ABSTRACT
The Gram-negative bacterium Pasteurella multocida is the causative agent of a number of economically important animal diseases, including avian fowl cholera. Numerous P. multocida virulence factors have been identified, including capsule, lipopolysaccharide (LPS), and filamentous hemagglutinin, but little is known about how the expression of these virulence factors is regulated. Hfq is an RNA-binding protein that facilitates riboregulation via interaction with small noncoding RNA (sRNA) molecules and their mRNA targets. Here, we show that a P. multocida hfq mutant produces significantly less hyaluronic acid capsule during all growth phases and displays reduced in vivo fitness. Transcriptional and proteomic analyses of the hfq mutant during mid-exponential-phase growth revealed altered transcript levels for 128 genes and altered protein levels for 78 proteins. Further proteomic analyses of the hfq mutant during the early exponential growth phase identified 106 proteins that were produced at altered levels. Both the transcript and protein levels for genes/proteins involved in capsule biosynthesis were reduced in the hfq mutant, as were the levels of the filamentous hemagglutinin protein PfhB2 and its secretion partner LspB2. In contrast, there were increased expression levels of three LPS biosynthesis genes, encoding proteins involved in phosphocholine and phosphoethanolamine addition to LPS, suggesting that these genes are negatively regulated by Hfq-dependent mechanisms. Taken together, these data provide the first evidence that Hfq plays a crucial role in regulating the global expression of P. multocida genes, including the regulation of key P. multocida virulence factors, capsule, LPS, and filamentous hemagglutinin.
INTRODUCTION
Small noncoding RNA (sRNA) molecules play essential roles in regulating the production of a wide range of proteins, including those involved in virulence; quorum sensing; and the metabolism of carbon, amino acids, and iron (1–5). Typically, sRNA molecules are between 50 and 400 nucleotides long and act primarily by interacting with one or more target mRNAs via short imperfect base pairing (6, 7). Protein production can be controlled by sRNAs via multiple mechanisms, including inhibition of translation, activation of translation, and alteration of transcript degradation rates (8, 9). The regulation of bacterial protein expression via sRNAs often requires the activity of Hfq, an RNA-binding chaperone protein that belongs to the Sm-like RNA-binding protein family and displays a highly conserved core sequence (6, 10). Hfq monomers form a homohexameric ring-shaped structure that preferentially binds to A/U-rich sequences on target RNA molecules and mediates the various posttranscriptional regulatory mechanisms of sRNAs (11–13). Hfq can also directly protect sRNA molecules from RNase E-mediated degradation (14) and can have an sRNA-independent role in the regulation of mRNA decay, via direct interaction with poly(A) polymerase (15).
Hfq homologues have been identified in a range of Gram-positive and Gram-negative bacterial species, including many pathogenic species (16). Most hfq mutants display a broad range of pleiotropic phenotypes, suggesting a global role for Hfq in bacterial physiology (6, 17, 18). These pleiotropic effects have been observed more commonly in Gram-negative species, including the pathogens Brucella abortus, Salmonella enterica, Yersinia pestis, Moraxella catarrhalis, Neisseria meningitidis, and Vibrio cholerae, but also in some Gram-positive species such as Listeria monocytogenes (18–24). Many hfq mutants show reduced growth rates, reduced virulence, and increased susceptibility to host defense mechanisms and environmental stresses (6, 16).
Pasteurella multocida is an encapsulated, Gram-negative, facultative anaerobe that is the causative agent of a number of important animal diseases, including fowl cholera (25). Fowl cholera can affect most avian species, including chickens, turkeys, ducks, and wild waterfowl, with infections resulting in high mortality rates and major economic losses to poultry industries worldwide (26, 27). P. multocida is a heterogeneous species, with strains being classified into five serogroups (serogroups A, B, D, E, and F) based on capsular composition (28) and into eight lipopolysaccharide (LPS) genotypes (L1 through to L8) based on the LPS outer core biosynthesis genes (29). A number of P. multocida virulence factors have been defined, including the polysaccharide capsule, LPS, filamentous hemagglutinin adhesins, iron-sequestering systems, and sialic acid uptake (30–33). The polysaccharide capsule is a critical virulence factor of P. multocida that allows the bacteria to evade host immune defense mechanisms (34, 35). Very little is known about how the expression of these virulence factors is regulated, as only one P. multocida regulatory protein, Fis, has been identified and characterized. Fis has been shown to be essential for the production of capsule, and a fis mutant displayed highly reduced expression levels of many genes, including those within the capsule biosynthesis locus and pfhB2, encoding filamentous hemagglutinin (36).
