Abstract
Haemophilus parasuis is an important opportunistic pathogen in swine of high health status, but to date no proven virulence factors have been described. As virulence factors are known to be regulated during disease, the objective of this study was to identify genes of a virulent serovar 5 strain with altered expression after iron restriction or in the presence of porcine cerebrospinal fluid (CSF), conditions that reflect in vivo growth conditions. Using differential-display reverse-transcriptase-mediated polymerase chain reaction, we found that homologues of genes encoding fructose bisphosphate aldolase (fba), adenylosuccinate synthetase (purA), 2′,3′-cyclic nucleotide phosphodiesterase (cpdB), lipoprotein signal peptidase (lspA), pyrophosphate reductase (lytB), superoxide dismutase (sodC), tyrosyl t-RNA synthetase (tyrS), cysteine synthetase (cysK), an unknown protein, and a homologue of a hydrolase of the haloacid dehydrogenase superfamily were upregulated in response to iron restriction. In addition, the purA, cpdB, lspA, lytB, and sodC homologues, cDNAs homologous with a Na+/alanine symporter, fatty acid ligase (fadD), diadenosine tetraphosphatase (apaH), and an unknown protein were upregulated in response to CSF. In screening for the presence of these differentially expressed genes to assess their usefulness as diagnostic markers of high virulence potential, we detected homologues of all of these genes in all of the reference strains of the 15 established serovars. The hydrolase homologue, however, was expressed only in representative H. parasuis strains associated with a high virulence potential, suggesting that this enzyme may play a role in pathogenesis.
Résumé
Haemophilus parasuis est un agent pathogène opportuniste important chez les porcs à statut sanitaire élevé mais à ce jour aucun facteur de virulence véritable n’a été décrit. Étant donné qu’il est connu que les facteurs de virulence ont tendance à être régulé durant une infection, l’objectif de la présente étude était d’identifier les gènes d’une souche virulente du sérovar 5 avec expression modifiée suite à une restriction en fer ou en présence de liquide céphalo-rachidien porcin (CSF), conditions qui reflètent les conditions de croissance in vivo. À l’aide d’une technique de représentation différentielle obtenue par transcriptase inverse suivie d’une réaction d’amplification en chaîne par la polymérase, nous avons trouvé que des homologues de l’aldolase du fructose bisphosphate (fba), de la synthétase de l’adénylsuccinate (purA), de la phosphodiestérase des nucléotides 2′,3′-cycliques (cpdB), de la peptidase signal des lipoprotéines (lspA), de la pyrophosphate réductase (lytB), de la superoxide dismutase (sodC), de la tyrosyl ARN-t synthétase (tyrS), de la cystéine synthétase (cysK), une protéine inconnue, et un homologue d’une hydrolase de la super-famille des haloacides déshydrogénases étaient tous régulés à la hausse en réponse à une carence en fer. De plus, les homologues de purA, cpdB, lspA, lytB, et sodC, des ADNc homologues à un cotransporteur Na+/alanine, une ligase des acides gras (fadD), la diadénosine tétraphosphatase (apaH), et une protéine inconnue étaient régulés à la hausse en réponse au CSF. Lorsque les souches de référence des 15 sérovars reconnus ont été éprouvées pour la présence de ces gènes afin d’évaluer leur utilité comme marqueurs diagnostiques des souches avec un potentiel de virulence élevé, nous avons détecté les homologues de tous les gènes exprimés de manière différentielle chez toutes les souches de référence. Toutefois, l’homologue de l’hydrolase n’était exprimé que chez les souches représentatives de H. parasuis associées à un potentiel de virulence élevé, ce qui suggère que cet enzyme pourrait jouer un rôle dans la pathogénie.
(Traduit par Docteur Serge Messier)
Introduction
Haemophilus parasuis is a common upper respiratory tract commensal that can be isolated from conventionally raised pigs of all ages (1). In these animals, it causes a sporadic stress-associated disease. In high health status herds, H. parasuis can also be associated with high morbidity and mortality rates in pigs of all ages (1,2). To date, 15 serovars have been described (3), but, depending on the geographic region and the typing method used, up to 25% of isolates may be nontypable (4,5).
From a limited number of challenge studies in pigs and guinea pigs, serovars 1, 5, 10, and 12 to 14 appear to be associated with a high virulence potential and serovars 2, 4, and 15 with intermediate virulence; serovars 3, 6 to 9, and 11 are reported to be avirulent (3,6). Consistent with these findings, Oliveira et al (7) reported that serovars 1, 2, 5, 12 to 14, and nontypable isolates are the most frequently recovered from systemic sites, whereas serovar 3 and nontypable strains are present in the upper respiratory tract of healthy swine in the United States. Serovars 2, 4, 5, 12, 14, and nontypable H. parasuis disease isolates have also been reported in Australia, China, Japan, Germany, and Spain (8,9). That said, no clear correlation between virulence and serovar has been demonstrated, and by a variety of methods considerable genetic heterogeneity can be detected, not only between but also within serovars (8).
To date, numerous attempts have been made to identify virulence factors or even virulence markers for virulent strains. Several putative virulence factors have been suggested, including fimbriae (10), neuraminidase (11), and transferrin-binding proteins (12), but, in the absence of genetic tools, their precise roles have yet to be confirmed. Furthermore, neuraminidase and likely fimbriae and the transferrinbinding proteins appear to be present in strains of all serovars of H. parasuis. Polyacrylamide gel electrophoresis profiles, particularly proteins of 36 to 38 kDa, have been suggested to be a marker for virulent strains, but, again, their role in virulence, if any, is unknown (8). The identification of genes that can be used to differentiate between harmless commensal H. parasuis strains and those with higher virulence potential is crucial for the effective management of Glasser’s disease.
