|Home | About | Journals | Submit | Contact Us | Français|
Studies on virulence factors and pathogenecity of Haemophilus parasuis have long been hindered by a lack of a consistent system for genetic manipulation. In this study, competence was induced by transferring H. parasuis from rich medium to starvation medium media-IV (M-IV) and iscR gene deficient mutants of H. parasuis were generated efficiently. Transformation frequency varied from 4.1 × 10−5 to 1.1 × 10−8 when using circular plasmid, and increased to about 2- to 31-fold when transformed using linearized plasmid. Allele replacement occurred efficiently in 6 strains, which are transformable using both circular and linearized pTRU, but not in another 2 strains which could only be transformed using linearized plasmid. The iscR mutants were stable for at least 20 passages in vitro. Haemophilus parasuis strains vary extensively in natural transformation efficiency and the method established here allows for transformation of a larger spectrum of strains with an easily accessed plasmid. This provides important tools for genetic manipulation of H. parasuis.
Les études sur les facteurs de virulence et la pathogénicité d’Haemophilus parasuis ont longtemps été limitées à cause de l’absence d’un système constant de manipulations génétiques. Dans la présente étude, la compétence a été induite en transférant H. parasuis d’un milieu de culture riche au milieu pauvre M-IV et les H. parasuis mutants déficients pour le gène iscR étaient générés efficacement. La fréquence de transformation variait de 4,1 × 10−5 à 1,1 × 10−8 lors de l’utilisation d’un plasmide circulaire, et augmentait d’un facteur variant de 2 à 31 lorsque la transformation utilisait un plasmide linéarisé. Le remplacement d’allèle est survenu efficacement chez 6 souches, qui étaient transformables en utilisant le pTRU circulaire ou linéarisé, mais pas chez 2 autres souches qui ne pouvaient être transformées qu’en utilisant le plasmide linéarisé. Les mutants iscR étaient stables pendant au moins 20 passages in vitro. Les souches d’H. parasuis varient énormément dans leur efficacité de transformation naturelle et la méthode développée ici permet la transformation d’un plus large spectre de souches avec un plasmide facilement accessible. Ceci fournit d’importants outils pour la manipulation génétique d’H. parasuis.
(Traduit par Docteur Serge Messier)
Haemophilus parasuis is the causative agent of Glässer’s disease, characterized by fibrinous polyserositis, arthritis, and meningitis. The bacterium is Gram-negative and a member of the family Pasteurellaceae. It can be a commensal bacterium colonized in the upper respiratory tract of clinically normal swine or a pathogen causing severe systemic infection (1). Therefore, it is very important to differentiate between virulent and non-virulent isolates for diagnosis and control of the disease. Current research on virulence factors of H. parasuis is mainly based on comparison of high and low virulence strains, and an increasing number of putative virulence factors have been found (2,3). However, the function of these putative virulence factors is still to be determined due to a lack of a genetic manipulation method for construction of deletion mutant.
Construction of a genetic mutant is an efficient method for determination of virulence factors. So far, 2 different systems have been developed for transformation of H. parasuis including electroporation and natural transformation. A modified endogenous plasmid was used for the genetic manipulation of H. parasuis, and 15 out of 16 strains were transformable when the plasmid was pretreated with cell-free extracts (4). Electroporation efficiencies up to 105 were achieved when done with a native plasmid by growing H. parasuis at low temperature (5). A Tn5-based random mutagenesis method for use in H. parasuis was developed for preparation of mutant pools (6). However, there have been no reports of exogenous plasmid transformed into H. parasuis by electroporation. Hence, this method was not widely used due to difficulty in access to endogenous plasmid.
A natural transformation method was developed and mutants were constructed, but only 1/11 tested strains were naturally transformable. Moreover, the presence of specific DNA uptake signal sequence (USS) is required for natural transformation of H. parasuis, but the effect of the secondary messenger cyclic adenosine monophosphate (cAMP) on natural transformation is controversial (7,8). Recently, a successive markerless mutation system in H. parasuis through natural transformation was established, but only 1 of 6 screened strains was transformable (9). Searching for a consistent method to improve transformation efficiency for generation of H. parasuis mutants is still a challenge (2).
