Characterization of DegQ_Vh, a Serine Protease and a Protective Immunogen from a Pathogenic Vibrio harveyi Strain

Abstract

Vibrio harveyi is an important marine pathogen that can infect a number of aquaculture species. V. harveyi degQ (degQ_Vh), the gene encoding a DegQ homologue, was cloned from T4, a pathogenic V. harveyi strain isolated from diseased fish. DegQ_Vh was closely related to the HtrA family members identified in other Vibrio species and could complement the temperature-sensitive phenotype of an Escherichia coli strain defective in degP. Expression of degQ_Vh in T4 was modulated by temperature, possibly through the σ^E-like factor. Enzymatic analyses demonstrated that the recombinant DegQ_Vh protein expressed in and purified from E. coli was an active serine protease whose activity required the integrity of the catalytic site and the PDZ domains. The optimal temperature and pH of the recombinant DegQ_Vh protein were 50°C and pH 8.0. A vaccination study indicated that the purified recombinant DegQ_Vh was a protective immunogen that could confer protection upon fish against infection by V. harveyi. In order to improve the efficiency of DegQ_Vh as a vaccine, a genetic construct in the form of the plasmid pAQ1 was built, in which the DNA encoding the processed DegQ_Vh protein was fused with the DNA encoding the secretion region of AgaV, an extracellular β-agarase. The E. coli strain harboring pAQ1 could express and secrete the chimeric DegQ_Vh protein into the culture supernatant. Vaccination of fish with viable E. coli expressing chimeric degQ_Vh significantly (P < 0.001) enhanced the survival of fish against V. harveyi challenge, which was possibly due to the relatively prolonged exposure of the immune system to the recombinant antigen produced constitutively, albeit at a gradually decreasing level, by the carrier strain.

Under stress conditions, such as those induced by elevated temperatures, extreme pH, osmolarity shock, and host immune responses, damaged and misfolded proteins are accumulated in bacteria; the removal and correction of these otherwise harmful proteins depend on the operation of a number of protease/chaperon systems, among which are the HtrA/DegP proteins (5, 6, 28). Initially discovered in Escherichia coli as an essential protein required for growth at high temperatures (22, 23, 41), HtrA (high temperature requirement A) homologues have been identified in diverse prokaryotic and eukaryotic organisms, including humans (10). These proteins possess the dual function of protease and chaperon and can switch roles according to the input of environmental stimuli (39). Members of the HtrA family, which include DegP (also known as DO), DegQ, and DegS, share a relatively high level of sequence identity and the common feature of an N-terminal protease domain followed by one or two PDZ domains. In E. coli, degQ and degS are located next to each other on the chromosome and are transcribed independently (44); their respective encoded proteins, DegQ and DegS, are ∼36% identical to DegP, whose coding sequence is physically separated from the degQ-degS cluster. DegP is involved in the refolding/degradation of abnormal proteins accumulated in the periplasm; degP mutation vitiates bacterial virulence (9, 13-15, 29, 45) and impairs growth under certain stringent conditions, notably those caused by heat shock (23, 30, 35, 40, 46). Mutation of degQ does not lead to any observable growth defect in E. coli but affects the survival of Salmonella enterica serovar Typhimurium in the host (8). Inactivation of degS attenuates virulence in a number of bacterial species, probably on account of the fact that DegS regulates the activity of σ^E, the alternate sigma factor that controls the expression of genes, such as degP, that are essential to the viability of cells under stress conditions (1, 10, 48).

Vibrio harveyi is a gram-negative bacterium of the family Vibrionaceae and one of the important pathogens of cultured marine animals. Owing to its broad host range, which includes peneid shrimp, sea cucumber, and fish of various species (31), V. harveyi has caused severe losses to aquiculture industries worldwide. Recently genome sequencing data have revealed the ubiquitous existence of HtrA homologues in the Vibrio species, yet no characterization of these proteins has been documented. By utilizing a secretion signal trap developed in our laboratory (49), we cloned the genetic cluster containing the degQ and degS paralogues from a virulent V. harveyi strain. Our aim was to investigate whether there was any difference in functional and biochemical aspects between V. harveyi DegQ (DegQ_Vh) and other known DegQ proteins and whether DegQ_Vh could be explored in some new directions that would lead to its application in disease control. Our results indicated that DegQ_Vh was an endoprotease and a vaccine candidate against V. harveyi infection.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains used in this study are listed in Table 1. Bacillus strain B187 was isolated from healthy Japanese flounder, whereas V. harveyi T4, Edwardsiella tarda TX1, and V. harveyi HH2 were isolated from diseased Japanese flounder at fish farms in north China. The genetic identities of the isolates were determined by 16S rRNA gene analysis. Using Japanese flounder (average weight, 11.4 g) as animal models, the 50% lethal dose (LD₅₀) of strain B187 was greater than 1 × 10⁹ CFU while those of strains T4, TX1, and HH2 were less than 3 × 10⁷ CFU. All E. coli strains were grown in Luria-Bertani lysis broth (LB) (33) at 37°C, and all other strains were grown in the same medium at 28°C. Appropriate antibiotics were supplemented at the following concentrations: ampicillin, 100 μg/ml; kanamycin, 50 μg/ml; chloramphenicol, 25 μg/ml; tetracycline, 15 μg/ml.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)	Source or reference
Bacillus sp. strain B187	Fish commensal isolate	This study
Edwardsiella tarda TX1	Fish pathogen	This study
Escherichia coli
DH5α	Host strain for general cloning	TaKaRa (China)
NK7047	F− λ− ΔlacX74 rpsL galOP308	42
EMZ1	NK7047 defective in degP	This study
S17-1λpir	Tpr SmrrecA hsdR M+ RP4-2-Tc::Mu-Km::Tn7λpir	Biomedal (Spain)
Vibrio harveyi
T4	Fish pathogen (LD50, ∼1.5 × 107 CFU)	49
HH2	Fish pathogen (LD50, ∼2.5 × 107 CFU)	This study
Plasmids
pACYC184	General cloning vector	New England BioLabs
pBU	Signal sequence trap carrying signal sequence-lacking agaV	49
pBRL	pBR322 derivative carrying Plac	This study
pBEP	pBRL carrying E. coli degP	This study
pBEQ	pBRL carrying E. coli degQ	This study
pBVQ1	pBRL carrying degQVh	This study
pBVQ2	pBRL carrying degQVh encoding PDZ1 and PDZ2 deletion	This study
pBT3	Expression plasmid with Ptrc	This study
pBQ	pBT3 carrying degQVh	This study
pBTA1	pBT3 encoding AgaV107	This study
pBTA2	pBT3 encoding AgaV195	This study
pAQ1	pBTA1 encoding AgaV107-DegQVh	This study
pAQ2	pBTA1 encoding AgaV195-DegQVh	This study
pEF	pACYC184 carrying fur	This study
pET258	Expression plasmid	49
pETQ	pET258 carrying degQVh	This study
pETQ83	pET258 carrying degQVh(H83A)	This study
pETQ113	pET258 carrying degQVh(D113A)	This study
pETQ188	pET258 carrying degQVh(S188A)	This study
pETQD2	pET258 carrying degQVh encoding PDZ2 deletion (DegQVh-PDZ2)	This study
pETQD12	pET258 carrying degQVh encoding PDZ1 and PDZ2 deletion (DegQVh-PDZ12)	This study
pGP704	Suicide plasmid	25
pMBP	pACYC184 carrying malE	This study

