Synchronous Gene Expression of the Yersinia enterocolitica Ysa Type III Secretion System and Its Effectors

Abstract

Type III secretion systems (T3SSs) are complex units that consist of many proteins. Often the proteins are encoded as a cohesive unit on virulence plasmids, but several systems have their various components dispersed around the chromosome. The Yersinia enterocolitica Ysa T3SS is such a system, where the apparatus genes, some regulatory genes, and four genes encoding secreted proteins (ysp genes) are contained in a single locus. The remaining ysp genes and at least one additional regulator are found elsewhere on the chromosome. Expression of ysa genes requires conditions of high ionic strength, neutral/basic pH, and low temperatures (26°C) and is stimulated by exposure to solid surfaces. The AraC-like regulator YsaE and the dual-function chaperone/regulator SycB are required to stimulate the sycB promoter, which transcribes sycB and probably yspBCDA as well. The putative phosphorelay proteins YsrRS (located at the distal end of the ysa locus) and RcsB, the response regulator of the RcsBCD phosphorelay system, are required to initiate transcription at the ysaE promoter, which drives transcription of many apparatus genes. In this work, we sought to determine which ysp genes were coordinately regulated with the genes within the ysa locus. We found that six unlinked ysp genes responded to NaCl and required YsaE/SycB, YsrRS, and RcsB for expression. Three ysp genes had unique patterns, one of which was unaffected by all elements tested except NaCl. Thus, while the ysp genes were likely to have been acquired independently, most have acquired a synchronous regulatory pattern.

Yersinia enterocolitica is a gram-negative enteric pathogen with a tropism for lymphoid tissues. Typically ingested through contaminated food or water, it is usually self-limiting to the gastrointestinal tract and mesenteric lymph node, causing diseases that include gastroenteritis and lymphadenitis. In children or immunocompromised adults, a systemic infection can develop that is associated with high mortality (6).

Y. enterocolitica strains are separated into several biovars based on biochemical properties, and there is variability in the severity of diseases caused by the different biovars. Biovars 2 to 5, considered the Old World strains, are found mostly in Europe and Japan and are classified as having low pathogenicity (42). Biovar 1A, also an Old World strain, has lost the virulence plasmid and is nonpathogenic (3). New World strains, isolated mostly in North America, are those of biovar 1B and are the highly pathogenic strains (42). All the pathogenic strains carry an ∼70-kb virulence plasmid, called pYV, that encodes the Ysc-Yop type III secretion system (T3SS), the adhesin YadA, as well as regulators of these important virulence genes (reviewed in reference 5).

The highly pathogenic biovar 1B strains contain a large region (∼200 kb) of the chromosome that is not found in other Yersinia strains (33). It varies considerably among different Y. enterocolitica isolates and has been termed the plasticity zone. Several defined and predicted virulence determinants are contained in the plasticity zone, including a region similar to the Yersinia pseudotuberculosis adhesion pathogenicity island, the Yts1 type II secretion system, and the Ysa T3SS (33). The two secretion systems have been shown to be required for full virulence in oral infection models (17, 18, 23). Studies using a mouse model have indicated about a 10-fold increase in the 50% lethal dose of ysa mutants compared to that of wild-type strains, and ysa mutants are defective for colonization of the intestinal tract (17, 23). This defect is observed only for oral inoculations; no attenuation has been observed in mice infected by intraperitoneal injection (17).

The ysa locus contains genes encoding all the components of a type III secretion apparatus, a chaperone (sycB), four secreted proteins (yspBCDA), and at least four regulatory proteins (ysaE, sycB, ysrR, and ysrS) (15, 17, 23, 38) (Fig. 1). Additional ysp genes are found dispersed around the chromosome (23, 41). This T3SS is a member of the Inv/Mxi-Spa family of T3SSs and shares not only homology but gene organization with the systems in Salmonella SPI-1 and Shigella spp. (36). In addition, closely related systems are found in Sodalis glossinidius, a plant symbiont (7), and in Chromobacterium violaceum, which is found primarily in soil in tropical regions but can cause lethal infections in humans (4).

FIG. 1.

Schematic representation of the ysa locus. The region presented is ∼30 kb and is comprised of genes encoding the apparatus (designated ysa; black boxes), four secreted proteins (ysp; white boxes), and four regulatory proteins (gray boxes). SycB has dual function of chaperone and regulator (striped box). Stippled boxes indicate genes whose role/function is unknown. The ysaE and sycB promoters are indicated by arrows.

The regulatory protein YsaE is a member of the AraC-like family and is required, along with SycB, for expression of a promoter upstream of sycB (38). SycB has similarity to chaperone proteins, and deletion of sycB was shown to result in smaller amounts of YspB, suggesting it indeed functions as a chaperone (15). Thus, SycB has dual functionality: transcriptional regulator and chaperone. This mechanism of regulation has been observed in Salmonella enterica serovar Typhimurium with InvF/SicA and in Shigella flexneri with MxiE/IpgC (9, 10, 24, 25). In both of these systems, direct interactions between the AraC-like protein and the chaperone have been observed (10, 31).