In this study, the role of Hfq in P. multocida gene expression and protein production was characterized by using an hfq mutant constructed in the P. multocida fowl cholera strain VP161. As Hfq is a crucial modulator of sRNA action, an understanding of the role of Hfq provides an initial indication of the significance of sRNAs in the regulation of P. multocida genes and will underpin future work on specific sRNAs in this bacterium. The VP161 hfq mutant displayed highly reduced hyaluronic acid (HA) capsule production as well as reduced in vivo fitness. RNA sequencing (RNA-Seq) and high-throughput quantitative liquid proteomics analyses indicated that Hfq plays an important role in controlling the expression of >100 genes/proteins. In particular, genes/proteins belonging to the capsule biosynthetic locus were expressed at reduced levels in the hfq mutant, consistent with the reduced levels of HA capsule production. These analyses also revealed that Hfq plays a role in the production of filamentous hemagglutinin and proteins involved in LPS biosynthesis. Taken together, these data indicate that Hfq plays an important role in the regulation of key P. multocida virulence factors and is essential for full in vivo fitness.
MATERIALS AND METHODS
Bacterial strains, plasmids, and culture conditions.The bacterial strains and plasmids used in this study are listed in Table 1. Escherichia coli was grown in lysogeny broth (LB), and P. multocida strains were grown in either heart infusion (HI) or brain heart infusion (BHI) broth (Oxoid). Liquid cultures were incubated at 37°C with shaking (200 rpm). Solid media were obtained by the addition of 1.5% agar. When required, the media were supplemented with antibiotics at the following concentrations: 10 μg/ml (Escherichia coli) or 2.5 μg/ml (P. multocida) tetracycline and 50 μg/ml kanamycin (P. multocida).
Bacterial strains and plasmids used in this study
Measurement of bacterial growth rates in rich medium.Cultures of P. multocida strains grown overnight were subcultured at a 1:50 dilution in BHI broth and grown until cultures reached an optical density at 600 nm (OD600) of 0.2 to 0.3. A 1-ml aliquot (standardized to an OD600 of 0.2) of each culture was transferred to 50 ml of sterile BHI broth and incubated at 37°C with shaking. To determine the growth rate, optical density measurements of the liquid cultures (at 600 nm) were taken at regular intervals by using the WPA CO8000 Biowave cell density meter, until the late exponential growth phase was reached.
Ethics statement.All animal experiments were carried out in accordance with the provisions of the Prevention of Cruelty to Animal Act, 1986; the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (37); and Monash University Animal Welfare Committee guidelines and policies. The protocols were approved by the Monash Animal Research Platform 2 (MARP-2) Animal Ethics Committee (AEC) of Monash University (AEC number MARP/2011/066, Understanding How Pasteurella multocida Causes Disease in Animals).
DNA manipulations.Restriction digests, ligations, and PCR amplifications were performed by using enzymes obtained from NEB or Roche, according to the manufacturer's instructions. Plasmid DNA was prepared by using the NucleoSpin plasmid kit (Macherey-Nagel), while genomic DNA was prepared by using the HiYield genomic DNA minikit (Real Genomics). PCR amplification of DNA was performed by using Taq DNA polymerase (Roche) or Phusion high-fidelity DNA polymerase (NEB) on an Eppendorf Mastercycler. Amplified DNA was purified by using the NucleoSpin gel or PCR cleanup kit (Macherey-Nagel). The primers used in this study are listed in Table 2. Column-purified DNA samples were quantified by using the NanoDrop 1000 spectrophotometer (Thermo Scientific) and/or by gel electrophoresis. Sequencing reactions were performed by using Applied Biosystems Prism BigDye terminator mix, version 3.1. Electropherograms were generated on a capillary platform Applied Biosystems 3730 genetic analyzer and analyzed by using Vector NTI Advance, version 11.5 (Life Technologies).