DNA and molecular techniques.

Plasmid preparation, PCR amplifications, purification of PCR products, and genomic DNA preparation were carried out as described previously (49). Restriction endonucleases and modifying enzymes were purchased from Fermentas (China) and used in accordance with the manufacturer's specifications.

Chemicals.

All chemicals used in this study were purchased from Sangon (China) except for azocasein (Sigma-Aldrich) and collagen (Worthington).

Plasmid and strain constructions.

The plasmids used in this study are listed in Table 1. The primers used in this study are listed in Table S1 in the supplemental material. To construct pBRL, P_lac of pEGFP (Clontech) (amplified by PCR with the primers LacPF1/LacPR1) was inserted into pBR322 between the EcoRI and BamHI sites. pBVQ1 and pBVQ2 were generated by ligating the wild-type and mutant degQ_Vh genes (amplified by PCR with the primer pairs VQF15/VQR5 and VQF15/VQR19, respectively) into pBRL at the BamHI site. pBEP and pBEQ were constructed by inserting the E. coli degP and degQ genes (amplified by PCR with the primer pairs EPF3/EPR3 and EQF3/EQR3, respectively) into pBRL at the BamHI site. pETQ was constructed by ligating degQ_Vh (amplified by PCR with the primers VQF25/VQR2) into pET258 between the NdeI/XhoI sites. To construct pBT3, the rrnB transcription terminator of pTrcHis (Invitrogen) (amplified by PCR with the primers RNBF1/RNBR1) was ligated to the 2.8-kb EcoRV/BsaBI fragment of pBR322, yielding pBRB, which was then cut with EcoRI/BamHI and ligated to the P_trc promoter of pTrcHis, resulting in pBT; the NdeI site of pBT was inactivated by filling in with a 36-bp oligonucleotide (5′-TACAATAGTATAGTGGTAGCTTGTAGATCTAGATAG-3′), and the resulting plasmid was digested with BamHI and ligated to a His-tagged NdeI-XhoI linker (5′-GATCCAAGGAGATATACATATGGATATCCTCGAGCACCACCACCACCACCACTAAG-3′), yielding pBT3. pBQ was constructed by inserting degQ_Vh of pETQ into the NdeI/XhoI sites of pBT3. To construct pBTA1 and pBTA2, the DNA encoding the N-terminal 107- and 195-amino-acid residues of AgaV (amplified by PCR with the primer pairs UAF7/UAR9 and UAF7/UAR11, respectively) was inserted into pBT3 at the BamHI site. pAQ1 and pAQ2 were generated by inserting degQ_Vh (amplified by PCR with the primers VQF3/VQR2) into the NdeI/XhoI sites of pBTA1 and pBTA2, respectively. To construct pEF and pMBP, lacO-less P_lac (generated by PCR with primers LacpF1/LacPR4) of pEGFP was inserted into pBRB between the EcoRI-BamHI sites, yielding pLS; the above-described His-tagged NdeI-XhoI linker was then inserted into pLS at the BamHI site, resulting in pL2; the E. coli ferric uptake regulator gene (fur; generated by PCR with primers EFF7/EFR11) and the maltose binding protein gene (malE; generated by PCR with primers MalF1/MalR1) were inserted into pL2 between the NdeI-XhoI sites, resulting in pLE and pLM, respectively, which were digested with SwaI, and the fragments containing P_lac-fur and P_lac-malE were ligated to pACYC184 linearized with EcoRV, resulting in pEF and pMBP, respectively.

For the construction of the E. coli strain EMZ1, the cat gene of pACYC184 was inserted into pGP704 between the PstI and SalI sites, resulting in p704C, which was linearized with EcoRV and ligated to an internal fragment of E. coli degP (amplified by PCR with the primers DegPF1/DegPR1), resulting in pGPD, which was introduced into S17-1λpir by transformation; the transformants were mated with N7K, an NK7047 variant that was resistant to kanamycin. The transconjugants were selected for chloramphenicol and kanamycin resistance on LB agar plates. One of the selected clones was named EMZ1, which was confirmed to have degP inactivated by PCR and growth analysis, the latter showing that EMZ1 could grow at 37°C but not at 42°C and its growth defect could be rescued by DegP (Fig. 1).

