Regulatory Evolution of the phoH Ancestral Gene in Salmonella enterica Serovar Typhimurium

ABSTRACT

One important event for the divergence of Salmonella from Escherichia coli was the acquisition by horizontal transfer of the Salmonella pathogenicity island 1 (SPI-1), containing genes required for the invasion of host cells by Salmonella. HilD is an AraC-like transcriptional regulator in SPI-1 that induces the expression of the SPI-1 and many other acquired virulence genes located in other genomic regions of Salmonella. Additionally, HilD has been shown to positively control the expression of some ancestral genes (also present in E. coli and other bacteria), including phoH. In this study, we determined that both the gain of HilD and cis-regulatory evolution led to the integration of the phoH gene into the HilD regulon. Our results indicate that a HilD-binding sequence was generated in the regulatory region of the S. enterica serovar Typhimurium phoH gene, which mediates the activation of promoter 1 of this gene under SPI-1-inducing conditions. Furthermore, we found that repression by H-NS, a histone-like protein, was also adapted on the S. Typhimurium phoH gene and that HilD activates the expression of this gene in part by antagonizing H-NS. Additionally, our results revealed that the expression of the S. Typhmurium phoH gene is also activated in response to low phosphate but independently of the PhoB/R two-component system, known to regulate the E. coli phoH gene in response to low phosphate. Thus, our results indicate that cis-regulatory evolution has played a role in the expansion of the HilD regulon and illustrate the phenomenon of differential regulation of ortholog genes.

IMPORTANCE Two mechanisms mediating differentiation of bacteria are well known: acquisition of genes by horizontal transfer events and mutations in coding DNA sequences. In this study, we found that the phoH ancestral gene is differentially regulated between Salmonella Typhimurium and Escherichia coli, two closely related bacterial species. Our results indicate that this differential regulation was generated by mutations in the regulatory sequence of the S. Typhimurium phoH gene and by the acquisition by S. Typhimurium of foreign DNA encoding the transcriptional regulator HilD. Thus, our results, together with those from an increasing number of studies, indicate that cis-regulatory evolution can lead to the rewiring and reprogramming of transcriptional regulation, which also plays an important role in the divergence of bacteria through time.

INTRODUCTION

The acquisition of foreign DNA through horizontal transfer events is one of the main genetic processes that influences the dynamics of bacterial genomes (1). Likewise, the adaptation of regulatory mechanisms to control the expression of acquired genes, or even to modify the expression pattern of ancestral genes, has been pivotal for evolution of bacteria, including pathogenic species (2–5). During its divergence from Escherichia coli, about 100 million years ago, the Gram-negative pathogen Salmonella enterica serovar Typhimurium (S. Typhimurium) shaped approximately a quarter of its genome with acquired DNA (6). Most of the genes for S. Typhimurium virulence have been acquired by horizontal transfer events and are located in discrete regions of the genome called Salmonella pathogenicity islands (SPI), which are not present in E. coli (3, 6). SPI-1, a cluster of 39 genes, is essential for the S. Typhimurium invasion of intestinal epithelium cells, which leads to enteritis (7, 8).

HilD is a member of the AraC/XylS family of transcriptional regulators, located in SPI-1, which directly or through distinct regulators positively controls the expression of the SPI-1 genes and several other related genes located outside this island (2, 3, 9–13). Interestingly, HilD has been shown to induce not only the expression of numerous acquired genes but also that of some ancestral genes (those present in E. coli and other phylogenetically distant bacteria). For instance, HilD positively regulates the expression of the flhDC ancestral operon, which encodes the master complex regulator of flagellar and chemotaxis genes, FlhD₄C₂ (14, 15). Furthermore, HilD induces the expression of the yobH ancestral gene that codes for the 79-amino-acid polypeptide YobH (11, 12, 16). Importantly, both the flagellar/chemotaxis genes and the yobH gene are necessary for the Salmonella invasion of host cells (16–20). Therefore, HilD has played a major role in the evolution of Salmonella pathogenicity by integrating the expression of tens of acquired genes and some ancestral genes into a big virulence regulon.

Previously, we and others found that HilD also induces the expression of the phoH ancestral gene under SPI-1-inducing conditions, such as the growth of bacteria in nutrient-rich lysogeny broth (LB) (10–12). Orthologs of the PhoH protein are present in E. coli and many other bacteria as well as in some archaea; however, the function of PhoH remains unknown in all microorganisms (21, 22). In E. coli, it has been demonstrated that the expression of phoH is induced by the PhoB/R two-component system in minimal medium containing a low concentration of phosphate; additionally, it was found that PhoH shows ATPase activity (23). The PhoB/R system positively regulates the expression of several other genes in response to low phosphate, including the pstS gene; the response regulator PhoB binds to a sequence denominated pho-box that is present on the regulatory regions of its target genes (24).

Remarkably, to our knowledge, in addition to the PhoB/R system (or orthologs), HilD is the only other regulator that has been involved in the expression of phoH. Hence, in this study we aimed to uncover the elements that led to the integration of the phoH gene into the HilD regulon.

By comparing the expression of the phoH gene from S. Typhimurium and E. coli, we show that the gain of HilD and cis-regulatory evolution led to the regulation of phoH by HilD in S. Typhimurium. Furthermore, our results support that HilD activates the expression of phoH under SPI-1-inducing conditions in part by acting as an anti-H-NS factor. Additionally, our results revealed differential regulation of the phoH gene in response to low phosphate; in E. coli this regulation is mediated by the PhoB/R system, whereas in S. Typhimurium it is independent of PhoB.

RESULTS

The S. Typhimurium phoH gene, but not that of E. coli K-12, is expressed and regulated by HilD under SPI-1-inducing conditions.

