and Coomassie-stained gel for loading comparison (bottom panel). 206 Mutations in Ftle affect chlorOpIast division 77'? A . 1' . 3' . T' . T. . f' . “ . 7' . 3' . 3° AIFtsZ1-1 --~---—-—--—- m:ssssum—--—-—m ‘ n figgfl :mmmfi : 1’".-. __.:- wigmml=fin :3 iififéizi ____a « A1FtsZ1-1 IIRm-‘l 3573’ FtsZZ ._ .M .5:- _J' ' 2 - ..-"x.ii itiin- it.» Manhatwnw. W ”‘1'“ “(rt-“*Li‘h 313).! 4n .rsmlvr .Anmcarnesp . 3'! -i !I|“-:n -"'" 1*; 2mm? amt-51mm ‘ - 11 “Emir“ III: 2‘50 ' ‘ t’" H {lR'lmLz' , i lillthWDPfll' r_ H‘VH mum - =.. ll‘lLVWDUUI‘. Min :uzorm' i: -' :i . wotvmrmmm; WV .11» :4 market Inn-Hum ‘ 'U'tll ~11th 'VH AlFtsZt-1 AlFtsZZ-‘l AtFtsZ2-2 EeFtsZ-1 MthsZ1-1 ss AIM-433 MARKERStW ‘ ' ' . I . ' AlFuZ1- 1 Anna:- 1 —'__ g A' 533:1 '2 o--. N -Tm.' m" u FtsZ1-1 — °°"‘ ”9”“ 1 c.— Spacer rogbn C-terminal conserved not! 82990 B _ . 118.”; G to A ‘ 0.695! (17678 3306/1 .1404 433 GIOGtoA Huntiton’ 156M110 C161tun'ufrlion 1 1 294 397 496 590 819909 1076 1183 1416 1513 1788 Fig. 2 (A) Amino acid alignment of AFtsZi-I (NP_200339), AtFtsZ2-l tNP_190843). AtFtsZZ-Z (NP_565839), EthsZ (AAC732065 and Mth52 (QS7816) proteins with predicted secondary structure. Dark gray bars represent rat-helices and light gray arrows represent B-strands, with all structures labeled based on the homologous structures from Methanocaccusjannaschr’i (Lowe, 1998). Black dots above the sequence indicate equ valent bacterial mutations D96A (Redick et al 2005') is a mutatior equivalent to 0159 and A239V represent both :5! in Bacillus subti'h's (M chie et al 2006) and MI. 863 in Escherichia coli (Addinall et al. 2005') The positiOns oi the five mutant alleles are indicated by black arrows and the amino ac d changes are indicated above the alignment. Black bars above the alignment are the tubulin motif (CCCGTGUSG) (RayChaudhur and Park 1992) and T7 loop. The C~1ermmal conserved motif is indicated by a black bar besow the alignment. The legend indicates the NTD, CTD, N-termnal region, core region and spacer region. (B) We transcriptor unit of .At5g55280, with the position of each mutation indicated above with black anows. At, Arabidopsis mair'ana; Ec, Escherichia cnii; Mi, Methanowccus jannaschi‘i. 207 “J ~J "JO Mutations in Ftle affect chloroplast division Table l atftle-l mutant phenotypes and morphologies in leaves from 3-wcck-old plants Linc Chloroplast morphology Socd Z filament morphology Mutation set (98)” Phenotype Mean Ranng number“ Col-0 Wild type - 41 30—50 97.6 Midplastid FtsZ rings — null Giant chloroplasts, 5 l—12 99.2 No Ftle protein, FtsZZ T-DNA insertion, some smaller filaments are long and exon 1 chloroplasts randomly distributed D159N Heterogeneous, one 14 7—20 99.] Long, possibly parallel Z EMS, lateral surface greatly enlarged with filaments. some numerous smaller midplastid rings in chloroplasts smaller chloroplasts 0267R Giant chloroplasts 1 l—--3 97.9 Reduced FtsZ] protein EMS, T7 loop levels, punctatc Ftle. long and randomly distributed FtsZZ filaments R2980 Heterogeneous 19 1245 96.1 Faint FtsZ] rings in young EMS. N-tcmu'nal tissue. midplastid FtsZZ of H9 rings and filaments; rcduccd Ftle protein level W's-2 Wild type — 55 — 99.2 Midplastid FtsZ rings - G3bbA Heterogeneous, one 23 —'i 98.2 Midplastid FtsZ filaments L'ntaggcd T-DNA, greatly enlarged with with some rings in smaller between S9 and 510 numerous smaller chloroplasts chloroplasts A404 Slightly enlarged 22 15—29 95.9 Truncated Ftle protein T-DNA inscrtion, with reduced protein level; 30 rcsiduc truncation midplastid FtsZ rings and filaments " Mean number of chloroplasts in 100 cells counted through a single median focal plane. ”The minimum and maximum number of chloroplasts detected in any one cell. “Percentage of m'ldtnx seeds. The difference from 100% roprcscnts aborted or abnormal seeds. Immunoblot analysis of protein extracts from mature leaf tissue of homozygous atftsZI-I-AI individuals was pcrfomrcd using AtFtle-l-Spccific antibodies (Stokes ct al. 2000) (Fig. 113). AtFtle-l protein was not detected in the mutant (Fig. 1E. lane 2'), demonstrating that arfrsZI-I-AI is a null allele ofArFrsZI-I. consistcnt with thc sitc of T—DNA inscrtion (Fig. 1A). Thcrc was no detectable different): in the AtFtsZZ-l protein level in this mutant when comparcd with the wild type (Fig. 1B, lanes 1 and 2). These data show that AtFtle-l is not essential for the survival of Arabidopsis or the propagation of plastids. The null allele of FtsZ} also provided a basis for phenotypic comparisons among the other F1521 mutant alleles described below and served as a negative control for Z filament morphology and immunoblot assays. Miriam alleles of AtFtsZ 1-1 In addition to arfrsZI-I—AI, we identified five new] mutant alleles of AtFtsZI-I from several mutant populations, {our recessive and one semi-dominant. The positions of mutations in the predicted structures of FtsZ1 and the ArFtsZI-I gene are shown (Figs. 2A, B. and 3:). Chloroplast phenotypes and FtsZ morphologies in the mutants were compared with those in their respective wild types—Col—O for azftsZI -I-A I, alflle-I (‘1)159N). arft.rZI-1{626?R) and atflsZI- IrR298Q) (Figs. 4A—M, Q, L', and 5A-—M, Q. L7), and Wassiljcvskija (W's-2) for arfrsZI-I ("03:66:11) and agfzszr—1m4m—433) (Figs. 4N-P, R—T, v—x, and 531- P, R-T, V—X). All phenotypic data have been compilcd in Table l. 208 Mutations in 9.le aflect chloroplast division 779 C-tornunus H7 G26? T7 loop N-torminus Fig. 3 Predicted thee-dimensmnal structure of Atltle—l based upon M. jannaschr'l' ruz protein. Nature 359. 251—254. RayChaudhuri. D. and Park. 151‘. (1994) A point mutation converts Escherichia coli" F tsZ septation GTPase to an ATPase. J. Biol Chem 269; 22941—22944. Redick. S.D.. Stricker. 3.. Briscoe, G. and Erickson. HP. (2005) Mutants of FtsZ targeting the protofilament interface; effects on cell tin'ision and GTPase activity. J. Bacteriol. 187. 2727—2736. 220 Mutations in Ftle affect chloroplast. division 791 Rensing, 3A., Knesslmg. 1.. Resld. R. and Decker, E L. (2.004) Diversification of ftsZ during, early land plant evolution. J Moi. Evol. 58: 154462. Rutherford. S.M. (1996) The genetic and physmal analysis of mutants of chloroplast number and size in Arubidc’wfitj‘ thaliana. DeparUnenl of Biology. University of York. Schcfiers. DJ.. dc Wit. 1.0.. den Blaauwm. T. and Driessen, A}. (2002) GTP hydrolysis of cell drvismn protein F tsZ: evraence that the active site is formed by the association of monomers. erhemu'm' 41. 