FIG. 1.

Effect of DegQ_Vh and the E. coli DegP and DegQ proteins on the growth of EMZ1 at high temperature. N7K and EMZ1 harboring pBRL derivatives were grown at 42°C in the absence or presence of different concentrations of IPTG; aliquots were taken at different time points for the measurement of absorbance at OD₆₀₀. Data are the means for at least three independent experiments and are presented as the means ± SE.

Generation of DegQ_Vh mutants.

Site-directed mutagenesis was performed using the method of overlap extension PCR (12). For H83A, D113A, and S188A mutations, the overlapping PCRs were performed with the primer pairs VQF25/VQR15 and VQF26/VQR2, VQF25/VQR16 and VQF27/VQR2, and VQF25/VQR17 and VQF28/VQR2, respectively; the fusion PCRs were performed with the primer pair VQF25/VQR2. The PCR products were ligated into pET258 between the NdeI/XhoI sites, resulting in pETQ83, pETQ113, and pETQ188, respectively.

pETQD2 and pETQD12 were created by ligating the PCR products generated with the primer pairs VQF25/VQR14 and VQF25/VQR13, respectively, into pET258 between the NdeI/XhoI sites.

Cloning of the degQ_Vh cluster.

T4 genomic DNA was digested with Sau3A1, and the fragments between 4 and 6 kb were recovered and ligated into pBU at the BamHI site; DH5α was transformed with the ligation mix, and the transformants were selected as described previously (49). The complete sequence of the degQ_Vh cluster was obtained by genome walking as described previously (49).

Expression and purification of wild-type and mutant recombinant DegQ_Vh.

BL21(DE3) was transformed with pET258 carrying degQ_Vh variants; the transformants were grown in LB medium to an optical density at 600 nm (OD₆₀₀) of 0.7, and the expression of the recombinant proteins was induced by adding to the culture 1 μM isopropyl-β-d-thiogalactopyranoside (IPTG). After an additional 5 h of growth, the cells were harvested and the recombinant proteins were purified with nickel-nitrilotriacetic acid agarose as described previously (49).

Purification of recombinant E. coli Fur.

The coding sequence of the E. coli fur gene (amplified by PCR with primers EFF7/EFR11) was inserted into pET258 between the NdeI-XhoI sites, resulting in pETEF. BL21(DE3) was transformed with pETEF, and the recombinant Fur was purified from BL21(DE3)/pETEF by using nickel-nitrilotriacetic acid beads.

Real-time RT-PCR.

Total RNA was extracted from cells grown in LB medium to various densities by using the SV total RNA isolation system (Promega). Real-time reverse transcriptase PCR (RT-PCR) was carried out in an ABI 7300 real-time detection system (Applied Biosystems) by using the Sybr ExScript RT-PCR kit (Takara, China). The primers used for RT-PCR of degQ_Vh were DF14 (5′-CACCCCAGTCACCGCAA-3′) and DR14 (5′-GACGCTCACGAGTTTGTTCTGT-3′). Each assay was performed in triplicate with the 16S rRNA as a control. The primers used for the RT-PCR of 16S rRNA were 933F (5′-GCACAAGCGGTGGAGCATGTGG-3′) and 16SRTR1 (5′-CGTGTGTAGCCCTGGTCGTA-3′). The amplification efficiencies of the control and the sample were 99.99 and 100%, respectively. Dissociation analysis of amplification products was performed at the end of each PCR to confirm that only one PCR product was amplified and detected. The comparative threshold cycle method (2^−ΔΔCT method) was used to analyze the mRNA level. All data were given in terms of relative mRNA, expressed as means plus or minus standard errors of the means (SE). Statistical analyses were performed by using the two-tailed Student t test.

Enzyme assay.

The protease activity of purified recombinant DegQ_Vh was analyzed by incubating the protein, which had been diluted to various concentrations, with β-casein (0.5 mg/ml) in the standard assay buffer (50 mM potassium phosphate [pH 7.4]) at 37°C for 1 h. The samples were then boiled for 2 min and electrophoresed in a 0.1% sodium dodecyl sulfate (SDS)-12% polyacrylamide gel. The effect of pH on the enzyme activity was determined by incubating the enzyme with azocasein (0.5%) at 50°C for 1 h in three different buffers: 50 mM citric acid-sodium phosphate (pH 4 to 6), 50 mM sodium phosphate (pH 6 to 9), and 50 mM glycine-NaOH (pH 9 to 11); the reaction was stopped by the addition of an equal volume of 10% trichloroacetic acid followed by incubation at 4°C for 15 min; the sample was then centrifuged, and the supernatant was used for the measurement of absorbance at 350 nm. The effect of temperature on the activity of DegQ_Vh was assayed as described above for the effect of pH except that the reaction was performed in the standard assay buffer at various temperature points.

Western and immunoblot analysis.

Cells were grown in LB medium to an OD₆₀₀ of ∼1.2 and harvested by centrifugation at 4°C. The extracellular proteins were prepared by concentrating the supernatant ∼100 times using Amicon Ultra-4 centrifugal filter devices (Millipore); the whole-cell proteins were prepared by lysing the cell pellet in the lysis buffer (100 mM NaH₂PO₄, 10 mM Tris-Cl, and 8 M urea; pH 8.0), followed by centrifugation at 4°C to remove the debris; the periplasmic proteins were prepared according to the method of Flannagan et al. (9). The proteins obtained were electrophoresed in 0.1% SDS-12% polyacrylamide gels. After electrophoresis, the proteins were transferred to nitrocellulose membranes. Immunoblotting was performed as described previously (24) using mouse anti-His monoclonal antibody (Tiangen, China) and anti-Fur and anti-DegQ_Vh polyclonal antibodies.