Primers used in this study
Construction of P. multocida hfq and hyaD mutants.The TargeTron (Sigma-Aldrich) mutagenesis system was used to separately inactivate hfq and, as a control for HA assays, the capsular biosynthesis gene hyaD in P. multocida A:1 strain VP161, as described previously (38). The group II intron within P. multocida TargeTron plasmid pAL953 (Table 1) was retargeted to hfq or hyaD according to the TargeTron user manual and with specific primers designed by using the TargeTron design site (Sigma-Aldrich) (Table 2). The retargeted TargeTron plasmids pAL1104 (hfq targeted) and pAL1069 (hyaD targeted) were then used to separately transform P. multocida VP161 by electroporation, and transformants were selected on solid agar containing kanamycin. Following transformation, strains were cured of free-replicating plasmid as described previously (38), and putative hfq and hyaD mutants were identified by patching for Kanr/Specs colonies. Colony PCRs using the TargeTron-specific primer EBS universal, together with an hfq- or hyaD-specific primer (BAP7400 and BAP2067, respectively) (Table 2), were used to initially identify TargeTron mutants (data not shown). To confirm that P. multocida hfq or hyaD mutants contained only a single TargeTron insertion, Southern hybridization was performed (data not shown), using a TargeTron intron-specific digoxigenin (DIG)-labeled probe. In addition, the exact position of the intron within the genome was determined by direct sequencing using genomic DNA isolated from each TargeTron mutant as the template with the intron-specific EBS universal primer (Table 2). Two mutants that generated unambiguous sequencing data identical to data for the correct target gene sequence (immediately adjacent to the intron insertion site) were selected for further study and designated AL2521 (VP161 hfq mutant) and AL2234 (VP161 hyaD mutant).
Complementation of the hfq mutant.For complementation of the hfq mutant, an intact copy of the hfq gene was amplified from VP161 genomic DNA by using primers BAP7400 and BAP7401 that contained BamHI and SalI restriction sites, respectively (Table 2). After digestion with BamHI and SalI, the amplified fragment was cloned into BamHI- and SalI-digested pAL99T (Table 1), such that the P. multocida constitutive promoter tpiA would drive the transcription of hfq. The ligated products were then used to transform E. coli strain DH5α. E. coli transformants containing the correct plasmid were identified by colony PCR using pAL99T-specific and hfq-specific primers BAP2679 and BAP7401, respectively. DNA sequencing was used to confirm the fidelity of the hfq sequence, and one correct recombinant plasmid was selected and designated pAL1108. The vector pAL99T and the complementation plasmid pAL1108 were then used to separately transform the VP161 hfq mutant (AL2521) via electroporation, generating strains AL2527 and AL2526, respectively (Table 1).
Quantitative hyaluronic acid capsule assay.The amount of HA capsular material produced by each P. multocida strain was determined as described previously (34). Student's unpaired t test was used to assess the differences in HA production between strains.
Competitive in vivo growth assays.Competitive growth assays in mice were performed to quantify the relative in vivo fitness of the hfq mutant and complemented strains compared with wild-type strain VP161, as described previously (39). To identify each bacterial population, after overnight incubation of input and output samples on HI medium, individual colonies were patched onto HI medium and either HI medium containing kanamycin (to select for the hfq mutant) or HI medium containing kanamycin and tetracycline (to select for the hfq mutant containing the plasmid). Mutants with a competitive index (CI) of <0.5 were identified as having significantly reduced in vivo fitness, and Student's unpaired two-tailed t test was also used to assess the statistical differences in relative CIs between each of the tested strains.