Materials and methods
Strains and growth conditions
A virulent serovar 5 strain, H. parasuis 1185 (HP1185) (13), was used for the studies involving differential-display reverse-transcription polymerase chain reaction (DD-RT-PCR). The 15 KRG serovar reference strains (3), obtained from Øystein Angen, Danish Veterinary Laboratory, Copenhagen, were screened for all differentially expressed genes. As well, 20 clinical isolates submitted for autogenous bacterin production (Gallant Custom Laboratories, Cambridge, Ontario), 48 clinical isolates from the Université de Montréal, Saint-Hyacinthe, Québec, and 46 H. parasuis-positive nasal swab cultures obtained during a survey of Ontario swine herds (14) were analyzed for the presence of the cpdB homologue.
The H. parasuis strains were routinely cultured in brain–heart infusion (BHI) broth (Fisher Scientific, Ottawa, Ontario) supplemented with 0.02% nicotinamide adenine dinucleotide (NAD). To examine the effects of iron restriction, HP1185 was passaged 3 times on iron-restricted medium to deplete internal iron reserves. In preliminary iron-restriction experiments, HP1185 was cultured in BHI-NAD medium containing 160 μM of the ferrous iron chelator 2,2′-dipyridyl and in BHI-NAD medium containing 200 μM 2,2′-dipyridyl and 210 μM FeSO4. In later experiments, HP1185 was cultured in HEPES-buffered tryptone–yeast extract (TYE) supplemented with NAD (TYE-NAD) (15), in TYE-NAD that was rendered iron-deficient by the addition of 40 μM of the ferric iron chelator ethylene diamine diacetic acid (EDDA), and in TYE-NAD with 55 μM EDDA and 20 μM FeCl3. To approximate conditions encountered during meningitis, HP1185 was cultured in 100% porcine cerebrospinal fluid obtained postmortem by cisternal or lumbar puncture with an 18-gauge syringe needle from pigs approximately 5 wk old and immediately centrifuged at 500 × g for 10 min to pellet any contaminating erythrocytes. The supernatant was removed to a fresh tube. Three batches of CSF, each from 3 different pigs, were pooled and stored at −70°C. Consistent with the findings of O’Reilly and Niven (16), NAD was found to be required for the growth of H. parasuis in CSF, suggesting that either the levels of NAD in CSF are growth-limiting or the NAD was degraded during the freezing and thawing. Accordingly, HP1185 was inoculated into 3.0 mL of CSF supplemented with 0.02% NAD and incubated overnight at 37°C with shaking at 200 rpm.
Isolation and RT of RNA
Following the manufacturer’s instructions, we isolated RNA from HP1185 using TRIzol reagent (Invitrogen Life Technologies, Burlington, Ontario) once an optical density (OD) at 600 nm of 0.2 to 0.3 was reached, as recommended for RNA isolation. For RT reactions, a mixture of 1 μg of RNA, 1.0 μL of random hexamers (3.0 μg/μL; Invitrogen), 2.0 μL of 10 mM deoxynucleotide triphosphates (dNTPs; Roche Applied Science, Laval, Québec), 4.0 μL of 5× first-strand buffer (250 mM Tris–HCl at pH 8.3, 375 mM KCl, and 15 mM MgCl2), 2.0 μL of 0.1 M dithiothreitol, 1.0 μL of RNase Inhibitor (Invitrogen), and 8 units of Moloney murine leukemia virus RT (Invitrogen) was brought up to a final volume of 20 μL with deionized water (dH2O) treated with diethylpyrocarbonate. The synthesis reaction was carried out for 60 min at 40°C in a T-gradient thermocycler (Biometra, Goettingen, Germany) and stopped by incubation at 75°C for 10 min.
Differential-display PCR
Arbitrary primers were obtained from GenHunter Corporation, Nashville, Tennessee, USA, and the PCRs were performed as described previously (17). Of more than 20 primer pairs tested by Hill et al (17), 10 were found to reliably give rise to amplicons. The DD-PCR products were resolved on a 7% polyacrylamide gel for 77 min at 100 V in 1× TAE buffer (0.04 M Tris-acetate and 0.001 M ethylene diamine tetraacetic acid, pH 8.0). For silver staining, the gels were first incubated with gentle agitation in a fixative solution consisting of 12% acetic acid (vol/vol) and 50% methanol (vol/vol) for 30 min. After being rinsed 3 times with dH2O (for 30 s each time), the gels were incubated for 30 min with gentle agitation in 0.1% AgNO3 (wt/vol) and 0.03% formaldehyde (vol/vol). The gels were then rinsed twice with dH2O and visualized for 2 to 3 min with a developer solution (3% Na2CO3 [wt/vol], 0.03% formaldehyde [vol/vol], and 0.0002% sodium thiosulfate [wt/vol]). The visualization process was stopped by incubation in the fixative solution for 20 min.
Once a differentially expressed band was identified, the silver-stained gel was rinsed twice with dH2O and the band “scratched” with a 25-gauge needle. The needle was placed in 10 μL of dH2O and incubated at 56°C for 10 min, and the DNA solution was then immediately used in a 30-μL reaction according to the protocol initially used for DD-PCR. The reamplified products were gel-purified from 1.5% agarose gels.
Cloning and sequencing
Before cloning by means of a TOPO cloning kit (Invitrogen), 3′ A overhangs were added to the gel-purified DD-PCR products and the cpdB gene fragment of 1463 base pairs (bp). In a 10-μL reaction, 5.0 μL of PCR product was incubated with 0.6 μL of 10× PCR buffer (Invitrogen), 0.4 μL of 1 mM dNTPs, and 3.4 units of Taq DNA polymerase (Invitrogen) at 72°C for 10 min. The product was subsequently ligated into pCR4–TOPO and transformed into chemically competent Escherichia coli One Shot TOP10 (Invitrogen) according to the manufacturer’s instructions. Putative recombinants were evaluated by the modified rapid screening procedure of Sambrook et al (18). Plasmids with inserts of the expected size were selected for further analysis.
