ABSTRACT
The production of type 1 fimbriae in Salmonella enterica serovar Typhimurium is controlled, in part, by three proteins, FimZ, FimY, and FimW. Amino acid sequence analysis indicates that FimZ belongs to the family of bacterial response regulators of two-component systems. In these studies, we have demonstrated that introducing a mutation mimicking phosphorylation of FimZ is necessary for activation of its target gene, fimA. In addition, the interaction of FimZ with FimW, a repressor of fimA expression, occurs only when FimZ is phosphorylated. Consequently, the negative regulatory effect of FimW is most likely due to downmodulation of the active FimZ protein. FimY does not appear to function as a response regulator, and its activity can be lost by mimicking the phosphorylation of FimY. Overproduction of FimY cannot alleviate the nonfimbriate phenotype in a FimZ mutant, whereas high levels of FimZ can overcome the nonfimbriate phenotype of a FimY mutant. It appears that FimY acts upstream of FimZ to activate fimA expression.
INTRODUCTION
Type 1 fimbria production by Salmonella serovars is proposed to play a significant role in the attachment of the bacteria to host cells and tissues (1, 2). These fimbriae mediate binding to mannosylated glycoconjugates located on eukaryotic cell membranes as well as soluble mannose-containing proteins. This attachment process represents the initial stage of colonization, growth, and, in some cases, biofilm formation in vivo. In addition to Salmonella strains, type 1 fimbriae are produced by a broad range of enterobacteria (3, 4). Their mechanism of assembly has been characterized in Escherichia coli, and the structural components of the fimbriae are subject to an ordered assembly process requiring a periplasmic chaperone and an outer membrane scaffolding (usher) protein. Analysis of the type 1 fimbrial gene clusters from other bacteria suggests that the chaperone/usher assembly pathway is invariably used to assemble these appendages (5–8). All of the type 1 fimbrial systems possess gene products that are predicted to encode these two assembly proteins.
Although the mechanism of assembly of type 1 fimbriae appears to be highly conserved among enterobacteria, the regulation of fimbria production at the genetic level is variable (3). Control of fimbrial gene (fim) expression in E. coli is primarily mediated by inversion of the fimS DNA switch possessing the promoter region of the fim operon (9–12). The rate of switching is modulated by two site-specific recombinases encoded by fimB and fimE that are located immediately adjacent to the fimA promoter region. In Klebsiella pneumoniae, fim gene expression is affected by an additional gene, fimK, that is not found in E. coli and is predicted to affect gene expression by modulating the intracellular levels of the second messenger, cyclic di-GMP (13). In Salmonella enterica serovar Typhimurium fim gene expression, there are no site-specific recombinases encoded by genes adjacent to the fim gene cluster and little evidence to indicate the presence of a functional fimS. The major regulatory proteins that control fim expression have been identified as FimZ, FimY, and FimW (14–18). In addition, a tRNA gene, fimU, has been suggested to affect the production of these regulators. FimZ and FimY are both necessary for positive regulation of fimA, which encodes the major structural component of the type 1 fimbriae. Initial comparison of FimZ to other bacterial regulatory proteins indicated that FimZ was an orphan response regulator of a two-component regulatory system (19–23). Subsequently, we and others have shown that FimZ is a DNA-binding protein, and the site of its binding to the promoter region of the fim gene cluster has been mapped (18). FimZ mutants of S. Typhimurium are phenotypically nonfimbriate but exhibit an increased swarming phenotype (14). FimY exhibits little relatedness to other bacterial gene products but is necessary for optimal fimA expression. As for FimZ, FimY mutants are phenotypically nonfimbriate, and both FimZ and FimY appear to be necessary for fimbria production (17, 18).
