Functional dissection of a cell-division inhibitor, SulA, of Escherichia coli and its negative regulation by Lon

SulA is induced in Escherichia coli by the SOS response and inhibits cell division through interaction with FtsZ. To determine which region of SulA is essential for the inhibition of cell division, we constructed a series of N-terminal and C-terminal deletions of SulA and a series of alanine substitution mutants. Arginine at position 62, leucine at 67, tryptophan at 77 and lysine at 87, in the central region of SulA, were all essential for the inhibitory activity. Residues 3–27 and the C-terminal 21 residues were dispensable for the activity. The mutant protein lacking N-terminal residues 3–47 was inactive, as was that lacking the C-terminal 34 residues. C-terminal deletions of 8 and 21 residues increased the growth-inhibiting activity in lon ⁺ cells, but not in lon ⁻ cells. The wild-type and mutant SulA proteins were isolated in a form fused to E. coli maltose-binding protein, and tested in vitro for sensitivity to Lon protease. Lon degraded wild-type SulA and a deletion mutant lacking the N-terminal 93 amino acids, but did not degrade the derivative lacking 21 residues at the C-terminus. Futhermore, the wild-type SulA and the N-terminal deletion mutant formed a stable complex with Lon, while the C-terminal deletion did not. MBP fused to the C-terminal 20 residues of SulA formed a stable complex with, but was not degraded by Lon. When LacZ protein was fused at its C-terminus to 8 or 20 amino acid residues from the C-terminal region of SulA the protein was stable in lon ⁺ cells. These results indicate that the C-terminal 20 residues of SulA permit recognition by, and complex formation with, Lon, and are necessary, but not sufficient, for degradation by Lon. Received: 8 October 1996 / Accepted: 27 November 1996     

Regulatory role of C-terminal residues of SulA in its degradation by Lon protease in Escherichia coli

The SulA protein is a cell division inhibitor in Escherichia coli, and is specifically degraded by Lon protease. To study the recognition site of SulA for Lon, we prepared a mutant SulA protein lacking the C-terminal 8 amino acid residues (SA8). This deletion protein was accumulated and stabilized more than native SulA in lon(+) cells in vivo. Moreover, the deletion SulA fused to maltose binding protein was not degraded by Lon protease, and did not stimulate the ATPase or peptidase activity of Lon in vitro, probably due to the much reduced interaction with Lon. A BIAcore study showed that SA8 directly interacts with Lon. These results suggest that SA8 of SulA was recognized by Lon protease. The SA8 peptide, KIHSNLYH, specifically inhibited the degradation of native SulA by Lon protease in vitro, but not that of casein. A mutant SA8, KAHSNLYH, KIASNLYH, or KIHSNAYH, also inhibited the degradation of SulA, while such peptides as KIHSNLYA did not. These results show that SulA has the specified rows of C-terminal 8 residues recognized by Lon, leading to facilitated binding and subsequent cleavage by Lon protease both in vivo and in vitro.     

Overproduction and purification of SulA fusion protein in Escherichia coli and its degradation by Lon protease in vitro

To overproduce extremely unstable SulA protein, which is the cell-division inhibitor of Escherichia coli, we fused the sulA gene to the maltose-binding protein (MBP) fusion vectors with or without the signal sequence (plasmids pMAL-p-SulA and pMAL-c-SulA respectively). The amount of the full-length fusion protein expressed from the plasmid pMAL-p-SulA (pre-MBP-SulA) in E. coli was much larger than that expressed from the plasmid pMAL-c-SulA (MBP-SulA). A major amount of the pre-MBP-SulA fusion protein was expressed in a soluble form and affinity-purified by amylose resin. Since site-specific cleavage of the fusion protein with factor Xa resulted in the precipitation of SulA protein, the pre-MBP-SulA fusion protein was used to study the degradation of SulA protein by E. coli Lon protease in vitro. It was found that only the SulA portion of the fusion protein was degraded by Lon protease in an ATP-dependent manner. This result provides direct evidence that Lon protease plays an important role in the rapid degradation of SulA protein in cells.     

The unique sites in SulA protein preferentially cleaved by ATP-dependent Lon protease from Escherichia coli.

SulA protein is known to be one of the physiological substrates of Lon protease, an ATP-dependent protease from Escherichia coli. In this study, we investigated the cleavage specificity of Lon protease toward SulA protein. The enzyme was shown to cleave approximately 27 peptide bonds in the presence of ATP. Among them, six peptide bonds were cleaved preferentially in the early stage of digestion, which represented an apparently unique cleavage sites with mainly Leu and Ser residues at the P1, and P1' positions, respectively, and one or two Gln residues in positions P2-P5. They were located in the central region and partly in the C-terminal region, both of which are known to be important for the function of SulA, such as inhibition of cell growth and interaction with Lon protease, respectively. The other cleavage sites did not represent such consensus sequences, though hydrophobic or noncharged residues appeared to be relatively preferred at the P1 sites. On the other hand, the cleavage in the absence of ATP was very much slower, especially in the central region, than in the presence of ATP. The central region was predicted to be rich in alpha helix and beta sheet structures, suggesting that the enzyme required ATP for disrupting such structures prior to cleavage. Taken together, SulA is thought to contain such unique cleavage sites in its functionally and structurally important regions whose preferential cleavage accelerates the ATP-dependent degradation of the protein by Lon protease.     