521-529. Schwedc. T., Kopp, 1.. Guex, N. and Peitsch, MC. (2003) SWISS- MODEL; an automated protein homology-modeling server. Nucieic Acids Res. 31; 3381—3335. Stokes. K.D., McAndrew, R.S.. Figueroa, K, Vitha, S. and Osteryoung. KW. (2000) Chloroplast (iivmon and morphology are differentially affected by overexpresmon of Poll and H.112 genes in Arabm’upyts. Plant Physiol. I24. lbbS—IG'FT’. Stokes. KI). and Osteryoung. KW. (2003) Fatty divergence of the Fle and FtsZZ plastid civision gene families in photosynttenc eukan'otes. Gene 320; 97-103. Suman. M.R.. Amasino. R.M., Young, J.C.. Krysan, PJ. and Austin-Phillips. S. (2000) The Arabidomsis knockout facility at the Umvemty of Wiscormn-Madison. Han: Physiol. 124; 1465—1467. Vaughan. 5.. Wickstead, B., Gull. K. and Addinall. 5.6. (2004'! Mclecular evolution of FtsZ protein sequences erurded within the genomes of archaea. baaer‘n, and eukaryota. J. Mol. Evol. 58: 19—29. Vitha. 5.. Froehlich J.E., Kokstarova. 0.. Pyke, K}... van Erp. H. and Osteryoung. KW. {2003) ARC6 is a J-doma'm plastid dJViswn protein and evolutionary descendant of the cyanobacterial cell drvmon protein Ftn2. PImt Cell 15. 1918-1933. Vitha, 8., McAndrew. RS. and Osteryoung, KW. (2001‘! FtsZ ring formation at the chloroplast divisron site in plants. 1. Cell 3101'. 1:3. Ill-119. Wang, 0.. Kong, D.. Wang. Y., Hu, Y.. He. Y. and Sun. J. (2003'; Isolation of two plastid division flsZ genes from Chlamm'omona: minim-dd: and its evolutionary implication for the role of FtsZ in plastid civmon J. Exp. Bot. 54: lllS-l I16- Wang. X.. Huang, 1., Mukherjee, A.. Caro. C. and Lutkenhaus. I, (1997) Analyst. of the interaction of FtsZ with itself, GTP. and FtsA. J. Bacteriol. 179; 5551—5559. (Received January 4, 2007; Accepted April 17. 2007') 221 REFERENCES Addinall, S.G., Bi, E. and Lutkenhaus, J. (1996) FtsZ ring formation in fts mutants. J. Bacteriol. 178: 3877-3884. Addinall, 5.6. and Lutkenhaus, J. (1996) FtsA is localized to the septum in an FtsZ- dependent manner. J. Bacteriol. 178: 7167-7172. Ahmad, FJ., Yu, W., McNally, F.J. and Baas, P.W. (1999) An essential role for katanin in severing microtubules in the neuron. J. Cell Biol. 145: 305-315. Aldridge, C., Maple, J. and Moller, 5.6. (2005) The molecular biology of plastid division in higher plants. J. Exp. Bot. 56: 1061-1077. Alonso, J.M., Stepanova, A.N., Leisse, T.J., Kim, C.J., Chen, H., et al. (2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 301: 65 3-657. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, DJ. (1990) Basic local alignment search tool. J. Mol. Biol. 215: 403-410. Anand, S., Rajeswari, H., Gupta, P., Srinivasan, R., lndi, S. and Ajitkumar, P. (2004) A C- terminal deletion mutant of Mycobacterium tuberculosis FtsZ shows fast polymerization in vitro. 150: 1119-1121. Anderson, D.E., Gueiros-Filho, H. and Erickson, H.P. (2004) Assembly dynamics of FtsZ rings in Bacillus subtilis and Escherichia coli and effects of FtsZ-regulating proteins. J. Bacteriol. 186: 5775-5781. Anderson, DJ. and Blobel, G. (1983) Immunoprecipitation of proteins from cell-free translations. Methods Enzymol. 96: 111-120. Andreu, J.M., Oliva, MA. and Monasterio, O. (2002) Reversible unfolding of FtsZ cell division proteins from archaea and bacteria. Comparison with eukaryotic tubulin folding and assembly. J. Biol. Chem. 277: 43262-43270. Archer, E.K. and Keegstra, K. (1993) Analysis of chloroplast transit peptide function using mutations in the carboxyl-terminal region. Plant Mol. Biol. 23: 1105-1115. Arnon, DJ. (1949) Copper enzymes in isolated chloroplasts. Polyphenoloxidase in Beta Vulgaris. Plant Physiol. 24: 1-15. 222 Barak, I. and Wilkinson, AJ. (2007) Division site recognition in Escherichia coli and Bacillus subtilis. FEMS Microbiol. Rev. 31: 311-326. Belmont, LD. and Mitchison, TJ. (1996) Identification of a protein that interacts with tubulin dimers and increases the catastrophe rate of microtubules. Cell 84: 623-631. Bernhardt, T.G. and de Boer, P.A.J. (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. and Lutkenha us, J. (19903) FtsZ regulates frequency of cell division in Escherichia coli. J. Bacteriol. 172: 2765-2768. Bi, E. and Lutkenhaus, J. (1990b) Interaction between the min locus and ftsZ. J. Bacteriol. 172: 5610-5616. Bi, E. and Lutkenhaus, J. (1991) FtsZ ring structure associated with division in Escherichia coli. Nature 354: 161-164. Bi, E. and Lutkenhaus, J. (1993) Cell division inhibitors SulA and MinCD prevent formation of the FtsZ ring. J. Bacteriol. 175: 1118-1125. Bleazard, W., McCaffery, J.M., King, E.J., Bale, S., Mozdy, A., Tieu, Q., Nunnari, J. and Shaw, J.M. (1999) The dynamin-related GTPase Dnm1 regulates mitochondrial fission in yeast. Nat. Cell Biol. 1: 298-304. Borhani, D.W. and White, EL. (2004) Polymerization of C-terminally truncated Mycobacterium tuberculosis FtsZ is unlikely to be physiologically relevant. Microbiology 150: 3903-3906. Bork, P., Sander, C. and Valencia, A. (1992) An ATPase domain common to prokaryotic cell cycle proteins, sugar kinases, actin, and hsp70 heat shock proteins. Proc. Natl. Acad. Sci. U. S. A. 89: 7290-7294. Bruce, B.D. (2000) Chloroplast transit peptides: structure, function and evolution. Trends in Cell Biology 10: 440-447. Bruce, B.D., Perry, 5., Froehlich, J. and Keegstra, K. (1994) In vitro import of proteins into chloroplasts. In Plant Molecular Biology Manual J1. Edited by Gelvin, 5.8. and Schilperoort, R.B. pp. 1-15. Kluwer Academic Publishers, Boston. Caplan, M. and Erickson, H.P. (2003) Apparent cooperative assembly of the bacterial cell-division protein FtsZ demonstrated by isothermal titration calorimetry. J. Biol. Chem. 278: 13784-13784. 