FimW negatively regulates fimA expression, and FimW mutants exhibited an increased fimbriate phenotype compared to wild-type bacteria and increased fimA gene expression (17). Although FimW was predicted to possess a DNA-binding domain, we were unable to demonstrate binding to target DNA molecules in vitro. However, our previous studies have demonstrated, using a bacterial two-hybrid system, that FimW interacts with FimZ (17). Such an interaction may lead to decreased activation by FimZ in the context of the fimA promoter region. Since FimW interacts with FimZ and both FimY and FimZ are necessary for fimA expression, we decided to investigate whether FimY influences the FimZ-FimW interaction. In addition, because FimZ is related to the family of response regulators, we constructed mutants that would mimic constitutive phosphorylation and no phosphorylation phenotypes in vivo to examine their effect on fimA expression and fimbria production. Finally, we also examined the role of a putative conserved aspartate residue within FimY in its ability to mediate fimbria production when this amino acid was replaced with a residue mimicking constitutive phosphorylation.
MATERIALS AND METHODS
Strains, plasmids, and media.Table 1 lists the bacterial strains and plasmids used in this study. All strains were grown on Luria-Bertani (LB) medium at 37°C, unless otherwise stated. Media were supplemented with ampicillin (100 μg/ml), kanamycin (25 μg/ml), or chloramphenicol (25 μg/ml), as necessary. Manipulation of DNA and construction of plasmids were carried out using standard techniques.
Bacterial strains and plasmids used in this study
Complementation of strains in a single copy was accomplished by integrating a plasmid containing the desired gene into the bacteriophage λ integration site on the strain's chromosome using the conditional-replication, integration, excision, and retrieval (CRIM) plasmid-host system devised by Haldimann and Wanner (24). Briefly, the desired gene was cloned with its native promoter into a pir-dependent CRIM plasmid. When this plasmid was transformed into a non-pir host containing a helper plasmid synthesizing Int, the entire CRIM plasmid was integrated onto the chromosome. The nucleotide primers used for the construction of specific strains can be found in Table S1 in the supplemental material. Primer pair pLAfimZfwd and pLAfimZrev was used to integrate the cloned wild-type fimZ as well as fimZ point mutations, and pLA2fimYfwd and pLAfimYrev were used to integrate wild-type fimY. Other fimZ constructs used for complementation were constructed using primer pair pBADfimZfwd and pBADfimZrev.
Construction of specific fim mutations.Nonpolar deletion mutants of S. Typhimurium were constructed using the method of Datsenko and Wanner (25). Briefly, oligonucleotides were designed to amplify the regions immediately flanking the gene to be deleted and possessing an antibiotic resistance marker and an FLP recognition target (FRT). Appropriate PCR products were directly transformed into S. Typhimurium strains expressing a plasmid-borne Red recombinase. The antibiotic resistance cassette was removed using a plasmid expressing the FLP recombinase, and these plasmids were subsequently cured from the deletion mutant by growth at 42°C. Confirmation that only appropriate sequences had been deleted from each mutant was performed by nucleotide sequence analysis (University of Iowa DNA Sequencing Facility).
Primer pair fimZdelfwd and fimZdelrev was used to delete fimZ, primer pair fimWdelfwd and fimWdelrev was used to delete fimW, primer pair SAZ113 and SAZ114 was used to delete fimY, and primer pair SAZ113 and SAZ136 was used to delete fimZ and fimY. Point mutations in fimZ were constructed by PCR mutagenesis. For each mutation, two PCRs were carried out: the first one with a forward primer flanking the gene and a reverse primer containing the point mutation and a second one with a reverse primer flanking the gene and a forward primer containing the point mutation. Resulting PCR products were ligated together in a subsequent PCR, and 1 μl of each of the original reaction mixtures was included in this step. Mutation-specific primers SAZ137 and SAZ138 were used for fimZL55A, SAZ67 and SAZ68 were used for fimZD56A, SAZ69 and SAZ70 were used for fimZD56E, and SAZ139 and SAZ140 were used for fimZI57A.
