Phenotype, Virulence and Immunogenicity of Edwardsiella piscicida Cyclic AMP Receptor Protein (Crp) Mutants in Catﬁsh Host.

Edwardsiella piscicida, a facultative aerobic pathogen belonging to the Enterobacteriaceae family, is the etiological agent of edwardsiellosis that causes significant economic loses in the aquaculture industry. cAMP receptor protein (CRP) is one of the most important transcriptional regulators, which can regulate large quantities of operons in different bacteria. Here we characterize the crp gene and report the effect of a crp deletion in E. piscicida. The crp-deficient mutant lost the capacity to utilize maltose, and showed significantly reduced motility due to the lack of flagella synthesis. We further constructed a ΔPcrp mutant to support that the phenotype above was caused by the crp deletion. Evidence obtained in fish serum killing assay and competitive infection assay strongly indicated that the inactivation of crp impaired the ability of E. piscicida to evade host immune clearance. More importantly, the virulence of the crp mutant was attenuated in both zebrafish and channel catfish, with reductions in mortality rates. In the end, we found that crp mutant could confer immune protection against E. piscicida infection to zebrafish and channel catfish, indicating its potential as a live attenuated vaccine.


Introduction
Edwardsiella piscicida is a facultative anaerobe and Gram-negative enteric pathogen that generally causes lethal edwardsiellosis, which is a systematic enterohemorrhagic septicemic disease that can lead to high stock mortality in fresh and marine fish [1,2]. E. piscicida infection induces symptoms including emphysematous putrefactive disease with swelling skin lesions, as well as gill ulceration and necrosis in internal organs such as kidney, liver, spleen, and musculature [2,3].
To date, antimicrobial susceptibility testing shows that E. piscicida strains isolated from different hosts and geographical regions are susceptible to most commonly used antibiotics for the treatment of edwardsiellosis [4]. However antibiotic abuse represents a problematic method of treating bacterial infections in the aquaculture industry and has led to the evolution of multi-drug resistance strains [5].

Ethics Statement
All animal experimental work was approved by the Laboratory Animal Monitoring Committee of Huazhong Agricultural University. Procedures conducted involving in animals were in accordance with the suggestions of the Guide for the Care and Use of Laboratory Animals of Hubei Province, China.

Bacterial Strains and Growth Conditions
The bacterial strains used in the study are listed in Table 1. E. piscicida was routinely cultured in Luria broth (LB) or on Luria agar (LA) plates (Difco, Detroit, MI, USA) at 28 • C. When necessary, 0.2% arabinose was supplemented to activate the P BAD . Escherichia coli χ7213 was used for mutant plasmid harvest and was cultured in LB broth at 37 • C, supplemented with diaminopimelic acid (50 g/L) (Sigma, St. Louis, MO, USA). When necessary, ampicillin (Sigma, St. Louis, MO, USA) and chloramphenicol (Sigma, St. Louis, MO, USA) were supplemented at final concentrations of 100 and 50 µg/mL, respectively.

Sequence Analysis
The crp nucleic acid and Crp amino acid sequences from various bacteria species were downloaded from NCBI. The accession numbers were provided in the supplementary data. MAFFT 7 and ESPript 3 were used for multiple sequence alignment. The neighbor-joining method in Molecular Evolutionary Genetics Analysis package (MEGA 7.0) was used to construct a phylogenetic tree. The evolutionary distances were computed using the p-distance model. The E. piscicida Crp 3D structure was predicted using the Iterative Threading ASSEmbly Refinement (I-TASSER) server [21].