A promoter upstream of ysaE likely drives transcription of all genes from ysaE to yspA (38) (Fig. 1). High levels of NaCl in the growth medium induce secretion by the Ysa T3SS (17), and it was subsequently shown that the ysaE promoter was activated by NaCl (37, 38). The Ysa T3SS is regulated by the putative phosphorelay proteins YsrS and YsrR that are encoded by genes adjacent to the ysa locus (Fig. 1) and by the known phosphorelay protein RcsB (37, 38). YsrS, a putative hybrid sensor kinase, was shown to activate the ysaE promoter to the same level as NaCl and is hypothesized to sense the NaCl levels (38). YsrR is a putative response regulator that was shown to be required for Ysa-dependent secretion (38). RcsB is the response regulator of the RcsCDB system and is a DNA binding protein that can activate transcription alone or as a heterodimer with RcsA (reviewed in reference 22). This phosphorelay system has been implicated in the regulation of T3SS and other virulence determinants in several gram-negative bacterial species (2, 21, 34, 39). In addition, CRP has been implicated in the positive regulation of the ysa locus, but the mechanism remains unclear (30). Furthermore, while expression from the ysaE promoter is dependent on high levels of NaCl, it can be further stimulated by exposure to solid surfaces (27).

Of the four Ysp proteins encoded within the ysa locus, YspB, YspC, and YspD are homologous to “translocon” proteins that form a pore in host cells to allow for translocation of effector proteins (15, 17). YspA is unique, although the proteins encoded by genes in this relative position in the Salmonella (SipA) and Shigella (IpaA) systems have been defined and shown to affect actin polymerization (35, 44). In addition to these four proteins, Matsumoto and Young have identified seven other proteins secreted by this system (23). Two Ysp proteins showed homology to known proteins, and their biochemical activities were demonstrated in vitro; YspP is a phosphatase and YspK is a serine/threonine kinase (23). The remaining ysp genes showed no homology to known genes, and only yspY had a homolog in other bacterial species (Yersinia pestis and Y. pseudotuberculosis) (23). Since type III-secreted proteins are often coordinately regulated with the apparatus genes (9, 10, 20, 24, 25), the focus of this study was to examine the transcriptional expression and regulatory patterns of the newly identified ysp genes. To this end, we determined the expression pattern of ysp genes in a wild-type strain and examined their expression levels in several regulatory mutant strains. We found that most, but not all, ysp genes displayed a common regulatory expression pattern with the apparatus genes.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Bacterial strains and plasmids used in this work are listed in Table 1 and described below. All cultures of Escherichia coli strains were grown in LB medium (170 mM NaCl; Difco) at 37°C. Cultures of Y. enterocolitica were grown at 26°C in LB medium, L broth (1% tryptone, 0.5% yeast extract, 0 mM NaCl), or L broth with 290 mM NaCl (referred to as LB-290). Antibiotics were added as needed at the following concentrations: kanamycin (Kan), 100 μg/ml; nalidixic acid, 20 μg/ml; chloramphenicol (Cm), 12.5 μg/ml; and spectinomycin (Sp), 100 μg/ml.

TABLE 1.

Strains and plasmids used in this work

Strain or plasmid	Relevant characteristics	Reference
Strains
E. coli
DH5α	F− φ80dΔlacZΔM15 Δ(lacZYA-argF)U169 deoP recA1 endA1 hsdR17(rK− mK−)	Invitrogen
S17-1λpir	Tpr StrrrecA thi pro hsdR hsdM+ RP4::2-Tc::Mu::Km Tn7 λpir lysogen	28
Y. enterocolitica
JB580v	8081v (r− m+ Nalr)	19
YVM932	JB580v ΔysaE	38
YVM981	JB580v ΔsycB	38
YVM969	JB580v ΔysrS	38
YVM1006	JB580v ΔysrR	38
YVM1236	JB580v ΔrcsB	41
YVM1321	YVM969 carrying pKW22 and pKW28	This work
YVM1349	JB580v carrying pWKS130 and pKW27	This work
YVM1350	JB580v carrying pKW22 and pKW28	This work
YVM1351	YVM969 carrying pWKS130 and pKW27	This work
YVM1352	YVM1006 carrying pWKS130 and pKW27	This work
YVM1353	YVM1006 carrying pKW22 and pKW28	This work
YVM1041	JB580v ysrR-lacZYA	This work
Plasmids
pPROBE-gfp[tagless]	Kanr; gfp transcriptional fusion reporter	29
pET24b	Kan; overexpression vector with C-terminal His6 tag	NEB
pSR47S	Kanr; MobRP4 oriR6K, cloning vector	26
pWKS130	Kanr; low-copy-number cloning vector	40
pKW27	Cmr Strr Spr; pACYC184 with Str/Sp cassette	38
pKN8	Cmr; MobRP4 oriR6K, transcriptional reporter vector	14
pKW22 (pYsaE)	ysaE coding sequence and promoter cloned into pWKS130	38
pKW28 (pSycB)	sycB coding sequence and promoter cloned into pKW27	38
pKW53	rcsB lacking codons 7-211 cloned into pSR47S	41
pKW71	pPROBE-gfp[tagless] with ysrR promoter region	This work
pKW15	ysrR promoter region cloned into pKN8	This work
pMWO-002	ysrR coding region cloned into pET24b	This work

To examine transcript levels during the course of the growth curve, triplicate-saturated cultures of wild-type Y. enterocolitica JB580v grown in L broth were diluted to an initial optical density at 600 nm (OD₆₀₀) of 0.2 in 100 ml fresh L broth or LB-290 and grown at 26°C with aeration (in 250-ml flasks). At the specified times, a volume of cells was transferred to Oak Ridge tubes, snap-cooled in liquid nitrogen, and collected by centrifugation, and the pellets were stored at −80°C. For determination of regulation by YsaE, SycB, YsrR, YsrS, and RcsB, strains harboring in-frame deletions of each gene were similarly prepared but subcultured only into LB-290. These same culturing conditions were performed to determine if YsaE and SycB could complement a ysrS or ysrR deletion, but only the 2-h samples were collected for transcript analysis.