Bacterial two-hybrid system.A LexA-based bacterial two-hybrid system using E. coli SU202 and plasmids pMS604 and pDP804 was used (26). The construction of plasmids pISF244fimY-lexA, pISF245fimZ-lexA408, and pISF248fimW-lexA has previously been described by our group (17). Additionally, pISF244fimY-lexA-fimZ, pISF245fimZ-lexA408-fimW, and pISF248fimW-lexA-fimY were constructed using primer pairs SAZ107 and SAZ108, SAZ109 and SAZ110, and SAZ105 and SAZ106, respectively. Combinations of these fim genes were cloned into pDP804 using primers SAZ111 and SAZ112. Appropriate transformants of E. coli SU202 were grown in static liquid LB broth for 48 h, in shaking liquid LB broth for 4 to 5 h, or on LB agar plates, all of which were incubated at 37°C and subsequently induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) before assaying for β-galactosidase production. As previously described, the interacting proteins Jun and Fos were used as positive controls for this system (26).
fim gene expression.RNA was extracted from bacterial strains grown under fimbria-inducing conditions (serial subculture in static liquid broth) using a modified procedure previously described by Chouikha and coworkers (27). After RNA isolation, residual DNA was removed using a DNA-free protocol (Ambion, Austin, TX). The absence of DNA was verified by conventional PCRs in the absence of reverse transcriptase (RT), and cDNA was synthesized from RNA using SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). Transcription of fimA was detected by RT-PCR, and amplicons were detected using conventional electrophoretic techniques through 2% agarose gels.
Production of surface-assembled type 1 fimbriae.Bacterial cultures were serially grown in static liquid broth at 37°C. Bacteria were harvested by centrifugation and resuspended in a small volume of the remaining supernatant, as previously described (4, 28, 29). Monospecific fimbrial antiserum was used to detect the presence of fimbriae on the bacterial surface using procedures described by our group (17, 29).
RESULTS
Aspartate 56 is a site of FimZ phosphorylation required for fimbria production.The amino sequence of FimZ exhibits relatedness to several two-component response regulators (19, 22, 23, 30). FimZ contains a conserved aspartate residue at amino acid 56 which may act as a site of phosphorylation conferring activation of FimZ in the positive control of fimA. In order to examine the effect of FimZ phosphorylation on type 1 fimbria production in SL1344, FimZ aspartate 56 was mutated to either an alanine or a glutamate by site-directed mutagenesis. These mutations have been constructed in several other response regulators to inhibit the ability of the regulator to be phosphorylated or to mimic constitutive phosphorylation, respectively (19, 21, 30–34). Mutated fimZ alleles were integrated onto the chromosome of SL1344ΔfimZ, and the mutants were examined for their ability to produce type 1 fimbriae. The results are shown in Table 2. Complementation of the fimZ mutant by the wild-type allele (fimZ) resulted in restoration of fimbria production under inducing conditions (serial broth cultures) but poor fimbriation when grown on agar. However, fimZ with a D-to-A change at position 56 (fimZD56A) was unable to complement the mutant, and the mutant exhibited a nonfimbriate phenotype even when grown under inducing conditions. The presence of the fimZD56E mutation in the FimZ mutant resulted in the ability of bacteria to produce type 1 fimbriae at levels similar to those for SL1344. In addition, fimbriae were detected on the surface of SL1344ΔfimZ::fimZD56E when it was grown under noninducing conditions (plate cultures) as well as inducing conditions, indicating that FimZ is constitutively active in this construct.
Type 1 fimbria production by S. Typhimurium FimZ strains
To examine further the effects of these mutations on fim gene expression, transcription of fimA and fimZ was investigated in strains SL1344ΔfimZ, SL1344ΔfimZ::fimZ, SL1344ΔfimZ::fimZD56A, and SL1344ΔfimZ::fimZD56E. The results of these assays are shown in Fig. 1. Transcription of fimA and fimZ was significantly decreased in the presence of the fimZD56A allele compared to the level of transcription observed in the other strains. Even though small amounts of the fimA and fimZ transcripts could be detected in the FimZ mutants, this did not correlate with the ability to produce surface-assembled type 1 fimbriae (Table 2).