ATP-dependent degradation of SulA, a cell division inhibitor, by the HslVU protease in Escherichia coli.

HslVU is an ATP-dependent protease consisting of two multimeric components, the HslU ATPase and the HslV peptidase. To gain an insight into the role of HslVU in regulation of cell division, the reconstituted enzyme was incubated with SulA, an inhibitor of cell division in Escherichia coli, or its fusion protein with maltose binding protein (MBP). HslVU degraded both proteins upon incubation with ATP but not with its nonhydrolyzable analog, ATPgammaS, indicating that the degradation of SulA requires ATP hydrolysis. The pulse-chase experiment using an antibody raised against MBP-SulA revealed that the stability of SulA increased in hsl mutants and further increased in lon/hsl double mutants, indicating that SulA is an in vivo substrate of HslVU as well as of protease La (Lon). These results suggest that HslVU in addition to Lon plays an important role in regulation of cell division through degradation of SulA.     

Determination of the cleavage sites in SulA, a cell division inhibitor, by the ATP-dependent HslVU protease from Escherichia coli

HslVU is an ATP-dependent protease from Escherichia coli and known to degrade SulA, a cell division inhibitor, both in vivo and in vitro, like the ATP-dependent protease Lon. In this study, the cleavage specificity of HslVU toward SulA was investigated. The enzyme was shown to produce 58 peptides with various sizes (3-31 residues), not following the 'molecular ruler' model. Cleavage occurred at 39 peptide bonds preferentially after Leu in an ATP-dependent manner and in a processive fashion. Interestingly, the central and C-terminal regions of SulA, which are known to be important for the function of SulA, such as inhibition of cell division and molecular interaction with certain other proteins, were shown to be preferentially cleaved by HslVU, as well as by Lon, despite the fact that the peptide bond specificities of the two enzymes were distinct from each other.     

The Escherichia coli replication inhibitor CspD is subject to growth-regulated degradation by the Lon protease

Post-translational proteolysis-dependent regulation of critical cellular processes is a common feature in bacteria. The Escherichia coli Lon protease is involved in the control of the SOS response, acid tolerance and nutritional deprivation. Moreover, Lon plays a role in the regulation of toxin-antitoxin (TA) systems and thereby is linked to persister cell induction. Persister cells represent a small subpopulation that has reversibly switched to a dormant and non-dividing state without genomic alterations. Formation of persister cells permits viability upon nutritional depletion and severe environmental stresses. CspD is a replication inhibitor, which is induced in stationary phase or upon carbon starvation and increases the production of persister cells. It has remained unknown how CspD activity is counteracted when growth is resumed. Here we report that CspD is subject to proteolysis by the Lon protease both in vivo and in vitro. Turnover of CspD by Lon is strictly adjusted to the growth rate and growth phase of E. coli, reflecting the necessity to control CspD levels according to the physiological conditions.     

Proteolysis and modulation of the activity of the cell division inhibitor SulA in Escherichia coli lon mutants.

Intracellular accumulation of the inducible cell division inhibitor SulA is modulated by proteases that ensure its degradation, namely, the Lon protease and another ATP-dependent protease(s). Lon- cells are UV sensitive because SulA is stable. We asked whether these ATP-dependent proteases are more active when lon cells are grown at high temperature or in synthetic medium since these conditions decrease the UV sensitivity of lon cells. We found that these growth conditions have no direct effect on Lon-independent degradation of SulA. They may, instead, decrease the SulA-FtsZ interaction.     

The HslU ATPase acts as a molecular chaperone in prevention of aggregation of SulA, an inhibitor of cell division in Escherichia coli

HslVU is an ATP-dependent protease consisting of two multimeric components: the HslU ATPase and the HslV peptidase. SulA, which is an inhibitor of cell division and has high tendency of aggregation, is degraded by HslVU protease. Here we show that HslU plays a role not only as a regulatory component for the HslV-mediated proteolysis but also as a molecular chaperone. Purified HslU prevented aggregation of SulA in a concentration-dependent fashion. This chaperone activity required oligomerization of HslU subunits, which could be achieved by ATP-binding or in the presence of high HslU protein concentrations. hsl mutation reduced the SulA-mediated inhibition of cell growth and this effect could be reversed upon overproduction of HslU, suggesting that HslU promotes the ability of SulA to block cell growth through its chaperone function. Thus, HslU appears to have two antagonistic functions: one as a chaperone for promotion of the ability of SulA in cell growth inhibition by preventing SulA aggregation and the other as the regulatory component for elimination of SulA by supporting the HslV-mediated degradation.     