223 Caplow, M. and Fee, L. (2002) Dissociation of the tubulin dimer is extremely slow, thermodynamically very unfavorable, and reversible in the absence of an energy source. Mol. Biol. Cell 13: 2120-2131. Catarecha, P., Segura, M.D., Franco-Zorrilla, J.M., Garcia-Ponce, B., Lanza, M., Solano, R., Paz-Ares, J. and Leyva, A. (2007) A mutant of the Arabidopsis phosphate transporter PHT1;1 displays enhanced arsenic accumulation. Plant Cell 19: 1123-1133. Chau, M.F., Radeke, M.J., de Ines, C., Barasoain, l., Kohlstaedt, LA. and Feinstein, SC (1998) The microtubule-associated protein tau cross-links to two distinct sites on each alpha and beta tubulin monomer via separate domains. Biochemistry 37: 17692-17703. Chen, Y., Anderson, D.E., Rajagopalan, M. and Erickson, H.P. (2007) Assembly dynamics of Mycobacterium tuberculosis FtsZ. J. Biol. Chem. 282: 27736-27743. Chen, Y., Bjornson, K., Redick, SD. and Erickson, H.P. (2005) A rapid fluorescence assay for FtsZ assembly indicates cooperative assembly with a dimer nucleus. Biophys. J. 88: 505-514. Chen, Y. and Erickson, H.P. (2005) Rapid in vitro assembly dynamics and subunit turnover of FtsZ demonstrated by fluorescence resonance energy transfer. J. Biol. Chem: M500895200. Chen, Y. and Erickson, H.P. (2008) In vitro assembly studies of FtsZ/Tubulin-like proteins (TubZ) from Bacillus Plasmids: Evidence for a capping mechanism. J. Biol. Chem. 283: 8102-8109. Chomczynski, P., Mackey, K., Drews, R. and Wilfinger, W. (1997) DNAzol: a reagent for the rapid isolation of genomic DNA. BioTechniques 22: 550-553. Chung, K.-M., Hsu, H.-H., Yeh, H.-Y. and Chang, B.-Y. (2007) Mechanism of regulation of prokaryotic tubulin-like GTPase FtsZ by membrane protein Eer. J. Biol. Chem. 282: 14891-14897. Cleveland, D.W. (1982) Treadmilling of tubulin and actin. Cell 28: 689-691. Cline, K., Werner-Washburne, M., Andrews, J. and Keegstra, K. (1984) Thermolysin is a suitable protease for probing the surface on intact pea chloroplasts. Plant Physiol. 75: 675-678. Clos, J. and Brandau, S. (1994) pJC20 and pJC40-two high-copy—number vectors for T7 RNA polymerase- dependent expression of recombinant genes in Escherichia coli. Protein Expr. Purif. 5: 133-137. 224 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—determining factor MinD mediates placement of the chloroplast division apparatus. Curr. Biol. 10: 507-516. Cordell, S.C., Anderson, RE. and Lowe, J. (2001) Crystal structure of the bacterial cell division inhibitor MinC. EMBO J. 20: 2454-2461. Dai, K. and Lutkenhaus, J. (1992) The proper ratio of FtsZ to FtsA is required for cell division to occur in Escherichia coli. J. Bacteriol. 174: 6145-6151. David-Pfeuty, T., Erickson, H.P. and Pantaloni, D. (1977) Guanosinetriphosphatase activity of tubulin associated with microtubule assembly. Proceedings of the National Academy of Sciences 74: 5372-5376. de Boer, P., Crossley, R. and Rothfield, L. (19923) The essential bacterial cell-division protein FtsZ is a GTPase. Nature 359: 254-256. de Boer, P.A., Crossley, RE. and Rothfield, L.|. (1989a) A division inhibitor and a topological specificity factor coded for by the minicell locus determine proper placement of the division septum in E. coli. Cell 56: 641-649. de Boer, P.A., Crossley, RE. and Rothfield, L.l. (1992b) Roles of MinC and MinD in the site-specific septation block mediated by the MinCDE system of Escherichia coli. J. Bacteriol. 174: 63-70. de Boer, P.A.J., Crossley, RE. and Rothfield, L.|. (1989b) A division inhibitor and a topological specificity factor coded for by the minicell locus determine proper placement of the division septum in E. coli. Cell 56: 641-649. den Blaauwen, T., Aarsman, M.E.G., Vischer, NE. and Nanninga, N. (2003) Penicillin- binding protein PBP2 of Escherichia coli localizes preferentially in the lateral wall and at mid-cell in comparison with the old cell pole. Mol. Microbiol. 47: 539-547. Derry, W.B., Wilson, L. and Jordan, MA. (1998) Low potency of taxol at microtubule minus ends: implications for its antimitotic and therapeutic mechanism. Cancer Res. 58: 1177-1184. Dewar, S.J., Begg, KJ. and Donachie, W.D. (1992) Inhibition of cell division initiation by an imbalance in the ratio of FtsA to FtsZ. J. Bacteriol. 174: 6314-6316. Diaz, J.F., Kralicek, A., Mingorance, J., Palacios, J.M., Vicente, M. and Andreu, J.M. (2001) Activation of cell division protein FtsZ. Control of switch loop T3 conformation by the nucleotide gamma-phosphate. J. Biol. Chem. 276: 17307-17315. 225 Dougherty, C.A., Himes, R.H., Wilson, L. and Farrell, K.W. (1998) Detection of GTP and Pi in wild-type and mutated yeast microtubules: implications for the role of the GTP/GDP- Pi cap in microtubule dynamics. Biochemistry 37: 10861-10865. Downing, K.H. and Nogales, E. (1998a) New insights into microtubule structure and function from the atomic model of tubulin. Eur Biophst 27: 431-436. Downing, K.H. and Nogales, E. (1998b) Tubulin and microtubule structure. Curr. Opin. Cell Biol. 10: 16-22. Downing, K.H. and Nogales, E. (1998c) Tubulin structure: insights into microtubule prOperties and functions. Curr. Opin. Struct. Biol. 8: 785-791. Du, Y. and Arvidson, CG. (2003) Identification of ZipA, a signal recognition particle- dependent protein from Neisseria gonorrhoeae. J. Bacteriol. 185: 2122-2130. Dutcher, S.K. (2001) The tubulin fraternity: alpha to eta. Curr. Opin. Cell Biol. 13:49-54. Edwards, D.H., Thomaides, H.B. and Errington, J. (2000) Promiscuous targeting of Bacillus subtilis cell division protein DileA to division sites in Escherichia coli and fission yeast. EMBO J. 19: 2719-2727. El-Kafafi, E.S., Mukherjee, S., El-Shami, M., Putaux, J.L., Block, M.A., Pignot-Paintrand, l., Lerbs-Mache, S. and Falconet, D. (2005) The plastid division proteins, Ftle and FtsZ2, differ in their biochemical properties and sub-plastidial localisation. Biochem. J. 387: 669-676. El-Kafafi, S., Karamoko, M., Pignot-Paintrand, I., Grunwald, D., Mandaron, P., Lerbs- Mache, S. and Falconet, D. (2008) Developmentally regulated association of plastid division protein Ftle with thylakoid membranes in Arabidopsis thaliana. Biochem. J. 409: 87-94. Emanuelsson, 0., 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. (1995) FtsZ, a prokaryotic homolog of tubulin? Cell 80: 367-370. Erickson, H.P. (1997) FtsZ, a tubulin homologue in prokaryote cell division. Trends Cell Biol 7: 362-367. Erickson, H.P. and Stoffler, D. (1996) Protofilaments and rings, two conformations of the tubulin family conserved from bacterial FtsZ to alpha/ beta and gamma tubulin. J. Cell Biol. 135: 5-8. 226 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 USA 93: 519-523. Fisher, CL. and Pei, GK. (1997) Modification of a PCR-based site-directed mutagenesis method. BioTechniques 23: 570-571, 574. Fitzpatrick, L.M. and Keegstra, K. (2001) A method for isolating a high yield of Ara bidopsis chloroplasts capable of efficient import of precursor proteins. PlantJ 27: 59- 65. Fribourg, S. and Conti, E. (2003) Structural similarity in the absence of sequence homology of the messenger RNA export factors Mtr2 and p15. EMBO Rep. 4: 699-703. Fu, X., Shih, Y.L., Zhang, Y. and Rothfield, LL (2001) The MinE ring required for proper placement of the division site is a mobile structure that changes its cellular location during the Escherichia coli division cycle. Proc. Natl. Acad. Sci. U. S. A. 98: 980-985. Fujiwara, M. and Yoshida, S. (2001) Chloroplast targeting of chloroplast division FtsZZ proteins in Arabidopsis. Biochem. Biophys. Res. Commun. 