Plasmid DNA was purified by mean of a QIAprep Spin Miniprep Kit (Qiagen, Mississauga, Ontario), and insert DNA was sequenced by dye terminator cycle sequencing with the use of a T7 primer (College of Biological Sciences DNA Facility, University of Guelph, Guelph, Ontario). Sequences were analyzed with the BLAST algorithm to identify homologues of the genes at GenBank, the US National Center for Biotechnology Information server (www.ncbi.nlm.nih.gov).
Confirmation of differential gene expression
Specific forward and reverse primers (Sigma-Genosys, Oakville, Ontario) were designed to each gene fragment identified by DD-RT-PCR (Table I) on the basis of the sequence of the cloned cDNA. The primer sequences for the apaH, fadD, and cysK homologues had been reported previously (17). The cDNA from each condition used in the initial PCR was then subjected to a 2nd round of more stringent PCR with these specific primers in a 20-μL reaction that contained 2.0 μL of 10× PCR buffer (200 mM Tris–HCl, pH 8.4, and 500 mM KCl), 0.8 μL of 50 mM MgCl2, 4.0 μL of 1.0 mM dNTPs, 1.0 μL each of the forward and reverse primers, and 0.1 μL of Taq DNA polymerase. The cycling parameters were as follows: 5 min at 95°C, 30 s at 94°C, 30 s at 50°C to 58°C, and 1 min at 72°C for 25 cycles to maintain the initial differences in starting template (19). A final elongation step for 5 min was performed at 72°C. The PCR products were resolved on a 7% TAE polyacrylamide gel and visualized by silver staining. To demonstrate a lack of DNA contamination, we used RNA as a negative control. The reactions were performed in triplicate with cDNA reverse-transcribed from at least 2 batches of RNA. The differences in expression of the genes were measured by densitometry with the use of GeneTools software (Syngene, Fisher Scientific) on images captured with a Gene Genius Bioimaging System (Syngene, Synoptics, Fisher Scientific) and GeneSnap software (Fisher Scientific).
Table I.
Specific primers used to confirm Haemophilus parasuis gene expression in experiments involving differential-display reverse-transcription polymerase chain reaction (DD-RT-PCR)
| Primer | Sequence (5′ – 3′) | % G + C | Annealing temperature (°C) |
|---|---|---|---|
| Dmg1-F | CGATTGGCTCTTCTGAAAGG | 50 | 56 |
| Dmg1-R | GCAAAAGAACACAACTTCGC | 45 | |
| Dmg2-F | TGAAATATCAGCCGTTATCG | 40 | 54 |
| Dmg2-R | GGCAGAGAATGGTATCATGA | 45 | |
| Dmg5-F | TTACCAGGTGCAGTGTTCTG | 50 | 58 |
| Dmg5-R | TACGTTTAATCCCACCGAAG | 45 | |
| Dmg6-F | ATAATCCGCAGATTTCAATG | 35 | 56 |
| Dmg6-R | ACGACCACAAAAACTGAAGG | 45 | |
| Dmg11-F | GAAGATTTACCGCAAAATGC | 40 | 58 |
| Dmg11-R | GGCTATTTTACACTACTTTGCC | 41 | |
| Dmg12-F | GCCTGTTTCATCAGTGCCAACG | 55 | 58 |
| Dmg12-R | GTGGTTGATGAAGATGGCGTG | 52 | |
| Dmg15-F | GATGCGGCGGAACAAGAG | 61 | 56 |
| Dmg15-R | CTTGTTGTGCGGAACGGATT | 50 | |
| Dmg16-F | GCAGGAAATAGTAAAGGGGC | 50 | 54 |
| Dmg16-R | GCTTTCTCTGGATCAATTTG | 40 | |
| Dmg17-F | CAGTTTTAACCCTTGCTTTA | 35 | 50 |
| Dmg17-R | TGATCAACCGCTAATGCAGG | 50 | |
| Dmg19a-F | TTGGCTATGTGGTTGATTTT | 35 | 50 |
| Dmg19a-R | CGATGTTCTAAAATGGATTC | 35 | |
| Dmg19b-F | ATGAACATTATTTTAGCGAATC | 27 | 50 |
| Dmg19b-R | AAGCTTTTGATCCGTCCACC | 50 | |
| Chg10-F | GCCCCAAAGCCTTCACATA | 53 | 50 |
| Chg10-R | GATGAACTTTGGCTAACGGG | 50 | |
| Chg15-F | GCGGTGGAGCAAAGTTGGAA | 55 | 50 |
| Chg15-R | GTCCCGTGCGTAAGAAACCA | 55 | |
| Chg27-F | GCTTACCAGGTCCACACAAT | 50 | 55 |
| Chg27-R | ATAACGTTCTGAGGCGGAAG | 50 |
F — forward; R — reverse.
Screening by PCR
We prepared DNA from the reference strains of the 15 KRG H. parasuis serovars using GenomicPrep Cells and tissue DNA isolation kits (Amersham Pharmacia Biotech, Piscataway, New Jersey, USA). The same PCR protocol used to confirm differential gene expression was used to determine the distribution of the differentially expressed genes in the reference strains. We also screened 68 clinical isolates and 46 mixed-culture nasal swabs from Ontario swine herds for the presence of the cpdB homologue. For these experiments, the PCR protocol used to confirm differential gene expression was again used, but the templates were boiled mixed-culture lysates.
Nucleotide sequence accession numbers
The GenBank accession numbers for the sequences reported in this paper are DQ173931 (fba homologue), DQ173933 (Na+/alanine symporter), DQ173944 (unknown protein), DQ173934 (purA homologue), DQ173935 (cpdB homologue), DQ173937 (hypothetical protein homologue), DQ173938 (tyrS homologue), DQ173939 (ldhA-like homologue), DQ173941 (hydrolase homologue), DQ173940 (sodC), DQ173942 (lspA), DQ173943 (lytB), AY204905 (apaH homologue), AY204906 (fadD homologue), and AY204909 (cysK homologue).