Natural competence has been extensively studied in Haemophilus influenzae, which is also a member of the family Pasteurellaceae. When H. influenzae was transferred from rich medium to starvation medium or from aerobic conditions to anaerobic conditions, competence was developed (10,11). The addition of cAMP to exponentially growing bacteria also induces competence (12). Transferring early exponentially growing bacteria to starvation medium also induces competence in Haemophilus parainfluenzae, Actinobacillus pleuropneumoniae, and Gallibacterium anatis (13–15).
The objective of this study was to develop a consistent method for transformation and construction of mutant H. parasuis. Natural transformation of H. parasuis was conducted using 10 different field isolates and an iscR (a HTH-type transcriptional regulator at the 3′ end of capsule gene cluster) deficient mutant was constructed in this study.
The H. parasuis strains used in this study are listed in Table I. All of the strains were isolated from different farms in Zhejiang province, China. The serovar of the strains were determined by a newly developed multiplex polymerase chain reaction (PCR) that distinguished between all previously described serovars except 5 and 12 (16). The strains were cultured at 37°C in brain heart infusion (BHI) or on Tryptic Soy Agar (TSA) supplemented with 0.025% nicotinamide adenine dinucleotide (NAD) and 2% bovine serum.
Primers used in this study are listed in Table II, and a 9 bp USS fragment of H. parasuis is present in primers 1 and 6. To construct a suicide vector for transformation and generate iscR mutants, a 973 bp fragment (containing BamH I site) upstream of the iscR ATG start codon and a 974 bp fragment (containing Hind III site) downstream of the iscR TAA stop codon were amplified from genome DNA of the strain ZJ1018; and a 816 bp fragment of kanamycin coding sequence was amplified from pET 28a vector. The 3 fragments were ligated by overlap PCR, as described (4), and cloned into pUC 18 vector (the pMB1 replicon carried in the plasmid allows it to replicate in E. coli but not H. parasuis) by BamH I and Hind III to form pTRU (Figure 1). Another plasmid containing the same gene with pTRU except the presence of the 9 bp USS fragment was constructed and designated as pTR (Figure 1). Plasmids were extracted from DH5α using a kit (Endofree Maxi Plasmid Kit; Tiangen, China), and DNA concentration was measured using spectrophotometer (NanoVue; GE Healthcare, Baie d’Urfé, Quebec).
Competence was induced by transferring H. parasuis from BHI to M-IV medium (11). Haemophilus parasuis strains were retrieved from −80°C and inoculated in BHI supplemented with NAD and bovine serum at 1:100 dilution. The bacteria were cultured at 37°C by rotating at 200 rounds per minute (rpm) in a shaker incubator. When the optical density at 600 nm (OD600) reaches 0.1~0.2, the bacteria were collected by centrifugation and washed once in equal volume of M-IV. The bacteria was resuspended in equal volume of M-IV and incubated at 37°C by rotating at 200 rpm for 100 min in a shaker incubator. Then cells were collected and resuspended in an equal volume of M-IV medium for the transformation assay.
Transformation with closed-circular or linearized plasmid DNA was done as described for H. influenzae with little modification (17). When transformed with closed-circular plasmid, 1 μg of circular recombinant plasmid pTR or pTRU was added to 1 mL of competent cells, mixed gently by inverting the tube 5 to 6 times, and then incubated at 37°C for 30 min. Then 80% glycerol was added to a final concentration of 30% to 32%, and incubated at 25°C for 10 min after inverting the tube 5 to 6 times. The bacteria were collected and resuspended in BHI before plating on TSA plus NAD and serum (containing 25 μg/mL kanamycin) immediately at proper dilution. The TSA plates were incubated at 37°C for 24 to 72 h. Transformation frequencies were calculated by CFU mL−1 on TSA with kanamycin divided by CFU mL−1 on TSA without kanamycin.
When transformed with linearized plasmid, 1 μg of BamH I linearized pTRU was mixed with 1 mL of competent cells, and the mixture was incubated at 37°C for 15 min. Cells were harvested and resuspended in BHI before plating on TSA plus NAD and serum (containing 25 μg/mL kanamycin) immediately at proper dilution. The TSA plates were incubated at 37°C for 24 to 72 h, and transformation frequencies were calculated as described above.