RNA extraction and purification, rRNA depletion, library preparation, and high-throughput RNA sequencing.P. multocida strains AL2521 (VP161 hfq mutant) and VP161 (wild type) were each grown in BHI broth to the mid-exponential growth phase (OD600 = 0.6; ∼1.0 × 109 CFU/ml). A total of 2.5 ml of ice-cold killing buffer (containing 0.05 M Tris-HCl [pH 7.5], 15 mg/ml sodium azide, and 0.6 mg/ml chloramphenicol) was added to each 25-ml culture, and the mixed sample was cooled on ice before centrifugation at 9,000 × g for 10 min (4°C). Pelleted cells were then resuspended in 1 ml RNAlater stabilization reagent (Qiagen) and incubated for 5 min at room temperature before centrifugation at 5,000 × g for 10 min. Total RNA was purified by using the RNeasy minikit (Qiagen) with DNase treatment according to the manufacturer's instructions. RNA samples were quantified and analyzed for DNA contamination by using the Qubit kit (Invitrogen) and visualized by gel electrophoresis. rRNA was removed by using the Ribo-Zero magnetic kit (Illumina) according to the manufacturer's instructions. The TruSeq RNA sample preparation kit (Illumina) was used to construct cDNA libraries, with library validation, normalization, and pooling performed by Micromon Services (Monash University). All libraries were combined and sequenced on a single lane of a HiSeq 2000 instrument (Macrogen, South Korea). Mapping of the RNA-Seq reads to the draft VP161 genome was carried out by using CLC Genomics Workbench v 7.0 (CLC Bio), and statistical analyses were carried out by using voom and limma as described previously (40). Raw combined data sets of 50,071,021 and 47,130,114 total paired reads were generated for AL2521 and VP161, respectively. Of these, 17,718,215 and 14,903,935 paired reads (for AL2521 and VP161, respectively) aligned uniquely to the P. multocida VP161 genome (∼30% unambiguous reads for each combined data set), giving high coverage of the transcriptome. Genes were identified as being differentially expressed if they displayed a ≥2.0-fold (≥1.0-log2) change in expression across the triplicate replicates at a false discovery rate (FDR) of <0.05. The differentially expressed VP161 genes identified were mapped to their closest orthologue in the fully annotated reference genome of P. multocida strain Pm70 (41) so that overrepresented gene ontology groups and pathway associations could be determined via BioCyc (42), as we reported previously (40).
Quantitative proteomics and mass spectrometry.Cultures of each P. multocida strain were grown to either the early exponential (OD600 = 0.2) or mid-exponential (OD600 = 0.6) growth phase and then centrifuged at 9,400 × g for 10 min. The pelleted cells were washed twice in 100 mM Tris-HCl (pH 7.5) and lysed via the addition of 0.35 ml of lysis buffer (4% SDS, 100 mM dithiothreitol [DTT], 100 mM Tris-HCl [pH 7.5]) and incubation for 10 min at 99°C. The protein concentration was determined by using the Bradford assay (43). Proteins (150 μg/sample) were purified and trypsin digested by using a filter-aided sample preparation (FASP) protein digestion kit (Expedeon) as described previously (44), except that triethyl ammonium bicarbonate was used instead of ammonium bicarbonate. The digested proteins were isotopically labeled by dimethylation using heavy and light formaldehyde as described previously (45). Heavy and light samples were then mixed and desalted by using an Empore 4-mm/1-ml extraction disk cartridge (C18-SD) (3M). Desalted samples were concentrated under a vacuum and resuspended in 0.1% formic acid and 2% acetonitrile. Peptides were resolved and analyzed by nano-liquid chromatography (LC)–tandem mass spectrometry using a Dionex ultra-high-performance liquid chromatography system coupled to a Thermo Scientific Q-Exactive Orbitrap mass spectrometer (located at the Monash Biomedical Proteomics Facility). Proteins were identified and quantified with MaxQuant (46) and Perseus software. For differential expression testing, the protein expression ratios were log transformed. These transformed ratios were roughly normally distributed, so limma (47) was used to fit and test for the significance of differential expression. Those proteins showing a ≥2.0-fold (≥1.0-log2) change in expression across the triplicate replicates at an FDR of <0.05 were considered to be differentially expressed.
Accession numbers.The RNA-Seq data are available at the NCBI Gene Expression Omnibus under accession number GSE77721. The proteomics data have been deposited in ProteomeXchange via the PRIDE database under accession number PXD003586.