Construction and Characterization of Crp Mutants
Primers and plasmids used in this experiment are listed in Tables 1 and 2. The procedures for mutant construction were described previously [22]. The Primers P1/P2 and P3/P4 were used in amplifications of upstream or downstream flanking fragments of crp via PCR respectively. Then, the upstream and downstream homologous arms were fused by overlap PCR using primers P1/P4, and ligated into the suicide plasmid pRE112 at the XbaI/SacI sites. The resulting plasmid pRE112-crp was transformed into E. coli χ7213 for mobilization into E. piscicida via conjugation. Strains containing single-crossover plasmid insertions were isolated on LA media containing chloramphenicol. Loss of the suicide vector after the second recombination between homologous regions was selected by using the sacB-based sucrose-sensitivity counter-selection system. Resulting strain was selected on LA plates containing chloramphenicol. PCR identification via primers P1/P4 and direct DNA sequencing of the mutation sites using genomic DNA preparations were conducted to verify that the construction of the crp gene deletion strain was completed.  We further constructed ∆P crp by replacing the promoter of crp gene with the arabinose-regulated araC P BAD promoter. The procedures were described previously [23]. A 557-bp DNA fragment containing the region upstream of the crp promoter was PCR amplified as a template with primers P5/P6 ( Table 2). The PCR-amplified fragment was digested with BglII/HindIII and cloned into the vector pYA3700 (Table 1). A 595-bp PCR fragment, including 583-bp of crp gene and its original Shine-Dalgarno sequence was amplified using primers P7/P8 ( Table 2). The PCR fragment was digested with XhoI/KpnI and inserted into the intermediate plasmid described above. The resulting construct was confirmed by DNA sequence analysis. Then, a 2.5-kb DNA fragment including araC P BAD , P crp 5' and 3' flanking regions were excised from the plasmid by digestion with KpnI/XmaI and inserted into pRE112, resulting in plasmid pRE112-∆P crp . The following procedure was the same as that in ∆crp mutant construction.
To determine the growth kinetics of different strains, 1:100 diluted overnight cultures were cultured in LB medium at 28 • C. Samples were taken hourly, and the optical densities were measured at 600 nm In order to study whether CRP is involved in the process of maltose utilization, the growth of crp mutant strains and parent strain were observed on MacConkey agar supplemented with maltose (1%). Swimming motility was measured on LA plates that were prepared with 0.3% (w/v) agar according to the method described previously [20]. To increase flagella synthesis, the bacterial samples were collected from motility agar plates. Negative staining method was used to observe the bacteria morphology under transmission electron microscopy (TEM; Hitachi H-7000FA, Tokyo, Japan). Ten microliters of the resuspension of each strain was applied to Formvar-coated copper grids and negatively stained for 1 min with 1% uranyl acetate [20].

Resistance Against Host Clearance of E. piscicida
To investigate resistance of the parent J118 strain and ∆crp against serum killing, a catfish serum survival assays was accomplished as previously described [24]. Blood was withdrawn from the caudal vein of channel catfish (average weight, 500 g) by sterile injector and placed on ice immediately. The blood was allowed for clotting over night at 4 • C. For serum survival assay, the serum was treated with or without heating at 56 • C for 1h. A mid-log growth phase inoculum of 5.0 × 10 5 CFU bacteria in 100 µL was mixed with 400 µL serum. The mixtures were incubated at 28 • C for 1h. Then, the numbers of CFU were determined by serial dilutions and plating. The survival percentage was subsequently calculated as follows: (CFUs after co-incubation /CFUs in PBS control group) × 100%. E. coli DH5α strain was used as a sensitive control strain in this assay. The experiments were repeated three times independently.
To determine the persistent carrier state of J118 and ∆crp strains in host, the competitive assay was performed as previously described [22]. J118 and ∆crp strains grown to mid-exponential phase in LB were harvested by centrifugation, washed and resuspended in PBS to a bacterial density of 5.0 × 10 6 CFU/mL, respectively. In three independent assays, a total of eleven channel catfish were injected by the intracoelomic (i.c.) route with 200 µL volume of 1:1 ∆crp mutant -J118 mixture (1.0 × 10 6 CFU/fish). Twenty-four hours later, fish were euthanized and the blood, ascites, liver, spleen, trunk kidney and head kidney samples were collected. Homogenates were serially diluted and plated on MacConkey agar with maltose (1%) at 37 • C for 24 h. The parent strain is red and the ∆crp mutant is white on the plate, respectively. Bacterial counts titers were calculated by dividing the weights of the tissues from the bacteria loads in the samples.

Virulence of E. piscicida ∆crp in Zebrafish and Catfish Fingerlings
The zebrafish (average weight, 300 mg) used in this study were from the Institute of Hydrobiology, Chinese Academy of Sciences (Wuhan, China). The zebrafish were acclimatized for 2 weeks before injection. They were fed with commercial feed for aquatic animal twice per day under natural photoperiod. The water temperature was maintained at 24-26 • C during cultivation. Zebrafish are divided into 12 groups, each group of 10, infected with J118 strain and ∆crp respectively (doses: 10 2 CFU/fish, 10 3 CFU/fish, 10 4 CFU/fish, 10 5 CFU/fish, 10 6 CFU/fish, and 10 7 CFU/fish). Chanel catfish (average weight, 800 mg) from a local aquaculture farm were acclimatized and monitored for 2 weeks before infection. Catfish were fed daily with commercially produced food pellets (Haida, China) under natural photoperiod. Water temperature was maintained at 23-25 • C. Catfish are divided into 10 groups, each group of 10, i.c. infected with J118 strain and the ∆crp mutant, respectively (doses: 10 4 CFU/fish, 10 5 CFU/fish, 10 6 CFU/fish, 10 7 CFU/fish, and 10 8 CFU/fish). The clinical symptoms and mortality of infected fish were observed for 14 days. The LD 50 was calculated by Karber's methods [25].