The phoH structural gene and its encoded protein, PhoH, share identities of 82% and 95%, respectively, between S. Typhimurium SL1344 and E. coli MG1655. In contrast, the intergenic sequences upstream of phoH show a lower identity (54%), and the genomic context is poorly conserved between these two bacteria (Fig. 1A). Therefore, we wondered whether, in addition to the gain of HilD by S. Typhimurium, differentiation of the phoH regulatory sequence in S. Typhimurium was also necessary for the adaptation of the regulation of phoH by HilD. To examine this idea, we analyzed the expression of the phoH gene from S. Typhimurium SL1344 (phoH_STm) and from E. coli MG1655 (phoH_Ec) in S. Typhimurium and E. coli strains expressing HilD or not and grown under SPI-1-inducing conditions (LB at 37°C). First, we constructed the phoH_Ec-cat transcriptional fusion (carrying the full-length intergenic sequence upstream of phoH_Ec), which is equivalent to the phoH_STm-cat transcriptional fusion that we previously used to show the regulation of the phoH_STm gene by HilD (10) (Fig. 1B, Table 1). The expression of both fusions, phoH_STm-cat and phoH_Ec-cat, then was quantified in the wild-type (WT) S. Typhimurium SL1344 strain and its derivative ΔhilD mutant, as well as in the WT E. coli MC4100 strain, in the presence of the pK6-HilD plasmid expressing HilD or the pMPM-K6 vector. Consistent with our previous study (10), the expression of the phoH_STm-cat fusion was significantly decreased in the ΔhilD+vector mutant and in the WT E. coli+vector strain with respect to the WT S. Typimurium+vector strain; furthermore, it was induced by the pK6-HilD plasmid in both the ΔhilD mutant and the WT E. coli strain, 4- and 8-fold, respectively (Fig. 2A). In contrast, the phoH_Ec-cat fusion showed a low expression level in all the strains tested (Fig. 2B), indicating that the expression of the phoH_Ec gene is not activated by HilD under the growth conditions tested. To confirm these results, we performed a Western blot analysis to monitor the chromosomal expression of the PhoH-FLAG protein (PhoH tagged with a 3×FLAG epitope) in the S. Typhimurium ΔhilD mutant and the WT E. coli strain carrying the pBAD-HilD1 plasmid, which expresses HilD from an arabinose-inducible promoter. In agreement with our results obtained with the transcriptional fusions, the S. Typhimurium PhoH-FLAG (PhoH_Stm-FLAG), but not the E. coli PhoH-FLAG (PhoH_Ec-FLAG), was detected under the conditions assessed; furthermore, the expression of HilD from the pBAD-HilD1 plasmid increased the amount of PhoH_Stm-FLAG in the ΔhilD mutant in a dose-dependent fashion (Fig. 2C). These results show that the phoH_STm gene, but not the phoH_Ec gene, is expressed and regulated by HilD under SPI-1-inducing conditions, which supports that both the gain of HilD and differentiation of the regulatory sequence led the phoH_STm gene to be expressed under conditions relevant for Salmonella virulence.

FIG 1.

Genomic context of the phoH_STm and phoH_Ec genes and schematic representation of the phoH_STm-cat and phoH_Ec-cat fusions. (A) Genomic context of the phoH gene from the S. Typhimurium SL1344 and E. coli K-12 MG1655 strains. Genes are shown by arrows. Homologous genes are depicted with the same color. (B) Schematic representation of the phoH_STm-cat and phoH_Ec-cat transcriptional fusions containing the full-length upstream region of phoH from S. Typhimurium SL1344 or E. coli K-12 MG1655 strains, respectively. Transcriptional start sites are shown by bent arrows. Positions indicated are relative to the transcriptional start site of promoter 1 of each gene.

TABLE 1.

Bacterial strains and plasmids used in this study^a

Strain or plasmid	Genotype or description	Reference or source
Strains
S. Typhimurium
SL1344	Wild type; xyl, hisG, rpsL; Smr	60
DTM142	phoH::3×FLAG-kan	This study
JPTM25	ΔhilD	47
DTM143	ΔhilD phoH::3×FLAG	This study
DTM144	ΔphoB::kan	This study
DTM145	ΔphoB	This study
DTM146	ΔphoB phoH::3×FLAG-kan	This study
DTM150	ΔphoH::kan	This study
DTM151	ΔsiiE::kan	This study
E. coli
MG1655	Wild type; prototrophic E. coli K-12	Michael Cashel
DH10β	Laboratory strain	Invitrogen
MC4100	F− (araD139) Δ(argF-lac)169λ− e14− flhD5301 Δ(fruK-yeiR) 725(fruA25) relA1 rpsL150(SmR) rbsR22 Δ(fimB-fimE)632(::IS1)deoC1	61
BL21/DE3	Strain for expression of recombinant proteins	Invitrogen
JPMC34	MC4100 Δhns	62
JW0389	BW25113 ΔphoB::kan	63
DTM147	phoH::3×FLAG-kan	This study
DTM148	BW25113 ΔphoB	This study
DTM149	BW25113 ΔphoB phoH::3×FLAG-kan	This study
Plasmids
pKK232-8	pBR322 derivative containing a promoterless chloramphenicol acetyltransferase (cat) gene, Apr	48
pphoHSTm-cat	pKK232-8 derivative containing a phoH-cat transcriptional fusion of S. Typhimurium from nucleotides −653 to +416	10
pphoHEc-cat	pKK232-8 derivative containing a phoH-cat transcriptional fusion of E. coli from nucleotides −406 to +406	This study
pP1phoHSTm-cat	pKK232-8 derivative containing a phoH-cat transcriptional fusion of S. Typhimurium from nucleotides −653 to +12	This study
pP2phoHSTm-cat	pKK232-8 derivative containing a phoH-cat transcriptional fusion of S. Typhimurium from nucleotides −5 to +416	This study
pP1phoHSTm-248 + 12-cat	pKK232-8 derivative containing a phoH-cat transcriptional fusion of S. Typhimurium from nucleotides −248 to +12	This study
pP1phoHSTm−67 + 3-cat	pKK232-8 derivative containing a phoH-cat transcriptional fusion of S. Typhimurium from nucleotides −67 to +3	This study
pphoHHilDmut-cat	pphoHSTm-cat derivative containing mutations in the predicted HilD-binding site	This study
pphoHP1mut-cat	pphoHSTm-cat derivative containing mutations in the P1 promoter	This study
ppstSSTm-cat	pKK232-8 derivative containing a pstS-cat transcriptional fusion of S. Typhimurium from nucleotides −232 to +352	This study
pBADMycHisC	Expression vector for constructing C-terminal MycHis fusions under an arabinose-inducible promoter, Apr	Invitrogen
pBAD-HilD1	pBADMycHis derivative expressing HilD-MycHis under an arabinose-inducible promoter	47
pMPM-K6Ω	p15A derivative cloning vector containing an arabinose-inducible promoter, Kanr	64
pK6-HilD	pMPM-K6Ω derivative expressing HilD under an arabinose-inducible promoter, Kanr	10
pMAL-HilD1	pMAL-c2X derivative expressing MBP-HilD from a lac promoter, Apr	2
pBAD-H-NS-FH	pBADMycHisC derivative expressing H-NS-FH from an arabinose-inducible promoter, Apr	57
pKD46	pINT-ts derivative expressing red recombinase under an arabinose-inducible promoter, Apr	49
pKD4	pANTsγ derivative template plasmid containing the kanamycin cassette for λRed recombination, Apr	49
pSUB11	pGP704 derivative template plasmid for FLAG epitope tagging	50
pFLP3	pFLP2 derivative template plasmid expressing yeast Flp recombinase, Apr, Tcr	51

^aThe coordinates for the cat fusions are indicated with respect to the transcriptional start site of pstS or that of the P1 promoter of phoH_STm or phoH_Ec. Ap^r, ampicillin resistance; Sm^r, streptomycin resistance; Kan^r, kanamycin resistance; Tc^r, tetracycline resistance.