To ascertain how quickly gene expression responded to NaCl, saturated overnight cultures of JB580v grown in L broth were diluted to an OD₆₀₀ of 0.2 in 250 ml fresh L broth and grown for 2 h at 26°C (in a 1-liter flask). The culture was split into two equal volumes, and the cells were collected by centrifugation. One pellet was resuspended in 125 ml fresh L broth and the other in 125 ml LB-290 (in 250-ml flasks), and the cultures were grown at 26°C with aeration. To avoid cold shock, the samples were kept at room temperature during all handling. Samples were taken from each set at 0, 20, 40, 60, 90, and 120 min after resuspension, snap-cooled, and collected by centrifugation, and the pellets were stored at −80°C.

Plasmid and strain construction.

The plasmids and strains used in this study are listed in Table 1, and the primers used are listed in Table 2. A ysrR-gfp transcriptional fusion was constructed as follows. About 500 bp of the putative ysrR promoter region was amplified with primers ysrR-delA and ysrR-delB, digested with SalI and BamHI, ligated into the same sites of pPROBE-gfp[tagless], and transformed into E. coli DH5α. The resulting plasmid, pKW71, was confirmed by digest and sequencing and transformed into the desired Y. enterocolitica strains by electroporation. This same fragment was cloned into pCR2.1-TOPO (Invitrogen), from which an XbaI-BglII fragment was excised and cloned into pKN8 to generate pKW15. pKW15 was conjugated into JB580v as previously described (38), resulting in strain YVM1041. Plasmid pMWO-002 was generated for the overexpression of YsrR-His. Primers MWO-002 and MWO-003 were used to amplify the coding region of ysrR; the resulting product was cleaved with NheI and XhoI and cloned into those sites of pET24b.

TABLE 2.

Primers used in this work

Namea	Sequenceb (5′ → 3′)
ysrR-delAc	ACGCGTCGACGCAGGATAATCCGATGAAATCTCG
ysrR-delBc	CGGGATCCCATCAGCGCAAGGCGACTGAAAGG
yspA F	GGAATTCCATATGCCTAATATCATGGCACCCAGTAC
yspA R	GCTCTAGAGTAGAATTCATTTGCTGCACCCATC
yspE F	CCGCTCGAGATGAGTAGGATAAGTCAAAGC
yspE R	GCTCAAGTTTATAACGGCCACC
yspF F	ATGACGCCAGCAAAAATTAGTTCGAG
yspF R	CTGAACTATTTCCTGTTGAATAGCAG
yspK F	CCGCTCGAGATGAAAATAACACCTACCATT
yspK R	GGAGTCCCGGGCACGCGTAGC
yspL F	ATGAATATTAAAGAGCGTATTAATTTC
yspL R	GCAATAATGCATATTCATCAGCAGAC
yspM F	CCGCTCGAGATGAGTATTAATTTTAACCACAAC
yspM R	GGTGAGGTCAGGAAATCAACCC
yspP F	ATGCTTACAAAAAACATCGCCCCAACACC
yspP R	ATTAGTATCACCAAGCGTACTGAG
yspY F	GGTTATTGTTATGGCCGATTATTTC
yspY R	CCTTTATTGGCTTGAGCTTCCCAC
ysaE F	GCCGGTAACGCCAGTGGCTCATTG
ysaE R	GTCGCATCATCTGGTGCATCAAGG
sycB F	GAACCAGAAACATGATGCGGCAG
sycB R	GCCAAAGAGTAAAGTTCCACTG
rcsB F	GACCATCCAATTGTGTTGTTTGGC
rcsB R	GCGTCGACCAGTTACCAGGAAACCTTCAGC
16S	CCGTCCGCCGCTCGCCGGCAA
MWO-002	CTACTAGCTAGCATGACACAAACGAAAACGC
MWO-003	GACTGCTCGAGTAGAGAAATTTCATGAGC

^aF, forward primers; R, reverse primers.

^bRestriction sites are underlined. Some primers used to generate products for Northern probes contain restriction sites but are not noted because the products were not digested.

^cThese primers were originally published in reference 38 and are included here for completeness.

RNA extraction and Northern blot analysis.