Antisera.

Antisera to recombinant DegQ_Vh and Fur were prepared by subcutaneously injecting adult rabbits (purchased from the Institute for Drug Control, Qingdao, China) with 200 μg of purified recombinant protein mixed in complete Freund's adjuvant, followed by a boost with the same amount of protein in incomplete Freund's adjuvant 3 weeks later. The rabbits were bled 12 days after the boosting, and the blood was collected, from which the sera were obtained by centrifugation and used for Western immunoblotting and enzyme-linked immunosorbent assay, which was performed according to the method of Kawai et al. (16).

Northern dot blotting.

Sixteen micrograms of RNA was spotted onto a Hybond-N+ nitrocellulose membrane (Amersham Biosciences), and hybridization was carried out at 42°C according to the protocol recommended by Amersham Biosciences using biotin-labeled DNA probes. The labeling of the probe was carried out by using the Randon Primer DNA labeling kit (Beyotime, China). After hybridization, the hybridized RNA was detected by using a chemiluminescent biotin-labeled nucleic acid detection kit (Beyotime, China). Densitometry was performed by using the SensiAnsys gel analysis system (Peiqing, China).

Fish vaccination and challenge.

Healthy Japanese flounder (average weight, 11 g) were purchased from a commercial fish farm in China. The fish were reared at 20 to 22°C in seawater and fed daily with commercial dry pellets. For the vaccination experiment, strain B187 was grown in LB medium to an OD₆₀₀ of 0.6; the cells were harvested and washed in phosphate-buffered saline (PBS). Washed B187 was resuspended in PBS to 2 × 10⁸ CFU/ml (calculated based on the result of a viable count which showed that at an OD₆₀₀ of 0.6, 1 ml of BB187 contained ∼4.8 × 10⁸ viable cells). Purified recombinant DegQ_Vh was mixed into the above B187-PBS solution to a concentration of 250 μg/ml. Eighty-four Japanese flounder were divided randomly into two groups (42 fish/group), designated A and B. Fish in group A were each injected intraperitoneally (i.p.) with 100 μl of B187-PBS-DegQ_Vh mix, whereas fish in group B were each injected i.p. with 100 μl of B187-PBS. To reduce the stress associated with injection, the fish were stunned by being immersed in ice-seawater immediately before the injection. Three weeks postimmunization, the fish of group A were boosted with 20 μg of purified DegQ_Vh diluted in PBS without B187 while the fish in group B were sham boosted with PBS. At the 13th day postboosting, the fish were challenged with strain T4, which had been cultured in LB medium, washed, and resuspended in PBS. Mortality was monitored over a period of 14 days after the challenge, and the relative percentage of survival (RPS) was calculated according to the following formula: RPS = [1 − (% mortality in vaccinated fish/% mortality in control fish)] × 100 (2). Vaccinations with DegQ_Vh against HH2 and TX1 challenges were carried out in the same fashion.

For vaccination with live recombinant DH5α, two groups (42 fish/group) of flounder were i.p. injected with either 100 μl of live recombinant DH5α that had been grown in LB medium to late logarithmic phase, washed, and resuspended in PBS to 2 × 10⁸ CFU/ml or, for the control group, 100 μl of PBS. The fish were challenged as described above, and mortality was monitored for 14 days postchallenge. The RPS was calculated as described above.

All vaccination experiments were repeated once. Statistical analysis was performed by using the chi-square test with Yates' correction.

Bacterial survival in vaccinated fish.

Japanese flounder were immunized via i.p. injection with DH5α carrying pAQ1, pBQ, and pBTA1; at least three fish were sacrificed at 2, 3, 4, 7, 10, and 13 days postimmunization, and the peritoneal fluids, guts, spleen, and liver were removed aseptically. The peritoneal fluids were diluted in PBS and plated on LB agar plates supplemented with ampicillin (marker for pAQ1, pBQ, and pBTA1); the organs were homogenized with glass homogenizers before being plated on LB plates containing ampicillin. The plates were incubated at 37°C for 24 h; the colonies that emerged were examined by PCR using the plasmid- and E. coli-specific primers and subsequent sequencing of the PCR products.

Database search.

The database search was conducted using the BLAST programs at the NCBI. The signal peptide search and predictions were performed using the SignalP 3.0 server.

Nucleotide sequence accession number.

The nucleotide sequence of the degQ_Vh (under the name vhdQ) cluster has been deposited in the GenBank database under the accession number EU344975.

RESULTS

Sequence characterization of the DegQ homologue of V. harveyi.

The gene encoding V. harveyi DegQ (named degQ_Vh) was cloned from T4, a pathogenic strain isolated from diseased fish at a fish farm in north China. The degQ_Vh cluster was obtained by using the pBU system, a signal sequence trap (49). Sequence analysis of the 4-kb DNA thus obtained revealed three tandem open reading frames with lengths of 444, 1,368, and 1,068 bp, respectively. orf444 encodes a hypothetical protein conserved in the Vibrio species. The protein encoded by orf1368 shares, respectively, 96, 92, and 83% overall sequence identity with the protease DO of V. harveyi strain BAA-1116, Vibrio parahaemolyticus, and Vibrio vulnificus (GenBank accession no. ABU69878, BAC58696, and AAO09118, respectively). The predicted amino acid sequence of ORF1068 is 96, 94, and 87% identical to the DegS sequences of the above-mentioned Vibrio species, respectively. Putative signal peptides were identified at the N termini of ORF1368 and ORF1068 by using the SignalP 3.0 server, with the most likely cleavage sites situated between A26-A27 and S28-N29, respectively. When the NCBI conserved domain search program was used, a trypsin-like protease domain (residues 77 to 255) and two PDZ domains (residues 267 to 356 and 361 to 454, respectively) homologous to those of DegQ were identified in ORF1368. ORF1068 was also predicted to be a trypsin family protease, possessing only one PDZ domain. A Q-linker (residues 68 to 88) of the length that is typical of DegQ (19, 47) was identified in ORF1368. Based on these data, orf1368 and orf1068 were considered the genes encoding the V. harveyi DegQ and DegS homologues, respectively, and were named degQ_Vh and degS_Vh, respectively.