FIG 2.

phoH_STm, but not phoH_Ec, is expressed and regulated by HilD under SPI-1-inducing conditions. CAT-specific activity of the phoH_STm-cat (A) and phoH_Ec-cat (B) transcriptional fusions, contained in the pphoH_STm-cat and pphoH_Ec-cat plasmids, respectively, was determined in the WT S. Typhimurium (STm) SL1344 strain, its derivative ΔhilD mutant, and the WT E. coli (Ec) MC4100 strain, in the presence of the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD from an arabinose-inducible promoter. Expression of HilD from the pK6-HilD plasmid was induced with 0.001% l-arabinose added to the medium at the beginning of cultures. Data represent the averages with standard deviations from three independent experiments performed in duplicate. Statistically different values are indicated (*, P < 0.05; ****, P < 0.0001). (C) Expression of the PhoH-FLAG protein in the S. Typhimurium ΔhilD mutant and the WT E. coli (Ec) MG1655 strain carrying the vector pBADMycHisC (−) or the pBAD-HilD1 (+) plasmid, which expresses HilD from an arabinose-inducible promoter, was analyzed by Western blotting using monoclonal anti-FLAG antibodies. As a control, the expression of GroEL was also detected using polyclonal anti-GroEL antibodies. Expression of HilD from the pBAD-HilD1 plasmid was induced with the indicated amounts of l-arabinose added to the medium 2 h postinoculation of cultures. CAT activity and detection of the PhoH-FLAG and GroEL proteins were determined from samples of bacterial cultures grown in LB at 37°C for 9 and 4 h, respectively.

HilD specifically induces the expression of the P1 promoter of phoH_STm.

Previous transcriptomics studies revealed two transcription start sites (TSSs) for the phoH_STm gene (25, 26). To determine whether these TSSs are really generated by two different promoters (P1 and P2) and whether HilD regulates both or only one of them, we constructed and analyzed the expression of cat transcriptional fusions only containing the putative promoter P1 (P1phoH_STm-cat) or P2 (P2phoH_STm-cat) (Fig. 3A). The expression of these fusions was quantified in the WT S. Typhimurium strain and its derivative ΔhilD mutant carrying the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD under SPI-1-inducing conditions. The two fusions showed expression in the WT+vector strain, indicating that both P1 and P2 promoters are active under the conditions tested, with the P1 promoter 2-fold more active than the P2 promoter (Fig. 3B). However, only the expression of the P1phoH_STm-cat fusion decreased in the ΔhilD+vector mutant and was induced by the presence of the pK6-HilD plasmid (Fig. 3B). In agreement with these results, in electrophoretic mobility shift assays (EMSAs), purified MBP-HilD protein bound the phoH_STm fragment contained in the P1phoH_STm-cat fusion but not that contained in the P2phoH_STm-cat fusion (Fig. 3C). Together, these results support that HilD regulates only the P1 promoter of the phoH_STm gene.

FIG 3.

HilD directly regulates the P1 promoter of phoH_STm. (A) Schematic representation of the P1phoH_STm-cat and P2phoH_STm-cat transcriptional fusions containing promoter 1 (P1) or 2 (P2) of phoH_STm, respectively. Transcriptional start sites (bent arrows) and promoters (blue and red boxes) are shown. Positions indicated are relative to the transcriptional start site of promoter 1. (B) CAT-specific activity of the P1phoH_STm-cat and P2phoH_STm-cat transcriptional fusions, contained in the pP1phoH_STm-cat and pP2phoH_STm-cat plasmids, respectively, was determined in the WT S. Typhimurium (STm) SL1344 strain and its derivative ΔhilD mutant carrying the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD from an arabinose-inducible promoter. Expression of HilD from the pK6-HilD plasmid was induced with 0.001% l-arabinose added to the medium at the beginning of cultures. CAT activity was quantified from samples of bacterial cultures grown in LB at 37°C for 9 h. Data represent averages with standard deviations from three independent experiments performed in duplicate. Statistically different values are indicated (****, P < 0.0001). (C) MBP-HilD binding to the phoH_STm fragments contained in the P1phoH_STm-cat (P1) and P2phoH_STm-cat (P2) fusions was analyzed by competitive nonradioactive EMSAs. The DNA-protein complexes are indicated by an asterisk. Nonspecific PCR products are indicated by an arrow.

Cis-regulatory evolution led to the regulation of the P1 phoH_STm promoter by HilD.

To delimit the sequence required for the control of P1phoH_STm by HilD, we constructed and analyzed the expression of two additional cat fusions carrying different 5′ and 3′ deletions regarding the P1phoH_STm-cat-653 + 12 fusion used initially (Table 1). The expression of these two new fusions, P1phoH_STm-cat-248 + 12 and P1phoH_STm-cat-67 + 3, was similarly decreased in the ΔhilD+vector mutant with respect to the WT+vector strain and was induced in the ΔhilD mutant by the presence of the pK6-HilD plasmid (Fig. 4A). These results reveal that HilD acts on the sequence between positions −67 and +3 of the P1 phoH_STm promoter (P1phoH_STm), the sequence contained in the P1phoH_STm-cat-67 + 3 fusion. In agreement with this conclusion, a HilD-binding site was predicted between positions −59 and −25 of P1phoH_STm, which shows 62% A+T content (Fig. 4B). Previous studies indicate that HilD binding sites have a high A+T content (14, 27).

FIG 4.

HilD induces the expression of phoH_STm by acting on a regulatory sequence differentiated between Salmonella and E. coli. (A) CAT-specific activity of the P1phoH_STm-cat-248 + 12 and P1phoH_STm-cat−67 + 3 transcriptional fusions, contained in the pP1phoH_STm-248 + 12-cat and pP1phoH_STm−67 + 3-cat plasmids, respectively, was determined in the WT S. Typhimurium (STm) SL1344 strain and its derivative ΔhilD mutant carrying the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD from an arabinose-inducible promoter. Expression of HilD from the pK6-HilD plasmid was induced with 0.001% l-arabinose added to the medium at the beginning of cultures. CAT activity was quantified from samples of bacterial cultures grown in LB at 37°C for 9 h. Data represent averages with standard deviations from three independent experiments performed in duplicate. Statistically different values are indicated (****, P < 0.0001). (B) Alignment of the P1 −67/+3 sequences of S. Typhimurium (STm) SL1344 and E. coli (Ec) MG1655. The predicted HilD-binding site on the STm sequence and the PhoB-binding site (pho-box) on the E. coli sequence as well as the mutations generated in the predicted HilD-binding site and in the −10 region of the P1 phoH_STm promoter (HilDmut and P1mut, respectively) are shown. Bent arrows indicate the transcriptional start site of the respective P1 promoter. Asterisks indicate conserved nucleotides. (C) Phylogenetic tree based on the P1 −67/+3 sequence of E. coli O104:H4 (EAEC), E. coli O157:H7 (EHEC), E. coli K-12 MG1655, E. coli O78:H11 (ETEC), E. coli O127:H6 (EPEC), E. coli STEC_7v (STEC), E. coli K-12 ATCC 1175, E. coli 26-1 (UPEC), S. bongori, S. Choleraesuis SC-B67, S. Gallinarum 9184, S. Enteritidis 22510-1, S. Paratyphi A ATCC 11511, S. Typhi Ty2, S. Typhimurium SL1344, and S. Heidelberg 41563. Identity with respect to the sequence of S. Typhimurium SL1344 is shown with a heat map.