287: 462-467. Fujiwara, M.T., Nakamura, A., ltoh, R., Shimada, Y., Yoshida, S. and Moller, S.G. (2004) Chloroplast division site placement requires dimerization of the ARC11/AtMinD1 protein in Arabidopsis. J. Cell Sci. 117: 2399-2410. Gaikwad, A., Babbarwal, V., Pant, V. and Mukherjee, S.K. (2000) Pea chloroplast FtsZ can form multimers and correct the thermosensitive defect of an Escherichia coli ftsZ mutant. Mol Gen Genet 263: 213-221. Gao, H., Kadirjan-Kalbach, D., Froehlich, J.E. and Osteryoung, K.W. (2003) ARCS, a cytosolic dynamin-like protein from plants, is part of the chloroplast division machinery. Proc. Natl. Acad. Sci. U. S. A. 100: 4328-4333. 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. Geissler, S., Siegers, K. and Schiebel, E. (1998) A novel protein complex promoting formation of functional alpha- and gamma-tubulin. EMBO J. 17: 952-966. Gerstein, M. and Levitt, M. (1997) A structural census of the current population of protein sequences. Proc. Natl. Acad. Sci. U. S. A. 94: 11911-11916. 227 Gish, W. and States, DJ. (1993) Identification of protein coding regions by database similarity search. Nat. Genet. 3: 266-272. Glynn, J.M., Miyagishima, S., Yoder, D.W., Osteryoung, K.W. and Vitha, S. (2007) Chloroplast division. Trafiic 8: 451-461. Gonzalez, J., Velez, M., Jimenez, M., Alfonso, C., Schuck, P., Mingorance, J., Vicente, M., Minton, A. and Rivas, G. (2005) Cooperative behavior of Escherichia coli cell-division protein FtsZ assembly involves the preferential cyclization of long single-stranded fibrils. 102: 1895-1900. Gonzalez, J.M., Jimenez, M., Velez, M., Mingorance, J., Andreu, J.M., Vicente, M. and Rivas, G. (2003) Essential cell division protein FtsZ assembles into one monomer-thick ribbons under conditions resembling the crowded intracellular environment. J. Biol. Chem. 278: 37664-37671. Gray, M.W. (1993) The endosymbiont hypothesis revisited. Int. Rev. Cytol 141: 233-357. Gueiros-Filho, FJ. and Losick, R. (2002) A widely conserved bacterial cell division protein that promotes assembly of the tubulin-like protein FtsZ. Genes Dev. 16: 2544-2556. Guex, N. and Peitsch, M.C. (1997) SWISS-MODEL and the Swiss-deViewer: an environment for comparative protein modeling. Electrophoresis 18: 2714-2723. Hale, CA. and de Boer, PA. (1999) Recruitment of ZipA to the septal ring of Escherichia coli is dependent on FtsZ and independent of FtsA. J. Bacteriol. 181: 167-176. Hale, CA. and de Boer, P.A.J. (1997) Direct binding of FtsZ to ZipA, an essential component of the septal ring structure that mediates cell division in E. coli. Cell 88: 175- 185. Hale, C.A., Meinhardt, H. and de Boer, PA. (2001) Dynamic localization cycle of the cell division regulator MinE in Escherichia coli. EMBO J. 20: 1563-1572. Haney, S.A., Glasfeld, E., Hale, C., Keeney, D., He, 2. and de Boer, P. (2001) Genetic analysis of the Escherichia coli FtsZ-ZipA interaction in the yeast ywo-hybrid system. Characterization of FtsZ residues essential for the interactions with ZipA and with FtsA. J. Biol. Chem. 276: 11980-11987. Harry, E.J., Rodwell, J. and Wake, R.G. (1999) Co-ordinating DNA replication with cell division in bacteria: 3 link between the early stages of a round of replication and mid-cell 2 ring assembly. Mol. Microbiol. 33: 33-40. 228 Hartman, J.J., Mahr, J., McNally, K., Okawa, K., lwamatsu, A., Thomas, S., Cheesman, S., Heuser, J., Vale, RD. and McNally, F.J. (1998) Katanin, a microtubule-severing protein, is a novel AAA ATPase that targets to the centrosome using a WD40-containing subunit. Cell 93: 277-287. Higashitani, A., lshii, Y., Kato, Y. and Koriuchi, K. (1997) Functional dissection of a cell- division inhibitor, SulA, of Escherichia coli and its negative regulation by Lon. Mol Gen Genet 254: 351-357. Hirata, D., Masuda, H., Eddison, M. and Toda, T. (1998) Essential role of tubulin-folding cofactor D in microtubule assembly and its association with microtubules in fission yeast. EMBO J. 17: 658-666. Howell, B., Larsson, N., Gullberg, M. and Cassimeris, L. (1999) Dissociation of the tubulin-seq uestering and microtubule catastrophe-promoting activities of oncoprotein 18/stathmin. Mol. Biol. Cell 10: 105-118. 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, 2. and Lutkenhaus, J. (2000) Analysis of MinC reveals two independent domains involved in interaction with MinD and FtsZ. J. Bacteriol. 182: 3965-3971. 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 FtsZ to prevent polymerization. Proc. Natl. Acad. Sci. U. S. A. 96: 14819-14824. Huang, J., Cao, C. and Lutkenhaus, J. (1996) Interaction between FtsZ and inhibitors of cell division. J. Bacteriol. 178: 5080-5085. Huecas, S. and Andreu, J. (2003) Energetics of the cooperative assembly of cell division protein FtsZ and the nucleotide hydrolysis switch.: 46146-46154. Huecas, S. and Andreu, J. (2004) Polymerization of nucleotide-free, GDP- and GTP-bound cell division protein FtsZ: GDP makes the difference. FEBS Lett. 569: 43-48. Huecas, S., Llorca, 0., Boskovic, J., Martin-Benito, J., Valpuesta, J.M. and Andreu, J.M. (2007a) Energetics and geometry of FtsZ polymers: nucleated self-assembly of single protofilaments. Biophys. J.: biophysj.107.115493. 229 Huecas, S., Schaffner-Barbero, C., Garcia, W., Yebenes, H., Palacios, J.M., Diaz, J.F., Menendez, M. and Andreu, J.M. (2007b) The interactions of cell division protein FtsZ with guanine nucleotides. J. Biol. Chem. 282: 37515-37528. lngerman, E. and Nunnari, J. (2005) A continuous, regenerative coupled GTPase assay for dynamin-related proteins. Methods Enzymol. 404: 611-619. lngerman, E., Perkins, E.M., Marina, M., Mears, J.A., McCaffery, J.M., Hinshaw, J.E. and Nunnari, J. (2005) Dnm1 forms spirals that are structurally tailored to fit mitochondria. J. Cell Biol. 170: 1021-1027. ltoh, 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. Jeong, K.J. and Lee, S.Y. (2003) Enhanced production of recombinant proteins in Escherichia coli by filamentation suppression. Appl. Environ. Microbial. 69: 1295-1298. Justice, S.S., Garcia-Lara, J. and Rothfield, LL (2000) Cell division inhibitors SulA and MinC/MinD block septum formation at different steps in the assembly of the Escherichia coli division machinery. Mal. Microbial. 