Dissecting the process of plastid division in higher plants using functional genomics approaches

Plastids arise by division from pre-existing plastids in the cytosol. Reminiscent of their prokaryotic ancestors, chloroplast division occurs via binary fission and due to the inherent similarities, bacterial cell division has been used as a paradigm for chloroplast division. In bacteria, cell division is initiated by FtsZ, a tubulin-like GTPase, which assembles into a cytokinetic ring (Z-ring) at midcell, resulting in the production of two identical daughter cells and correct Z-ring placement depends on the proteins MinC, MinD and MinE. Most plants contain nuclear genes with high similarity to FtsZ, MinE and MinD and the Arabidopsis homologues AtFtsZ1-1, AtFtsZ2-1, AtMinE1, and AtMinD1 all have roles in chloroplast division. For example AtMinE1 (Fig. 1B) or AtMinD1 (Fig. 1C) overexpression in transgenic Arabidopsis results in larger and fewer chloroplasts compared to wild- type (Fig. 1A) whilst AtFtsZ deficiency results in one chloroplast/cell.

It is becoming increasingly clear that the different division proteins do not act in isolation but in protein complexes. We have shown that during chloroplast division AtFtsZ1-1 and AtFtsZ2-1 form homo and heterodimers and that AtFtsZ2-1 interacts specifically with ARC6, a protein similar to the cyanobacterial cell division protein Ftn2. We have demonstrated that AtMinE1 forms homodimers, acting as a topological specificity factor and that AtMinE1 interacts with AtMinD1, which also dimerizes, and that this complex localizes to discrete intraplastidic regions (Fig. 2A). Furthermore, we have shown that AtMinD1 is a Ca²⁺-dependent ATPase and ATP hydrolysis is important for localization and interaction with AtMinE1. Homology searches also led to the identification of GIANT CHLOROPLAST 1 (GC1) which localizes as a dimer to the chloroplast inner envelope. GC1 is essential in that deficiency results in plant cells having 1-2 giant chloroplasts (Fig. 1D).

Although the Arabidopsis division machinery is evolutionary conserved division is controlled by coordinated action of both prokaryote- and eukaryote-derived proteins: ARC5, a dynamin-like protein and ARC3, a FtsZ-PIP5 kinase are of eukaryotic origin and important components during plastid division. Based on our findings to date we have constructed a working model of the molecular machinery that controls plastid division in Arabidopsis thaliana (Fig. 2B).

The principal objective of our research is to dissect the process of plastid division in higher plants and to determine the integral nature of the division process within plant cells employing several functional genomics approaches combining protein-protein interaction studies, microarray and proteomics analysis and reverse genetics. Our sub-goals are four-fold:

1) Identify new plastid division proteins using already characterized plastid division proteins as bait in yeast two-hybrid screens and immunoprecipitation experiments followed by cell biological approaches and reverse genetics to establish their role in the plastid division process.

2) Establish the effect of different plastid division states on nuclear gene expression dynamics using microarray analysis on plastid division mutants with varying degrees of division defects.

3) Use comparative quantitative proteome analysis on isolated chloroplasts from wild-type (WT) Arabidopsis and various plastid division arc mutants to determine what proteins are recruited/de-recruited to chloroplasts and/or stabilized/destabilized within chloroplasts during the division process. This data will complement the microarray data.

4) To understand the coordination between the stromal plastid division machinery and the cytosolic plastid division machinery across a double-membraned structure (Fig 3).

Figure 2. Model of protein-protein interactions within the plastid division machinery. (A) Bimoecular flourescence complementation assays have been sucessfuly used to confirm protein interations in planta. Assays were preformed by coexpressing stromal plastid division components fused to the N-terminal (NY) or C-terminal (CY) half of YFP and reconstituted YFP fluorophore detected by epifluorescence microscopy. Abrevtations used: E = AtMinE1, D = AtMinD1, F1 = AtFtsZ1-1, F2 = AtFtsZ2-1. (B) Working model for plastid division showing the identified protein components to date, their localization patterns and protein-protein interaction properties. AtMinE1 and AtMinD1 localise to discrete sopts at the chloroplast membrane and interact to form a complex. MSL2 and MSL3 (MSL) are predicted transmembrane proteins and colocalise with AtMinE1 to the poles of the chloropalst. GC1 localizes to the stromal side of the inner envelope membrane and forms dimers but is unable to interact with other plastid division components. AtFtsZ1-1 (F1) and AtFtsZ2-1 (F2) form a ring-like structure at the chloropaslt mid point and can form homodimers and heterodimers. AtFtsZ1-1 interacts with ARC3 (3) and AtFtsZ2-1 interacts with ARC6 (6). ARC3 and ARC6 both localise to ring-like structures and both can dimerise. ARC5 localises to a ring-like structures on the cytosolic surface of the outer envelope membrane.

Figure 3. Stromal and cytosolic plastid division machineries. As the first step of stromal division machinery assembly AtFtsZ1-1 (F1) and AtFtsZ2-1 (F2) form a Z-ring at the centre of chloroplasts. ARC6 (6) and ARC3 (3) are recruited to the Z-ring through specific interactions with AtFtsZ2-1 and AtFtsZ1-1, respectively. AtCDT1 also interacts with ARC6, although the localisation of AtCDT1 is not known. The placement of the Z-ring requires the combined action of AtMinE1, AtMinD1 and possibly ARC3, which form a complex and can localise to plastid poles. GC1 localizes to the stromal side of the inner envelope membrane and forms dimers. PDV1 and PDV2 localise to ring-like structures on the cytosolic surface of the outer envelope membrane and recruit ARC5 to the division site to constitute the cytosolic division machinery. The inner and outer PD rings are not shown. Coordination and signalling between the two division machineries may occur through a direct interaction between known proteins (e.g. between ARC6 and PDV1), may require as yet unidentified intermembrane space proteins (x) or through the action of signalling components (black spots).

Recent publications

Jodi Maple and Simon Geir Møller (2007). Plastid division coordination across a double-membraned structure. FEBS Letters, in press

Jodi Maple, Lea Vojta, Jurgen Soll and Simon Geir Møller (2007) ARC3 is a stromal Z-ring accessory protein essential for plastid division. EMBO Rep. in press

Cassie Aldridge and Simon Geir Møller (2005) The plastid division protein AtMinD1 is a Ca2+-ATPase stimulated by AtMinE1. J. Biol. Chem. 280. 31673-31678

Jodi Maple and Simon Geir Møller (2005). An emerging picture of plastid division in higher plants. Planta, 223, 1-4.

Jodi Maple, Cassie Aldridge and Simon Geir Møller (2005) Plastid division is mediated by combinatorial assembly of plastid division proteins. Plant J. 43. 811-823

Jodi Maple, Cassie Aldridge and Simon Geir Møller (2005) The molecular biology of plastid division. J. Ex. Bot.

56, 1061-1077.

Makoto T. Fujiwara, Ayako Nakamura, Ryuuichi Itoh, Yukihisa Shimada, Shigeo Yoshida and Simon Geir Møller (2004) Chloroplast division site placement requires dimerisation of the ARC11/ AtMinD1 protein in Arabidopsis. J. Cell Science, 117, 2399-2410.

Jodi Maple, Makoto T. Fujiwara, Nobutaka Kitahata, Tracey Lawson, Neil Baker, Shigeo Yoshida and Simon Geir Møller (2004) GIANT CHLOROPLAST 1 is essential for correct plastid division in Arabidopsis. Current Biology. 14, 776-781.