To confirm that HilD mediates expression of phoH_STm by acting on the P1 promoter of this gene, the predicted HilD-binding site or the −10 region of the P1 promoter were mutated, as indicated in Fig. 4B, in the phoH_STm-cat-653 + 416 fusion carrying the full-length regulatory region of phoH_STm (contains the P1 and P2 promoters), generating the phoH_HilDmut-cat and phoH_P1mut-cat fusions, respectively. The expression of the WT and its derivative mutated transcriptional fusions was quantified in the WT S. Typhimurium strain and the ΔhilD mutant carrying the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD under SPI-1-inducing conditions. As previously observed (Fig. 2A), the WT phoH_STm-cat fusion showed regulation by HilD, that is, its expression was decreased and increased in the ΔhilD+vector and ΔhilD+pK6-HilD strains, respectively (Fig. 5A). In contrast, both the phoH_HilDmut-cat and phoH_P1mut-cat fusions showed a low expression level that was similar in the three strains tested (Fig. 5A), which indicates that the regulation of phoH_STm by HilD is lost when the predicted HilD-binding site or the P1 promoter are mutated. EMSAs confirmed that HilD indeed binds the predicted site upstream of the P1 promoter (Fig. 5B to D). The MBP-HilD protein used in the EMSAs apparently forms overly large complexes that remained near the wells of the gel, with or without the tested DNA fragments; the WT P1phoH-157/+12 fragment and that carrying the mutated P1 promoter, but not that carrying the mutated HilD-binding sequence, were bound by MBP-HilD (Fig. 5B to D). Together, these results show that HilD positively and specifically regulates the P1 promoter of phoH_STm by binding a site that overlaps the −35 region of this promoter.

FIG 5.

HilD positively regulates phoH_STm by binding a site that overlaps the P1 promoter of this gene. (A) CAT-specific activity of the phoH_STm-cat, phoH_HilDmut-cat, and phoH_P1mut-cat transcriptional fusions, contained in the pphoH_STm-cat, pphoH_HilDmut-cat, and pphoH_P1mut-cat plasmids, respectively, was determined in the WT S. Typhimurium (STm) SL1344 strain and its derivative ΔhilD mutant carrying the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD from an arabinose-inducible promoter. Expression of HilD from the pK6-HilD plasmid was induced with 0.001% l-arabinose added to the medium at the beginning of cultures. CAT activity was quantified from samples of bacterial cultures grown in LB at 37°C for 9 h. Data represent the averages with standard deviations from three independent experiments performed in duplicate. Statistically different values are indicated (*, P = 0.0148; ****, P < 0.0001). MBP-HilD binding to the WT P1phoH-157/+12 fragment (B), the P1phoH_HilDmut-157/+12 fragment mutated in the predicted HilD-binding site (C), and the P1phoH_P1mut-157/+14 fragment mutated in the P1 promoter (D) was analyzed by nonradioactive EMSAs. The upper panels show the free DNA and the DNA-protein complexes stained with ethidium bromide, and the lower panels show the immunodetection of MBP-HilD at the top of the lines using anti-MBP monoclonal antibodies. The DNA-protein complexes are indicated by an asterisk.

Two promoters have also been reported for the phoH_Ec gene (23) that are located in positions similar to those of the phoH_STm gene (Fig. 1B). As shown in Fig. 2, the phoH_Ec gene is not regulated by HilD, which indicates that it lacks the HilD-acting site located upstream of P1phoH_STm. To investigate this notion, we performed an alignment of the corresponding −67/+3 sequence of the phoH_STm and phoH_Ec P1 promoters (P1 −67/+3 sequences). Interestingly, these sequences share only 45% identity; moreover, the nucleotides of the putative HilD-binding site located on the S. Typhimurium sequence are poorly conserved in the E. coli sequence (Fig. 4B). Consistent with this, a HilD-binding site was not predicted in the E. coli P1 −67/+3 sequence. To explore whether the differences shown by the P1 −67/+3 sequence of phoH_STm and phoH_Ec are a product of an evolutionary divergence between Salmonella and E. coli or are specific to the bacterial strains tested, we extended the analysis of the P1 −67/+3 sequence of phoH to other S. enterica serovars, the S. bongori species, and other K12 strains and pathotypes of E. coli. As shown in Fig. 4C, phylogenetic analysis based on the P1 −67/+3 sequence of phoH grouped the Salmonella and E. coli strains tested in two different clades. In agreement with these results, the S. Typhimurium P1 −67/+3 sequence is better conserved in the S. enterica serovars (94 to 100% identity), and to a lower extent in S. bongori (73% identity), than in the E. coli strains assessed (54 to 56% identity) (Fig. 4C). These results suggest that evolution led to the differentiation of the P1 −67/+3 sequence between Salmonella and E. coli.

Together, our results support that the regulation of the phoH_STm gene by HilD was adapted by the gain of HilD and the generation of a HilD-acting site through cis-regulatory evolution.

phoH_STm and phoH_Ec are differentially affected by the H-NS repressor.

HilD induces the expression of target genes by counteracting H-NS-mediated repression; in the absence of H-NS activity, these genes are expressed independently of HilD (2, 3, 16, 27–30). We sought to determine whether HilD follows a similar mechanism to induce the expression of the phoH_STm gene. Salmonella hns mutants show severe growth defects due to the overexpression of SPI-1 and other related genes, and suppressor mutations are generated (31, 32). Thus, to analyze the effect of H-NS on the phoH_STm gene, we used the WT E. coli MC4100 strain and its derivative Δhns mutant that does not show growth defects; the E. coli genetic background has been useful to study the regulation of other Salmonella genes by H-NS (H-NS of S. Typhimurium SL1344 and E. coli MC4100 share 95% identity) (16, 30). On the other hand, HilD induces the expression of phoH_STm in E. coli K-12 (Fig. 2A), indicating that no other Salmonella-specific regulator is required for the HilD-mediated expression of this gene. We quantified the expression of the phoH_STm-cat fusion in the WT E. coli MC4100 strain and its derivative Δhns mutant carrying the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD, grown under SPI-1-inducing conditions. As a control, the expression of this fusion was also tested in the WT S. Typhimurium strain carrying the pMPM-K6 vector. In line with the results described above (Fig. 2A), the expression of the phoH_STm-cat fusion was induced in the WT S. Typhimurium+vector strain as well as in the WT E. coli+pK6-HilD strain but not in the E. coli+vector strain (Fig. 6A). Notably, in the E. coli Δhns+vector strain the expression of the phoH_STm-cat fusion reached levels similar to those in the WT S. Typhimurium strain (Fig. 6A). These results show that H-NS represses the expression of phoH_STm and that in the absence of H-NS this gene is expressed independently of HilD, which supports that HilD induces expression of phoH_STm by counteracting H-NS-mediated repression. Nevertheless, the presence of the pK6-HilD plasmid further increased (2-fold) the expression of the phoH_STm-cat fusion in the E. coli Δhns mutant (Fig. 6A), indicating that HilD also induces partial expression of the phoH_STm gene by a mechanism independent of H-NS.