37:410-423. Katayama, N., Takano, H., Sugiyama, M., Takio, S., Sakai, A., Tanaka, K., Kuroiwa, H. and Ono, K. (2003) Effects of Antibiotics that Inhibit the Bacterial Peptidoglycan Synthesis Pathway on Mass Chloroplast Division. Plant Cell Physiol. 44: 776-781. Koksharova, CA. and Walk, C.P. (2002) A novel gene that bears a DnaJ motif influences cyanobacterial cell division. J. Bacterial. 184: 5524-5528. Kollman, J.M., Zelter, A., Muller, E.G.D., Fox, B., Rice, L.M., Davis, TN. and Agard, DA. (2008) The structure of the 'y-tubulin small complex: implications of its architecture and flexibility for microtubule nucleation. Mal. Biol. Cell 19: 207-215. Krause, G.H. and Heber, U. (1976) Topics in Photosynthesis. Edited by Barber, J. pp. 171- 214. Elsevier. Kuroiwa, H., Mori, T., Takahara, M., Miyagishima, S.Y. and Kuroiwa, T. (2002) Chloroplast division machinery as revealed by immunofluorescence and electron microscopy. Planta 215: 185-190. Laemmli, U.K. (1970) Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature 227: 680-685. 230 Larsen, R.A., Cusumano, C., Fujioka, A., Lim-Fong, G., Patterson, P. and Pogliano, J. (2007) Treadmilling of a prokaryotic tubulin-like protein, TubZ, required for plasmid stability in Bacillus thuringiensis. Genes Dev. 21: 1340-1352. Leech, R.M., Thomson, W.W. and PIatt-Aloia, K.A. (1981) Observations on the mechanism of chloroplast division in higher plants. New Phytol. 87: 1-9. Levin, P.A., Kurtser, LG. and Grossman, AD. (1999) Identification and characterization of a negative regulator of FtsZ ring formation in Bacillus subtilis. Proc. Natl. Acad. Sci. U. S. A. 96: 9642-9647. Levin, P.A., Schwartz, R.L. and Grossman, AD. (2001) Polymer stability plays an important role in the positional regulation of FtsZ. J. Bacterial. 183: 5449-5452. Lewis, S.A., Tian, G. and Cowan, NJ. (1997) The (x- and B-tubulin folding pathways. Trends in Cell Biology 7: 479-484. Li, H., DeRosier, D.J., Nicholson, W.V., Nogales, E. and Downing, K.H. (2002) Microtubule structure at 8 A resolution. Structure (Camb) 10: 1317-1328. Lin, D.C., Levin, PA. and Grossman, AD. (1997) Bipolar localization of a chromosome partition protein in Bacillus subtilis. Proc. Natl. Acad. Sci. U. S. A. 94: 4721-4726. Liu, Z., Mukherjee, A. and Lutkenhaus, J. (1999) Recruitment of ZipA to the division site by interaction with FtsZ. Mal. Microbial. 31: 1853-1861. Lohret, T.A., McNally, FJ. and Quarmby, L.M. (1998) A role for katanin-mediated axonemal severing during Chlamydomonas deflagellation. Mol. Biol. Cell 9: 1195-1207. Lohse, S., Hause, B., Hause, G. and Fester, T. (2006) FtsZ characterization and immunolocalization in the two phases of plastid reorganization in arbuscular mycorrhizal roots of Medicago truncatula. Plant Cell Physiol. 47: 1124-1134. Lowe, J. (1998) Crystal structure determination of FtsZ from Methanocaccus jannaschii. J Struct Biol 124: 235-243. Lowe, J. and Amos, LA. (1998) Crystal structure of the bacterial cell-division protein FtsZ. Nature 391: 203-206. Lowe, J. and Amos, LA. (1999) Tubulin-like protofilaments in CaZ+-induced FtsZ sheets. EMBO J. 18: 2364-2371. Lu, C. and Erickson, H.P. (1998) Purification and assembly of FtsZ. Methods Enzymol. 298: 305-313. 231 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, C., Stricker, .I. and Erickson, H.P. (1998) FtsZ from Escherichia coli, Azatobacter vinelandii, and Thermataga maritima—q uantitation, GTP hydrolysis, and assembly. Cell Motility And The Cytaskeletan 40: 71-86. Lu, C., Stricker, J. and Erickson, H.P. (2001) Site-specific mutations of FtsZ--effects on GTPase and in vitro assembly. BMC Microbial 1: 7. Lutkenha us, J. (2007) Assembly dynamics of the bacterial MinCDE system and spatial regulation of the 2 ring. Annu. Rev. Biochem. 76: 539-562. Ma, X., Ehrhardt, D.W. and Margolin, W. (1996) Colocalization of cell division proteins FtsZ and FtsA to cytoskeletal structures in living Escherichia coli cells by using green fluorescent protein. Proc. Natl. Acad. Sci. U. S. A. 93: 12998-13003. Machida, M., Takechi, K., Sato, H., Chung, S.J., Kuroiwa, H., Takio, S., Seki, M., Shinoza ki, K., Fujita, T., Hasebe, M. and Takano, H. (2006) Genes for the peptidoglycan synthesis pathway are essential for chloroplast division in mass. Proceedings of the National Academy of Sciences 103: 675 3-6758. Maple, J., Aldridge, C. and Moller, S.G. (2005) Plastid division is mediated by combinatorial assembly of plastid division proteins. The PlantJaurnal 43: 811-823. Maple, J., Chua, N.H. and Moller, S.G. (2002) The topological specificity factor AtMinE1 is essential for correct plastid division site placement in Arabidopsis. PlantJ 31: 269-277. Maple, J., Fujiwara, M., Kitahata, N., Lawson, T., Baker, N., Yoshida, S. and Moller, S. (2004) GIANT CHLOROPLAST 1 is essential for correct plastid division in Arabidopsis. Curr. Biol. 14: 776-781. Maple, J. and Moller, S.G. (2007a) Plastid division coordination across a double- membraned structure. FEBS Lett. Maple, J. and Moller, S.G. (2007b) Plastid division: evolution, mechanism and complexity. Ann Bot (Land) 99: 565-579. Maple, J., Vojta, L., Soll, J. and Moller, S.G. (2007) ARC3 is a stromal Z-ring accessory protein essential for plastid division. EMBO Rep. 8: 293-299. Margolin, W. (2000) Themes and variations in prokaryotic cell division. FEMS Microbiol. Rev. 24: 531-548. 232 Margolin, W. (2002) Bacterial sporulation: FtsZ rings do the twist. Curr. Biol. 12: R391- 392. Margolin, W. (2005a) Bacterial mitosis: actin in a new role at the origin. Curr. Biol. 15: R259-261. Margolin, W. (2005b) FtsZ and the division of prokaryotic cells and organelles. Nat. Rev. Mol. Cell Biol. 6: 862-871. Margolin, W. (2007) Bacterial cytoskeleton: not your run-of-the-mill tubulin. Curr. Biol. 17: R633-636. Margolis, R.L. and Wilson, L. (1998) Microtubule treadmilling: what goes around comes around. BiaEssays 20: 830-836. Margulis, L. (1971) The origin of plant and animal cells. Am Sci 59: 230-235. Margulis, L. (1975) Symbiotic theory of the origin of eukaryotic organelles; criteria for proof. Symp Soc Exp Biol: 21-38. 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. Mal. Microbial. 