RT-PCR of RNA from SL1344 strains. Primers specific for rpoD (positive control), fimA, and fimZ were used in these assays. Lanes 1, no template; lanes 2, amplicons from SL13344 genomic DNA as the template; lanes 3 to 10, PCR products obtained using RNA prepared from the strains indicated in the following lane descriptions: lanes 3 and 4, SL1344ΔfimZ with and without RT, respectively; lanes 5 and 6, SL1344ΔfimZ::fimZ with and without RT, respectively; lanes 7 and 8, SL1344ΔfimZ::fimZD56A with and without RT, respectively; lanes 9 and 10, SL1344ΔfimZ::fimZD56E with and without RT, respectively.
Additionally, the fimA promoter activity in the four strains was determined using the fimA lacZ promoter fusion plasmid pISF145 (Fig. 2). The activity of the fimA promoter is greatly decreased when FimZ aspartate 56 is mutated to an alanine (SL1344ΔfimZ::fimZD56A). No β-galactosidase activity was detected in the FimZ mutant, but the strains possessing the fimZD56E mutation exhibited β-galactosidase activity at levels similar to those of the parental strain, SL1344.
β-Galactosidase production by the reporter plasmid pfimZlacZ transformed into SL1344 and its derivatives. Statistically significance differences between strains were determined using Student's t test. A significant difference (***, P < 0.001) in enzyme expression between the two FimZ mutants carrying the D56A and D56E alleles was observed. No significant difference in galactosidase production by the FimZ mutant carrying the wild-type (WT) and D56E alleles was detected.
FimZ aspartate 56 mutants alter the ability of FimZ to interact with FimW.We have previously reported that FimZ interacts with FimW to modulate fimA gene expression (17). The ability of the mutations in fimZ to alter FimZ's ability to bind to FimW was investigated, and the results are shown in Fig. 3. In the LexA-based bacterial two-hybrid system, protein interactions are detected by repression of β-galactosidase activity, as indicated in Fig. 3, using the strongly interacting proteins Jun and Fos as positive controls (26). The fimZD56A mutation results in the production of a FimZ molecule that has a significantly decreased ability to interact with FimW (Fig. 3). However, FimZD56E retained the ability to interact with FimW. In order to confirm that the decreased interaction of FimZD56A is specific for this amino acid, we also constructed mutations (FimZL55A and FimZI57A) immediately adjacent to aspartate 56 and examined their ability to interact with FimW. As can be seen in Fig. 3, these two mutations did not alter the FimZ-FimW interaction. In all cases, the presence of single fim alleles (fimW or fimZ) in the two-hybrid systems resulted in β-galactosidase activity equivalent to that observed in the presence of Jun or Fos alone.
Bacterial two-hybrid assay of FimW and FimZ protein interactions. Repression of β-galactosidase activity indicates interaction between pairs of proteins. The results for plasmids encoding single proteins also demonstrate the lack of repression due to the presence of the recombinant plasmid alone. The only significant difference in the interaction between FimZ and FimW was detected using the FimZD56A mutant with FimW, as determined by the Student t test (***, P < 0.001). None of the other combinations of FimW with FimZ mutants exhibited any significant differences compared to the FimW and FimZ wild-type (FimW-FimZ) interaction.
FimY does not affect the FimZ-FimW interaction.In previous studies, we have shown that FimZ and FimW are the only two of the three regulatory proteins in which an interaction can be demonstrated in pairwise combinations (17). Therefore, we investigated the ability of FimY to affect the FimZ-FimW interaction. To examine the effect of FimY on this interaction, fimY was cloned downstream of fimW in pMS604. Also, fimZ was cloned with fimW and fimY, and finally, fimW was cloned with fimY and fimZ. In all cases, expression of the fimY and the fimZ genes in these constructs was confirmed by complementation of SL1344ΔfimY or SL1344ΔfimZ, respectively, to restore the fimbrial phenotype, as previously described by us (16–18). The results of these assays are shown in Fig. 4. In all cases examined, the presence of a functional fimY gene did not affect the FimZ-FimW interaction. As previously reported by us, no demonstrable interaction between FimZ-FimY or FimW-FimY was observed. Even in the presence of fimW, no interaction of FimZ with FimY can be shown (Fig. 4).