FIG 6.

HilD antagonizes H-NS-mediated repression on phoH_STm. (A) CAT-specific activity of the phoH_STm-cat transcriptional fusion, contained in the pphoH_STm-cat plasmid, was evaluated in the WT S. Typhimurium (STm) SL1344 strain as well as in the WT E. coli (Ec) MC4100 strain and its derivative Δhns mutant in the presence of the pMPM-K6 vector or the pK6-HilD plasmid expressing HilD from an arabinose-inducible promoter. CAT-specific activity of the phoH_Ec-cat transcriptional fusion, contained in the pphoH_Ec-cat plasmid, was evaluated in the WT E. coli (Ec) MC4100 strain and its derivative Δhns mutant. Expression of HilD from the pK6-HilD plasmid was induced with 0.001% l-arabinose added to the medium at the beginning of cultures. CAT activity was quantified from samples of bacterial cultures grown in LB at 37°C for 9 h. Data represent the average with standard deviations from three independent experiments performed in duplicate. Statistically different values are indicated (****, P < 0.0001). (B) H-NS-FH binding to the phoH fragments contained in the phoH_STm-cat and phoH_Ec-cat fusions, carrying the full-length regulatory region of phoH_STm and phoH_Ec, respectively, was analyzed by competitive nonradioactive EMSAs. The DNA-protein complexes are indicated by an asterisk. Nonspecific PCR products are indicated by an arrow. (C) Competitive nonradioactive EMSAs between HilD-MBP and H-NS-FH on the P1phoH-157/+12 fragment of phoH_STm. Purified H-NS-FH protein was added at 4 μM (lanes 3 to 6), and purified MBP-HilD protein was added at 0.4, 0.8, and 1.2 μM (lanes 4 to 6, respectively). No protein was added in lane 1, and MBP-HilD was added at 1.2 μM in lane 2. In EMSAs, the upper panels show the free DNA and the DNA-protein complexes stained with ethidium bromide, whereas the lower panels show the immunodetection of H-NS-FH in the DNA-protein complexes at the top (B) or within the lanes of the gels (C), using anti-FLAG monoclonal antibodies.

We next assessed whether H-NS also represses the expression of the phoH_Ec gene. For this, the expression of the phoH_Ec-cat fusion was quantified in the WT E. coli and E. coli Δhns strains grown under SPI-1-inducing conditions. The phoH_Ec-cat fusion showed very low expression levels in both E. coli strains (Fig. 6A), showing that, in contrast to that observed for phoH_STm, the expression of phoH_Ec is not induced by the absence of H-NS under the conditions tested.

EMSAs were performed to determine whether H-NS regulates phoH_STm directly. As could be expected, purified H-NS-FH (H-NS-Flag-His₆) protein bound the DNA fragment spanning the entire regulatory region of phoH_STm, but not that of phoH_Ec, at the concentrations tested (Fig. 6B). The DNA-H-NS-FH complexes formed with the entire regulatory region of phoH_STm seem to be very large, since they were retained near the wells of the gel (Fig. 6B); with a shorter fragment of the regulatory region of phoH_STm, the DNA-H-NS-FH complexes were detected within the gel (Fig. 6C). H-NS is known to bind A+T-rich sequences (33); intriguingly, both the regulatory region of phoH_STm and that of phoH_Ec show high A+T content, 57% and 55%, respectively. Next, we analyzed by competitive EMSAs whether MBP-HilD can displace H-NS-FH from phoH_STm. Binding reactions containing the P1phoH-157/+12 fragment of phoH_STm were first incubated with a constant concentration of H-NS-FH, and then increasing concentrations of MBP-HilD were added. Binding reactions containing only H-NS-FH or MBP-HilD were also tested. The DNA-H-NS-FH complex was shifted by MBP-HilD to a slower-migrating complex that remained near the wells of the gel, similar to the complex formed only by MBP-HilD; furthermore, MBP-HilD decreased the amount of H-NS-FH bound to the P1phoH-157/+12 fragment (Fig. 6C). These results indicate that HilD displaces H-NS from phoH_STm.

In all, these results support that cis-regulatory evolution led to the repression of phoH_STm by H-NS, which is involved in the regulation of this gene by HilD.

phoH_STm and phoH_Ec are differentially regulated in response to inorganic phosphorus.

The phoH_Ec gene belongs to the phosphate regulon; its expression is induced by the PhoR/B two-component system in response to inorganic phosphorus (P_i) starvation (22, 23). It was therefore of interest to test whether the phoH_STm gene that was recruited into the HilD regulon conserves the ancestral regulation by the PhoR/B system in response to low P_i. To investigate this notion, we quantified the expression of the phoH_STm-cat fusion in the WT S. Typhimurium strain and its isogenic ΔphoB mutant grown in N-minimal medium (N-MM) containing low or high concentrations of P_i. Surprisingly, the expression of the phoH_STm-cat fusion was induced by low P_i in both the WT strain and the ΔphoB mutant (Fig. 7A). Similar results were obtained by analyzing the expression of the PhoH_Stm-FLAG protein (Fig. 7B). In contrast, the expression of the PhoH_Ec-FLAG protein or the phoH_Ec-cat fusion was induced by low P_i in the WT E. coli strain but not, or only slightly, in its derivative ΔphoB mutant (Fig. 7C and D). In agreement with these results, an in silico analysis identified the previously reported pho-box (PhoB-binding site) on the regulatory sequence of phoH_Ec but did not predict a pho-box on the corresponding sequence of phoH_STm (Fig. 4B) (23). The S. Typhimurium sequence conserves only 10 of the 18 nucleotides forming the E. coli pho-box (Fig. 4B). These results indicate that both phoH_Ec and phoH_STm are similarly regulated by P_i but through different mechanisms; the one for phoH_Ec involves PhoB and that for phoH_STm does not.

FIG 7.

Expression of phoH_STm is activated by low P_i independently of either PhoB or HilD. CAT-specific activity of the phoH_STm-cat (A) and pstS_STm-cat (E) transcriptional fusions, contained in the pphoH_STm-cat and ppstS_STm-cat plasmids, respectively, was determined in the WT S. Typhimurium (STm) SL1344 strain and its derivative ΔphoB and ΔhilD mutants. pstS is known to be regulated by PhoB; thus, it was tested as a positive control. (C) CAT-specific activity of the phoH_Ec-cat transcriptional fusion, contained in the pphoH_Ec-cat plasmid, was determined in the WT E. coli (Ec) BW25113 strain and its derivative mutant. Data represent the average with standard deviation from three independent experiments performed in duplicate. Statistically different values are indicated (*, P < 0.0108; ****, P < 0.0001). The expression of the PhoH_STm-FLAG protein in the WT S. Typhimurium (STm) SL1344 strain and its derivative ΔphoB and ΔhilD mutants (B), as well as the expression of the PhoH_Ec-FLAG protein in the WT E. coli (Ec) MG1655 strain and its derivative ΔphoB mutant (D), was analyzed by Western blotting using monoclonal anti-FLAG antibodies. As a control, the expression of GroEL was also detected using polyclonal anti-GroEL antibodies. CAT activity and detection of the PhoH-FLAG and GroEL proteins were determined from samples of bacterial cultures grown for 16 h at 37°C in N-MM containing low (L) or high (H) P_i.