33: 84-96. Marston, A.L., Thomaides, H.B., Edwards, D.H., Sharpe, M.E. and Errington, J. (1998) Polar localization of the MinD protein of Bacillus subtilis and its role in selection of the mid-cell division site. Genes Dev. 12: 3419-3430. Matsuo, Y., Yamada, A., Tsukamoto, K., Tamura, H.O., lkeawa, H., Nakamura, H. and Nishikawa, K. (1996) A distant evolutionary relationship between bacterial sphingomyelinase and mammalian DNase I. Protein Sci. 5: 2459-2467. Mazouni, K., Domain, F., Cassier-Chauvat, C. and Chauvat, F. (2004) Molecular analysis of the key cytokinetic components of cyanobacteria: FtsZ, ZipN and MinCDE. Mol. Microbiol. 52: 1145-1158. McAndrew, R.S., Froehlich, 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 Ftle and FtsZZ 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 viva quantitative relationship between plastid 233 division proteins FtsZ1 and FtsZZ and identification of ARC6 and ARC3 in a native FtsZ complex. Biochem. J. In Press. McNally, F.J., Okawa, K., lwamatsu, A. and Vale, RD. (1996) Katanin, the microtubule- severing ATPase, is concentrated at centrosomes. J. Cell Sci. 109 ( Pt 3): 561-567. McNally, F.J. and Thomas, S. (1998) Katanin is responsible for the M-phase microtubule- severing activity in Xenopus eggs. Mal. Biol. Cell 9: 1847-1861. Melki, R., Fievez, S. and Carlier, M.F. (1996) Continuous monitoring of Pi release following nucleotide hydrolysis in actin or tubulin assembly using 2-amino-6-mercapto- 7-methylpurine ribonucleoside and purine-nucleoside phosphorylase as an enzyme- linked assay. Biochemistry 35: 12038-12045. Menendez, M., Rivas, G., Diaz, J.F. and Andreu, J.M. (1998) Control of the structural stability of the tubulin dimer by one high affinity bound magnesium ion at nucleotide N- site. J. Biol. Chem. 273: 167-176. Michie, K. and Lowe, J. (2006) Dynamic filaments of the bacterial cytoskeleton. Annu. Rev. Biochem. 75: 467-492. Mingorance, J., Rueda, S., Gomez-Puertas, P., Valencia, A. and Vicente, M. (2001) Escherichia coli FtsZ polymers contain mostly GTP and have a high nucleotide turnover. Mal. Microbial. 41: 83-91. Miroux, B. and Walker, J.E. (1996) Over-production of proteins in Escherichia coli: mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J. Mol. Biol. 260: 289-298. Miyagishima, S., Kuroiwa, H. and Kuroiwa, T. (2001a) The timing and manner of disassembly of the apparatuses for chloroplast and mitochondrial division in the red alga Cyanidioschyzon merolae. Planta 212: 517-528. Miyagishima, S., Nishida, K. and Kuroiwa, T. (2003a) An evolutionary puzzle: chloroplast and mitochondrial division rings. Trends Plant Sci. 8: 432-438. Miyagishima, S., Takahara , M. and Kuroiwa, T. (2001b) Novel filaments 5 in diameter constitute the cytosolic ring of the plastid division apparatus. The Plant Cell 13: 707-721. Miyagishima, S., Takahara , M., Mori, T., Kuroiwa, H., Higashiyama, T. and Kuroiwa, T. (2001c) Plastid division is driven by a complex mechanism that involves differential transition of the bacterial and eukaryotic division rings. Plant Cell 13: 2257-2268. 234 Miyagishima, S., Wolk, C. and Osteryoung, K. (2005) Identification of cyanobacterial cell division genes by comparative and mutational analyses. Mal. Microbial. 56: 126-143. Miyagishima, S.Y., Froehlich, J.E. and Osteryoung, K.W. (2006) PDV1 and PDV2 mediate recruitment of the dynamin-related protein ARCS to the plastid division site. Plant Cell 18: 2517-2530. Miyagishima, S.Y., Nishida, K., Mori, T., Matsuzaki, M., Higashiyama, T., Kuroiwa, H. and Kuroiwa, T. (2003b) A plant-specific dynamin-related protein forms a ring at the chloroplast division site. Plant Cell 15: 655-665. Moritz, M., Braunfeld, M.B., Sedat, J.W., Alberts, B. and Agard, DA. (1995) Microtubule nucleation by gamma-tubulin-containing rings in the centrosome. Nature 378: 638-640. Mosyak, L., Zhang, Y., Glasfeld, E., Haney, S., 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. May, F.J., Glasfeld, E., Mosyak, L. and Powers, R. (2000) Solution structure of ZipA, a crucial component of Escherichia coli cell division. Biochemistry 39: 9146-9156. Mukherjee, A., Cao, C. and Lutkenhaus, J. (1998) Inhibition of FtsZ polymerization by SulA, an inhibitor of septation in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 95: 2885- 2890. 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. S. A. 90: 1053-1057. Mukherjee, A. and Lutkenhaus, J. (1994) Guanine nucleotide-dependent assembly of FtsZ into filaments. J. Bacteriol. 176: 2754-2758. Mukherjee, A. and Lutkenha us, J. (1998) Dynamic assembly of FtsZ regulated by GTP hydrolysis. EMBO J. 17: 462-469. Mukherjee, A. and Lutkenhaus, J. (1999) Analysis of FtsZ assembly by light scattering and determination of the role of divalent metal cations. J. Bacterial. 181: 823-832. Musser, SM. and Theg, SM. (2000) Proton transfer limits protein translocation rate by the thylakoid ApH/T at machinery. Biochemistry 39: 8228-8233. Nielsen, E., Akita, M., Davila-Aponte, J. and Keegstra, K. (1997) Stable association of chloroplastic precursors with protein translocation complexes that contain proteins from both envelope membranes and a stromal Hsp100 molecular chaperone. EMBO J. 16: 935-946. 235 Nishida, K., Takahara, M., Miyagishima, S.Y., Kuroiwa, H., Matsuzaki, M. and Kuroiwa, T. (2003) Dynamic recruitment of dynamin for final mitochondrial severance in a primitive red alga. Proc. Natl. Acad. Sci. U. S. A. 100: 2146-2151. Nogales, E. (1999) A structural view of microtubule dynamics. Cell. Mal. Life Sci. 56: 133- 142. Nogales, E. (2000) Structural insights into microtubule function. Annu. Rev. Biochem. 69: 277-302. Nogales, E., Downing, K.H., Amos, LA. and Lowe, J. (19983) Tubulin and FtsZ form a distinct family of GTPases. Nat Struct Biol 5: 451-458. Nogales, E. and Wang, H.-W. (2006) Structural intermediates in microtubule assembly and disassembly: how and why? Curr. Opin. Cell Biol. 18: 179-184. Nogales, E., Whittaker, M., Milligan, RA. and Downing, K.H. (1999) High-resolution model of the microtubule. Cell 96: 79-88. Nogales, E., Wolf, S.G. and Downing, K.H. (1998b) Structure of the alpha beta tubulin dimer by electron crystallography. Nature 391: 199-203. Oliva, M., Cordell, S. and Lowe, J. (2004) Structural insights into FtsZ protofilament formation. Nat. Struct. Mol. Biol. Oliva, M.A., Huecas, S., Palacios, J.M., Martin-Benito, J., Valpuesta, J.M. and Andreu, J.M. (2003) Assembly of archaeal cell division protein FtsZ and a GTPase-inactive mutant into double-stranded filaments. J. Biol. Chem. 278: 33562-33570. Olson, B.J.S.C. and Markwell, JP (2007) Assays for determination of protein concentration. In Current Protocols in Protein Science. Edited by Coligan, J.E., Dunn, B.M., Speicher, D.W. and Wingfield, P.T. pp. 3.4.1-3.4.29. John Wiley 8:. Sons, Inc. Olson, B.J.S.C. and Osteryoung, K.W. (2007) Biochemical analysis of Ftle and FtsZZ. In Plant Biology and Botany 2007 Joint Congress, Chicago, IL. Osawa, M., Anderson, DE. and Erickson, H.P. (2008) Reconstitution of contractile FtsZ rings in liposomes. Science In press. Osteryoung, K.W. (2000) Organelle fission: crossing the evolutionary divide. Plant Physiol. 