DegQ_Vh could complement the mutant phenotype of an E. coli strain defective in degP.

In E. coli, degP mutation is known to affect bacterial growth at elevated temperatures. Since we failed to generate a T4 isogenic strain with a mutated degQ_Vh locus after several attempts, we were unable to study the function of DegQ_Vh with the native genetic background. We therefore adopted the complementation approach, through which the effect of degQ_Vh on an E. coli degP mutant, EMZ1, was analyzed. For this purpose, the plasmids pBVQ1, pBVQ2, pBEQ, and pBEP were constructed, in which wild-type degQ_Vh, a degQ_Vh mutant bearing a deletion of the PDZ domains, the E. coli degQ gene, and the E. coli degP gene are expressed under the lactose promoter P_lac and thus are inducible by IPTG. The E. coli strain EMZ1, which was unable to grow at 42°C as a result of degP inactivation, was transformed with pBVQ1, pBVQ2, pBEQ, pBEP, and the control plasmid pBRL; the transformants were examined for growth at 42°C in the presence or absence of IPTG. The result showed that EMZ1 harboring pBVQ1 was able to grow at the temperature that was otherwise lethal to the host strain (Fig. 1); furthermore, the growth of EMZ1/pBVQ1 at 42°C was a function of the induction level of degQ_Vh and dependent on the protease activity of DegQ_Vh, since pBVQ2, which carries a proteolytically defective degQ_Vh mutant, failed to rescue EMZ1 grown at 42°C. Under the condition of full induction (i.e., in the presence of 100 μM IPTG), EMZ1/pBVQ1 grown at 42°C exhibited a slower doubling time and a lower maximum cell density than EMZ1/pBEQ and more so than EMZ1/pBEP (Fig. 1), suggesting that the complementation ability of DegQ_Vh was less than that of E. coli DegQ and less than that of E. coli DegP. Amino acid sequence comparison revealed that DegQ_Vh is 57 and 56% identical to E. coli DegP and DegQ, respectively, and most of the identities are clustered at the protease domain, especially around the catalytic site.

Expression of degQ_Vh was modulated by temperature.

It is known that expression of the E. coli degP gene is regulated by temperature (44); to investigate whether temperature had any effect on the expression of degQ_Vh, T4 was grown in LB medium at 28°C to an OD₆₀₀ of 0.2; the cells were then divided into two parts; the first part was cultured continuously at 28°C, while the second part was cultured at 37°C (sublethal temperature) for 5, 10, and 30 min. Total RNA was extracted from cells harvested at different growth points and used for real-time RT-PCR analysis of degQ_Vh expression and for Northern blotting analysis of 16S rRNA transcription. The result of Northern dot blotting showed that the differences in the transcription level of 16S rRNA were between 1- and 1.4-fold among the cells grown to an OD₆₀₀ of 0.2 to 0.8 at 28°C and at 37°C (data not shown), suggesting that for T4 the above-specified temperatures and growth phase had only minor effects on the transcription of 16S rRNA. Hence, 16S rRNA was used as an internal control in the real-time RT-PCR analysis of degQ_Vh. The result showed that shifting of the temperature from 28°C to 37°C caused an immediate 10.3-fold increase in the expression of degQ_Vh (Fig. 2); however, the enhancement effect of high temperature appeared to be transient, since prolonged (10 and 30 min) incubation at 37°C drastically reduced degQ_Vh expression (Fig. 2). Inspection of the sequence between orf444 and degQ_Vh revealed a potential rho-independent transcription terminator, followed by a putative promoter (named P_degQ) (Fig. 3) with sequence modules resembling those of the E. coli σ^E, σ³², and degP promoters, which are known to be σ^E dependent and inducible by heat shock (7, 22, 32). The putative −35 (GAACTT) and −10 (TCTAA) elements of P_degQ are perfect matches to those of the P2 promoter of E. coli σ^E, though the space between the −35 and −10 elements is 1 bp longer in P_degQ. These results suggested the possibility that in T4, expression of degQ_Vh was regulated by a σ^E-like factor under conditions of elevated temperature.

FIG. 2.

Expression of degQ_Vh was modulated by temperature. T4 was grown in LB medium at 28°C to an OD₆₀₀ of 0.2; half of the cell culture was removed and grown separately at 37°C for 5, 10, and 30 min while the remaining half of the cell culture was grown continuously at 28°C. Total RNA was extracted from cells grown at 28°C (black bar) or 37°C (white bar) at different growth points and used for real-time RT-PCR. The degQ_Vh mRNA level was normalized to that of 16S rRNA. Data are the means for three independent experiments and are presented as the means ± SE. *, P < 0.001.

FIG. 3.

Nucleotide sequence of the intergenic region between orf444 and degQ_Vh. The translation stop and start codons of orf444 and degQ_Vh, respectively, are in bold capital letters; the putative rho-independent transcription terminator is in italics; the putative −35 and −10 elements of P_degQ are in bold and underlined.

Recombinant DegQ_Vh was an active serine protease whose activity required the integrity of the catalytic triad and PDZ domains.