RESULTS
The hfq mutant produces reduced hyaluronic acid capsule.To investigate the role of Hfq in P. multocida, an hfq mutant was generated in the highly virulent strain VP161. The mutant grew indistinguishably from the wild-type strain in BHI liquid broth at 37°C in vitro (see Fig. S1 in the supplemental material), but on solid medium, it produced smaller and less mucoid colonies than those of wild-type strain VP161, consistent with reduced HA capsular expression. The amount of HA was quantified in mid-exponential-phase cultures of wild-type strain VP161, the hfq mutant (AL2521), the complemented hfq mutant (AL2526), and the hfq mutant containing the vector only (AL2527). The hfq mutant produced significantly less HA than did the wild-type strain, and HA production was restored to near-wild-type levels when the mutant was complemented with an intact copy of hfq (Fig. 1A). The control hyaD mutant produced no detectable HA (data not shown).
Hyaluronic acid (HA) polysaccharide capsule produced by Pasteurella multocida strains. (A) Amounts of HA produced at the mid-exponential growth phase by wild-type strain VP161 (WT), the hfq mutant, the complemented hfq mutant (hfq[hfq]), and the hfq mutant with the empty vector (hfq[vector]). (B) Amounts of HA detected (per 109 cells) in wild-type (●) and hfq mutant (▲) cells at the early exponential (OD600 = 0.20 to 0.25), mid-exponential (OD600 = 0.65 to 0.75), and late exponential (OD600 = 1.10 to 1.40) growth phases. The average amount of HA under each condition is indicated by the horizontal bars, ± standard deviations of the means. *, P value of <0.05 as determined by an unpaired t test.
The amount of HA capsule produced by the hfq mutant was also assessed at different growth phases in vitro. The wild-type strain produced small amounts of HA at the early exponential growth phase, maximal amounts of HA at the mid-exponential growth phase, and intermediate levels of HA at the late exponential growth phase (Fig. 1B). No detectable HA was produced by the hfq mutant at the early exponential growth phase, and significantly reduced levels of HA were produced at the mid- and late exponential growth phases (Fig. 1B). Again, the hyaD acapsular control strain produced no detectable HA at any of the tested growth phases (data not shown). These data indicate that capsule production in P. multocida is growth phase dependent, and inactivation of hfq results in significantly reduced capsule expression during all tested growth phases in vitro.
The hfq mutant displays reduced in vivo fitness in mice.As the HA capsule is a known virulence factor of P. multocida serogroup A strains (34), competitive growth assays were performed in mice to compare the in vivo fitness of the wild-type strain with that of the hfq mutant, the hfq mutant complemented with an intact copy of hfq, and the hfq mutant containing the vector only. The hfq mutant displayed a 4-fold reduction in in vivo growth compared to that of the wild-type parent strain VP161 (CI = 0.25 ± 0.05) (Fig. 2). Importantly, this loss of in vivo fitness was restored to near-wild-type levels when the hfq mutant was provided with an intact copy of hfq in trans (CI = 0.68 ± 0.06) (Fig. 2). These data indicate that the P. multocida VP161 Hfq is important for in vivo fitness in mice.
The hfq mutant is attenuated for growth in mice. Shown are competitive growth indices for the P. multocida strains in a mouse infection model. Each data point shows the individual CI for one pair of strains in one mouse. The average CI for each pair of competing strains is indicated by the horizontal bars, ± standard deviations of the means. * indicates a P value of <0.05 as determined by a two-tailed unpaired t test.
Global RNA expression changes in the P. multocida hfq mutant.Hfq is an RNA chaperone that has been shown in many other bacterial species to be essential for the action of most trans-encoded regulatory sRNA molecules (16). Hfq and associated sRNAs can act at the level of mRNA stability or at the level of translational efficiency (11). To identify changes in transcript abundance, the transcriptomes of the wild-type strain and the hfq mutant (grown to the mid-exponential growth phase) were compared by using high-throughput RNA-Seq. The RNA-Seq reads were mapped to the P. multocida VP161 genome, and 128 genes were identified as being differentially expressed. More than 85% of the differentially expressed genes (109 of 128 genes) showed increased expression levels compared to those in the wild type (see Table S1 in the supplemental material), suggesting that Hfq and its associated sRNAs generally act to decrease transcript levels, in keeping with the most common mode of sRNA/Hfq action at the transcriptional level, which is to increase RNase E turnover of transcripts.