123: 1213-1216. Osteryoung, K.W. (2001) Organelle fission in eukaryotes. Curr Opin Microbial 4: 639-646. 236 Osteryoung, K.W. and McAndrew, R.S. (2001) The plastid division machine. Annu Rev Plant Physiol Plant Mol Biol 52: 315-333. Osteryoung, K.W. and Pyke, K.A. (1998) Plastid division: evidence for a prokaryotically derived mechanism. Curr Opin Plant Biol 1: 475-479. 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: 473-474. Pichoff, S. and Lutkenhaus, J. (2001) Escherichia coli division inhibitor MinCD blocks septation by preventing Z-ring formation. J. Bacterial. 183: 6630-6635. 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. Pradel, N., Santini, C.-L., Bernadac, A., Fukumori, Y. and Wu, L.-F. (2006) Biogenesis of actin-like bacterial cytoskeletal filaments destined for positioning prokaryotic magnetic organelles. PNAS 103: 17485-17489. 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., Marrison, J.L. and Leech, RM. (1991) Temporal and spatial development of the cells of the expanding first leaf of Arabidopsis thaliana (L.) Heynh. J. Exp. Bot. 42: 1407-1416. Raskin, D.M. and de Boer, P.A. (1997) The MinE ring: an FtsZ-independent cell structure required for selection of the correct division site in E. coli. Cell 91: 685-694. Raskin, D.M. and de Boer, P.A. (1999a) MinDE-dependent pole-to-pole oscillation of division inhibitor MinC in Escherichia coli. J. Bacterial. 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. RayChaudhuri, D. and Park, LT. (1992) Escherichia coli cell-division gene ftsZ encodes a novel GTP-binding protein. Nature 359: 251-254. 237 RayChaudhuri, D. and Park, J.T. (1994) A point mutation converts Escherichia coli FtsZ septation GTPase to an ATPase. J. Biol. Chem. 269: 22941-22944. Raynaud-Messina, B. and Merdes, A. (2007) y-tubulin complexes and microtubule organization. Curr. Opin. Cell Biol. 19: 24-30. Raynaud, C., Cassier-Chauvat, C., Perennes, C. and Bergounioux, C. (20043) An Arabidopsis homolog of the bacterial cell division inhibitor SuIA is involved in plastid division. Plant Cell 16: 1801-1811. Raynaud, C., Cassier-Chauvat, C., Perennes, C. and Bergounioux, C. (2004b) An Arabidopsis homolog of the bacterial cell division inhibitor SulA is involved in plastid division. Plant Cell 16: 1801-1811. Redick, S.D., Stricker, J., Briscoe, G. and Erickson, H.P. (2005) Mutants of FtsZ targeting the protofilament interface: effects on cell division and GTPase activity. J. Bacteriol. 187: 2727-2736. Regamey, A., Harry, E.J. and Wake, R.G. (2000) Mid-cell Z ring assembly in the absence of entry into the elongation phase of the round of replication in bacteria: co-ordinating chromosome replication with cell division. Mal. Microbial. 38: 423-434. Rensing, S.A., Kiessling, J., Reski, R. and Decker, E.L. (2004) Diversification of ftsZ during early land plant evolution. J. Mal. Evol. 58: 154-162. Reski, R. (2002) Rings and networks: the amazing complexity of FtsZ in chloroplasts. Trends Plant Sci. 7: 103-105. Rivas, G., Fernandez, J.A. and Minton, AP (2001) Direct observation of the enhancement of noncooperative protein self- assembly by macromolecular crowding: Indefinite linear self-association of bacterial cell division protein FtsZ. Proc. Natl. Acad. Sci. U. S. A. 98: 3150-3155. Rivas, G., Lopez, A., Mingorance, J., Ferrandiz, M.J., Zorrilla, S., Minton, A.P., Vicente, M. and Andreu, J.M. (2000) Magnesium-induced linear self-association of the FtsZ bacterial cell division protein monomer. The primary steps for FtsZ assembly. J. Biol. Chem. 275: 11740-11749. Robertson, EJ., Rutherford, SM. and Leech, RM. (1996) Characterization of chloroplast division using the Arabidopsis mutant arc5. Plant Physiol. 112: 149-159. Romberg, L. and Levin, P.A. (2003) Assembly dynamics of the bacterial cell division protein FtsZ: poised at the edge of stability. Annu. Rev. Microbial. 57: 125-154. 238 Romberg, L. and Mitchison, TJ. (2004) Rate-limiting guanosine 5'-triphosphate hydrolysis during nucleotide turnover by FtsZ, a prokaryotic tubulin homologue involved in bacterial cell division. Biochemistry 43: 282-288. Romberg, L., Simon, M. and Erickson, H.P. (2001) Polymerization of FtsZ, a bacterial homolog of tubulin. is assembly cooperative? J. Biol. Chem. 276: 11743-11753. Rothfield, L., Justice, S. and Garcia-Lara, J. (1999) Bacterial cell division. Annu. Rev. Genet. 33: 423-448. Rothfield, L., Taghbalout, A. and Shih, Y.-L. (2005) Spatial control of bacterial division- site placement. Nat Rev Micro 3: 959-968. Rowland, S.L., Fu, X., Sayed, M.A., Zhang, Y., Cook, W.R. and Rothfield, LL (2000) Membrane redistribution of the Escherichia coli MinD protein induced by MinE. J. Bacterial. 182: 613-619. Rueda, S., Vicente, M. and Mingorance, J. (2003) Concentration and assembly of the division ring proteins FtsZ, FtsA, and ZipA during the Escherichia coli cell cycle. J. Bacterial. 185: 3344-3351. Sackett, D.L. (1995) Structure and function in the tubulin dimer and the role of the acidic ca rboxyl terminus. Subcell Biochem 24: 255-302. Sackett, D.L., Bhattacharyya, B. and Wolff, J. (1985) Tubulin subunit carboxyl termini determine polymerization efficiency. J. Biol. Chem. 260: 43-45. Santra, M.K. and Panda, D. (2003) Detection of an intermediate during unfolding of bacterial cell division protein FtsZ: loss of functional properties precedes the global unfolding of FtsZ. J. Biol. Chem. 278: 21336-21343. Scheffers, D. and Driessen, A.J. (2001) The polymerization mechanism of the bacterial cell division protein FtsZ. FEBS Lett. 506: 6-10. Scheffers, D.J., de Wit, J.G., den Blaauwen, T. and Driessen, A.J. (2002) GTP hydrolysis of cell division protein FtsZ: evidence that the active site is formed by the association of monomers. Biochemistry 41: 521-529. Scheffers, D.J., den Blaauwen, T. and Driessen, A.J. (2000) Non-hydrolysable GTP- gamma-S stabilizes the FtsZ polymer in a GDP-bound state. Mal. Microbial. 35: 1211- 1219. Scheffers, DJ. and Driessen, A.J. (2002) Immediate GTP hydrolysis upon FtsZ polymerization. Mal. Microbial. 