To determine whether the cloned degQ_Vh gene encoded a functional protease, the gene was subcloned into the expression plasmid pET258 and expressed in the E. coli strain BL21(DE3). The recombinant protein was purified and analyzed for protease activity using β-casein, bovine serum albumin (BSA), and collagen as potential substrates. The results showed that recombinant DegQ_Vh could degrade β-casein but not BSA or collagen; with β-casein as the substrate, DegQ_Vh acted as an endoprotease whose activity was sensitive to phenylmethanesulfonyl fluoride (Fig. 4), an inhibitor of serine protease. DegQ_Vh possesses a catalytic site composed of H83, D113, and S188 of the processed protein. To examine their potential functional importance, these three residues were individually mutated to alanine. In addition, two other DegQ_Vh mutants, DegQ_Vh-PDZ2 and DegQ_Vh-PDZ12, which bear deletions of the last PDZ domain and both PDZ domains, respectively, were also constructed. The mutant proteins were expressed in and purified from BL21(DE3) (see Fig. S1 in the supplemental material). A subsequent enzyme assay showed that compared to the wild-type protein, the H83A, D113A, S188A, DegQ_Vh-PDZ2, and DegQ_Vh-PDZ12 mutants exhibited, respectively, 93.4, 97.5, 91.8, 87, and 91.7% reductions in enzyme activity (Table 2). These results demonstrated that both the catalytic site and the PDZ regions were essential to the function of DegQ_Vh as a protease.

FIG. 4.

Enzymatic analysis of purified recombinant DegQ_Vh. The assay was performed at 37°C in the standard assay buffer containing various concentrations of purified recombinant DegQ_Vh and β-casein with or without PMSF (10 mM). After 1 h of reaction, the samples were boiled for 2 min and electrophoresed in a 0.1% SDS-12% polyacrylamide gel. After electrophoresis the gel was stained with Coomassie blue.

TABLE 2.

Effects of mutation of the catalytic site and deletion of the PDZ domains on protease activity of purified recombinant DegQ_Vh

DegQVh variant	Characteristic(s)	Relative activity (%)a
DegQVh	Wild type	100
DegQVh-H83A	DegQVh with H83A mutation	6.6 ± 0.1
DegQVh-D113A	DegQVh with D113A mutation	2.5 ± 0.8
DegQVh-S188A	DegQVh with S188A mutation	8.2 ± 0.9
DegQVh-PDZ2	DegQVh bearing PDZ2 deletion	13 ± 1.0
DegQVh-PDZ12	DegQVh bearing PDZ1 and PDZ2 deletion	8.3 ± 1.4

^aActivities of the purified proteins were determined by using azocasein as a substrate and are expressed as percentages of the activity of wild-type DegQ_Vh. Data are the means for at least three independent experiments and are presented as means ± standard deviations.

Effects of pH, temperature, and metal ions on the activity of recombinant DegQ_Vh.

To examine the enzymatic characteristics of the recombinant DegQ_Vh protein, the purified enzyme was subjected to activity assays under various conditions. The results showed that the optimal pH and temperature of the purified protein were 8.0 and 50°C, respectively; the protein exhibited more than 70% of the maximum activity over the pH range of 5 to 9 and the temperature range of 40 to 55°C (Fig. 5). The activity of the enzyme declined rapidly when the pH and temperature exceeded 9 and 55°C, respectively. Thermostability analysis showed that purified DegQ_Vh was stable over the temperature range of 10 to 50°C and retained 81% activity after incubation at 50°C for 1 h. To determine the effect of various cations on DegQ_Vh activity, Na⁺, K⁺, Ca²⁺, Co²⁺, Mg²⁺, and Mn²⁺ were each added to the standard assay buffer at two different concentrations and their effects on the proteolytic activity of DegQ_Vh were examined. The results (Table 3) showed that except for Mn²⁺, which heightened the activity of the enzyme at the concentration of 10 mM, all other tested metal ions, as well as EDTA, had no apparent effect on the enzyme activity.

FIG. 5.

Effect of pH (A) or temperature (B) on the activity of purified recombinant DegQ_Vh. (A) The effect of pH was determined in three different buffers using azocasein (0.5%) as a substrate: 50 mM citric acid-sodium phosphate (pH 4 to 6; •), 50 mM sodium phosphate (pH 6 to 9; □), and 50 mM glycine-NaOH (pH 9 to 11; ▴). (B) The effect of temperature (▴) was determined in the standard assay buffer at temperatures between 10 and 70°C with azocasein (0.5%) as a substrate. Thermostability (□) was determined by preincubating the enzyme in the standard assay buffer at the indicated temperature for 1 h before initiating the enzymatic reaction by the addition of azocasein (0.5%). Data are the means for at least three independent experiments and are presented as means ± SE.

TABLE 3.

Effects of cations and EDTA on activity of purified recombinant DegQ_Vh

Cation or EDTA	Relative activity (%) of DegQVh at cation or EDTA concn (mM) ofa:
	1	10
Na+	100 ± 4.0	101 ± 3.8
K+	99 ± 2.5	95 ± 1.9
Ca2+	109 ± 3.0	110 ± 3.5
Co2+	87 ± 2.8	89 ± 1.5
Mg2+	103 ± 2.6	107 ± 4.7
Mn2+	115 ± 2.5	142 ± 4.0
EDTA	103 ± 3.0	118 ± 3.8

^aActivities are expressed as percentages of the enzyme activity measured in 50 mM citric acid-sodium phosphate (pH 6.0). Data are the means for three independent experiments and are presented as means ± standard deviations.

The recombinant DegQ_Vh protein was a protective immunogen that could afford protection against infection by V. harveyi.