Total cellular RNA was extracted from cell pellets that had been stored at −80°C using TRIzol LS reagent (Invitrogen). TRIzol was added to the frozen cell pellet to a density of 10 OD₆₀₀ units/ml, and the remaining steps of the extraction were performed as described by the supplier. Total RNA was treated with DNase using DNA-free, following the manufacturer's protocol (Ambion). For Northern analysis gels, 10 μg RNA was separated in 1% formaldehyde agarose gels and transferred to nitrocellulose membranes by capillary action (32). For dot blots, 5 μg RNA was mixed with 50 μl of denaturing buffer (50% formamide, 7% formaldehyde, 1× SSC; 1× SSC is 0.18 M NaCl, 10 mM NaH₂PO₄, and 1 mM EDTA [pH 7.7]), and transferred to nitrocellulose membranes by vacuum suction using a dot blot apparatus (32). Gene-specific PCR products were labeled with [α-³²P]dATP by random-primed labeling following the supplier's protocol (Roche), and blots were hybridized in 50% formamide buffer. To control for equivalent loading, the 16S RNA was probed. An oligonucleotide probe (KW157) was labeled with [γ-³²P]ATP by T4 polynucleotide kinase. Hybridization and washes were conducted in sodium pyrophosphate buffer (hybridization buffer, 0.1% sodium pyrophosphate, 4× SSC, 5× Denhardt's reagent, 0.1% sodium dodecyl sulfate [SDS], 100 μg/ml salmon sperm DNA; and wash buffer, 0.1% sodium pyrophosphate, 4× SSC, 0.1% SDS) (Stephanie Strand, personal communication). Specific mRNA signals were quantified using a FLA-5000 phosphorimager and ImageGauge v4.22 software (Fuji Film Medical Systems, Stamford, CT). The gene-specific signal was normalized by dividing it by the 16S signal. Triplicate values were averaged and normalized to the peak of expression (for time course analysis) or to the wild-type levels (for regulation analysis). The expression differences relative to those of the wild type (set to 100) were determined by dividing the value obtained from the mutant strain by that from the wild-type strain. For the only exception to this protocol, see Fig. 4; the data were not normalized to 16S RNA and are presented as qualitative only.

FIG. 4.

Regulation of ysp expression by YsrRS is indirect. Triplicate-saturated cultures of the wild-type (WT; JB580v), ΔysrS (YVM969), and ΔysrR (YVM1006) strains alone, carrying plasmids pWKS130 and pKW27 (vectors), or pYsaE and pSycB were grown as described in the legend to Fig. 2 for 2 h in LB-290. Total RNA was extracted and used for qualitative Northern dot blot analysis as described in Materials and Methods. pYsaE is pKW22, and pSycB is pKW28.

Green fluorescent protein assays.

Saturated cultures of Y. enterocolitica strains carrying pKW71 or pPROBE-gfp[tagless] grown in L broth were diluted to an OD of 0.2 in 100 ml LB-290 or L broth and cultured at 26°C with shaking. At various times, samples were removed for measurement of absorbance and fluorescence. Fluorescence (excitation at 485 nm, emission at 528 nm) was measured using a Synergy HT plate reader (BioTek, Winooski, VT), and absorbance at 600 nm was measured with a spectrophotometer. Fluorescence values in relative light units (RLU) were divided by the culture OD to generate RLU/OD and averaged. The average RLU/OD for the vector control was subtracted from those of strains carrying the promoter fusion.

β-Galactosidase assays.

Saturated cultures grown overnight in L broth were diluted into fresh L broth or LB-290 to an initial OD₆₀₀ of 0.2 and grown at 26°C with shaking. Assays were performed as described previously (32).

Western blot analysis.

Saturated cultures of JB580v grown in L broth were subcultured into 100 ml fresh L broth or LB-290 and grown with aeration at 26°C. Samples were collected at 2, 4, 6, and 8 h and pelleted. Cell pellets were resuspended in 1× sample buffer (32), boiled, and loaded onto SDS-polyacrylamide gel electrophoresis gels. Proteins were transferred to nitrocellulose for Western blot analysis using a Bio-Rad Trans-Blot SD semidry transfer apparatus. Blots were blocked in phosphate-buffered saline with 0.1% Tween 20 (PBST) and 5% skim milk for 1 h at room temperature. Primary antibody directed against YsrR (1:500) was diluted in PBST-5% milk and allowed to react overnight at 4°C. The membranes were washed several times with PBST and then incubated with goat anti-rabbit immunoglobulin G-horseradish peroxidase at 1:30,000 in PBST-milk for 1 h at room temperature. The membranes were washed again in PBST and proteins were detected by chemiluminescence (Amersham).

RESULTS

Expression of the ysp genes is tightly growth phase regulated.

In order to fully analyze the regulation of the individual ysp genes, it was first necessary to determine the time at which each gene was maximally expressed. Previous studies had indicated that several genes within the ysa locus were maximally expressed during early to mid-log phase using β-galactosidase reporters (37). However, since reporter data does not always agree with transcript levels, we chose to expand our analysis by examining expression of the ysp genes and related genes over the course of a growth curve by Northern analysis. All the ysp genes were quickly upregulated upon subculturing into LB-290 (L broth with 290 mM NaCl) and peaked in logarithmic phase, and levels were lower to undetectable by early stationary phase (Fig. 2). No mRNA from any gene probed could be detected from a 24-h culture, so this time point was omitted from the graphs. While nearly all the genes appeared to be very tightly growth phase regulated, nevertheless, two patterns could be observed. Several genes (yspE, yspA, yspP, and yspM) were maximally expressed at the first time point examined, 2 h (Fig. 2). Also peaking at this time point were the regulators ysaE and sycB. The second group, consisting of yspK, yspL, yspI, yspF, and yspY, was maximally expressed at 4 h. These data indicate that there are two clusters of temporal ysp expression, suggesting that there may be some regulatory hierarchy or independent mechanisms of gene activation.