Immune Protection of E. piscicida ∆crp Mutant in Zebrafish and Catfish Fingerlings
We further evaluated the immune protection mediated by ∆crp in zebrafish and catfish fingerlings. Fish were divided into immunized group and control group, with 10 zebrafish in each group and 20 catfish fingerling in each group. Zebrafish and catfish were i.c. immunized with 10 3 CFU/fish and 10 4 CFU/fish of ∆crp, respectively. Subsequent booster immunization was given at day 14 post-primary immunization. At 2 weeks post-booster immunization, zebrafish and catfish were challenged i.c. with 10-times LD 50 E. piscicida J118. The fish were observed for clinical symptoms and mortalities for 14 days.

Statistical analysis
Statistical analysis was performed by GraphPad Prism 6 (Graph Pad Software, Inc., San Diego, CA, USA). Survival data were analyzed with the log-rank (Mantel-Cox) test. The statistical p values were calculated by the two-tailed Mann-Whitney t test. Differences were considered significant at p < 0.05 and highly significant at p < 0.01.

Sequence Analysis
The catabolic and virulence regulator crp gene is wide spread not only between enteric pathogens, including E. piscicida, but also within other bacterial domains ( Figure 1). The phylogenetic tree clearly showed crp genes in Enterobacteriaceae, Aeromonadaceae, Vibrionaceae and Pasteurellaceae, belonging to γ-Proteobacteria, were clustered in clade A. The closely related clade B includes other Proteobacteria, while the clade C includes Cyanobacteria, Actinobacteria, Bacteroidetes and Firmicutes, with a more distantly evolutionary relationship. Especially, crp genes in enteric fish pathogens including two Aeromonas species and two Edwardsiella species showed close genetic relationship according to this phylogenetic tree.
Sequence and structural alignment between functional representative bacterial Crp proteins revealed that 189 amino acid residues (90%) are strictly conserved out of 210 residues in E. piscicida Crp ( Figure 2A). E. piscicida Crp has 100.0%, 99.0%, 98.6%, and 89.0% amino acid similarity to the Crp of Escherichia coli, E. ictaluri, Yersinia pestis and Aeromonas hydrophila, respectively. The conserved cAMP binding amino acid residues: Gly72, Glu73, Arg83, Ser84, Thr128, and Ser129 ( Figure 2A) of E. piscicida Crp is similar to other Gram-negative Crp-family members. The E. piscicida Crp 3-D predicted structure exhibits the flexible hinge required for Crp dimerization and the F-helix & C-helix that interacts with the DNA ( Figure 2B). Thus, above results shows that Crp is conserved and indicates its function is conserved through the evolution of the Enterobacteriaceae family, including enteric fish pathogens.

Construction and Characterization of Crp Mutants
To investigate the role of Crp in the pathogenesis of E. piscicida, a crp deletion mutant was constructed by allelic replacement in parent strain J118 ( Figure 3A). The growth of E. piscicida strains was quantified over time by measuring the OD 600nm and CFUs in cultures ( Figure 3B). Although the OD 600nm growth curve showed growth defect of ∆crp in LB broth, the CFUs showed no differences between ∆crp mutant strain and parent strain, which indicated that the mutation of crp affected the secretion of bacterial extracellular products. The phenotype by growth on MacConkey agar supplemented with maltose (1%) shows that the ∆crp mutant does not utilize maltose, indicating that the genes related to maltose utilization is under the positive regulation by Crp ( Figure 4A). Except for involvement in catabolic functions, the Crp regulatory complex is also necessary for the flagella synthesis. We evaluated the motility of the ∆crp mutant in LB broth supplemented with 0.3% agar and flagella synthesis by TEM. The results showed that the ∆crp mutant loses motility ( Figure 4B) due to the lack of flagella synthesis ( Figure 4C), indicating that Crp positively regulates flagella synthesis in E. piscicida.

