GTP-dependent heteropolymer formation and bundling of chloroplast FtsZ1 and FtsZ2

Abstract

Bacteria and chloroplasts require the ring-forming cytoskeletal protein FtsZ for division. Although bacteria accomplish division with a single FtsZ, plant chloroplasts require two FtsZ types for division, FtsZ1 and FtsZ2. These proteins colocalize to a mid-plastid Z ring, but their biochemical relationship is poorly understood. We investigated the in vitro behavior of recombinant FtsZ1 and FtsZ2 separately and together. Both proteins bind and hydrolyze GTP, although GTPase activities are low compared with the activity of Escherichia coli FtsZ. Each protein undergoes GTP-dependent assembly into thin protofilaments in the presence of calcium as a stabilizing agent, similar to bacterial FtsZ. In contrast, when mixed without calcium, FtsZ1 and FtsZ2 exhibit slightly elevated GTPase activity and coassembly into extensively bundled protofilaments. Coassembly is enhanced by FtsZ1, suggesting that it promotes lateral interactions between protofilaments. Experiments with GTPase-deficient mutants reveal that FtsZ1 and FtsZ2 form heteropolymers. Maximum coassembly occurs in reactions containing equimolar FtsZ1 and FtsZ2, but significant coassembly occurs at other stoichiometries. The FtsZ1:FtsZ2 ratio in coassembled structures mirrors their input ratio, suggesting plasticity in protofilament and/or bundle composition. This behavior contrasts with that of alpha- and beta-tubulin and the bacterial tubulin-like proteins BtubA and BtubB, which coassemble in a strict 1:1 stoichiometry. Our findings raise the possibility that plasticity in FtsZ filament composition and heteropolymerization-induced bundling could have been a driving force for the coevolution of FtsZ1 and FtsZ2 in the green lineage, perhaps arising from an enhanced capacity for the regulation of Z ring composition and activity in vivo.

FIGURE 1.

GTP hydrolysis rates of FtsZ1 and FtsZ2 assayed separately and together. GTPase activities of 5 μm FtsZ1 (□, solid lines), 5 μm FtsZ2 (▵, dashed and dotted lines), and 2.5 μm each FtsZ1 plus FtsZ2 (○, dotted lines) were assayed at the indicated GTP concentrations. A, 25 °C. B, 37 °C.

FIGURE 2.

GTP-dependent assembly of FtsZ1 and FtsZ2 monitored by sedimentation. 5 μm FtsZ1 (lanes 2–5) or FtsZ2 (lanes 7–10) was incubated at room temperature (25 °C) for 3 or 30 min in the indicated buffer. Assembly was initiated by the addition of nucleotide. Supernatant (S) and pellet (P) fractions were obtained by centrifugation and analyzed by SDS-PAGE, Coomassie staining, and densitometry. Input, FtsZ1 (Z1, lanes 1) or FtsZ2 (Z2, lanes 6) in the starting reactions. The percentage of input protein in the pellet fraction ± S.E. (n = 4) (%P) is shown below the relevant lane. The buffers used were HMK (A), HMK-GTP (B), HMKCa-GDP (C), and HMKCa-GTP (D). FtsZ1 and FtsZ2 migrate at ∼40 and 45–46 kDa, respectively (20).

FIGURE 3.

Assembly of FtsZ1 and FtsZ2 separately and together. The reactions were performed at room temperature with 5 μm total protein. Assembly was initiated by the addition of nucleotide. A–C, electron micrographs of reaction aliquots taken after 10 min. A, FtsZ1 in HMKCa-GTP. B, FtsZ2 in HMKCa-GTP. C, FtsZ1 plus FtsZ2 (2.5 μm each) in HMK-GTP. D, time course of assembly of FtsZ1 in HMK-GTP (gray triangle) and HMKCa-GTP (gray diamond), FtsZ2 in HMK-GTP (inverted black triangle), and HMKCa-GTP (black diamond), and FtsZ1 plus FtsZ2 (2.5 μm each) in HMK-GTP (black square) or HMK-GDP (black circle) monitored by light scattering. E, coassembly of FtsZ1 and FtsZ2 (2.5 μm each) in HMK-GTP (lanes 2–5) or HMK-GDP (lanes 6–9) monitored by sedimentation as in Fig. 2. Input, FtsZ1 (Z1) or FtsZ2 (Z2) in the starting reactions (lane 1). The percentages of input FtsZ1 (Z1%P) and FtsZ2 (Z2%P) in the pellet fraction ± S.E. (n = 4) are shown below the relevant lanes. S, supernatant; P, pellet. F, relative maximum assembly rates for reactions in D estimated by taking the average of the maximum slopes from three independent light scattering traces.

FIGURE 4.