FIG. 2.

Growth phase expression of the ysa and ysp genes. Triplicate saturated cultures of JB580v were diluted to a starting OD of 0.2 in L broth (open squares, dotted line) or LB290 (solid squares, solid line) and grown at 26°C. Total RNA was extracted from cells grown for the indicated times and used for Northern analysis. Individual genes were probed with ∼300-bp fragments internal to the predicted coding region. The maximum signal intensity for each data set was assigned a value of 100, and all other signal intensities within each data set were normalized to this value. The panel in the bottom right corner shows the growth curves of the cultures from which the RNA was isolated. The data for ysaE are from the experiment presented in the legend to Fig. 3, which was conducted in identical fashion except that only LB-290 was used.

Because several of the ysp genes were maximally expressed at the first time point examined, we decided to test how quickly they responded to NaCl. Cultures of wild-type cells were subcultured into L broth, grown for 2 h, and then pelleted and resuspended in LB-290. Samples were collected every 20 min, and gene expression was monitored by Northern blot analysis. The data showed that the genes that peaked at 2 h had a rapid response to NaCl, reaching maximal levels by 20 or 40 min (data not shown). As expected, the group that peaked at 4 h showed a slower response; it had elevated mRNA levels by 20 min but could not reach maximal levels until 40 or 60 min. The results of this experiment demonstrate that ysp expression responds immediately to NaCl but that other factors may control the time to maximum expression levels. The differences in the time to peak expression in this experiment are more subtle than in the experiment presented in Fig. 2 and are most likely a consequence of the different culturing methods. The culture density of the 20-min time point was much higher than the 2-h point in the first experiment (OD₆₀₀, >1.0 versus 0.4 to 0.5). Nonetheless, this experiment supports the previous analysis that the response to NaCl is rapid for all genes and that there are indeed two separate groups of temporal expression.

Effect of NaCl on ysp gene expression.

Several previous studies showed that secretion of Ysp proteins and expression of selected ysa genes are induced by NaCl (17, 37, 38, 43). We therefore wanted to examine the impact of NaCl on ysp gene expression. With the exception of yspM, the expression profiles of the ysp genes in L broth (0 mM NaCl) did not appear to change in response to growth phase, and most showed little or no expression at all (Fig. 2). By comparing the relative expression between L broth and LB-290 at the time of peak expression in LB-290, we determined the fold change in expression upon growth in inducing conditions. Three distinct groups with respect to dependence on NaCl could be observed. Three genes, yspY, yspI, and yspM, were not induced or were only slightly induced by NaCl (1.7-, 2.0-, and 2.6-fold, respectively). The genes yspA, yspK, and yspE were moderately induced by NaCl 4-, 8.7-, and 12-fold, respectively. A third group contains genes whose expression is heavily dependent on NaCl. yspP and yspF were regulated 39- and 45-fold, respectively, and yspL absolutely required NaCl for expression; no signal for yspL could be detected from cultures grown in L broth. The regulatory gene sycB was also shown to be activated by NaCl 33-fold, confirming the results previously reported using β-galactosidase assays (37, 38). These results indicate that, like the apparatus genes, most ysp genes are at least partially dependent on NaCl for expression.

Effect of YsaE/SycB on ysp gene expression.

Previous work from our lab showed that the sycB promoter was stimulated by YsaE and SycB during conditions of active secretion (i.e., high NaCl levels) (38). This regulatory mechanism is similar to that observed in the Salmonella SPI-1 regulation of the sicA promoter by InvF/SicA and the Shigella Mxi-Spa regulation of several ipa and osp genes by MxiE/IpgC (8-10, 24, 25). In both of these systems, genes encoding secreted proteins located outside the locus containing the apparatus genes are also coordinately regulated by InvF/SicA or MxiE/IpgC (10, 25), which implies that YsaE/SycB may regulate nonlinked ysp genes. To determine the influence of YsaE/SycB on ysp gene expression, we performed Northern analysis using RNA extracted from strains with an in-frame deletion for ysaE (YVM932) or sycB (YVM981). Six of the nine genes examined were dependent on both YsaE and SycB for full expression levels (Fig. 3). yspP and yspE were the most highly dependent, each being reduced over 100-fold in the ΔysaE and ΔsycB strains. yspF, yspL, and yspA were reduced from 20- to 75-fold, while yspM was modestly reduced by about sixfold. The three genes that showed little or no dependence on YsaE or SycB for expression were yspK, yspI, and yspY (Fig. 3). Curiously, expression of yspK was actually somewhat elevated (less than threefold) in the absence of YsaE/SycB. These results indicate that YsaE and SycB are key regulators of many ysp genes. Furthermore, it draws another parallel with InvF/SicA and MxiE/IpgC, indicating that this regulatory mechanism may be a common theme for controlled expression of type III effector genes.

FIG. 3.

Effect of YsaE, SycB, YsrR, YsrS, and RcsB on ysp gene expression. Triplicate-saturated cultures of the wild-type (WT; JB580v), ΔysaE (YVM932), ΔsycB (YVM981), ΔysrS (YVM969), ΔysrR (YVM1006), and ΔrcsB (YVM1236) strains were diluted to a starting OD of 0.2 in LB-290 and grown at 26°C. Total RNA was extracted and used in Northern dot blot analysis. At the time of peak expression determined in the legend to Fig. 2, the signal intensity obtained in the mutant strains was divided by the signal intensity obtained for the wild-type strain. Wild-type expression levels were normalized to 100. The data for yspM have been presented elsewhere (41), and are included here for completeness. ND, not determined.