Sequence Analysis
The catabolic and virulence regulator crp gene is wide spread not only between enteric pathogens, including E. piscicida, but also within other bacterial domains (Figure 1). The phylogenetic tree clearly showed crp genes in Enterobacteriaceae, Aeromonadaceae, Vibrionaceae and Pasteurellaceae, belonging to γ-Proteobacteria, were clustered in clade A. The closely related clade B includes other Proteobacteria, while the clade C includes Cyanobacteria, Actinobacteria, Bacteroidetes and Firmicutes, with a more distantly evolutionary relationship. Especially, crp genes in enteric fish pathogens including two Aeromonas species and two Edwardsiella species showed close genetic relationship according to this phylogenetic tree. Sequence and structural alignment between functional representative bacterial Crp proteins revealed that 189 amino acid residues (90%) are strictly conserved out of 210 residues in E. piscicida Crp (Figure 2A). E. piscicida Crp has 100.0%, 99.0%, 98.6%, and 89.0% amino acid similarity to the Crp of Escherichia coli, E. ictaluri, Yersinia pestis and Aeromonas hydrophila, respectively. The conserved  Figure 2B). Thus, above results shows that Crp is conserved and indicates its function is conserved through the evolution of the Enterobacteriaceae family, including enteric fish pathogens.

Construction and Characterization of Crp Mutants
To investigate the role of Crp in the pathogenesis of E. piscicida, a crp deletion mutant was constructed by allelic replacement in parent strain J118 ( Figure 3A). The growth of E. piscicida strains was quantified over time by measuring the OD600nm and CFUs in cultures ( Figure 3B). Although the OD600nm growth curve showed growth defect of Δcrp in LB broth, the CFUs showed no differences     We further constructed a ∆P crp mutant by replacing the promoter of crp gene with the arabinose-regulated araC P BAD promoter ( Figure 3A). The transcription of crp is dependent upon arabinose availability. The ∆P crp showed similar phenotypes with ∆crp mutant in the absence of arabinose ( Figure 3B, Figure 4A,B). However, when supplemented with arabinose, the defects in growth, maltose utilization and motility were all well complemented ( Figure 3B, Figure 4A,B). Above results verify that crp mutation, instead of other random mutation, contributes to the phenotypes differences.

Resistance Against Host Clearance of E. piscicida ∆crp
To investigate the role of crp in E. piscicida resistance against innate immune responses, we first examined the abilities of the strains to survive in a channel catfish serum survival assay. When E. piscicida J118 strain was incubated with catfish serum, the survival rate of the bacteria was found to be 104.3%, revealing that parent strain resisted the bactericidal effect of catfish serum. ∆crp mutant had showed a significantly reduced survival rate (40.0%) compared with the parent strain ( Figure 5A). The survival rate of E. coli DH5α, a serum-sensitive laboratory strain incubated under the same condition, was 0%. When incubated with inactivated serum, the ∆crp mutant showed an increased survival rate of 71.6%, but still lower than parent strain. Above results suggested that the crp mutant could poorly resist the killing by bactericidal substances in fish serum, and then could not effectively evade the host clearance. We further constructed a ΔPcrp mutant by replacing the promoter of crp gene with the arabinoseregulated araC PBAD promoter ( Figure 3A). The transcription of crp is dependent upon arabinose availability. The ΔPcrp showed similar phenotypes with Δcrp mutant in the absence of arabinose ( Figures 3B, 4A, 4B). However, when supplemented with arabinose, the defects in growth, maltose utilization and motility were all well complemented ( Figures 3B, 4A, 4B). Above results verify that crp mutation, instead of other random mutation, contributes to the phenotypes differences.

Resistance Against Host Clearance of E. piscicida Δcrp
To investigate the role of crp in E. piscicida resistance against innate immune responses, we first examined the abilities of the strains to survive in a channel catfish serum survival assay. When E. piscicida J118 strain was incubated with catfish serum, the survival rate of the bacteria was found to be 104.3%, revealing that parent strain resisted the bactericidal effect of catfish serum. Δcrp mutant had showed a significantly reduced survival rate (40.0%) compared with the parent strain ( Figure  5A). The survival rate of E. coli DH5α, a serum-sensitive laboratory strain incubated under the same condition, was 0%. When incubated with inactivated serum, the Δcrp mutant showed an increased survival rate of 71.6%, but still lower than parent strain. Above results suggested that the crp mutant could poorly resist the killing by bactericidal substances in fish serum, and then could not effectively evade the host clearance.  To test this idea, we performed a competitive infection assay. An approximately 1:1 mixture of J118 and ∆crp bacteria was inoculated by the intracoelomic (i.c.) route into channel catfish. Twenty-four hours later, blood, ascites, liver, spleen, trunk kidney and head kidney samples were collected after the fish were euthanized. Comparing to the parent strain, ∆crp colonized at significantly lower levels in all the samples ( Figure 5B), which confirmed that the crp mutation impaired the resistance ability of E. piscicida against innate immune clearance in vivo.