Functional groups overrepresented in the upregulated gene set included those involved in cytochrome complex assembly (GO:0017004; 6 genes; P = 2 × 10−6), organic substance transport (GO:0071702; 13 genes; P = 6 × 10−4), nitrogen compound transport (GO:0071705; 7 genes; P = 8 × 10−4), and protein-binding transcription factors (GO:0000988; 3 genes; P = 0.001). Among the upregulated genes were pcgB, pcgD, PM1042 (PMVP_1156, PMVP_1157, and PMVP_1057, respectively), encoding proteins involved in LPS biosynthesis (48; M. Harper and J. D. Boyce, unpublished data), and hsf and PM0855 (PMVP_0693 and PMVP_0855, respectively), predicted to encode the Hsf adhesin and a Flp pilus-like adhesin, respectively (39). The periplasmic nitrate reductase (nap) genes (napFDAGHBC) were all strongly upregulated (7.5- to 11.6-fold) in the hfq mutant, as were a number of genes encoding other putative electron transport proteins, including ccmABCDEF (2.0- to 4.1-fold) and nrfBCD (3-fold). In addition, numerous genes encoding proteins involved in heat shock and stress responses were also expressed at higher levels in the hfq mutant, including htpX, uspG, htrA, dnaK, rpoH, clpB, rseC, rseB, mclA, and rpoE (PMVP_0444, PMVP_0450, PMVP_0713, PMVP_0715, PMVP_1638, PMVP_1759, PMVP_1836, PMVP_1837, PMVP_1838, and PMVP_1839, respectively). An association of Hfq with the regulation of stress response genes/proteins has been seen in other bacteria, including Y. pestis, Clostridium difficile, and S. enterica (20, 49, 50).
Nineteen genes displayed decreased transcript levels in the hfq mutant compared to the wild-type strain (see Table S2 in the supplemental material). Functional groups overrepresented in this set included polysaccharide transport (GO:0015774; two genes; P = 4 × 10−4) and pyridoxal phosphate metabolism (GO:0042822; two genes; P = 0.001). Importantly, 5 of the 10 capsule biosynthesis genes showed significantly reduced transcript expression in the hfq mutant (P < 0.0001) (see Table 5), correlating with the reduced HA capsule production observed for the hfq mutant.
Global changes in protein production in the P. multocida hfq mutant.Hfq often regulates protein production by altering mRNA translational efficiency via sRNA binding (7). Therefore, the proteomes of the VP161 hfq mutant and wild-type strain VP161 were compared during both early exponential and mid-exponential growth phases by using quantitative high-throughput liquid proteomics. Totals of 1,147 and 1,041 proteins were identified during early exponential and mid-exponential growth, respectively. During the early exponential growth phase of the hfq mutant, 85 proteins were detected at increased levels and 21 proteins were detected at reduced levels compared to those in the wild-type strain (see Tables S3 and S4, respectively, in the supplemental material). During the mid-exponential growth phase, 78 proteins were identified as being differentially expressed in the hfq mutant: 49 at increased levels (see Table S5 in the supplemental material) and 29 at decreased levels (see Table S6 in the supplemental material). Comparing the data from both growth phases, 37 proteins were expressed at increased levels at both the early exponential and mid-exponential growth phases (Table 3), and 8 proteins were expressed at decreased levels at both growth phases (Table 4). Five of the capsule biosynthesis proteins (HyaE, HyaD, HyaC, HexD, and HexC) showed between 10- and 20-fold-reduced production at the early exponential growth phase and between 1.8- and 2.7-fold-reduced production at the mid-exponential growth phase. These data correlated well with the decreased transcript levels observed for the capsule biosynthesis genes as measured by RNA-Seq (Table 5) and with the reduced HA capsule levels expressed by this strain. In addition, the known P. multocida virulence factor PfhB2 and its secretion partner LspB2 were also present at reduced levels in the hfq mutant at both growth phases (Table 4). In contrast, the levels of production of the hemoglobin-binding proteins PMVP_0265 (PM0300) and PMVP_0304 (PM0337) were increased in the hfq mutant at both the early exponential and mid-exponential growth phases (Table 3).