Previous study indicated that DegQ of Photobacterium damselae subsp. piscicida (formerly Pasteurella piscicida) was one of the major antigens identified by using a rabbit polyclonal antiserum (11). Similarly, in a separate study in our laboratory of in vivo-induced immunogens of V. harveyi, DegQ_Vh was found to appear among the antigenic proteins that reacted with the rabbit antiserum raised against T4 (X. Jiao and L. Sun, unpublished data). Enzyme-linked immunosorbent assay analysis and Western immunoblotting showed that immunization of rabbits with the purified recombinant DegQ_Vh protein resulted in the production of DegQ_Vh-specific antibodies (data not shown). These findings promoted us to investigate the potential of DegQ_Vh as a protective immunogen. For this purpose, purified DegQ_Vh was mixed with B187, a fish commensal isolate of the Bacillus sp. used here as an adjuvant; the DegQ_Vh-B187 mix was administered via i.p. injection into Japanese flounder. The fish were boosted with DegQ_Vh without B187 3 weeks later and challenged with T4 at the 13th day postboosting. The fish were monitored for cumulative mortality, which showed that compared with the control fish, which displayed 88.1% accumulated mortality, the fish immunized with DegQ_Vh displayed 30.9% accumulated mortality, which yielded an RPS of 64.9%. A similar level of immunoprotection (RPS, 62.2%) was also observed with DegQ_Vh against another pathogenic V. harveyi strain, HH2, which is of a different serotype from T4. In contrast, vaccination with DegQ_Vh elicited no protection (RPS, 14.6%) in fish exposed to TX1, a virulent E. tarda strain. Taken together, these results demonstrated that vaccination with the purified recombinant DegQ_Vh protein could induce protection against V. harveyi infection.

Construction and delivery of a secreted form of DegQ_Vh as a more effective vaccine candidate.

Western immunoblotting analysis showed that in DH5α the recombinant DegQ_Vh protein, guided by its native secretion signal, was localized in the periplasmic space (Fig. 6). To facilitate the application and improve the efficiency of DegQ_Vh as a vaccine, we wanted to engineer a genetic system that could express and deliver DegQ_Vh as a secreted immunogen. For this purpose, we built several translational fusions consisting of the mature DegQ_Vh protein preceded by different secretion segments of AgaV, an extracellular β-agarase described previously (49). The fusions were inserted into the expression plasmid pBT3 and expressed constitutively under the P_trc promoter. DH5α harboring one of the constructs, pAQ1, in which the mature DegQ_Vh protein was fused to the N-terminal (1 to 107) region of AgaV (AgaV107), was able to secret the chimeric AgaV107-DegQ_Vh protein into the culture supernatant (Fig. 6), whereas DH5α harboring pAQ2, in which the mature DegQ_Vh protein was fused to the N-terminal (1 to 195) region of AgaV (AgaV195), retained the chimeric AgaV195-DegQ_Vh protein in the periplasm. To examine whether the detected proteins in the supernatant and periplasm were due to contaminations of cytoplasmic/periplasmic proteins, the plasmids pEF and pMBP were introduced separately into DH5α/pAQ1 and DH5α/pAQ2. pEF constitutively expresses the E. coli ferric uptake regulator (Fur), a cytoplasmic protein, and pMBP constitutively expresses a His-tagged E. coli periplasmic maltose binding protein (MalE). The transformants DH5α/pAQ1/pEF, DH5α/pAQ2/pEF, DH5α/pAQ1/pMBP, and DH5α/pAQ2/pMBP were cultured to an OD₆₀₀ of 0.9 in LB medium; proteins in the periplasm and the supernatant of the cultures were prepared and subjected to Western immunoblotting using antibodies against DegQ_Vh, Fur, and the histidine tag. The result showed that immunoblotting using anti-Fur antibodies produced no positive result; immunoblotting using anti-His antibodies detected MalE only in the periplasmic preparations of DH5α/pAQ1/pMBP and DH5α/pAQ2/pMBP; immunoblotting using anti-DegQ_Vh antibodies detected DegQ_Vh in the periplasmic preparations of all the cultures but only in the supernatants of DH5α/pAQ1/pEF and DH5α/pAQ1/pMBP (data not shown). These results demonstrated that while DH5α harboring pAQ2 could produce and transport the chimeric DegQ_Vh protein into the periplasm, DH5α harboring pAQ1 could secret the chimeric DegQ_Vh protein into the culturing supernatant. Vaccination study showed that compared to vaccination with the purified DegQ_Vh (RPS of 64.9%), vaccination with live DH5α/pAQ1 significantly (P < 0.001) increased the survival rate of fish against T4 (RPS of 94.6%), whereas vaccination with live DH5α harboring pBQ (the plasmid that expresses degQ_Vh) resulted in an RPS of 86.5%. In contrast, immunization with live DH5α harboring pBTA1 (the control vector) had no apparent effect (RPS of <15%), although mortality started 1 day later than that in fish immunized with PBS. Therefore, as a protective immunogen, DegQ_Vh delivered by live DH5α in the secreted chimeric form appeared to be more effective than DegQ_Vh in the periplasm-localized form, and both were more effective than DegQ_Vh administered in the form of purified recombinant protein.

FIG. 6.

Subcellular localization of the wild type and chimeric DegQ_Vh in DH5α. Cells were grown in LB medium to an OD₆₀₀ of ∼1.2 and used for the preparation of the periplasmic, extracellular, and whole-cell proteins. The proteins were electrophoresed in a 0.1% SDS-12% polyacrylamide gel. For proteins from the same cellular compartment, equal volumes were loaded into the gel. After electrophoresis, the proteins were transferred to a nitrocellulose membrane and immunoblotted with mouse anti-His monoclonal antibody.