Interactions for FimW, FimZ, and FimY. The proteins (separated by slashes) encoded by the pair of plasmids in the two-hybrid system are shown. For plasmids possessing two fim genes (FimYZ, FimWY, and FimZW), the second gene was cloned downstream of the first fim gene that exists as a lex-fim fusion. A functional gene product, i.e., FimZ in the FimYZ construct, FimY in the FimWY construct, and FimW in the FimZW construct, was confirmed by complementation of the appropriate deletion mutant. A significant interaction (**, P < 0.05) of FimZ with FimW was observed only when FimZ was produced as a Lex-Fim fusion molecule (FimW-FimZY or FimWY-FimZ) in this system, and no interaction could occur when it was produced as a native protein (FimW-FimYZ).
FimY acts upstream of FimZ.The roles of FimY and FimZ in relation to each other were further examined using the single-deletion mutants and also the SL1344ΔfimZY double mutant. All three mutants are nonfimbriate. Integration of either fimZ or fimY onto the chromosome of the homologous mutant resulted in restoration of the fimbrial phenotype (Table 3). The presence of both fimZ and fimY as single copies was required for complementation of the double mutant, whereas neither gene alone carried on the chromosome could complement this strain (Table 3). The fimZD56E allele that confers constitutive fimbria production on the FimZ mutant (see above) did not restore the ability of either SL1344ΔfimY or SL1344ΔfimZY to produce type 1 fimbriae, indicating that FimY must be present for its activity. However, if either fimZ or fimZD56E was introduced into SL1344ΔfimY or SL1344ΔfimZY carried on multicopy plasmids (pfim plasmids), these transformants were fimbriate (Table 3). Overexpression of FimY carried on a multicopy plasmid restored the fimbrial phenotype only in a FimY mutant and not in the other two strains (SL1344ΔfimZ and SL1344ΔfimZY). To further investigate whether overproduction of FimZ could alleviate fimbria production in a FimY mutant, the expression of fimZ was placed under the control of an arabinose-inducible promoter on plasmid pBADfimZ. In the presence of arabinose, SL1344ΔfimY transformants were strongly fimbriate, whereas in the absence of arabinose, these transformants were phenotypically nonfimbriate (Table 3).
Type 1 fimbria production by S. Typhimurium regulatory mutants and complemented strains
The predicted amino acid sequence of FimY exhibits little relatedness to other bacterial proteins. Recently, two hypothetical proteins of the recently sequenced genome of Enterobacter have been suggested to belong to the family of response regulators, and these two proteins possess a limited degree of sequence similarity to FimY. Our analysis of the amino acid sequence of FimY does not indicate that it is closely related to the family of response regulators. However, the site of an aspartate at position 58 in FimY is the only location of the conserved phosphorylation site for these proteins. Therefore, a FimYD58E mutation was constructed and integrated onto the chromosome of both the SL1344ΔfimY and SL1344ΔfimZ mutants. In both cases, the strains were unable to produce type 1 fimbriae (data not shown).
The absence of FimW does not relieve the necessity for FimZ and FimY coactivation.We have previously demonstrated that FimW is a repressor of fimA expression and type 1 fimbria production and interacts with FimZ (17). Therefore, to determine if the absence of this repressive activity could alleviate the necessity of FimZ and FimY to function as coactivators, we constructed mutants unable to produce either FimW and FimY or FimW and FimZ (SL1344ΔfimZW and SL1344ΔfimYW, respectively). Neither of these strains exhibited the ability to produce fimbriae. However, type 1 fimbriae could be produced by these strains upon the introduction of a functional fimY or fimZ gene into the appropriate strain (Table 3).