Next, we asked whether the regulation by P_i independent of PhoB also acts on the pstS gene, which has been shown to belong to the phosphate regulon and to be regulated by PhoB in different bacteria; this gene encodes a component of a high-efficiency P_i transporter (24). To know this, a cat transcriptional fusion of the S. Typhimurium SL1344 pstS (pstS_STm) gene was constructed and assessed in the same bacterial strains and growth conditions as in the previous experiment. The expression of the pstS_STm-cat fusion was induced by low P_i in the WT S. Typhimurium strain but not in its derivative ΔphoB mutant (Fig. 7E), indicating that the pstS_STm gene is regulated by P_i starvation through PhoB. Consistent with this, a computational analysis identified on pstS_STm (data not shown) the pho-box previously reported for the E. coli gene (34).

As expected, according to different studies indicating that HilD does not affect expression of target genes when S. Typhimurium is grown in minimal medium (2, 25), the regulation of phoH_STm and pstS_STm by P_i was similar in the WT strain and its derivative ΔhilD mutant (Fig. 7A, B, and E).

Taken together, these results support that the phoH_STm gene evolved to be regulated by P_i through a PhoB-independent mechanism.

DISCUSSION

Differential regulation of homologous genes is an important mechanism for phenotypic variability among closely related bacteria (4, 35–38). In this study, we further define the mechanism by which HilD induces the expression of phoH_STm and show that both the gain of HilD and cis-regulatory evolution adapted this regulation in S. Typhimurium, leading to differential expression of phoH between S. Typhimurium and E. coli K-12.

Our data indicate that in the absence of the H-NS repressor, the phoH_STm gene reached WT expression levels independently of HilD. We found that H-NS binds to and represses the expression of the phoH_STm gene but not the phoH_Ec gene. Moreover, we determined that HilD displaces H-NS from the phoH_STm gene. These results support that HilD induces the expression of phoH_STm by antagonizing the negative effect of H-NS, a mechanism that HilD follows to induce the expression of several other target genes (2, 9, 16, 27–29, 39). Interestingly, we observed that the overexpression of HilD further increased the expression of the phoH_STm gene in the absence of H-NS, which suggests that HilD, besides acting as an anti-H-NS factor, could antagonize the action of an additional repressor or improve the RNA polymerase binding on the promoter of this gene. In agreement with this notion, a previous report indicates that HilD counteracts repression exerted by the nucleoid-associated protein Hha on the rtsA gene (27). Furthermore, evidence from another study suggests that HilD recruits the RNA polymerase on the hilA promoter (40).

A previous transcriptome sequencing (RNA-seq) analysis revealed two transcriptional start sites for the phoH_STm gene (25, 26). Accordingly, our results show that two promoters (P1 and P2) transcribe this gene under SPI-1-inducing conditions. We found that HilD positively regulates the P1 promoter, but not the P2 promoter, by binding a site that overlaps the −35 region of the P1 promoter. To our knowledge, not one regulatory mechanism other than that mediated by the PhoB/R system (see below) had been defined for the phoH gene in any other bacteria.

We detected only background expression levels for the phoH_Ec gene under SPI-1-inducing conditions, even in the presence of HilD. In agreement with this finding, computational analyses revealed that the regulatory sequence required for the HilD-mediated expression of the phoH_STm gene (P1 −67/+3 sequence) is poorly conserved in the phoH_Ec gene, including the HilD-binding sequence on phoH_STm. Moreover, a phylogenetic analysis based on the P1 −67/+3 sequence of phoH grouped Salmonella and E. coli strains in two different clades. Thus, our results support that both the acquisition of HilD and regulatory evolution of the P1 −67 to +3 sequence adapted the regulation of phoH by HilD in S. Typhimurium, which led to this gene being expressed under growth conditions relevant for Salmonella virulence. Interestingly, cis-regulatory evolution also adapted repression of phoH_STm by H-NS, which is involved in the HilD-mediated expression of this gene. It is tempting to speculate that the regulation by HilD of other ancestral genes, like flhDC and yobH, was also adapted by cis-regulatory evolution. On the other hand, cis-regulatory changes in S. Typhimurium with respect to S. bongori led the srfN, ugtL, and flhDC genes to be regulated by the SPI-2 transcriptional regulator SsrB and thus to have a role in S. Typhimurium virulence (36–38). Thus, these reports indicate that cis-regulatory changes have played an important role in the evolution of Salmonella pathogenicity. It is reasonable to think that the integration of the phoH_STm gene into the HilD regulon was favorable for S. Typhimurium virulence. Our results indicate that the deletion of phoH_STm slightly decreases the early intestinal colonization but does not affect the systemic infection of mice by S. Typhimurium (Fig. 8A and B). We observed similar phenotypes with the deletion of the siiE gene (Fig. 8), which encodes a giant nonfimbrial adhesin shown to be necessary for intestinal colonization (41, 42). In contrast, the deletion of phoH_STm did not affect the S. Typhimurium invasion of HeLa cells or RAW264.7 mouse macrophages (data not shown). Further studies into the possible role of phoH in S. Typhimurium pathogenicity are required. Even when phoH is conserved in many bacteria, its biological function remains unknown in all of them.

FIG 8.

Effect of phoH_STm in the intestinal colonization and systemic infection of mice by S. Typhimurium. Groups of 12 streptomycin-pretreated mice were orally inoculated with the WT S. Typhimurium strain or its derivative ΔphoH or ΔsiiE mutant. (A) Numbers of CFU/g from feces were determined at 1, 2, and 3 days postinfection (dpi). (B) Numbers of CFU/g from the spleen were determined at 1 day postinfection. Bars denote the standard errors of the averages for each experimental group. Statistically different values are indicated (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

We found an additional trait for phoH_STm that was also generated by cis-regulatory evolution: the expression of this gene was induced by low P_i independently of the PhoR/B two-component system. A previous report indicates that only a small number of the S. Typhimurium genes that are induced by low P_i belong to the PhoB/R system regulon (43). Thus, S. Typhimurium seems to have a regulatory system alternative to PhoB/R to active phoH_STm and many other genes in response to P_i starvation. Our results indicate that HilD, as well as PhoP and SsrB, two transcriptional regulators that induce expression of Salmonella virulence genes in N-MM containing low P_i, are not involved in the activation of phoH_STm by low P_i (data not shown). Transcriptomic studies have shown that the expression of phoH_STm is induced by various types of stress that S. Typhimurium faces inside hosts, such as high concentrations of NaCl, low Fe²⁺, anaerobiosis, and exposure to peroxide or nitric oxide, among others (25). Various regulators could coordinate to activate the expression of phoH_STm in response to different cues, which remains to be determined.