Proteins showing increased production in the P. multocida hfq mutant strain in comparison to wild-type strain VP161 during both early exponential and mid-exponential growth phases as measured by high-throughput liquid proteomics
Proteins showing decreased production in the P. multocida hfq mutant strain in comparison to wild-type strain VP161 during both early exponential and mid-exponential growth phases as measured by high-throughput liquid proteomics
Differential expression of capsule biosynthesis genes/proteins in the P. multocida hfq mutant strain in comparison to wild-type strain VP161
Comparison of the proteomic and transcriptomic data.The proteomic analysis of the hfq mutant at the mid-exponential growth phase (see above) identified 78 proteins as being differentially expressed, with 49 being identified at increased levels and 29 at decreased levels. Of the 49 proteins produced at higher levels in the hfq mutant, 13 also showed increased transcript levels as measured by RNA-Seq; all 13 of these proteins also showed increased levels (as measured by proteomics) at the early exponential growth phase (see Table S7 in the supplemental material). The 13 genes/proteins that showed increased expression levels by both techniques included 2 outer membrane lipoproteins (Plp4 and PlpB), 3 components of the periplasmic nitrate reductase system, the universal stress protein UspG, and the oligopeptide transporter DppA. Similarly, of the 29 proteins measured at reduced levels in the hfq mutant by high-throughput proteomics, 6 were also downregulated as determined by RNA-Seq (see Table S8 in the supplemental material), and 5 of these also showed decreased expression at the early exponential growth phase (see Table S8 in the supplemental material). The six genes/proteins determined to be downregulated in the hfq mutant by both proteomics and transcriptomics analyses included three capsule biosynthesis proteins, the filamentous hemagglutinin secretion partner LspB, the global regulator of carbon metabolic flux FruR (51), and the imidazole glycerol phosphate synthase, which is a key enzyme that links amino acid and nucleotide biosynthesis pathways (52). Interestingly, 59 of the 78 proteins identified as being differentially expressed in the hfq mutant by proteomics analysis were not determined to be differentially expressed by transcriptomics analysis. It is likely that many of these proteins are regulated by Hfq-sRNA interactions that act posttranscriptionally, a common mechanism of sRNA action (7).
DISCUSSION
Inactivation of hfq in different bacterial species has resulted in a wide range of phenotypic effects. For some species, such as Brucella melitensis, Legionella pneumophila, and Salmonella enterica, hfq mutants show a reduced growth rate with a longer lag phase in certain in vitro media (53–55). In other bacterial species, such as Listeria monocytogenes, Staphylococcus aureus, Haemophilus influenzae (24, 56, 57), and P. multocida (as shown in this study for strain VP161), the growth rate in vitro in rich medium is unaffected by hfq inactivation. Despite this normal in vitro growth rate, competition assays in mice revealed that the P. multocida hfq mutant showed significantly reduced fitness in vivo. Reduced in vivo fitness has also been demonstrated for hfq mutant strains of Y. pestis, Klebsiella pneumoniae, Francisella tularensis, and Actinobacillus pleuropneumoniae (20, 58–60).
One of the critical P. multocida virulence factors is the polysaccharide capsule (34). The P. multocida hfq mutant showed significantly reduced HA capsular polysaccharide (CPS) production (up to 14-fold) at all tested growth phases. Transcriptomic and proteomic analyses of the P. multocida hfq mutant confirmed the phenotypic data, showing highly reduced transcript and protein levels for many of the capsule biosynthesis genes/proteins (Table 5). Thus, Hfq plays a crucial role in positively regulating capsule production. The reduced level of capsule biosynthesis gene transcripts suggests that Hfq-mediated regulation acts primarily at the level of transcription; however, we cannot rule out an additional role for posttranscriptional mechanisms. In contrast to the effect of Hfq on P. multocida capsule production, a K. pneumoniae hfq mutant is hypermucoid, with increased levels of K2-specific CPSs (58). The reduction in capsule production in the P. multocida hfq mutant is likely to play a major role in the reduced in vivo fitness observed for this strain.