DISCUSSION
Type 1 fimbriae are produced by many strains of enterobacteria and have been implicated in facilitating attachment to mannosylated glycoconjugates in order to facilitate colonization of host cells and tissues (1, 29, 35, 36). The control of production of these appendages occurs primarily at the level of gene expression and involves distinct regulatory properties, depending upon the bacterial species. In S. Typhimurium, fimA expression is controlled by the products of three fimbrial genes, fimZ, fimY, and fimW. FimZ is a positive activator of fimA expression and has been shown to bind to the promoter region of fimA (18). FimY is a coactivator of fimA expression but bears little relatedness to other transcriptional activators and has not been demonstrated to be a DNA-binding protein. FimW represses fimbria production and interacts with FimZ but not FimY (17). Although FimW possesses a helix-turn-helix DNA-binding motif in its C-terminal region, we and others have not been able demonstrate DNA binding to fim targets using purified FimW alone (data not shown) (15). Possibly, if FimW is a DNA-binding protein, it binds to its target only in the context of interacting with FimZ. In order to more fully understand the relationship between the three fimbrial regulatory proteins, we investigated whether FimY influences the ability of FimW to interact with FimZ and also to determine if FimY is necessary for FimZ activity. In addition, FimZ has been identified as being closely related to response regulators of two-component systems using amino acid homology analyses. Therefore, in these studies, we also examined the role of a conserved phosphorylation site on FimZ activity.
In response regulators, the phosphorylation of a conserved aspartate residue in the N-terminal region of these proteins has been shown to be important in facilitating transcriptional control of target genes (19, 20, 30–33). FimZ possesses an aspartate (D56) in the region frequently associated with the phosphorylation sites in other regulatory proteins, and we mutagenized this site using substitutions that have been used by many groups to mimic constitutive phosphorylation (D56E) or a null phenotype (D56A) in this group of proteins (20, 32, 34). Subsequently, the mutated fimZ alleles were integrated onto the chromosome of S. Typhimurium strains to obviate the effect of these genes being present in multiple copies. As expected, a FimZ mutant possessing a wild-type copy of fimZ was fully fimbriate and produced type 1 fimbriae under the same conditions as the parental strain, SL1344. A strain possessing the D56E mutation was constitutively fimbriate even under noninducing conditions on solid medium, whereas the D56A mutation abrogated fimbria production completely. These results strongly suggest that the phosphorylation of FimZ is necessary for its activity, although we cannot rule out the possibility that the latter mutation destabilizes FimZ. However, we think that this is unlikely, since genes with the D-to-E mutation and mutations immediately adjacent to the D56 amino acid retain functional activity and are therefore stable. FimZ is an orphan response regulator, and its cognate sensor kinase has yet to be identified. Examination of the S. Typhimurium genome for gene products possessing conserved regions associated with the histidine kinase activity of sensor kinases indicates the presence of at least 31 possible sensory kinases. To date, we have constructed mutations in four of these genes (yfhK, vgiY, ssrA, and torS), but all mutants are fully fimbriate. Consequently, the precise signaling molecule that mediates the phosphorylation of FimZ has yet to be defined, and therefore, the environmental signal that modulates FimZ activity is unknown.
The effect of phosphorylation on the interaction of FimZ with FimW was investigated using a bacterial two-hybrid system. Initially, we believed that the interaction of FimW with FimZ may be most stable when FimZ is not activated (FimZD56A). However, this was not observed, with a significant decrease in the interaction between these two proteins compared to that between FimW and the wild-type FimZ or FimZD56E being found. This would suggest that FimW has a higher affinity for the phosphorylated or activated FimZ. Possibly, in its inactive state (FimZD56A), modulation of FimZ activity by FimW is not necessary, but once it is phosphorylated, FimW may decrease its activity by interacting with FimZ. As mentioned above, FimW has not been shown to bind to any DNA targets, and its repressive activity may be one of inhibiting FimZ alone to activate fimA expression by facilitating binding of the FimZ-FimW complex to a different site. We are currently attempting to purify FimZ-FimW complexes to investigate this possibility.
