The Arabidopsis thaliana chloroplast division protein FtsZ1 counterbalances FtsZ2 filament stability in vitro

Abstract

Bacterial cell and chloroplast division are driven by a contractile "Z ring" composed of the tubulin-like cytoskeletal GTPase FtsZ. Unlike bacterial Z rings, which consist of a single FtsZ, the chloroplast Z ring in plants is composed of two FtsZ proteins, FtsZ1 and FtsZ2. Both are required for chloroplast division in vivo, but their biochemical relationship is poorly understood. We used GTPase assays, light scattering, transmission electron microscopy, and sedimentation assays to investigate the assembly behavior of purified Arabidopsis thaliana (At) FtsZ1 and AtFtsZ2 both individually and together. Both proteins exhibited GTPase activity. AtFtsZ2 assembled relatively quickly, forming protofilament bundles that were exceptionally stable, as indicated by their sustained assembly and slow disassembly. AtFtsZ1 did not form detectable protofilaments on its own. When mixed with AtFtsZ2, AtFtsZ1 reduced the extent and rate of AtFtsZ2 assembly, consistent with its previously demonstrated ability to promote protofilament subunit turnover in living cells. Mixing the two FtsZ proteins did not increase the overall GTPase activity, indicating that the effect of AtFtsZ1 on AtFtsZ2 assembly was not due to a stimulation of GTPase activity. However, the GTPase activity of AtFtsZ1 was required to reduce AtFtsZ2 assembly. Truncated forms of AtFtsZ1 and AtFtsZ2 consisting of only their conserved core regions largely recapitulated the behaviors of the full-length proteins. Our in vitro findings provide evidence that FtsZ1 counterbalances the stability of FtsZ2 filaments in the regulation of chloroplast Z-ring dynamics and suggest that restraining FtsZ2 self-assembly is a critical function of FtsZ1 in chloroplasts.

Keywords: Arabidopsis thaliana; FtsZ dynamics; GTPase; chloroplast; cytoskeleton; plant biochemistry.

Figure 1

AtFtsZ1 and AtFtsZ2 hydrolyze GTP, but only AtFtsZ2 exhibits GTP-dependent assembly in vitro.A, schematics of the complete AtFtsZ1-1 (At5g55280) and AtFtsZ2-1 (At2g36250) gene products. The constructs used in this study encoded the full-length mature proteins lacking their transit peptides (AtFtsZ1, aa 58–433; AtFtsZ2, aa 49–478) or the conserved core regions (AtFtsZ1_core, aa 73–376; AtFtsZ2_core, aa 119–423). The latter are composed of only the GTP binding (blue) and GTPase activating (pink) domains. TP, chloroplast transit peptide (green); CTP, AtFtsZ2 C-terminal peptide (aa 459–467, yellow). B, SDS-PAGE of purified AtFtsZ1 (Z1) and AtFtsZ2 (Z2) proteins. The gel was stained with Coomassie. Note that purified AtFtsZ2 runs as a doublet, as does endogenous AtFtsZ2-1 in plant extracts (60). Markers (kDa) are shown in the left. C, GTPase activities assayed in 500 μM GTP at 25 °C at protein concentrations ranging from 0.2 to 3 μM for AtFtsZ1 (blue) and from 0.5 to 6 μM for AtFtsZ2 (green). A representative set of assays is shown. The GTPase activity is the slope of the regression line above the Cc (Table 1). D and E, negative-stain TEM of 3 μM AtFtsZ2 or AtFtsZ1 incubated for 5 min at room temperature after addition of 500 μM GTP or GDP. Scale bars are as indicated. D, AtFtsZ2. The average width of AtFtsZ2 bundles in GTP was 60.15 ± 20.5 nm (SD; n = 32); Fig. S5. E, AtFtsZ1. F and G, sedimentation assays. Reactions containing 5 μM AtFtsZ2 or AtFtsZ1 were incubated for 30 min at room temperature after addition of 500 μM GTP or GDP, then centrifuged at 80,000g for 30 min at 25 °C. F, SDS-PAGE of proteins in the total (T), pellet (P), and supernatant (S) fractions. Representative gels stained with Coomassie are shown. Markers (kDa) are shown on the left. G, fraction of AtFtsZ protein in the pellet (n = 4).

Figure 2

AtFtsZ2 assembles stable protofilaments. Assembly reactions were performed at room temperature, initiated by addition of nucleotide, and repeated at least twice with similar results. A, assembly of AtFtsZ2 monitored by light scattering (LS) at the indicated protein concentrations after addition of 10 μM GTP. B, initial rates of AtFtsZ2 assembly for the LS traces in A (see Experimental procedures; initial rates determined for other LS replicates showed similar trends). C, LS assays of 5 μM AtFtsZ2 after addition of GTP or GDP at the indicated nucleotide concentrations. Arrowheads show the predicted times of GTP depletion for the reactions initiated with 5 μM GTP (purple) and 2.5 μM GTP (pink). Inset displays an extended assay initiated with 5 μM GTP for 5000 s. The y-axis (PMT counts) is the same as in the larger plot, while each tick on the x-axis denotes 1000 s. LS of 5 μM AtFtsZ1 after addition of 500 μM GTP is also shown (dark blue). D, initial rates of 5 μM AtFtsZ2 assembly for the LS traces in C. E, negative-stain TEM of 5 μM AtFtsZ2 incubated for 500 s (left) or 5000 s (right) after addition of 5 μM GTP. F, GTPase activity of AtFtsZ2 at different GTP concentrations. For assays in 2.5, 5, 10, 100, and 500 μM GTP, n = 5, 4, 4, 6, and 4, respectively. There were no significant differences between any two means (p > 0.7) as determined by a one-way ANOVA using Tukey’s multiple comparison test.