Beyond examples for specific genes, differential regulation of multiple genes mediated by a transcriptional regulator has been described among closely related bacteria. For instance, the SlyA and Ler regulons are different between S. Typhimurium and E. coli (44) and between enteropathogenic and enterohemorrhagic E. coli strains (45). Moreover, mechanisms underlying transcriptional control evolved from prokaryotes to eukaryotes (46).

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Bacterial strains used in this study are shown in Table 1. Bacterial cultures for the determination of chloramphenicol acetyltransferase (CAT) activity or for the Western blot assays were grown in 250-mL flasks containing 50 mL of LB or N-minimal medium (N-MM), which were inoculated with a 100-fold dilution of an overnight culture, as described previously (2, 47). To evaluate the low and high P_i conditions, KH₂PO₄ in N-MM was used at concentrations of 64 μM and 640 μM, respectively. The antibiotics ampicillin (200 μg/mL), streptomycin (100 μg/mL), and kanamycin (20 μg/mL) were added to growth media as needed.

Construction of plasmids.

Plasmids and primers used in this study are listed in Tables 1 and 2, respectively. To construct the pphoH_Ec-cat plasmid, the intergenic sequence upstream phoH_Ec was amplified by PCR with the primer pair phoH-Fw41/phoH-Rv42, using chromosomal DNA from the WT E. coli K-12 MG1655 strain as the template. To construct the pP1phoH_STm-cat, pP2phoH_STm-cat, and pphoH_STm-248 + 12-cat plasmids, the corresponding sequence upstream of phoH_STm (−653 to +12, −5 to +416, and −248 to +12, respectively, with respect to the P1 promoter) was amplified by PCR with the primer pairs phoH-Fw/phoH-Rv14, phoH-Fw15/phoH-Rv, and phoH-Fw27/phoH-Rv14, respectively, using chromosomal DNA from WT S. Typhimurium SL1344 strain as the template. The pphoH_HilDmut-cat and pphoH_P1mut-cat plasmids, carrying mutations in the predicted HilD-binding site and the P1 promoter, respectively, were constructed by overlapping PCR using chromosomal DNA from the WT S. Typhimurium strain as the template. First, two PCR products were obtained for each plasmid, with the primer pairs phoH-Fw/phoH-Rv80 and phoH-Fw79/phoH-Rv for pphoH_HilDmut-cat and with the primer pairs phoH-Fw/phoH-Rv84 and phoH-Fw83/phoH-Rv for pphoH_P1mut-cat. The PCR products were purified, and those for each plasmid were mixed and used as the template for a second PCR with the external primers phoH-Fw and phoH-Rv. To construct the ppstS_STm-cat plasmid, the intergenic sequence upstream of pstS_STm was amplified by PCR with the primer pair pstS-Fw71/pstS-Rv72, using chromosomal DNA from WT S. Typhimurium SL1344 strain as the template. All the PCR products were purified and digested whit BamHI and HindIII enzymes and then cloned into the same restriction sites of the pKK232-8 vector, which carries a promoterless cat gene (48). The pphoH_STm−67 + 3-cat plasmid was constructed with the complementary primers phoH-Fw37 and phoH-Rv38. Briefly, these primers were mixed in a 1:1 ratio, heated to 95°C for 10 min, and then cooled to room temperature for 1 h. The obtained double-strand product carries cohesive ends for its cloning into the BamHI and HindIII restriction sites of the pKK232-8 vector (48).

TABLE 2.

Primers used in this study

Primer	Sequencea (5′–3′)	Target gene	REb
For cat transcriptional fusion
phoH-Fw41	CGAGGATCCTGGATATGGGGTTGCTGTTTG	phoHEc	BamHI
phoH-Rv42	CGAAAGCTTCATGGAGAGCACCTTGAGTTG	phoHEc	HindIII
phoH-Fw	CGAGGATCCAATATGGCTGGCTGGATCTG	phoHS Tm	BamHI
phoH-Rv14	CGAAAGCTTGTCAGTCTCTTACAGAAAGATTAC	phoHS Tm	HindIII
phoH-Fw15	CGAGGATCCTGTAAGAGACTGACAATGACGC	phoHS Tm	BamHI
phoH-Rv	CGAAAGCTTCCATGGATAGCACCTTGAGT	phoHS Tm	HindIII
phoH-Fw27	CGAGGATCCTGACGCAATAGAGTAATGACAAAA	phoHS Tm	BamHI
phoH-Fw37	GATCCCCCGCAGTAGCTAATGATTATCTTTTTTAGTCTCCTGCCGATGAAATAATCGTGTAATCTTTCTGTAAGAA	phoHS Tm
phoH-Rv38	AGCTTTCTTACAGAAAGATTACACGATTATTTCATCGGCAGGAGACTAAAAAAGATAATCATTAGCTACTGCGGGG	phoHS Tm
phoH-Fw79	CAGCATTAATACCCGCAGTAGAAAGCACACAGCTTTTTTAGTCTCCTGCCGATG	phoHS Tm
phoH-Rv80	CATCGGCAGGAGACTAAAAAAGCTGTGTGCTTTCTACTGCGGGTATTAATGCTG	phoHS Tm
phoH-Fw83	CTGCCGATGAAATAATCGTGCGGCTCTTCTGTAAGAGACTGACAATG	phoHS Tm
phoH-Rv84	CATTGTCAGTCTCTTACAGAAGAGCCGCACGATTATTTCATCGGCAG	phoHS Tm
phoH-Fw17	CGAGGATCCAGCGATGGGAGAGAGGACAC	phoHS Tm	BamHI
pstS-Fw71	CGAGGATCCCTCTTTGTCCTGGCGATCCC	pstSS Tm	BamHI
pstS-Rv72	CGAAAGCTTGAACAGGCCTTCCTGATTCAG	pstSS Tm	HindIII
For gene deletion
phoH-H1P1	GGCTTACAAGGAAGCCAACCCTCAGATGTGTGTGCGCATAATTGTAGGCTGGAGCTGCTTCG	phoHS Tm
phoH-H2P2	CGGGCTCTGAAAGTCAATGCTATATTAGCTGTATGCGTGAAGCATATGAATATCCTCCTTAG	phoHS Tm
siiE-H1P1	ATGGGAAATAAAAGCATACAAAAGTTTTTTGCCGATCAAAATTGTAGGCTGGAGCTGCTTCG	siiE
siiE-H2P2	TTATGCGTGTTCTTCTTGATTATCTACAGGTAGCGTAACTTCCATATGAATATCCTCCTTAG	siiE
SphoB H1P1	CGACTAACAGGGCAAATTATGGCGAGACGTATTCTGGTCGTATGTAGGCTGGAGCTGCTTCG	phoBS Tm
SphoB H2P2	CAACTGGCGGAAAAGGCATTAAAAGCGGGTCGAAAAACGATACATATGAATATCCTCCTTAG	phoBS Tm
For gene FLAG tagging
phoHSFlagFw	GCGCTCGGCGCTTTGTCAGCGAACGCTTCACGCATACAGCGACTACAAAGACCATGACGGT	phoHS Tm
phoHSFlagRv	ACAAAGCCCGGTTTCCCGGGCTCTGAAAGTCAATGCTATACATATGAATATCCTCCTTAGTTC	phoHS Tm
phoHEcFlagFw	CGTTCGGCACTTTGCCAACGTACGCTGCATGCCTACAGTGACTACAAAGACCATGACGGT	phoHEc
phoHEcFlagRv	ACAAAGCCCGGTTCGCCCGGGCTCTGCACCGATAACACACCATATGAATATCCTCCTTAGTTC	phoHEc

^aUnderlined letters indicate the respective restriction enzyme site in the primer. The sequences corresponding to the template pKD4 or pSUB11 plasmids (Table 1) are in italics.