We have demonstrated that when they are present in a single copy, both FimZ and FimY are necessary for type 1 fimbria production. One possible role of FimY in this process is to modulate the ability of FimZ to interact with FimW. This was investigated by introduction of a functional FimY into the two-hybrid system encoding FimZ and FimW. Regardless of the plasmid carrying fimY in this system, there was no observable decrease in FimZ-FimW interaction. FimY is produced in this system, since the plasmids carrying fimY could be used to complement a FimY mutant of SL1344. Recently, whole-genome sequencing of an Enterobacter strain has indicated the limited relatedness of FimY to two proteins (NCBI locus numbers YP_003611795.1 and YP_001175729.1) that were putatively identified as belonging to the family of sensory regulators. Examination of the amino acid sequence of FimY does not indicate any known receiver domains or DNA-binding domains. Also, mutation of an aspartate residue at the site in this protein most likely to be phosphorylated to mimic phosphorylation (FimYD58E) resulted in a protein that lost the ability to activate fimA. Therefore, from our studies, the role of FimY in type 1 fimbria production does not appear to be one of acting as a response regulator or influencing the FimZ-FimW interaction. Regardless, FimY may act upstream of FimZ, since overproduction of FimZ in SL1344ΔfimY could overcome the lack of FimY. However, overproduction of FimY could not compensate for the lack of FimZ even when FimW was absent. Whether under normal physiological conditions FimY is involved in the phosphorylation of FimZ has yet to be determined.
In addition to regulating fimA expression, it has been suggested that FimZ and FimY are autoregulatory proteins and also can affect expression of fimY and fimZ, respectively, as well as that of fimW (15). Consequently, the regulatory circuit of type 1 fimbria production in S. Typhimurium appears to involve complex interactions involving feedback loops and activating/inhibitory modulations (3, 15, 18). To date, these regulators have not been detected in fim systems other than the fim system in Salmonella and most likely represent a pathway that has evolved only in this genus. Also, the molecules involved in the transduction of environmental signals and cues remain to be elucidated. FimZ has been shown not only to play a role in type 1 fimbria production by S. Typhimurium but also to affect flagella and virulence gene expression in these bacteria (14, 37). Considering the interaction of FimZ with FimW, it is possible that FimW may also have a wider regulatory role beyond its involvement in adhesin production.
Finally, our investigations indicate that phosphorylated FimZ is necessary for optimal activation of fimA and that this activation may be modulated by interaction with FimW. Also, we suggest that FimY acts upstream of FimZ in the regulatory pathway but does not interact directly with FimZ or FimW. However, it has been shown in many enterobacterial systems that control of fimbrial gene expression involves numerous regulatory pathways (9, 12, 13). Therefore, the functions of FimZ, FimY, and FimW must be considered in the context of integrated regulatory circuits controlling type 1 fimbria production in Salmonella. For example, studies by Baek et al. (38) clearly demonstrate that FimZ activity is affected by the intracellular concentration of Lrp that may have both a positive and a negative regulatory effect on type 1 fimbria production in Salmonella. Since Lrp and H-NS can also modulate fimA expression, the regulatory circuit of fim expression will involve both specific fimbrial regulators and global regulatory proteins that act in concert to modulate adherence factor production.
ACKNOWLEDGMENTS
This work was supported in part by NIH grants RO1 RO1GM084318 and RO1AI074693 to S.C.
FOOTNOTES
- Received 25 June 2013.
- Returned for modification 3 August 2013.
- Accepted 10 September 2013.
- Accepted manuscript posted online 16 September 2013.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00795-13.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