Early stages in development of the Escherichia coli cell-division site

Development of the Escherichia coli cell division site was studied in wild-type cells and in non-septate filaments of ftsZ null and ftsZTs mutant cells. Localized regions of plasmolysis were used as markers for the positions of annular structures that are thought to be related to the periseptal annuli that flank the ingrowing septum during cytokinesis. The results show that these structures are localized at potential division sites in non-septate filaments of FtsZ^- cells, contrary to previous reports. The positions of the structures along the long axis of the cells in both wild-type cells and FtsZ^- filaments were unaffected by the presence of plasmolysis bays at the cell poles. These results do not agree with a previous suggestion that the apparent association of plasmolysis bays with future division sites was artefactual. They support the view that division sites begin to differentiate before the initiation of septal ingrowth and that plasmolysis bays and the annular attachments that define them, mark the locations of these early events in the division process.     

Induction of the SOS Response in Ultraviolet-Irradiated Escherichia coli Analyzed by Dynamics of LexA, RecA and SulA Proteins

The SOS response in Escherichia coli is induced after DNA-damaging treatments including ultraviolet light. Regulation of the SOS response is accomplished through specific interaction of the two SOS regulator proteins, LexA and RecA. In ultraviolet light-treated cells, nucleotide excision repair is the major system that removes the induced lesions from the DNA. Here, induction of the SOS response in Escherichia coli with normal and impaired excision repair function is studied by simulation of intracellular levels of regulatory LexA and RecA proteins, and SulA protein. SulA protein is responsible for SOS-inducible cell division inhibition. Results of the simulations show that nucleotide excision repair influences time-courses of LexA, RecA and SulA induction by modulating the dynamics of RecA protein distribution between its normal and SOS-activated forms.     

Saturation and specificity of the Lon protease of Escherichia coli.

Lon is an ATP-dependent protease of Escherichia coli. The lon mutation has a pleiotropic phenotype: UV sensitivity, mucoidy, deficiency for lysogenization by bacteriophage lambda and P1, and lower efficiency in the degradation of abnormal proteins. All of these phenotypes are correlated with the loss of protease activity. Here we examine the effects of overproduction of one Lon substrate, SulA, and show that it protects two other substrates from degradation. To better understand this protection, we mutagenized the sulA gene and selected for mutants that have partially or totally lost their ability to saturate the Lon protease and thus can no longer protect another substrate. Some of the SulA mutants lost their ability to protect RcsA from degradation but could still protect the O thermosensitive mutant protein (Ots). All of the mutants retained their capacity to induce cell division inhibition. It was also found that deletion of the C-terminal end of SulA affected its activity but did not affect its susceptibility to Lon. We propose that Lon may have more than one specificity for peptide cleavage.     

Functional groups in the activity and regulation of Escherichia coli citrate synthase

1. Citrate synthase has been purified from Escherichia coli and shown to exist at an equilibrium between three forms: monomer (mol.wt. 57000), tetramer (mol.wt. 230000) and, possibly, octamer. Modification of the enzyme by photo-oxidation and by treatment with specific chemical reagents has been carried out to gain information on the amino acid residues involved in enzymic activity and in the inhibition of activity by NADH and alpha-oxoglutarate. 2. Several photo-oxidizable amino acids appear to be involved in activity. The nature of the pH-dependence of their rates of photo-oxidation with Methylene Blue suggests that these are histidines, a conclusion supported by the greater rate of photo-inactivation with Rose Bengal and the destruction of activity by diethyl pyrocarbonate. 3. The participation of histidine at the alpha-oxoglutarate effector site is indicated by photo-oxidation and the participation of cysteine at the NADH effector site suggested by photo-oxidation is confirmed by the desensitization to NADH produced by treatment with 5,5'-dithiobis-(2-nitrobenzoate). Inactivation of the enzyme after modification with this reagent suggests the additional involvement of cysteine in catalytic activity. 4. Amino acid analyses of native and photo-oxidized enzyme are consistent with these conclusions. 5. Modification with 2-hydroxy-5-nitrobenzyl bromide indicates the participation of tryptophan in the activity of the enzyme.     

Properties of a gonococcal inhibitor produced by Escherichia coli.

Strains of Escherichia coli can inhibit the in vitro growth of Neisseria gonorrhoeae. One E. coli strain released a potent agar-diffusible gonococcal growth inhibitor which was extracted and assayed in an agar well assay system. The culture conditions necessary to produce the inhibitor were determined. The inhibitor was bacteriostatic, in most cases, for N. gonorrhoeae. Based on ultrafiltration and column chromatography, the inhibitor appeared to have a molecular weight in the range of 1200 to 2000. Evidence that the molecule contained charged sites was obtained by membrane binding and column chromatography. The inhibitor was stable to extremes of heat, cold and pH. It was not volatile or susceptible to proteolytic enzymes, lysozyme, lipase, DNAase, RNAase or certain chelating agents. Its activity was completely blocked by ferric ammonium citrate. This inhibitor is dissimilar to previously reported gonococcal inhibitors of bacterial origin.