Results
Analysis of iron chelators
The iron-restricted media were initially prepared with 2,2′-dipyridyl. To confirm that the effects of this ferrous iron chelator on gene expression were due to a reduction in iron concentration, H. parasuis was cultured in a medium containing a concentration of chelator that inhibited growth, and FeSO4 was added to reverse the inhibition. We isolated RNA from cells grown under these conditions and performed DD-RT-PCR. No differentially expressed bands unique to the iron-restricted condition were observed, and all bands observed in this condition were also present in the condition that included the addition of FeSO4 (Figure 1). The iron-replete phenotype could not be restored by the addition of FeSO4 or various concentrations of other divalent cations, such as Zn2+, Mg2+, and Ca2+. In contrast, the iron-replete phenotype could be restored when ferric iron (FeCl3) was chelated with EDDA (Figure 2).
Figure 1.
Silver-stained, TAE-buffered 7% polyacrylamide gel of arbitrarily primed differential-display reverse-transcription polymerase chain reaction (DD-RT-PCR) products of Haemophilus parasuis 1185 (HP1185) after growth in (1) brain–heart infusion (BHI) broth supplemented with 0.02% nicotin-amide adenine dinucleotide (NAD), (2) BHI, NAD, and 160 μM 2,2′-dipyridyl, or (3) BHI, 200 μM 2,2′-dipyridyl, 210 μM FeSO4, and NAD. Markers of molecular sizes (in base pairs) indicated on the left side.
Figure 2.
Silver-stained, buffered 7% polyacrylamide gel of arbitrarily primed DD-RT-PCR products of HP1185 after growth in (1) HEPES-buffered tryptone–yeast extract (TYE) and NAD (TYE-NAD), (2) TYE-NAD and 40 μM of ethylene diamine diacetic acid (EDDA), or (3) TYE-NAD, 55 μM EDDA, and 20 μM FeCl3.
Identification of differentially expressed genes
Ten genes identified either by this work or in a previous DD-RT-PCR study (16) were upregulated in response to iron restriction, and 9 genes were upregulated in response to CSF (Table II). A broad range of levels of upregulation of expression of homologues was observed in HP1185 in response to iron restriction, from 1.65 for the tyrS homologue to 16.52 for the fba homologue. The fba homologue was upregulated to a high level of expression in all the strains tested.
Table II.
Results of BLASTX homology searches for the gene fragments identified by DD-RT-PCR and fold upregulation in the expression of homologues
| % of amino acids
|
||||||
|---|---|---|---|---|---|---|
| Fragment (no. of base pairs) | Homologue source | Description | GenBank accession no. | Identical | Conserved | Fold upregulation, mean (and standard deviation)a |
| DMG1 (474) |
P. multocida A. pleuropneumoniae H. influenzae |
Fructose bisphosphate aldolase (fba) | NP_246800.1
ZP_00348247.1 ZP_00156350.2 |
88
88 88 |
96
96 94 |
IR, 16.52 (3.50) |
| DMG5 (430) |
H. somnus A. pleuropneumoniae H. ducreyi |
Na+/ala symporter | ZP_00122619
ZP_00134784 AAP96397 |
79
79 71 |
93
88 86 |
CSF, 2.23 (0.37) |
| DMG6 (165) | No known homology | CSF, 2.29 (0.69) | ||||
| DMG11 (260) |
H. influenzae A. pleuropneumoniae H. ducreyi |
Adenylosuccinate synthetase (purA) | ZP_00157071
ZP_00135197 NP_874168.1 |
86
84 79 |
90
94 88 |
IR, 2.39 (0.33);
CSF, 2.45 (0.53) |
| DMG12 (394) |
A. pleuropneumoniae H. influenzae P. multocida |
2′,3′-cyclic nucleotide phosphodiesterase (cpdB) | ZP_00133994
ZP_00154334 AAK04098 |
76
72 70 |
90
83 79 |
IR, 4.11 (0.68);
CSF, 2.11 (0.33) |
| DMG14 (450) |
A. succinogenes V. cholerae |
Hypothetical protein | ZP_00733436.1
ZP_00746907.1 |
96
94 |
96
94 |
IR, 3.55 (0.83) |
| DMG15 (200) |
H. influenzae A. pleuropneumoniae H. ducreyi |
Tyrosyl t-RNA synthetase (tyrS) | AAX88252.1
ZP_00134668 AAP959611 |
96
94 88 |
98
98 90 |
IR, 1.65 (0.08) |
| DMG16 (98) |
A. pleuropneumoniae H. ducreyi H. somnus |
HAD-family hydrolase | ZP_00348255
AAP96161 ZP_00133462 |
73
68 60 |
83
72 73 |
IR, 4.65 (1.39) |
| DMG17 (383) |
H. influenzae A. pleuropneumoniae H. ducreyi |
Superoxide dismutase (sodC) | AAQ12654
CAA67771.1 AAP95739 |
80
79 64 |
86
87 70 |
IR, 6.23 (0.49);