Effect of YsrRS and RcsB on ysp gene expression.

YsrR, YsrS, and RcsB have been shown to be required for activation of the ysaE promoter (37, 38; K. A. Walker, S. Mildiner-Earley, and V. L. Miller, unpublished data), and we therefore wanted to examine if these proteins are required for expression of any ysp genes. Similar Northern blot analyses were performed as described above using strains with an in-frame deletion of ysrR (YVM1006), ysrS (YVM969), or rcsB (YVM1236). In agreement with previous studies using lacZ reporters (37, 38), ysaE and sycB were found to be dependent on YsrR, YsrS, and RcsB for full expression by Northern analysis. Levels of ysaE mRNA were decreased seven- to 13-fold relative to those of the wild-type strain, while those for sycB were decreased 65- to 165-fold (Fig. 3). The same ysp genes that are activated by YsaE/SycB also required all three phosphorelay proteins for expression: yspP, yspE, yspL, yspF, yspA, and yspM (Fig. 3). The genes yspP, yspE, yspL, and yspF were heavily dependent on these proteins, with expression reduced about 35- to 110-fold relative to that of the wild type. The genes yspA and yspM were somewhat less dependent, with expression decreased about 15- to 20-fold for yspA and about two- to fourfold for yspM.

As was observed in the ΔysaE and ΔsycB strains, expression of yspK was slightly elevated in the ΔysrS, ΔysrR, and ΔrcsB strains (Fig. 3). Levels were increased by about twofold, indicating that its activation is fully independent of the known proteins regulating expression of the Ysa T3SS, even though its expression was dependent on NaCl. Curiously, the results obtained for yspI seem to indicate that YsrRS-RcsB do not always regulate ysa and ysp genes jointly (Fig. 3). YsrS and RcsB were required for full expression, but deletion of ysrR resulted in a subtle increase in yspI expression (2.2-fold). The role of these phosphorelay proteins also varies with yspY and yspM. With both of these ysp genes, the loss of YsrS reduced expression by more than threefold compared to the wild type, and the loss of RcsB reduced expression by two- to threefold. However, the loss of YsrR resulted in only a slight reduction in expression (less than twofold), similar to the trend observed for yspI. Using low-copy-number gfp reporter plasmids, similar trends of regulation have been observed for yspK, yspI, and yspM (data not shown). These regulatory patterns are somewhat surprising, as we would have predicted that YsrRS and RcsB were functioning together or in tandem to regulate ysa and ysp expression. These results suggest that there may be a third response regulator that is part of the YsrS phosphorelay cascade or that RcsB alone can partially compensate for the loss of YsrR.

Regulation of ysp genes by YsrRS/RcsB is probably not direct.

All the genes that required YsrRS and RcsB were also dependent on YsaE/SycB. Because YsrRS and RcsB are required for expression of ysaE, and YsaE is required for expression of several ysp genes, it seemed likely that the impact observed by YsrRS/RcsB on ysp gene expression is not direct but a consequence of the loss of YsaE/SycB. To test this hypothesis, we transformed strains lacking ysrR or ysrS with plasmids carrying ysaE (pKW22) and sycB (pKW28) or vectors (pWKS130, pKW27) and then collected RNA from these strains grown for 2 h in LB-290. The results for this experiment are only qualitative but showed that for each gene tested, expression of ysaE and sycB together suppressed the deletions of ysrS and ysrR, whereas the vectors alone did not (Fig. 4). This is consistent with data previously obtained for sycB, where sycB-lacZ expression in a ΔysrS strain was restored only if both pKW22 and pKW28 were present (38). These data indicate that the regulatory effect observed in the ΔysrS and ΔysrR strains is indirect at these promoters and is most likely a consequence of reduced levels of YsaE and SycB.

ysrRS and rcsB expression patterns do not appear to dictate ysa and ysp expression patterns.

Two-component regulator systems are thought to be constitutively expressed at least at low levels so that the proteins are readily available when their specific environmental signal is sensed. Many, like the PhoRB system (16), are autoupregulated when their specific signal is detected to produce higher protein levels to adequately respond to environmental conditions. To ascertain if changes in ysrRS or rcsB transcription dictated ysp gene expression, we examined the expression patterns of these putative phosphorelay genes. Northern analysis of rcsB indicated that it is expressed in the absence of NaCl (Fig. 5A) and that it is expressed at near-maximum levels over the course of a growth curve in both the presence and absence of NaCl. There does appear to be some growth phase-dependent regulation in the presence of NaCl, but the peak (at 4 h) was less than twofold higher than that in the absence of NaCl. Judging by the intensity of hybridization, rcsB mRNA levels are abundant throughout the course of growth even though there is fluctuation. Therefore, our data suggest that changes in expression of rcsB do not account for the salt dependence or growth phase regulation of the ysp genes.

FIG. 5.