Three H. parasuis strains were selected for optimizing transformation conditions. To determine the optimum growth stage for development of competence, H. parasuis competence was induced at various growth stages measured by OD600, and competent cells were transformed using pTRU. To determine the effect of USS sequence on transformation frequency, H. parasuis competence was induced when OD600 reached 0.1~0.2, and competent cells were transformed using pTR or pTRU. Comparison of transformation efficiency using pTR or pTRU was analyzed by one-way ANOVA using computer software (SPSS Statistics, version 17.0; IBM, USA), data is considered statistically significant when P ≤ 0.05.
To measure the occurrence of natural competence among H. parasuis strains, 10 strains were selected randomly, and competence was induced at OD600 of 0.2, then transformed with circular or BamH I linearized pTRU.
To check the reliability of homologous recombination in H. parasuis, 6 strains were transformed using circular or linearized pTRU, and genome DNA of 20 kanamycin resistant colonies were extracted by proteinase K digestion for each strain (18). Primers P3/P4, P11/P12 were used to check the insertion of kanamycin gene and deletion of iscR gene in genome DNA, and M13 primers were used to check the absence of pTRU in the bacteria. Replacement of iscR gene by the kanamycin gene in the chromosome was further confirmed by sequencing the PCR amplicon with primers P9 and P10, which are located at 1349 bp ahead of the upstream homologous arm and 278 bp behind the downstream homologous arm, respectively.
Competence of H. parasuis strains were induced at various cell concentrations (OD600 0.1 to 1.0) and competent cells were transformed using pTRU. Transformation frequency was highest at OD600 0.1 to 0.2, and dropped continuously with the increase in cell concentration (Figure 2). Extensive loss of competence development ability was observed when OD600 reached 0.8 to 1.0.
Two recombinant plasmid pTRU (containing 2 extra USS) and pTR were transformed into H. parasuis under the same conditions. Transformation efficiency of pTRU was slightly higher than that of pTR in ZJ0901 and ZJ1017, and lower in ZJ1307 (Figure 3). No significant difference was found between the 2 plasmids in transformation efficiency (P > 0.05). Analysis of the pTR gene using computer software (MegAlign, DNASTAR Lasergene; DNASTAR, Madison, Wisconsin, USA) revealed no homologous sequence of any of the 2 reported USS sequences (7,8) and its complement sequence. The presence of the USS in pTRU did not facilitate the transformation of H. parasuis.
Ten randomly selected field strains were transformed with circular or linearized pTRU. It shows that 6 out of 10 strains could be transformed with both circular and linearized pTRU by using the M-IV procedure. Transformation frequency varies from 4.1 × 10−5 to 1.1 × 10−8 when using circular plasmid, and increases about 2- to 31-fold when transforming with linearized plasmid (Figure 4). Among the rest, 4 strains that could not be transformed with circular plasmid, 2 of them (ZJ1004 and ZJ1209) could be transformed using linearized pTRU at very low efficiency. Extensive difference in transformation efficiency was observed among strains (Figure 4).
The deletion of iscR and insertion of kanamycin gene were determined by PCR. It shows that the iscR gene was replaced by the kanamycin gene in the H. parasuis chromosome, and the iscR mutant shows genetic stability for at least 20 passages on TSA in the presence or absence of kanamycin (Figure 5). Twenty kanamycin resistant clones from each of the 6 strains transformed with circular or linearized pTRU were checked, and over 90% of the clones were iscR mutant. Interestingly, targeted gene replacement only took place in the 6 strains that could be transformed using both circular and linearized plasmid, but not in strains (ZJ1004 and ZJ1209), which could be only transformed using linearized plasmid. In these 2 strains (ZJ1004 and ZJ1209), incomplete homologous recombination may occur, and the kanamycin gene and iscR gene were both detected in the recombinants obtained from the linearized plasmid. In addition, the suicide of the pTRU in H. parasuis was confirmed by PCR with M13 primers (Figure 5).
Although increases in potential virulence factors have been found, pathogenesis of H. parasuis is poorly understood. Identification of these factors involved in pathogenesis has been limited due to a lack of general methods for producing mutants. In this study, we found the applicability of the defined M-IV medium in inducing competence of H. parasuis, and developed a simple method for generating mutants. The M-IV medium was first developed for inducing competence in H. influenzae (11). Competence development of H. parasuis was better at an early exponential growth stage (OD600 0.1 to 0.2), which is similar to the conditions known to be optimal for induction of competence in H. influenzae (17).