CSF, 3.98 (1.12) |
| DMG19a (151) |
A. pleuropneumoniae H. ducreyi H. influenzae |
Lipoprotein signal peptidase (lspA) | ZP_00135015
AAP95075 NP_439167 |
79
75 69 |
87
85 83 |
IR, 3.76 (0.80); |
| DMG19b (150) |
A. pleuropneumoniae H. influenzae H. ducreyi |
Pyrophosphate reductase (lytB) | ZP_00135016
AAC22668.1 AAP95076 |
96
95 92 |
98
97 96 |
CSF 3.75 (0.38) |
| CHG10 (142) |
H. ducreyi A. actinomycetemcomitans H. somnus |
Diadenosine tetraphosphatase (apaH) | NP_873702.1
AAC00202.1 |
73
75 66 |
96
87 84 |
CSF, 4.27 (2.23) |
| CHG15 (304) |
H. influenzae P. multocida A. pleuropneumoniae |
Long-chain fatty acid coenzyme A ligase (fadD) | ZP_0131943.1
NP_438175.1 AAK03009.1 ZP_00135531.2 |
87
84 84 |
92
91 89 |
CSF, 3.24 (0.99) |
| CHG27 (398) |
A. pleuropneumoniae H. influenzae H. ducreyi |
Cysteine synthetase (cysK) |
ZP_00135187.2 NP_439260.1 NP_873391.1 |
91
88 84 |
94
92 91 |
IR, 4.49 (1.01) |
P. — Pasteurella; A. — Actinobacillus; H. — Haemophilus; V. — Vibrio; HAD — haloacid dehydrogenase.
Based on mean levels of expression, as measured by densitometry, in reactions performed in triplicate in media with iron restriction (IR) or supplementary cerebrospinal fluid (CSF).
Distribution of the cpdB homologue
We performed PCR analysis to screen the reference strains of the 15 known serovars of H. parasuis for the presence of homologues of the differentially expressed genes to determine their usefulness as diagnostic markers of virulence. With only 1 exception, an amplicon of the predicted size was detected in all of the reference strains. The expected 394-bp cpdB fragment, however, was amplified from only 5 of the 15 reference strains; a 1463-bp fragment was amplified from the remainder. Except for the serovar 11 reference strain (H465), a 394-bp fragment was amplified from all the reference strains isolated from diseased pigs (serovars 5 [Nagasaki], 12 [H425], 13 [84–17975], 14 [84–22113], and 15 [84–15995]), whereas the 1463-bp fragment was amplified from the strains recovered from healthy pigs (serovars 1 [No. 4], 2 [SW140], 3 [SW114], 4 [SW124], 6 [131], 7 [174], 8 [C5], 9 [D74], and 10 [H555]). Cloning and sequencing of the 1463-bp fragment showed that its additional DNA corresponded to the 5′-nucleotidase domain of the functional 2′,3′-cyclic nucleotide phosphodiesterase.
To further evaluate the significance of the cpdB homologue as a marker of virulence, we screened 2 additional collections of H. parasuis isolates by PCR amplification. With H. parasuis-positive lysates of mixed nasal-swab cultures from healthy pigs, the 394-bp fragment was amplified from the samples from 24 farms, and the 1463-bp fragment from the samples from 5 farms; amplicons of both sizes were detected in samples from the remaining 16 farms. No amplicons were obtained in samples from pigs found to be H. parasuis-negative by a 16S rDNA PCR assay. The 394-bp fragment was also amplified from 32 (47%) of the 68 clinical isolates evaluated; the 1463-bp fragment was amplified from the remaining 36 isolates (53%). Among the 46 clinical isolates whose site of isolation was known, there initially appeared to be a correlation between the 1463-bp amplicon and internal sites such as joints, heart, liver, spleen, peritoneal cavity, pericardium, and meninges; however, this association was not statistically significant. Of the 29 lung isolates, the 1463-bp fragment was amplified from 16 (55%) and the 394-bp fragment from the remaining 13 (45%). The 1463-bp fragment was amplified from 15 (88%) of the other 17 strains isolated from internal sites and 5 (23%) of the 22 strains isolated from unknown sites.
Gene expression in strains with high or low virulence potential
The expression levels of the iron-responsive genes were compared in H. parasuis strains associated with high or low virulence potential (Table III). Strains Nagasaki (serovar 5), D2K-525 (serovar 4), D2K-472 (serovar 4), and D02-507 (serovar 12) were chosen as representing strains of high virulence potential because these serovars have been implicated in severe Glasser’s disease. Strains SW114 (serovar 3), 131 (serovar 6), H465 (serovar 11), and D74 (serovar 9) were chosen as representing strains with a low virulence potential because these serovars have historically been associated with a low virulence potential (3). The pattern of expression observed in strain SW114 was similar to that of strain HP1185 except for the hypothetical protein and hydrolase homologues. Expression of these gene fragments was evaluated in further strains. Although the predicted pattern of expression of the hypothetical protein was observed in strains D2K-525 and 131, it was not observed in the Nagasaki strain. As with strain HP1185, expression of the hydrolase gene was upregulated in all of the virulent strains tested, and, as with strain SW114, hydrolase expression was not detected in the other 3 avirulent strains tested.
Table III.
Comparison of the fold upregulation of the gene fragments identified by DD-RT-PCR under iron-restricted conditions in strains associated with high and low virulence potential
| Virulence potential; strain (and serovar); mean fold upregulation
|
|||||||||
|---|---|---|---|---|---|---|---|---|---|
| High
|
Low
|
||||||||
| Homologue | HP1185 (5) | Nagasaki (5) | D2K-525 (4) | D2K-472 (4) | D02-507 (12) | SW114 (3) | 131 (6) | H465 (11) | D74 (9) |
| fba | 16.52 | 14.98 | 10.42 | NT | NT | 20.17 | 8.34 | NT | NT |
| Symporter | — | NT | NT | NT | NT | — | NT | NT | NT |
| Unknown protein | — | NT | NT | NT | NT | — | NT | NT | NT |
| purA | 2.39 | NT | NT | NT | NT | 6.58 | NT | NT | NT |
| cpdB | 4.11 | 6.36 | 4.09 | NT | NT | 7.58 | 3.20 | NT | NT |
| Hypothetical protein | 3.55 | — | 2.02 | NT | NT | — | — | NT | NT |
| tyrS | 1.65 | NT | NT | NT | NT | 1.72 | NT | NT | NT |
| Hydrolase | 4.65 | 3.03 | 3.65 | 2.02 | 2.69 | —a | —a | —a | —a |
| sodC | 6.23 | NT | NT | NT | NT | 3.10 | NT | NT | NT |
| lspA/lytB | 3.76 | NT | NT | NT | NT | 2.61 | NT | NT | NT |
| apaH | — | NT | NT | NT | NT | — | NT | NT | NT |
| fadD | — | NT | NT | NT | NT | — | NT | NT | NT |
| cysK | 4.49 | NT | NT | NT | NT | 5.83 | NT | NT | NT |
NT = not tested; — = same level of expression in all conditions.