Expression of rcsB and ysrR. (A) Growth phase expression of rcsB; data are from the experiment shown in Fig. 2. (B) Growth phase expression of ysrR monitored by fluorescence using a ysrR-gfp reporter. (C) Growth phase expression of ysrR monitored by a β-galactosidase assay using a ysrR-lacZ reporter (YVM1041). (D) Expression of rcsB in ΔysrS and ΔysrR strains; data were obtained as shown in Fig. 3. (E) Expression of ysrR-gfp in ΔysrS, ΔysrR, and ΔrcsB strains. All cultures were grown as described in the legend to Fig. 2, and data were obtained as described in Materials and Methods. For panels B and C, inset graphs show the growth curves for these experiments. For panels A, B, and C, expression in L broth is presented as a dotted line and expression in LB-290 as a solid line.

Northern analysis could not be used to examine ysrR expression. A second putative two-component system located ∼10 kb from the ysa locus has an exceptionally high degree of identity, making it difficult to design a ysrR-specific probe. Therefore, we constructed a ysrR-gfp reporter and examined its expression by fluorescence. The low-copy-number plasmid was transformed into JB580v, and the absorbance and fluorescence were measured over the course of a growth curve. The levels of ysrR-gfp gradually increased with time and were still elevated at the time points of 6 and 8 h, when most ysp genes showed notable decreases in mRNA levels (Fig. 5B). We did not observe a strong influence of NaCl, calculating a less than twofold difference between green fluorescent protein levels of cultures grown in L broth compared to those in LB-290 at 8 h. Since these data contradict what was published by Venecia and Young using a ysrS-lacZ reporter (37), we also examined expression using a lacZ reporter, with a transcriptional fusion to ysrR at amino acid 23. The β-galactosidase assay data are consistent with the fluorescence data, showing no dependence on NaCl (Fig. 5C). No growth phase-dependent regulation was observed using the ysrR-lacZ reporter, but the increases observed with the ysrR-gfp reporter in late log and early stationary phase are subtle (less than threefold higher than at the 4-h time point). Furthermore, preliminary Western blot analysis of YsrR protein levels under these same conditions showed that YsrR is not detectable until 8 h, a time when very few of the ysp genes had any detectible transcripts (not shown). These results suggest that ysrRS expression does not exhibit the same salt- and growth phase-dependent expression pattern as the ysa and ysp genes and therefore does not account for fluctuations in ysa and ysp expression.

Given that our data on rcsB and ysrR expression differed considerably from those of Venecia and Young (37), we continued our investigation by examining the impact of YsrR and RcsB regulating each other. By Northern blot analysis, transcript levels of rcsB in the ΔysrR and ΔysrS strains were not significantly different than those of the wild type (less than a 2.5-fold change) (Fig. 5D). Using the ysrR-gfp reporter, no change in fluorescence was observed in ΔrcsB, ΔysrR, or ΔysrS strains (Fig. 5E). The data presented here showed that the loss of YsrRS did not result in a decrease in rcsB mRNA as was reported by Venecia and Young (37) and that the loss of RcsB did not result in a decrease in ysrRS mRNA. Furthermore, there did not appear to be any feedback regulation of ysrRS. The data generated in this study suggest that the involvement of RcsB and YsrR in ysa and ysp gene activation is not due to a transcriptional regulatory cascade.

DISCUSSION

T3SSs are a well-established virulence determinant and are found in a wide variety of gram-negative bacterial plant and animal pathogens as well as symbionts (36). The genes encoding the secretion apparatus are usually found as a single cluster. The Shigella Mxi-Spa and Yersinia Ysc-Yop T3SSs are encoded on virulence plasmids which harbor not only the apparatus genes but also the secreted effectors and transcriptional regulators. However, in other T3SSs such as Salmonella SPI-1 and E. coli LEE, the genes encoding the apparatus, secreted effectors, and transcriptional regulators are scattered around the chromosome (1, 11, 12). In addition, transcriptional regulators encoded with or proximal to the apparatus genes have also been shown to be required for expression of unlinked effector genes (9, 10, 12, 13). In these situations, the system components are speculated to have been acquired in several independent transfer events and must have incorporated the careful coordination of expression during the evolution of the organism (1).

Synchronous regulation of Ysa T3SS components.

Components of two phosphorelay systems, YsrRS and RcsB, are required for expression of ysaE as well as several ysp genes (37, 38). While the specific mechanism(s) by which they function has yet to be determined, it can safely be concluded that they are requisite factors in the expression of the Ysa T3SS. Activation of the ysaE promoter by the sensor YsrS requires elevated levels of NaCl; no expression is observed in the absence of NaCl or in the absence of YsrS (37, 38). The transcript initiated at the ysaE promoter likely results in expression of about 18 genes (ysaE through yspA) (Fig. 1). We previously showed that YsaE and SycB are both required for stimulation of the sycB promoter, which leads to upregulation of sycB and probably at least some of the downstream ysp genes (yspB to yspA) (38). This body of work sought to determine if any of these known regulators were also involved in the expression of the unlinked ysp genes that were recently reported by Matsumoto and Young (23) and Witowski et al. (41). A group of six unlinked ysp genes was determined to be regulated by all five regulatory proteins examined, as follows: yspP, yspL, yspE, yspF, yspA, and yspM. In addition, all of the genes required NaCl for activation. Thus, these six genes are all coordinately regulated with the apparatus itself and, as such, should be produced simultaneously. While sycB mRNA was determined to require all of these elements, we further showed that the effect of YsrRS on sycB mRNA levels was not directly at the sycB promoter. This is consistent with previous reports showing that transformation of a ΔysrS strain with plasmids expressing YsaE and SycB resulted in high levels of sycB expression (38). Similarly, the ysp genes listed above were determined to be indirectly regulated by YsrRS, as their expression was also restored in ΔysrS and ΔysrR strains carrying pYsaE and pSycB. While it has not been tested, we would predict a similar trend in the ΔrcsB strain carrying pYsaE and pSycB. These data suggest that the regulation of this group of ysp genes is dictated by levels of YsaE/SycB, which presumably form a complex, as was shown for the analogous regulators InvF/SicA and MxiE/IpgC (10, 31). Several promoters regulated by InvF and MxiE have a conserved sequence that appears to be the DNA binding site for these regulators (10, 20, 25). Based on this sequence, we examined the region upstream of genes regulated by YsaE and have identified putative binding sites in four promoter regions (sycB, yspF, yspL, and yspE), and a weaker site was identified upstream of yspP. No putative YsaE binding site could be found upstream of yspA, but it may be transcribed as part of the sycB-yspBCD operon. These observations are based purely on promoter gazing and await formal testing. Nonetheless, they support the model that these promoters are directly regulated by YsaE/SycB.