The DNA uptake signal sequence is highly over-represented in genomes of several members of Pasteurellaceae, and is evolutionary conserved (20). One USS signal sequence of H. parasuis, identical to that found in Actinobacillus pleuropneumoniae and Mannheimia haemolytica, was identified in the SH0165 genome (21). The effect of USS on natural transformation of H. parasuis has been confirmed previously, although H. parasuis strains may differ in requirements of the USS nucleotide sequence (7,8). But our result shows that the reported USS of H. parasuis has no effect on transformation efficiency of M-IV medium-induced competent cells. Similarly, it is believed that the USS does not create a practical barrier to efficient transformation in Gallibacterium anatis (G. anatis), and plasmids with no USS sequences could be well transformed into the bacteria (15).
It is believed that the competence development of H. influenzae is due to a starvation response. In H. influenzae, competence is induced with the starvation medium and absolutely requires an increase in cAMP, an established indicator of nutritional stress, but excessive cAMP may interfere with competence induction (22). In H. parasuis, the effect of cAMP on natural transformation is controversial. Increasing transformation efficiency was observed when the concentration of cAMP supplementation rising from 0.05 to 8 mM, but no significant difference in transformation efficiency was observed in another report under similar cAMP concentration and transformation procedure (7,8). The genetic difference of strains used in a previous report may be responsible for the sensitivity of cAMP in inducing competence, and it had been shown that an icc deletion mutant of H. influenzae was much more sensitive to exogenous cAMP (22). We also tested the effect of cAMP on transformation efficiency in M-IV medium induced competent cells, and supplementation of 1 mM cAMP in either competence induction or transformation stage did not increase the transformation frequency. This may be because sufficient cAMP has been produced when cultured in starvation medium, extra cAMP is not necessary. It has been demonstrated that E. coli cAMP levels peak during the transition to starvation (23–25).
The variation in transformability was reported in several species. Only 18 of 31 serotype b isolates of H. influenzae were transformable, while 29 of 34 strains were transformable when the tested isolates were dominated by non-serotype b strains (26,27). In G. anatis, all 9 tested strains were transformable with chromosomal DNA or linearized gene disruption constructs, but only 6 of 9 were transformable using circular plasmid (15). Ten randomly selected H. parasuis strains were subjected to natural transformation, and 6 of them were found transformable using both circular and linearized plasmid. Moreover, 2 of the remaining 4 strains could be transformed using linearized plasmid. The increase of transformation frequencies was observed when using linearized instead of the circular pTRU, and the increase of transformants generated by linear plasmid was also observed when using a different natural transformation procedure (9). Only 1 of the 10 strains (ZJ1208) were transformable when subjected to the previously reported transformation procedure, which similarly reported that 1 of 11 tested strains were transformable (8). Electroporation was used to transform H. parasuis, but no kanamycin resistant clone was observed under our experiment conditions.
In this study, 2 serovar 7 strains were transformable using both circular and linearized plasmid; all 3 serovar 4 strains were transformable with linearized plasmid, but only 2 of them were transformable using circular plasmid. Only 1 of the 3 serovar 13 strains could be transformed using both forms of plasmid, while the other 2 strains were non-transformable with both forms of the plasmid. Hence, natural competence of these strains shows no strict association with serovar, but it is impossible to carry on a statistical analysis due to the limited number of strains tested in this study. Analysis with more strains in each serotype is needed to determine the relationship between natural competence and serovar.
Study of H. parasuis virulence factors has long been hindered by problems in generating mutants. The homologous recombination was efficient in M-IV induced competent H. parasuis, and the iscR mutants were genetically stable. Moreover, another wza (capsular export protein) mutant was also generated by the procedure established in this study in our laboratory (28). Hence, the method established here allows the transformation of a larger spectrum of strains with an easily accessed plasmid, which will facilitate the identification of virulence factors and the characterization of pathogenicity of H. parasuis.
This study was financially supported by National Natural Science Foundation of China (Grant No. 31201934).