No expression of the gene fragment detected in any condition.
Discussion
Iron is an essential nutrient for almost all bacteria; therefore, survival in vivo, when iron is largely sequestered, depends on the development of iron-obtaining mechanisms. As well, H. parasuis encounters a variety of other nutritional limitations and stresses in host fluids, such as CSF, during systemic infection. Genes expressed in response to these conditions may encode virulence factors that can be exploited to better understand pathogenesis and to develop vaccines or therapeutic agents.
In response to iron restriction and CSF supplementation, 10 and 9 gene fragments, respectively, were upregulated in HP1185, the virulent serotype 5 strain of H. parasuis (Figure 3). As well, 3 genes previously shown to be upregulated by an elevated temperature were induced by iron restriction (1 gene fragment) or CSF supplementation (2 gene fragments). Although induction of iron-uptake systems was not detected under iron-restricted conditions, all of the induced genes are likely important in host nutrient acquisition or, in the case of sodC, in protecting the organism from host defenses. Several gene fragments were upregulated in response to both iron restriction and CSF supplementation, reflecting the fact that CSF is iron-restricted. It may be postulated that the Na+/ala symporter and the unknown protein homologue, which were upregulated only in response to CSF, may be important for regulating intracellular ion levels or nutrient uptake in the CSF, which has low levels of amino acids, glucose, and calcium.
Figure 3.
Distribution of homologues upregulated in response to iron restriction, elevated temperature (40°C) during growth in 50% swine serum (17), or supplementation with cerebrospinal fluid (CSF).
Except for the cpdB homologue, all the upregulated gene fragments were amplified in almost all of the H. parasuis reference strains, regardless of the predicted virulence potential, which suggests that these genes were not necessarily associated with high virulence. Depending on the serovar, 2 variants of cpdB were amplified. A dual-function periplasmic enzyme, CpdB catalyzes the hydrolysis of 2′,3′-cyclic phosphates of adenosine, cytosine, uridine, and guanosine and has a nucleotidase function that is responsible for the conversion of ribonucleotides to ribonucleosides (20). It has been hypothesized that CpdB functions to hydrolyze nontransportable nucleotides so they can cross the cytoplasmic membrane and be used as a carbon and energy source (21). Therefore, CpdB would be involved in scavenging and facilitating the use of exogenously supplied 2′,3′-cyclic nucleotides (22). Despite the apparent deletion of 1100 bp from the middle of the cpdB gene, the sequence remains in the same reading frame, and gene expression is upregulated in response to iron restriction, which suggests the production of a functional diesterase protein. Currently, there are no other reports of similar deletions occurring in 2′,3′-cyclic nucleotide phosphodiesterases or related enzymes.
To further evaluate the significance of the cpdB homologue as a marker of virulence, we screened collections of H. parasuis clinical isolates by PCR analysis. Despite the correlation between the 394-bp fragment and the reference strains of the most virulent serovars, no statistically significant correlation of the 394-bp cpdB amplicon and isolates associated with severe invasive disease was found. Moreover, when PCR analysis was performed on H. parasuis-positive mixed-culture cell lysates from healthy pigs, the 394-bp fragment alone was amplified from samples from 24 of 46 farms, whereas the 1463-bp fragment was found only in samples from 5 farms; both amplicons were detected in samples from a further 16 farms. These results were somewhat surprising, since multiple serovars are usually found on a single farm or even in the same animal and could reflect a predominance of 1 serovar at a particular time. Accordingly, further investigation is needed to determine the role of the different cpdB variants, if any, in the pathogenesis of diseases caused by H. parasuis.
Although homologues of all of the upregulated genes were present in all the serovars, regardless of pathogenic potential, differences in their expression could play a role in virulence. Accordingly, expression of all of the iron-regulated gene fragments was examined in H. parasuis SW114, a serovar 3 strain predicted to have low virulence potential. After 2 rounds of analysis, only 1 of the iron-regulated gene fragments appeared to be associated with high virulence potential. In addition to H. parasuis 1185, used for the DD-RT-PCR experiments, expression of the hydrolase homologue was detected in 4 strains predicted to have high virulence potential, but no expression of this gene fragment was detected in 4 strains of serovars generally considered to be associated with a low virulence potential (3). The DMG16 gene fragment is predicted to encode a hydrolase of the haloacid dehydrogenase (HAD) superfamily. Few of these enzymes have been functionally characterized, and those that have fall into 1 of 5 groups: haloacid dehydrogenases, phosphohydrolases, phosphatases, phosphoglucomutases, and adenosine triphosphatases (23). Although inactivation of the gene encoding the phosphoglucomutase of Vibrio furnissii led to a decrease in serum resistance and attenuation in vivo (24), there have been no reports of the association of HAD-superfamily phosphohydrolases with either adaptation to iron restriction or pathogenesis. Expression of a gene fragment only in strains known to be involved in clinical disease and absence of expression in strains not associated with disease suggest that the gene may be involved in pathogenesis. These hydrolases have not previously been reported to have any clear involvement in virulence, so further investigation into the role this gene may play in the pathogenesis of disease caused by H. parasuis is warranted.