To determine whether the heightened protection of DH5α/pAQ1 and DH5α/pBQ was the result of prolonged existence of the immunogen, the in vivo survival of the carrier strain was investigated. To this end the peritoneal fluids, guts, spleen, and liver of the fish immunized with DH5α/pAQ1, DH5α/pBQ, and DH5α/pBTA1 were examined at various time points postvaccination for viable plasmid-harboring DH5α. The result showed that plasmid-retaining DH5α cells could be recovered from the peritoneal fluids but not from the guts, spleen, or liver of the immunized fish as late as 7 days postimmunization, with the number of viable cells being highest at the early days and declining with time (Fig. 7). This result suggested that there was a prolonged, though relatively short, duration of the presence of low-level DegQ_Vh-producing DH5α cells in the vaccinated fish, which may have contributed to the enhanced protective effect observed with DH5α/pAQ1 and DH5α/pBQ.

FIG. 7.

Survival of antigen carrier strains in vaccinated fish. The peritoneal fluids of the fish immunized with DH5α carrying pBQ, pAQ1, and pBTA1 were taken at different days postvaccination and examined for viable plasmid-retaining E. coli cells. The data are means for at least three independent assays and are presented as the means ± SE.

DISCUSSION

In E. coli, DegP has two close homologues, DegQ and DegS. While DegS has a function that is distinct from that of DegP and cannot complement mutation in the former, overexpression of degQ can rescue the growth defect resulting from degP deficiency (44). In our study we found that the V. harveyi DegQ protein could complement, to a certain extent, the mutant phenotype of a degP-defective E. coli strain. The fact that complementation by DegQ_Vh required the catalytic site suggested that it was probably in the role of a protease, not of a chaperon, which functions independently of the protease activity (21), that DegQ_Vh acted while surrogating as DegP. This finding implied that there is at least a partial overlap in the targets of the proteases DegQ_Vh and E. coli DegP. Unlike the expression of E. coli degQ, which is not heat inducible (44), the expression of degQ_Vh was modulated by temperature. This observation, together with the facts that DegQ_Vh could operate as a substitute for E. coli DegP and that DegQ_Vh shares a higher sequence identity with E. coli DegP than with E. coli DegQ, placed DegQ_Vh in a position that is closer to that of E. coli DegP than to that of E. coli DegQ. It is known that the hyperthermophilic bacterium Thermotoga maritima has only one HtrA homologue, DegQ, which functions as both a protease and a chaperon (17). A survey of the available genome sequences of the Vibrio species revealed that V. harveyi, V. parahaemolyticus, V. vulnificus, V. cholerae, and V. fischeri all possess only two members of HtrA family proteins, i.e., DegS and the DegQ_Vh counterpart. It is therefore reasonable to speculate that in V. harveyi DegQ_Vh may play the roles that are assumed by DegP and DegQ in E. coli.

Studies of E. coli DegP and DegQ reveal that they exhibit very similar substrate specificities and degrade proteins that are transiently or irreversibly denatured (20). The known substrates for DegP include β-casein (43), MalS (38), PhoA (37), maltose binding protein (4), OmpF (26), and PapA pilin (15), which are naturally or unnaturally misfolded. The protease activity of E. coli DegP manifests only when the temperature is above 20°C and increases rapidly between 37 and 55°C (36, 39). In our study we found that DegQ_Vh could degrade the loosely structured β-casein but not the well-structured BSA or collagen. Like the E. coli DegP protein, DegQ_Vh was proteolytically active over a wide pH range and in the presence of different cations. In line with the observation that it could rescue the temperature-sensitive defect of an E. coli DegP mutant, recombinant DegQ_Vh was most active over the temperature range of 40 to 50°C and was thermostable. As has been observed with the HtrA protein of E. coli, Burkholderia cepacia, and T. maritime (9, 18, 34), the proteolytic activity of DegQ_Vh required the presence of both PDZ1 and PDZ2. The PDZ domains of E. coli DegP have been proposed to be involved in vital enzymatic processes, including temperature-dependent substrate binding and translocation of the bound substrate to the catalytic cavity (5, 34). Recent studies by Kim and Kim (18) demonstrated that PDZ2 of T. maritime HtrA, the only DegQ homologue with the protease domain analyzed at the crystal structure level (17, 18), functions in the formation of the oligomeric protein complex (19). The essentialness of the PDZ domains of DegQ_Vh suggested that they may play a role analogous to that played by the PDZ domains of the E. coli and T. maritime HtrA proteins.

E. coli serving as an expression and delivery vehicle for exogenous antigens has been reported previously by Onate et al. (27) and Andrews et al. (3), who showed that vaccination of mice with live E. coli producing the Cu/Zn superoxide dismutase of Brucella abortus can elicit a strong protective immune response. Similarly, we found that immunization of fish with viable E. coli expressing degQ_Vh induced effective protection. The greater protective effect of the chimeric over the periplasmic DegQ_Vh protein was probably due to the soluble nature of the former, which may increase the accessibility of DegQ_Vh by host immune systems; the carrier protein, AgaV, may also contribute to the protection by playing a structural role, so that DegQ_Vh in the chimeric form was more effectively presented as an immunogen. The inability of DH5α carrying the control vector to induce any protection ruled out the possibility that the protection observed with DH5α/pAQ1 and DH5α/pBQ was due to nonspecific immune responses elicited by the carrier strain, DH5α, or expression of the backbone plasmid components; rather, it suggested that the protection was mediated specifically by DegQ_Vh in its respective form. However, the immune responses elicited by the carrier strain, DH5α, may have facilitated the induction or boosted the scale of the DegQ_Vh-specific immune response, which could in part account for the observation that DegQ_Vh presented by live DH5α cells was a more potent immunogen than DegQ_Vh in the form of a purified subunit vaccine. The nature of the immune responses induced by the purified recombinant DegQ_Vh protein and by DH5α/pAQ1 is currently under study in our laboratory. It is a possibility that compared to vaccination with purified DegQ_Vh, vaccination with a genetically altered form of DegQ_Vh presented by live E. coli cells may lead to alterations in certain aspects of the host immune responses.