Critical concentration for FtsZ1 and FtsZ2 coassembly. Coassembly of equimolar FtsZ1 and FtsZ2 in HMK-GTP at room temperature was monitored by light scattering at the indicated total FtsZ concentration. The relative maximum assembly rate for each reaction, estimated by taking the maximum slope of the corresponding light scattering trace, was plotted against protein concentration (35). All of the data points except those below 1.0 μm FtsZ, where assembly was not observed, were fit with the linear regression equation shown. The x intercept indicated a C_c of ∼0.75 μm for FtsZ1/FtsZ2 coassembly, similar to the C_c of ∼1 μm reported for EcFtsZ (10, 64).

FIGURE 5.

Coassembly of FtsZ1 and FtsZ2 at various stoichiometries. All of the reactions were performed at room temperature with 5 μm total protein. Assembly was initiated by the addition of nucleotide. A and B, time course of coassembly of FtsZ1 (Z1) and FtsZ2 (Z2) at the indicated molar ratio monitored by light scattering. The experiment was repeated three times; data from a single experiment are shown. The control reactions (red circles) were in HMK-GDP; all others were in HMK-GTP. A, reactions with excess FtsZ2. B, reactions with excess FtsZ1. C, relative maximum assembly rates ± S.E. (n = 3) derived by averaging the maximum slopes of the light scattering traces from three experiments. D and E, coassembly of 1:5 (D) or 5:1 (E) FtsZ1:FtsZ2 in HMK-GTP (lanes 2–5) or HMK-GDP (lanes 6–9) monitored by sedimentation as described in the legend for Fig. 2. Input, FtsZ1 (Z1) and FtsZ2 (Z2) in the starting reactions (lanes 1). Percentages of input FtsZ1 (Z1 %P) and FtsZ2 (Z2 %P) in the pellet fraction ± S.E. (n = 4) are shown below the relevant lanes. S, supernatant; P, pellet.

FIGURE 6.

Coassembly of FtsZ1 and FtsZ2 T7 loop mutants and effect on GTPase activity of the wild type proteins. A, electron micrograph of FtsZ1D275A and FtsZ2D322A (2.5 μm each) coassembled HMK-GTP at room temperature. B, GTPase activities of FtsZ1 assayed at 25 °C in the presence of increasing FtsZ2D322A (Z1+Z2D322A, black bars) and FtsZ2 assayed in the presence of increasing FtsZ1D275A (Z2+Z1D275A, gray bars). The total protein concentration was kept at 5 μm, and each assay was performed three times.

FIGURE 7.

Comparison of untagged and His-tagged FtsZ1 and FtsZ2. A, SDS-PAGE and silver staining of assembly-purified untagged FtsZ1 (lanes 1–6) and FtsZ2 (lanes 7–12). Lanes 1 and 7, uninduced crude cell extracts; lanes 2 and 8, crude cell extracts from isopropyl β-d-thiogalactopyranoside-induced cultures; lanes 3 and 9, soluble fractions from isopropyl β-d-thiogalactopyranoside-induced cultures; lanes 4 and 10, insoluble fractions from isopropyl β-d-thiogalactopyranoside-induced cultures (inclusion bodies); lanes 5 and 6, FtsZ1 before and after assembly purification, respectively; lanes 11 and 12, FtsZ2 before and after assembly purification, respectively. B, GTP hydrolysis rates of His-tagged and untagged FtsZ1 (Z1) and FtsZ2 (Z2) assayed individually (5 μm) or mixed 1:1 (2.5 μm each; Z1/Z2) with 1 mm (black bars) or 3 mm (gray bars) GTP in HMK at 25 °C. At 3 mm GTP, the activity of equally mixed His-tagged FtsZ1 + FtsZ2 is significantly higher than that of FtsZ1 (p = 0.05, n = 6) and FtsZ2 (p = 0.01, n = 6), and the activity of equally mixed untagged FtsZ1 + FtsZ2 is significantly higher than that of FtsZ1 (p = 0.01, n = 3) and FtsZ2 (p = 0.01, n = 3). C, electron micrographs of coassembled untagged FtsZ1 and FtsZ2 (2.5 μm each) in HMK with 1 mm GTP. Scale bars, 2 μm and 100 nm in panels 1 and 2, respectively. D, coassembly of His-tagged (lanes 2–5) and untagged (lanes 7–10) FtsZ1 and FtsZ2 (2.5 μm each) in HMK with 1 mm (lanes 2, 3, 7, and 8) or 3 mm (lanes 4, 5, 9, and 10) GTP monitored by sedimentation as in Fig. 2. Input, FtsZ1 (Z1) and FtsZ2 (Z2) in the starting reactions (lanes 1 and 6); S, supernatant; P, pellet. Percentages of input FtsZ1 (Z1%P) and FtsZ2 (Z2%P) in the pellet fraction ± S.E. (n = 2) are shown below the relevant lanes.