Abstract
The FtsZ protein forms a dynamic polymeric ring structure, which functions as the guiding scaffold for septal invagination at the mid-cell site during cytokinesis in bacterial cells (Bi & Lutkenhaus, 1991). FtsZ forms polymers in a GTP-dependent manner in vivo and under different cationic conditions in vitro (Andreu et al., 2002; Beuria et al., 2003; Bi & Lutkenhaus, 1991; Bramhill & Thompson, 1994; Caplan & Erickson, 2003; Diaz et al., 2001; Erickson et al., 1996; Lu et al., 1998; Mingorance et al., 2001; Mukherjee & Lutkenhaus, 1994, 1999; Rivas et al., 2000; Romberg et al., 2001; Scheffers et al., 2001; Wang et al., 1997; White et al., 2000; Yu & Margolin, 1997). The rate of polymerization of the FtsZ protein of Mycobacterium tuberculosis H37Rv (MtFtsZ) is remarkably slow when compared to that of the Escherichia coli FtsZ protein (EcFtsZ) (Bramhill & Thompson, 1994; Mukherjee & Lutkenhaus, 1999; White et al., 2000). While polymerization of EcFtsZ reaches steady state in 30 s after the addition of GTP in vitro (Bramhill & Thompson, 1994; Mukherjee & Lutkenhaus, 1999), MtFtsZ takes nearly 10 min to reach the same state (White et al., 2000). The rate of depolymerization of MtFtsZ was also found to be slow, in conformity with its slow GTPase activity (White et al., 2000).
Although such striking differences exist in their polymerization and GTPase activities, MtFtsZ shows notable conservation with EcFtsZ at the N terminus, with the regions involved in FtsZFtsZ interactions (Wang et al., 1997) being more or less identical. However, the C terminus of MtFtsZ shows remarkable divergence from that of EcFtsZ. This structural variation tempted us to speculate that the differences in the biochemical properties of MtFtsZ and EcFtsZ might be due to the sequence divergence at the C terminus. Structural modelling of protofilaments of Methanococcus jannaschii FtsZ (MjFtsZ) (Lowe & Amos, 1999), based on the three-dimensional structure of MjFtsZ (Lowe & Amos, 1998), and site-directed mutagenesis studies on EcFtsZ (Lu et al., 2001), had indicated that at least four interacting faces, two longitudinal faces and two lateral faces of a FtsZ monomer might be engaged in the formation of protofilaments and polymers (Erickson, 1998; Lu et al., 2001; Romberg et al., 2001). One of these interacting faces maps to the divergent C terminus of the protein, thereby indicating involvement of this region in polymerization. Taking these correlations into consideration, we reasoned that the residues constituting the divergent C terminus of MtFtsZ might be responsible for imposing slow polymerization. If this contention is true, then an MtFtsZ mutant that lacks the divergent C-terminal residues might form polymers in much shorter time than the full-length MtFtsZ would do. In order to verify this premise, we determined the earliest measurable time point at which the deletion mutant MtFtsZ-ΔC169, which lacked 169 residues spanning the divergent C terminus, and full-length MtFtsZ could form equivalent-type long polymers. The time points at which full-length MtFtsZ and MtFtsZ-ΔC169 samples were examined for polymers were 30 s, 1 min, 2 min, 4 min and 10 min after the addition of GTP.
The earliest time at which long polymers of full-length MtFtsZ could be found was 10 min after the addition of GTP (Fig. 1a). Even very short polymers of full-length MtFtsZ could be found only from 2 min after the addition of GTP. By contrast, the MtFtsZ-ΔC169 deletion mutant formed equivalent-type long polymers as early as 30 s after the addition of GTP (Fig. 1b). The polymeric structures of MtFtsZ-ΔC169 protein were stable and could be observed at the end of 10 min incubation as well. Even at 10 min or 30 min, neither full-length MtFtsZ nor MtFtsZ-ΔC169 mutant formed polymers in the absence of GTP. The polymers formed by MtFtsZ-ΔC169 were approximately 1·5 times thicker than those formed by full-length MtFtsZ. Taken together, our results showed that MtFtsZ protein, when devoid of C-terminal 169 residues, could form long polymers in dramatically shorter time, as compared to the full-length MtFtsZ protein. This observation indicates that the residues responsible for slow polymerization of MtFtsZ are located in the region spanning the divergent C terminus of the protein. A systematic mutational analysis of the C-terminal region might enable identification of residues that impose slow polymerization on MtFtsZ. Our observations also imply that among the 379 residues of the MtFtsZ protein, a stretch of 210 residues from the N terminus is sufficient for in vitro polymerization of MtFtsZ, which may hold true for FtsZ proteins of other bacteria as well owing to their homology at the N terminus.