^bRE, restriction enzyme for which a site was generated in the primer.

Construction of chromosomal deletion and 3×FLAG-tagged mutant strains.

The phoB, phoH, and siiE genes were replaced with a selectable kanamycin resistance cassette in the S. Typhimurium SL1344 strain by the λRed recombinase system, as reported previously (49), using the primers shown in Table 2, generating the DTM144, DTM150, and DTM151 strains, respectively. The chromosomal phoH gene was 3×FLAG-tagged in the WT S. Typhimurium SL1344 strain as well as in the JPTM25, DTM145, MG1655, and DTM148 strains, using a previously reported method based on the λRed recombinase system (50), and the primers are shown in Table 2, generating the DTM142, DTM143, DTM146, DTM147, and DTM149 strains, respectively. The DTM145 and DTM148 strains were generated by removing the kanamycin resistance cassette from the DTM144 and JW0389 strains, respectively, using the pFLP3 plasmid expressing the FLP recombinase, as described previously (51). All modified strains were verified by PCR amplification and sequencing.

CAT assays.

The assays for CAT-specific activity were performed as described previously (52).

Western blotting.

Western blot assays were conducted as described previously (47), with minor modifications. Briefly, to detect PhoH-FLAG, sonicated cell extracts were prepared from samples of bacterial cultures grown in LB or N-MM at the indicated time. Proteins were separated on SDS–15% polyacrylamide gels. Immunoblots were performed with monoclonal anti-FLAG M2 (Sigma) or polyclonal anti-GroEL (StressGen) antibodies at 1:5,000 and 1:100,000 dilutions, respectively. To detect MBP-HilD or H-NS-FH in the EMSAs, monoclonal anti-MBP (Sigma) or monoclonal anti-FLAG M2 (Sigma) antibodies were used at a dilution of 1:3,000. Horseradish peroxidase-conjugated anti-mouse and anti-rabbit (Pierce) diluted at 1:10,000 were used as secondary antibodies. Reaction bands on membranes were developed with the Western Lightning chemiluminescence reagent plus (Perkin-Elmer) and exposition to KodaK X-Omat films.

Sequence analysis.

BLASTP and Clustal Omega2 (EMBL-EBI) were used to compare amino acid and nucleotide sequences, respectively. The phylogenetic tree based on the P1phoH−67 + 3 sequence was generated with the IQ-TREE (53) and interactive Tree of Life (v6) (54) software using default parameters. Prediction of HilD-binding sites was performed with the previously reported position-specific scoring matrices, representing HilD-binding consensus sequences (10), using the matrix-scan program (55) with a P value of 1e−3. Prediction of PhoB-binding site was accomplished with the BPROM tool of the Softberry software (56).

Expression and purification of MBP-HilD and H-NS-FH.

The maltose binding protein (MBP)-HilD and the H-NS-FLAG-His₆ (H-NS-FH) protein were expressed in the E. coli BL21/DE3 strain containing the pMAL-HilD1 or pBAD-H-NS-FH plasmid and purified by using amylose-agarose or Ni²⁺-nitrilotriacetic-agarose affinity columns, respectively, as described previously (2).

EMSAs.

The fragments P1 and P2, spanning the phoH_STm sequences contained in the P1phoH_STm-cat and P2phoH_STm-cat fusions, were obtained by PCR amplification using the primer pairs phoH-Fw/phoH-Rv14 and phoH-Fw15/phoH-Rv, respectively, and chromosomal DNA from the WT S. Typhimurium SL1344 strain as the template. The WT and mutated P1phoH-157/+12 fragments were obtained by PCR using the primer pairs phoH-Fw17/phoH-Rv14 and phoH-Fw17/phoH-Rv84 and the DNA from the pphoH_STm-cat (WT), pphoH_HilDmut-cat (mutations in the predicted HilD-binding site), or pphoH_P1mut-cat (mutations in the P1 promoter) plasmids as the template. The fragments spanning the phoH_STm or phoH_Ec full-length regulatory region, contained in the phoH_STm-cat or phoH_Ec-cat fusions, were obtained by PCR using the primer pairs phoH-Fw/phoH-Rv and phoH-Fw41/phoH-Rv42 and chromosomal DNA from WT S. Typhimurium SL1344 and E. coli MG1655 strains as the template, respectively. The final PCR products were purified with the Zymo Research DNA Clean & Concentrator kit (ZYMO RESARCH). Protein-DNA binding reactions with purified MBP-HilD and/or H-NS-FH proteins were analyzed as described previously (2, 57). For competitive EMSAs, the WT P1phoH-157/+12 fragment was first incubated with 4 μM H-NS-FH for 15 min and then incubated with increasing concentrations of MBP-HilD for an additional 20 min. Binding mixtures were electrophoretically separated in 6% nondenaturing acrylamide in 0.5× Tris-borate-EDTA buffer at room temperature. DNA bands were visualized by staining with ethidium bromide in an Alpha-Imager UV transilluminator (Alpha Innotech Corp.). MBP-HilD and H-NS-FLAG were detected on the DNA/protein complexes by Western blotting.

Mouse infection assays.

Mouse experiments were conducted according to the standard operating protocols approved by the International Committee for Animal Care and Use from CICUAL-UNAM and by the Official Mexican Norm NOM-062-Z00-1999. Pathogen-free BALB/c female mice (6 to 7 weeks old) were obtained from the Experimental Medicine Research Unit, School of Medicine, UNAM, Mexico. Maintenance, treatment with 50 mg of streptomycin, infection, and euthanasia of the mice, as well as the counting of the number of CFU per gram from feces or the spleen were performed as described previously (57, 58). Groups of 12 streptomycin-pretreated mice were infected by the orogastric route with bacterial suspensions containing 1.0 × 10⁶ CFU of the WT S. Typhimurium strain or its derivative ΔphoH and ΔsiiE mutants, prepared in sterile 1× phosphate-buffered saline.

Cell infection assays.

Invasion of HeLa cells or RAW264.7 mouse macrophages was tested by gentamicin assays, as described previously (57, 59).

Statistical analysis.

Statistical analyses were performed with the GraphPad Prism 6.0 or 9.2.0 software (GraphPad Inc., San Diego, CA), using the one-way analysis of variance (ANOVA) or Student's t test (unpaired). P values of <0.05 were considered statistically significant.