We have previously shown that the global transcriptional regulator Fis is essential for the expression of P. multocida capsule biosynthesis genes (36). Our proteomic and transcriptomic analyses indicated that Fis was produced at unchanged (proteomics data) or slightly increased (transcriptomics data) levels (1.8-fold; FDR = 0.03) in the hfq mutant compared to the wild-type strain, implying that the loss of capsule expression in the hfq mutant was not the result of a loss of Fis production. However, Fis may play a role in the regulation of hfq expression, as RNA-Seq analysis of a P. multocida VP161 fis mutant during the early exponential growth phase revealed a 2.2-fold decrease in hfq expression (FDR = 0.001) compared to that of the wild-type strain (Harper and Boyce, unpublished). These data suggest that Fis positively regulates Hfq, which in turn positively regulates capsule expression. However, mutation of fis alone has a much larger effect on capsule expression than mutation of hfq alone, suggesting that Fis primarily acts separately from Hfq.
P. multocida expresses two filamentous hemagglutinins, PfhB1 and PfhB2, that are important for in vivo fitness (31, 61). Our proteomic analyses showed that the loss of Hfq leads to a reduction in the amounts of PfhB2 and the secretion partner LspB2 during both early exponential and mid-exponential growth. It is possible that the reduced expression of PfhB2 may have contributed, along with reduced capsule expression, to the reduced in vivo fitness of the hfq mutant. Our transcriptomic analyses showed that genes encoding these proteins were also likely downregulated (PfhB2 was 1.4-fold downregulated, with an FDR of 0.056, and LspB2 was 2.3-fold downregulated, with an FDR of 0.01), indicating that the Hfq-associated regulation of these genes is also likely to be at the transcriptional level.
LPS is another crucial P. multocida virulence factor; strains expressing truncated LPS are highly attenuated for virulence (62, 63). Three genes involved in LPS biosynthesis (pcgB, pcgD, and PM1042) were expressed at increased levels in the hfq mutant. PcgB and PcgD are required for the biosynthesis and transfer of phosphocholine residues to the terminal galactose residues of the LPS produced by strain VP161 (48), and a functional PM1042 protein is required for the addition of phosphoethanolamine to lipid A (Harper and Boyce, unpublished). Both phosphoethanolamine and phosphocholine are important charged LPS substituents that can affect the interaction of host antimicrobial peptides with the bacterial surface (48). Numerous genes/proteins involved in electron transport were also upregulated in the hfq mutant, including napFDAGHBC, ccmABCDEF, and nrfBCD. Interestingly, the Nap system is predicted to reduce nitrate to nitrite during anaerobic growth (64), and a previous microarray study of P. multocida (strain X73) showed that the expression of the nap genes was significantly upregulated during in vivo growth in the chicken host, a relatively anaerobic niche (65). Thus, appropriate regulation of these genes appears critical for full in vivo fitness.
In conclusion, this study defined a clear role for Hfq in the regulation of P. multocida gene and protein expression and allowed us to identify genes controlled (either directly or indirectly) by Hfq-sRNA interactions at both the transcriptional and translational levels. Together, transcriptional and proteomic analyses of the hfq mutant supported our hypothesis that Hfq is involved in the regulation of virulence- and fitness-associated genes of P. multocida. Hfq is clearly involved in regulating the production of the known P. multocida virulence factors HA capsule, filamentous hemagglutinin, and LPS as well as accessory virulence genes such as those encoding iron uptake proteins, stress response proteins, and bacterial adhesins. We are currently focusing on identifying specific sRNA-mRNA interactions and in particular the specific mechanism(s) by which Hfq/sRNA action regulates the expression of HA capsule, LPS, and filamentous hemagglutinin.
ACKNOWLEDGMENTS
We thank Deanna Deveson Lucas for assistance with RNA-Seq sample preparation and Marietta John for construction of the VP161 hyaD mutant.
This work was supported in part by the Australian Research Council.
FOOTNOTES
- Received 10 February 2016.
- Accepted 11 February 2016.
- Accepted manuscript posted online 16 February 2016.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00122-16.
- Copyright © 2016, American Society for Microbiology. All Rights Reserved.