In most studies attempting to identify bacterial virulence factors, a large percentage of the genes preferentially expressed either in vivo or in conditions that simulate those encountered in a host during disease encode metabolic and biosynthetic proteins, which suggests that these genes are likely involved in most adaptive responses to different environments. Genes encoding proteins involved in such pathways may play a more significant role in pathogenesis than previously thought. For example, Silo-Suh et al (25) reported the critical involvement of zwf, which encodes glucose-6-phosphate dehydrogenase, an enzyme involved in carbon catabolism, in the pathogenesis of lung infections due to Pseudomonas aeruginosa in cystic fibrosis patients. In agreement with the work of Silo-Suh et al and similar studies, most of the genes upregulated in response to the in vivo conditions used in the current work encode proteins traditionally associated with “housekeeping” functions rather than known virulence factors. Melnikow et al (26) recently developed a microarray for H. parasuis and used it to study gene expression in response to a variety of in vitro growth conditions. In addition to acidic pH, increased temperature, and low oxygen, iron restriction was investigated. Most of the genes identified were also functionally classified as encoding biosynthetic and metabolic proteins; however, no specific genes were found in common with those found in this study. The lack of shared genes upregulated in response to iron restriction may be explained in part by differences between these studies. Melnikow et al identified genes upregulated in response to iron depletion after 10, 30, and 60 min versus overnight growth in our study, and they did not deplete internal iron reserves. Although serovar 5 isolates were used in both studies, there may have been genetic differences between the strains. In addition, Melnikow et al used the ferrous iron chelator 2,2′-dipyridyl to establish iron restriction. We were unable to reverse the effects on gene expression of the iron chelator by restoring iron to the iron-depleted medium. Despite its frequent use, at least with the conditions in this study, and consistent with an observation by Boyer et al (27), 2,2′-dipyridyl appears to have the potential to nonspecifically chelate other divalent cations, and caution should be used in interpreting the results when using this chelator to establish iron restriction.
In summary, 11 H. parasuis gene fragments were identified that were upregulated in response to iron restriction (4 fragments), the presence of CSF (2 fragments), or both (5 fragments). In addition, 3 gene fragments previously shown to be upregulated by elevated temperature and serum were also shown to be regulated by iron restriction or the presence of CSF (1 and 2 gene fragments, respectively). Homologues of all of these genes were detected in the reference strains of the 15 known serovars of H. parasuis, but in the case of the cpdB homologue, an amplicon of 394 bp was associated with serovars believed to have high virulence potential, whereas a 1463-bp amplicon was detected in serovars with a low virulence potential. After the analysis of 68 clinical samples and mixed cultures from nasal swabs of healthy animals from 46 farms, no statistically significant correlation could be made. It was possible, however, to demonstrate that a hydrolase homologue was expressed only in strains associated with a high virulence potential, which suggests a role for this enzyme in pathogenesis. Knowledge of the genes involved in adaptation to environments encountered during disease will help elucidate the mechanisms of pathogenesis for this economically significant bacterium. Experiments are under way using a recently developed H. parasuis transformation method to see if it is possible to create isogenic mutants to study the role of these putative virulence factors (28).
Acknowledgments
This work was funded by the Natural Sciences and Engineering Research Council of Canada as part of the Research Network on Bacterial Diseases of Swine. We thank Jackie Gallant for the H. parasuis bacterin strains. We also thank Drs. Gaylan Josephsen, Robert Friendship, and Alexandre Loretti for CSF collection, Donald Tremblay for PCR analysis of the H. parasuis isolates from the Université de Montréal, and Dr. Abdul Lone for helpful technical suggestions.
References
- 1.Nicolet J. Haemophilus parasuis. In: Diseases of Swine. Ames, Iowa: Iowa State University Press, 1992:526–527.
- 2.Takahashi K, Nagai S, Yagihashi T, et al. A cross-protection experiment in pigs vaccinated with Haemophilus parasuis serovars 2 and 5 bacterins, and evaluation of a bivalent vaccine under laboratory and field conditions. J Vet Med Sci. 2001;63:487–491. doi: 10.1292/jvms.63.487. [DOI] [PubMed] [Google Scholar]
- 3.Kielstein P, Rapp-Gabrielson V. Designation of 15 serovars of Haemophilus parasuis on the basis of immunodiffusion using heat-stable antigen extracts. J Clin Microbiol. 1992;30:862–865. doi: 10.1128/jcm.30.4.862-865.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Turni C, Blackall PJ. Comparison of the indirect haemagglutination and gel diffusion test for serotyping Haemophilus parasuis. Vet Microbiol. 2005;106:145–151. doi: 10.1016/j.vetmic.2004.12.019. [DOI] [PubMed] [Google Scholar]
- 5.Tadjine M, Mittal KR, Bourdon S, Gottschalk M. Development of a new serological test for serotyping Haemophilus parasuis isolates and determination of their prevalence in North America. J Clin Microbiol. 2004;42:839–840. doi: 10.1128/JCM.42.2.839-840.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Amano H, Shibata M, Kajio N, Morozumi T. Pathologic observations of pigs intranasally inoculated with serovar 1, 4 and 5 of Haemophilus parasuis using immunoperoxidase method. J Vet Med Sci. 1994;56:639–644. doi: 10.1292/jvms.56.639. [DOI] [PubMed] [Google Scholar]
- 7.Oliveira S, Blackall PJ, Pijoan C. Characterization of the diversity of Haemophilus parasuis field isolates by use of serotyping and genotyping. Am J Vet Res. 2003;64:435–442. doi: 10.2460/ajvr.2003.64.435. [DOI] [PubMed] [Google Scholar]
- 8.Oliveira S, Pijoan C. Computer-based analysis of Haemophilus parasuis protein fingerprints. Can J Vet Res. 2004;68:71–75. [PMC free article] [PubMed] [Google Scholar]
- 9.Cai X, Chen H, Blackall PJ, et al. Serological characterization of Haemophilus parasuis isolates from China. Vet Microbiol. 2005;111:231–236. doi: 10.1016/j.vetmic.2005.07.007. [DOI] [PubMed] [Google Scholar]
- 10.Munch S, Grund S, Kruger M. Fimbriae and membranes on Haemophilus parasuis. J Vet Med. 1992;39:59–64. doi: 10.1111/j.1439-0450.1992.tb01138.x. [DOI] [PubMed] [Google Scholar]
- 11.Lichtensteiger C, Vimr E. Purification and renaturation of membrane neuraminidase from Haemophilus parasuis. Vet Microbiol. 2003;93:79–87. doi: 10.1016/s0378-1135(02)00443-1. [DOI] [PubMed] [Google Scholar]
- 12.Charland N, D’Silva C, Dumont R, Niven D. Contact-dependent acquisition of transferrin-bound iron by two strains of Haemophilus parasuis. Can J Microbiol. 1995;41:70–74. doi: 10.1139/m95-009. [DOI] [PubMed] [Google Scholar]
- 13.MacInnes JI, Paradis M-A, Vessie GH, et al. Efficacy of prophylactic tilmicosin in the control of experimentally induced Haemophilus parasuis infection in pigs. Swine Health Prod. 2003 Jul–Aug;:174–180. [Google Scholar]
- 14.MacInnes J, Lone A, Metcalf D, et al. A survey of Ontario swine herds for the presence of Actinobacillus pleuropneumoniae, A. suis, Haemophilus parasuis, Pasteurella multocida, and Streptococcus suis In: Proceedings of the 84th Annual Meeting of Conference of Research Workers in Animal Diseases; Chicago. 2003 Nov 9–11:57.
