Mot protein assembly into the bacterial flagellum: a model based on mutational analysis of the motB gene.

The 308 residue MotB protein anchors the stator complex of the Escherichia coli flagellar motor to the peptidoglycan of the cell wall. Together with MotA, it comprises the transmembrane channel that delivers protons to the motor. At the outset of the mutational analysis of MotB described here, we found that the non-motile phenotype of a DeltamotAB strain was rescued better by a pmotA(+)B(+) plasmid than the non-motile phenotype of a DeltamotB strain was rescued by a pmotB(+) plasmid. Transcription in each case was from the inducible tac promoter but relied on the native ribosome-binding site (RBS). This result confirms that translational coupling to motA is important for normal translation of the motB mRNA, since overproduction of MotA in trans did not improve complementation by pmotB. However, introduction of an optimized RBS into pmotB (to generate pmotB(o)) did. To dissect the function of the periplasmic domain of MotB, site-directed mutagenesis was used to replace Gln, Ser, and Tyr codons scattered throughout motB with amber (UAG) codons. Plasmid-borne motB(am) genes were introduced into sup(o), supE, and supF strains to see what motility defects were imposed by particular amber mutations and whether the defects could be suppressed by amber-suppressor tRNAs inserting the native or heterologous amino acids. Amber mutations at codon 268 or earlier in pmotB, and at codon 261 or earlier in pmotB(o) or pmotAB, eliminated motility. Thus, in agreement with the deletion analysis of motB by another laboratory, we conclude that the portion of MotB carboxyl-terminal to its peptidoglycan-binding motif (residues 161 to 264) is not essential. In strains containing supE or supF alleles, motility defects associated with motB(am) mutations were suppressed weakly, if at all, in pmotB. In contrast, motility defects conferred by most motB(am) mutations in pmotB(o) or pmotAB could be suppressed to a significant extent. However, the S18(am), Q100(am), Q112(am), Q124(am), Y201(am), and Y208(am) mutations were still suppressed extremely poorly. Full-length MotB was present at very low levels in suppressor strains containing the first four mutations, but Y201(am) and Y208(am) were suppressed efficiently at the translational level. We suggest that a translational pause by suppressor tRNAs reading UAG at these two positions may divert the nascent polypeptide into an alternative folding pathway that traps MotB in a non-functional conformation. We further propose that MotA and MotB form a stable pre-assembly complex in the membrane. In this complex, MotB exists in a form that cannot associate with peptidoglycan and blocks the proton-conducting channel. Opening of the channel and attachment to the cell wall may occur when the complex collides with a flagellar basal body and MotA makes specific contacts with the C ring and/or the MS ring.

Function of proline residues of MotA in torque generation by the flagellar motor of Escherichia coli.

Bacterial flagellar motors obtain energy for rotation from the membrane gradient of protons or, in some species, sodium ions. The molecular mechanism of flagellar rotation is not understood. MotA and MotB are integral membrane proteins that function in proton conduction and are believed to form the stator of the motor. Previous mutational studies identified two conserved proline residues in MotA (Pro 173 and Pro 222 in the protein from Escherichia coli) and a conserved aspartic acid residue in MotB (Asp 32) that are important for function. Asp 32 of MotB probably forms part of the proton path through the motor. To learn more about the roles of the conserved proline residues of MotA, we examined motor function in Pro 173 and Pro 222 mutants, making measurements of torque at high load, speed at low and intermediate loads, and solvent-isotope effects (D2O versus H2O). Proton conduction by wild-type and mutant MotA-MotB channels was also assayed, by a growth defect that occurs upon overexpression. Several different mutations of Pro 173 reduced the torque of the motor under high load, and a few prevented motor rotation but still allowed proton flow through the MotA-MotB channels. These and other properties of the mutants suggest that Pro 173 has a pivotal role in coupling proton flow to motor rotation and is positioned in the channel near Asp 32 of MotB. Replacements of Pro 222 abolished function in all assays and were strongly dominant. Certain Pro 222 mutant proteins prevented swimming almost completely when expressed at moderate levels in wild-type cells. This dominance might be caused by rotor-stator jamming, because it was weaker when FliG carried a mutation believed to increase rotor-stator clearance. We propose a mechanism for torque generation, in which specific functions are suggested for the proline residues of MotA and Asp32 of MotB.

FlbD has a DNA-binding activity near its carboxy terminus that recognizes ftr sequences involved in positive and negative regulation of flagellar gene transcription in Caulobacter crescentus.