Figure 3

AtFtsZ1 reduces assembly of AtFtsZ2.A, light scattering (LS) assays of 5 μM AtFtsZ2 mixed with AtFtsZ1 at various ratios, after addition of 10 μM GTP. B, initial rates of assembly for the LS traces in A. C, negative-stain TEM of 5:0 μM (top) and 5:5 μM (bottom) AtFtsZ2:AtFtsZ1 assembled for 2000 s after addition of 10 μM GTP. D, LS assays of 5 μM AtFtsZ2 mixed with AtFtsZ1 at the indicated ratios after addition of excess GTP (500 μM). E, initial rates of assembly for the LS traces in D.

Figure 4

AtFtsZ1 GTPase activity is required to reduce AtFtsZ2 assembly.A, B, D and E, sedimentation assays of AtFtsZ2 mixed with either AtFtsZ1 (A and B) or AtFtsZ1_D275A (D and E) at the indicated protein concentrations (μM). Assays were initiated by addition of 500 μM GTP or GDP and performed as described in Figure 1, F and G. A and D, SDS-PAGE of proteins in the total (T), pellet (P), and supernatant (S) fractions. Representative gels stained with Coomassie are shown. Markers (kDa) are indicated on the left. B and E, fraction of AtFtsZ in the pellet (n = 4 for all reactions except for AtFtsZ2:AtFtsZ1 mixed at 6:3 μM and 5:5 μM, where n = 3). The 5:0 μM and 0:5 μM data in B are repeated from Figure 1G. C, slopes of the light scattering (LS) traces of preassembled AtFtsZ2 after addition of buffer (green), AtFtsZ1 (purple), or AtFtsZ1_D275A (orange). Following each addition, the final concentration of AtFtsZ2, AtFtsZ1, AtFtsZ1_D275A, and GTP were 5 μM, 2.5 μM, 2.5 μM, and 500 μM, respectively. See Fig. S4 for explanation of how slopes were determined. F, LS assays of 5 μM AtFtsZ2 (Z2) mixed with AtFtsZ1 (Z1) or AtFtsZ1_D275A (Z1_D275A) at the indicated ratios. Assays were initiated by addition of 500 μM GTP.

Figure 5

The distinct assembly dynamics of AtFtsZ1 and AtFtsZ2 are determined primarily by their conserved core regions.A, SDS-PAGE of purified AtFtsZ1_core (Z1_c) and AtFtsZ2_core (Z2_c). The gel was stained with Coomassie. Markers (kDa) are shown on the right. B, GTPase activities assayed in 500 μM GTP at 25 °C at protein concentrations ranging from 0.5 to 6 μM for AtFtsZ1_core (blue) and from 1 to 6 μM for AtFtsZ2_core (green). A representative set of assays is shown. The GTPase activity is the slope of the regression line above the Cc. C, sedimentation assays of 3 μM AtFtsZ1_core (blue) or AtFtsZ2_core (green) after addition of 500 μM GTP or GDP. Assays were performed as described in Figure 1, F and G, centrifuged at 4 °C, and the fraction of protein in the pellet is shown (n = 3). A representative Coomassie-stained gel is shown in Figure 6D. D and E, negative-stain TEM of 3 μM AtFtsZ_core proteins incubated for 5 min after addition of 500 μM GTP or GDP. D, AtFtsZ2_core. The region in the red square is shown at higher magnification in the middle panel. The average width of AtFtsZ2_core bundles assembled in GTP was 25.84 ± 13.1 nm (SD; n = 32; Fig. S5). E, AtFtsZ1_core in GTP or GDP. F, light scattering (LS) assays of AtFtsZ2_core at the indicated concentrations after addition of 10 μM GTP. G, initial rates of assembly for the LS traces in F. H, LS of 5 μM AtFtsZ2_core after addition of GTP or GDP at the indicated nucleotide concentrations. LS of 5 μM AtFtsZ1_core after addition of 500 μM GTP is also shown (dark blue). I, initial rates of assembly for the LS traces in H.

Figure 6

AtFtsZ1_corereduces the overall assembly of AtFtsZ_coreproteins.A, light scattering (LS) assays of 5 μM AtFtsZ2_core mixed with AtFtsZ1_core at the indicated ratios after addition of 10 μM GTP. B, initial rates of assembly for the LS traces in A. C, negative-stain TEM of 5:0 μM (top) and 5:5 μM (bottom) AtFtsZ2_core:AtFtsZ1_core assembled for 2000 s after addition of 10 μM GTP. D and E, sedimentation assays containing AtFtsZ2_core mixed with AtFtsZ1_core at the indicated concentration ratios. Assays were initiated by addition of 500 μM GTP or GDP, performed as described in Figure 1, F and G, and centrifuged at 4 °C. D, SDS-PAGE of proteins in the total (T), supernatant (S), and pellet (P) fractions. Representative gels stained with Coomassie are shown. Markers (kDa) are indicated on the left. E, fraction of total AtFtsZ_core in the pellet (n = 3). The 3:0 and 0:3 μM data in E are repeated from Figure 5C.