- 15.Niven D, Donga J, Archibald F. Responses of Haemophilus pleuropneumoniae to iron restriction: changes in the outer membrane protein profile and the removal of iron from porcine transferrin. Mol Microbiol. 1989;3:1083–1089. doi: 10.1111/j.1365-2958.1989.tb00258.x. [DOI] [PubMed] [Google Scholar]
- 16.O’Reilly T, Niven DF. Levels of nicotinamide adenine dinucleotide in extracellular body fluids of pigs may be growth-limiting for Actinobacillus pleuropneumoniae and Haemophilus parasuis. Can J Vet Res. 2003;67:229–231. [PMC free article] [PubMed] [Google Scholar]
- 17.Hill C, Metcalf D, MacInnes J. A search for virulence genes of Haemophilus parasuis using differential display RT-PCR. Vet Microbiol. 2003;96:189–202. doi: 10.1016/s0378-1135(03)00212-8. [DOI] [PubMed] [Google Scholar]
- 18.Sambrook J, Fritsch E, Maniatis T. Plasmid vectors. In: Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press, 1989:1.32.
- 19.Mathieu-Daude F, Trenkle T, Welsh J, Jung B, Vogt T, McClelland M. Identification of differentially expressed genes using RNA fingerprinting by arbitrarily primed polymerase chain reaction. Methods Enzymol. 1999;303:309–324. doi: 10.1016/s0076-6879(99)03020-7. [DOI] [PubMed] [Google Scholar]
- 20.Trulzsch K, Roggenkamp A, Pelludat C, Rakin A, Jacobi C, Heesemann J. Cloning and characterization of the gene encoding periplasmic 2,3-cyclic phosphodiesterase of Yersinia entercolitica O:8. Microbiology. 2001;147:203–213. doi: 10.1099/00221287-147-1-203. [DOI] [PubMed] [Google Scholar]
- 21.Liu J, Beacham IR. Transcription and regulation of the cpdB gene in Escherichia coli K12 and Salmonella typhimurium LT2: evidence for modulation of constitutive promoters by cyclic AMP–CRP complex. Mol Gen Genet. 1990;222:161–165. doi: 10.1007/BF00283039. [DOI] [PubMed] [Google Scholar]
- 22.Cannistraro V, Kennell D. Rnase I, a form of Rnase I*, and mRNA degradation in Escherichia coli. J Bacteriol. 1991;173:4653–4659. doi: 10.1128/jb.173.15.4653-4659.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Proudfoot M, Kuznetsova E, Brown G, et al. General enzymatic screens identify three new nucleotidases in Escherichia coli: biochemical characterization of SurE, YfbR, and YjjG. J Biol Chem. 2004;279:54687–54694. doi: 10.1074/jbc.M411023200. [DOI] [PubMed] [Google Scholar]
- 24.Kim S, Ahn S, Lee J, Kim N, Park K, Kong I. Genetic analysis of phosphomannomutase/phosphoglucomutase from Vibrio furnissii and characterization of its role in virulence. Arch Microbiol. 2003;180:240–250. doi: 10.1007/s00203-003-0582-z. [DOI] [PubMed] [Google Scholar]
- 25.Silo-Suh L, Suh S, Phibbs P, Ohman D. Adaptations of Pseudomonas aeruginosa to the cystic fibrosis lung environment can include deregulation of zwf, encoding glucose-6-phosphate dehydrogenase. J Bacteriol. 2005;87:7561–7568. doi: 10.1128/JB.187.22.7561-7568.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Melnikow E, Dornan S, Sargent C, et al. Microarray analysis of Haemophilus parasuis gene expression under in vitro growth conditions mimicking the in vivo environment. Vet Microbiol. 2005;110:255–263. doi: 10.1016/j.vetmic.2005.08.007. [DOI] [PubMed] [Google Scholar]
- 27.Boyer E, Bergevin I, Gros P, Cellier M. Acquisition of Mn (II) in addition to Fe (II) is required for full virulence of Salmonella enterica serovar Typhimurium. Infect Immun. 2002;70:6032–6042. doi: 10.1128/IAI.70.11.6032-6042.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bigas A, Garrido M, de Rozas A, Badiola I, Barbe J, Llagostera M. Development of a genetic manipulation system for Haemophilus parasuis. Vet Microbiol. 2005;105:223–228. doi: 10.1016/j.vetmic.2004.10.015. [DOI] [PubMed] [Google Scholar]