FlbD is a transcriptional regulatory protein that negatively autoregulates fliF, and it is required for expression of other Caulobacter crescentus flagellar genes, including flaN and flbG. In this report we have investigated the interaction between carboxy-terminal fragments of FlbD protein and enhancer-like ftr sequences in the promoter regions of fliF, flaN, and flbG. FlbDc87 is a glutathione S-transferase (GST)-FlbD fusion protein that carries the carboxy-terminal 87 amino acids of FlbD, and FlbDc87 binds to restriction fragments containing the promoter regions of fliF, flaN, and flbG, whereas a GST-FlbD fusion protein carrying the last 48 amino acids of FlbD failed to bind to these promoter regions. DNA footprint analysis demonstrated that FlbDc87 is a sequence-specific DNA-binding protein that makes close contact with 11 nucleotides in ftr4, and 6 of these nucleotides were shown previously to function in negative regulation of fliF transcription in vivo (S. M. Van Way, A. Newton, A. H. Mullin, and D. A. Mullin, J. Bacteriol. 175:367-376, 1993). Three DNA fragments, each carrying an ftr4 mutation that resulted in elevated fliF transcript levels in vivo, were defective in binding to FlbDc87 in vitro. We also found that a missense mutation in the recognition helix of the putative helix-turn-helix DNA-binding motif of FlbDc87 resulted in defective binding to ftr4 in vitro. These data suggest that the binding of FlbDc87 to ftr4 is relevant to negative transcriptional regulation of fliF and that FlbD functions directly as a repressor. Footprint analysis showed that FlbDc87 also makes close contacts with specific nucleotides in ftr1, ftr2, and ftr3 in the flaN-flbG promoter region, and some of these nucleotides were shown previously to be required for regulated transcription of flaN and flbG (D. A. Mullin and A. Newton, J. Bacteriol. 175:2067-2076, 1993). Footprint analysis also revealed a new ftr-like sequence, ftr5, at -136 from the transcription start site of flbG. Our results demonstrate that FlbD contains a sequence-specific DNA-binding activity within the 87 amino acids at its carboxy terminus, and the results suggest that FlbD exerts its effect as a positive and negative regulator of C. crescentus flagellar genes by binding to ftr sequences.

Identification of the promoter and a negative regulatory element, ftr4, that is needed for cell cycle timing of fliF operon expression in Caulobacter crescentus.

The fliF operon of Caulobacter crescentus, which was previously designated the flaO locus, is near the top of the flagellar-gene regulatory hierarchy, and it is one of the earliest transcription units to be expressed in the cell cycle. In this report, we have identified two cis-acting sequences that are required for cell cycle regulation of fliF transcription. The first sequence was defined by the effects of three 2-bp deletions and five point mutations, each of which greatly reduced the level of fliF operon transcript in vivo. These eight mutations lie between -37 and -22 within an 18-bp sequence that matches, at 11 nucleotides, sequences in the 5' regions of the flaQR (flaS locus) and fliLM operons, which are also expressed early and occupy a high level in the regulatory hierarchy (A. Dingwall, A. Zhuang, K. Quon, and L. Shapiro, J. Bacteriol. 174:1760-1768, 1992). We propose that this 18-bp sequence contains all or part of the fliF promoter. We have also identified a second sequence, 17 bp long and centered at -8, which we have provisionally designated ftr4 because of its similarity to the enhancer-like ftr sequences required for regulation of sigma 54 promoters flaN and flbG (D. A. Mullin and A. Newton, J. Bacteriol. 171:3218-3227, 1989). Six of the seven mutations in ftr4 examined resulted in a large increase in fliF operon transcript levels, suggesting a role for ftr4 in negative regulation. A 2-bp deletion at -12 and -13 in ftr4 altered the cell cycle pattern of fliF operon transcription; the transcript was still expressed periodically, but the period of its synthesis was extended significantly. We suggest that the ftr4 sequence may form part of a developmental switch which is required to turn off fliF operon transcription at the correct time in the cell cycle.

Characterization of the Caulobacter crescentus flbF promoter and identification of the inferred FlbF product as a homolog of the LcrD protein from a Yersinia enterocolitica virulence plasmid.

We have investigated the organization and expression of the Caulobacter crescentus flbF gene because it occupies a high level in the flagellar gene regulatory hierarchy. The nucleotide sequence comprising the 3' end of the flaO operon and the adjacent flbF promoter and structural gene was determined, and the organization of transcription units within this sequence was investigated. We located the 3' ends of the flaO operon transcript by using a nuclease S1 protection assay, and the 5' end of the flbF transcript was precisely mapped by primer extension analysis. The nucleotide sequence upstream from the 5' end of the flbF transcript contains -10 and -35 elements similar to those found in promoters transcribed by sigma 28 RNA polymerase in other organisms. Mutations that changed nucleotides in the -10 or -35 elements or altered their relative spacing resulted in undetectable levels of flbF transcript, demonstrating that these sequences contain nucleotides essential for promoter function. We identified a 700-codon open reading frame, downstream from the flbF promoter region, that was predicted to be the flbF structural gene. The amino-terminal half of the FlbF amino acid sequence contains eight hydrophobic regions predicted to be membrane-spanning segments, suggesting that the FlbF protein may be an integral membrane protein. The FlbF amino acid sequence is very similar to that of a transcriptional regulatory protein called LcrD that is encoded in the highly conserved low-calcium-response region of virulence plasmid pYVO3 in Yersinia enterocolitica (A.-M. Viitanen, P. Toivanen, and M. Skurnik, J. Bacteriol. 172:3152-3162, 1990).