43: 1517-1521. 239 Shih, Y.L., Fu, X., King, G.F., Le, T. and Rothfield, L. (2002) Division site placement in E.cali: mutations that prevent formation of the MinE ring lead to loss of the normal midcell arrest of growth of polar MinD membrane domains. EMBO J. 21: 3347-3357. 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-Terminal tail of FtsZ. Biochemistry 46: 11013-11022. Small, E. and Addinall, S.G. (2003) Dynamic FtsZ polymerization is sensitive to the GTP to GDP ratio and can be maintained at steady state using a GTP-regeneration system. Microbiology 149: 2235-2242. Small, E., Marrington, R., Rodger, A., Scott, DJ., Sloan, K., Roper, D., Dafforn, T.R. and Addinall, S.G. (2007) FtsZ polymer-bundling by the Escherichia coli ZapA orthologue, Yng, involves a conformational change in bound GTP. J. Mol. Biol. 369: 210-221. Sassong, T.M., Jr., Brigham-Burke, M.R., Hensley, P. and Pearce, K.H., Jr. (1999) Self- activation of guanosine triphosphatase activity by oligomerization of the bacterial cell division protein FtsZ. Biochemistry 38: 14843-14850. 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 FtsZZ 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: 97-108. 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. Stricker, J. and Erickson, H.P. (2003) In viva characterization of Escherichia califtsZ mutants: effects on Z-ring structure and function. J. Bacterial. 185: 4796-4805. Stricker, J., Maddox, P., Salmon, ED. and Erickson, H.P. (2002) Rapid assembly dynamics of the Escherichia coli FtsZ-ring demonstrated by fluorescence recovery after photobleaching. Proc. Natl. Acad. Sci. U. S. A. 99: 3171-3175. 240 Swarbreck, D., Wilks, C., Lamesch, P., Berardini, T.Z., Garcia-Hernandez, M., et al. (2007) The Arabidopsis Information Resource (TAIR): gene structure and function annotation. Nucl. Acids Res.: gkm965. Szeto, T.H., Rowland, S.L. and King, G.F. (2001) The dimerization function of MinC resides in a structurally autonomous C-terminal domain. J. Bacterial. 183: 6684-6687. Takada, A., Nagai, K. and Wachi, M. (2005) A decreased level of FtsZ is responsible for inviability of RNase E-deficient cells. Genes Cells 10: 733-741. Tanaka-Takiguchi, Y., ltoh, TJ. and Hotani, H. (1998) Visualization of the GDP-dependent switching in the growth polarity of microtubules. J. Mal. Biol. 280: 365-373. Tian, G., Bhamidipati, A., Cowan, NJ. and Lewis, SA. (1999) Tubulin folding cofactors as GTPase-activating proteins. GTP hydrolysis and the assembly of the alpha/beta-tubulin heterodimer. J. Biol. Chem. 274: 24054-24058. Tian, G., Huang, Y., Rommelaere, H., Vandekerckhove, J., Ampe, C. and Cowan, NJ. (1996) Pathway leading to correctly folded beta-tubulin. Cell 86: 287-296. Tian, G., Lewis, S.A., Feierbach, B., Stearns, T., Rommelaere, H., Ampe, C. and Cowan, NJ. (1997) Tubulin subunits exist in an activated conformational state generated and maintained by protein cofactors. J. Cell Biol. 138: 821-832. Tieu, Q. and Nunnari, J. (2000) Mdv1p is a WD repeat protein that interacts with the dynamin-related GTPase, Dnm1p, to trigger mitochondrial division. J. Cell Biol. 151: 353- 366. Tieu, O., Okreglak, V., Naylor, K. and Nunnari, J. (2002) The WD repeat protein, Mdv1p, functions as a molecular adaptor by interacting with Dnm1p and Fislp during mitochondrial fission. J. Cell Biol. 158: 445-452. van den Ent, F. and Lowe, J. (2000) Crystal structure of the cell division protein FtsA from Thermataga maritima. EMBO J. 19: 5300-5307. Vandecandelaere, A., Brune, M., Webb, M.R., Martin, SR and Bayley, PM. (1999) Phosphate release during microtubule assembly: what stabilizes growing microtubules? Biochemistry 38: 8179-8188. Vaughan, S., Wickstead, B., Gull, K. and Addinall, S.G. (2004) Molecular evolution of FtsZ protein sequences encoded within the genomes of archaea, bacteria, and eukaryota. J. Mol. Eval. 58: 19-29. 241 Vitha, S., Froehlich, J.E., Koksharova, O., Pyke, K.A., van Erp, H. and Osteryoung, K.W. (2003) ARC6 Is a J-Domain plastid pivision protein and an evolutionary descendant of the cyanobacterial cell division protein Ftn2. Plant Cell 15: 1918-1933. Vitha, S., Holzenburg, A. and Osteryoung, K.W. (2005) Assembly dynamics of FtsZ and ARC6 plastid-dividing proteins. Microsc. Micraanal. 11 (Suppl 2): 1154. Vitha, S., McAndrew, RS. and Osteryoung, K.W. (2001) FtsZ ring formation at the chloroplast division site in plants. J. Cell Biol. 153: 111-119. Wang, X., Huang, J., Mukherjee, A., Cao, C. and Lutkenhaus, J. (1997) Analysis of the interaction of FtsZ with itself, GTP, and FtsA. J. Bacteriol. 179: 5551-5559. Wang, X. and Lutkenhaus, J. (1993) The FtsZ protein of Bacillus subtilis is localized to the division site and has a GTPase activity that is dependent upon the FtsZ concentration. Mal. Microbiol. 9: 435-442. Waterman-Storer, CM. and Salmon, ED. (1998) How microtubules get fluorescent speckles. Biophys. J. 75: 2059-2069. Weigel, D. and Glazebrook, J. (2002) Arabidopsis: A laboratory manual. Cold Spring Harbor Press, Cold Spring Harbor, New York. White, E.L., Ross, L.J., Reynolds, R.C., Seitz, L.E., Moore, G.D. and Borhani, D.W. (2000) Slow polymerization of Mycobacterium tuberculosis FtsZ. J. Bacterial. 182: 4028-4034. Woldringh, C.L., Mulder, E., Valkenburg, J.A., Wientjes, F.B., Zaritsky, A. and Nanninga, N. (1990) Role of the nucleoid in the toporegulation of division. Res. Microbial. 141: 39- 49. Xu, Y. and Xu, D. (2000) Protein threading using PROSPECT: design and evaluation. Proteins 40: 343-354. Yan, K., Pearce, K.H. and Payne, DJ. (2000) A conserved residue at the extreme C- terminus of FtsZ is critical for the FtsA-FtsZ interaction in Staphylococcus aureus. Biochem. Biophys. Res. Commun. 270: 387-392. Yoder, D.W., Kadirjan-Kalbach, D., Olson, BJ.S.C., Miyagishima, S.Y., Deblasio, S.L., Hangarter, R.P., Osteryoung, K.W. and Vitha, S. (2007) Effects of mutations in Arabidopsis Ftle on plastid division, FtsZ ring formation and positioning, and FtsZ filament morphology in vivo. Plant Cell Physiol. 48: 775-791. 242 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. Yu, X.C. and Margolin, W. (1997) Ca2+-mediated GTP-dependent dynamic assembly of bacterial cell division protein FtsZ into asters and polymer networks in vitro. EMBO J. 16: 5455-5463. Zheng, Y., Wang, M.L., Alberts, B. and Mitchison, T. (1995) Nucleation of microtubule assembly by a gamma-tubulin-containing ring complex. Nature 378: 578-583. 243 V IIIIIIII IIIIIIIIIIIII 93 02