Noncoordinately regulated genes.

Three ysp genes showed regulatory patterns distinct from those described above. While yspY and yspI did display growth phase regulation, there appears to be no involvement of YsaE or SycB and only a slight dependence on YsrR and RcsB. However, both genes required YsrS for expression. These results suggest that the regulation by phosphorelay proteins is quite complex and probably involves more than the Ysr and Rcs systems. In addition, neither yspY nor yspI was strongly dependent on NaCl levels. While there are many models to explain these data, the simplest explanation is that a third response regulator works in concert with YsrS and that YsrS senses a signal other than NaCl (or general ionic strength) and phosphorylates this regulator in response to that signal. The study of ysa and ysp gene regulation could provide interesting insights into the complexity of phosphorelay regulatory circuits.

The levels of yspK mRNA were quite abundant in all experiments. While it displayed growth phase-dependent regulation and required high levels of NaCl, it does not require any of the protein regulators examined in this study. Furthermore, levels of yspK mRNA were somewhat elevated in the mutant strains examined (two- to threefold), suggesting that it may actually be slightly repressed by one or more of these factors. We previously observed that ysaE-lacZ expression was equally low when measured from wild-type cells grown in L broth (no salt) or from ΔysrS cells grown in either L broth or LB-290 (38). Thus, we predicted that YsrS was sensing NaCl concentrations and initiating a phosphorelay event in response to high levels of NaCl. While this may still be the case for the genes regulated by YsrRS, yspK likely derives its salt dependence from a different mechanism. YspK may represent an example of an effector whose expression is such that it is producing protein to be ready for secretion at the first moment that the bacterium meets its target host cell. It has been shown to possess serine/threonine kinase activity and is speculated to play a role in altering the innate immune response (23). If this is indeed its function, it would be consistent with it being immediately available and one of the first proteins secreted through the Ysa T3SS.

An additional point of curiosity regarding the expression of two effectors not coordinately regulated with the apparatus is that yspI and yspK are nearby neighbors on the chromosome, encoded by open reading frames YE2444 and YE2447, respectively (23). These genes are transcribed in opposite directions and thus are not in an operon. However, their proximity to each other could indicate they were acquired during a single DNA transfer event. It is tempting to speculate that they were acquired together and are still in the process of developing coordinated regulation; yspI has acquired dependence on YsrS and yspK has acquired dependence on NaCl.

Regulation of phosphorelay components.

The transcriptional expression data reported here suggests that ysrRS is not strongly growth phase dependent and not affected by high NaCl levels as are many of the ysp genes. Expression of rcsB is somewhat growth phase dependent but not regulated by NaCl. In addition, neither YsrR nor RcsB appear to regulate each other, and ysrRS is not autoregulated. Venecia and Young reported that expression of ysrRS and rcsB was dependent on NaCl and that rcsB expression was dependent on YsrS (37). A plausible source of these differences is the use of different strains and methodology. Venecia and Young used an rcsB-lacZYA fusion generated by transposon mutagenesis in which the transposon insertion disrupts the gene, presumably rendering it nonfunctional. This can be problematic when studying transcriptional regulators because many are autoregulated. In this study, we examined rcsB mRNA levels, which can be a more direct measurement of transcriptional activity. In addition, all our mutant strains are constructed as in-frame deletions rather than insertional mutants as utilized by Venecia and Young. However, the observed differences in expression patterns beg for a more detailed analysis. While the regulation of these genes differs, our data agree that YsrRS and RcsB are required for ysa gene expression. A more complete set of strains, including deletions of rcsC and rcsD, and more direct probing of the involved promoters will be required to fully ascertain the regulatory properties of the genes encoding these phosphorelay proteins. Current research in our lab will provide the necessary strains for such an investigation. While the model built by Venecia and Young that implies that regulation of ysrRS largely dictates the expression of downstream genes, and ultimately the ysa genes, is a logical model based on their extensive experiments (37), our data do not support it. Based on the data we have obtained, we hypothesize that the Ysr and Rcs systems are not part of a regulatory cascade but rather are working together to function in an integrated mechanism to regulate expression of the ysa and ysp genes. The exact mechanism by which they function in this capacity is the focus of the current investigation.