Virulence of E. piscicida ∆crp in Zebrafish and Catfish Fingerling
The ideal live attenuated bacterial vaccine should be totally attenuated. Zebrafish have been developed as an easy and powerful model to test pathogenesis of E. piscicida. The LD 50 of the parent strain J118 was 7.94 × 10 3 CFU, while the LD 50 of the ∆crp mutant increased up to 3.16 × 10 5 CFU, nearly 40-fold compared to the parental strain (Table 3). For the zebrafish infected with J118, the death occurred immediately during the first 24 h, while it was quite delayed in the ∆crp infection group ( Figure 6A). The E. piscicida ∆crp mutant was further evaluated in the channel catfish host. We found that the ∆crp mutant applied by the i.c. route was attenuated with an estimated 10-times LD 50 increase over the parent strain (Table 4). Catfish fingerlings that were infected with J118 developed the typical symptoms of enteric septicemia, skin lesions and distended abdomen, while the ones infected with the ∆crp mutant exhibited mild symptoms ( Figure 6B). Taken together, these results indicate that Crp significantly contributed to the virulence of E. piscicida.

Immune Protection of E. piscicida ∆crp Mutant in Zebrafish and Catfish Fingerling
Previous results indicate that the E. piscicida ∆crp mutant had reduced virulence and that the host is able to control the infection. We further evaluated the immune protection mediated by the ∆crp mutant in zebrafish and catfish fingerlings. At 2 weeks post-booster immunization, zebrafish were challenged i.c. with 10-times LD 50 E. piscicida J118. The ∆crp mutant applied by the i.c. route was immune protective in the zebrafish host, with 70% survival while only 10% survival in the PBS control group ( Figure 7A). A challenge study was also performed to determine whether the ∆crp mutant induces protective immunity in catfish fingerlings. At 2 weeks post-booster immunization, catfish fingerlings were challenged i.c. with 2-times LD 50 E. piscicida J118. The ∆crp mutant showed a 90% survival rate against J118 infection, which is significantly higher than that of PBS ( Figure 7B).

Immune Protection of E. piscicida Δcrp Mutant in Zebrafish and Catfish Fingerling
Previous results indicate that the E. piscicida Δcrp mutant had reduced virulence and that the host is able to control the infection. We further evaluated the immune protection mediated by the Δcrp mutant in zebrafish and catfish fingerlings. At 2 weeks post-booster immunization, zebrafish were challenged i.c. with 10-times LD50 E. piscicida J118. The Δcrp mutant applied by the i.c. route was immune protective in the zebrafish host, with 70% survival while only 10% survival in the PBS control group ( Figure 7A). A challenge study was also performed to determine whether the Δcrp mutant induces protective immunity in catfish fingerlings. At 2 weeks post-booster immunization, catfish fingerlings were challenged i.c. with 2-times LD50 E. piscicida J118. The Δcrp mutant showed a 90% survival rate against J118 infection, which is significantly higher than that of PBS ( Figure 7B).

Discussion
cAMP receptor protein (CRP; also known as catabolite activator protein, CAP) is a regulatory protein in bacteria. The binding with cAMP causes a conformation change that allows CRP to bind tightly to the DNA site in the promoters of the genes it controls and interact with RNA polymerase [26,27]. Crp is widely distributed among the bacteria and relatively conserved, indicating its regulation of similar gene families related to carbohydrate metabolism [10,28], competence, growth, and virulence determinants [20,[29][30][31].
E. piscicida is a facultative aerobic pathogen belonging to the Enterobacteriaceae family, with a wide ecological niche and host range including various species of fish [1]. In this study, a crp gene deletion mutant (Δcrp) and a crp promoter mutant (ΔPcrp) of E. piscicida were constructed in order to investigate the role of the crp gene in E. piscicida physiology, virulence and ability to confer immune protection to fish hosts. Replacement of the promoter of crp gene by the arabinose-regulated araC PBAD promoter, yielding ΔPcrp in which crp transcription was arabinose dependent. In this study, ΔPcrp was used as complement strain when the arabinose was added. Firstly, the Δcrp mutant exhibited impaired growth in vitro. The growth delay in crp mutants seem common in diverse bacteria species, relating to its role as a global regulator [9]. Secondly, lacking of Crp, E. piscicida could not utilize maltose, which means that Crp does regulate the genes related to carbohydrate utilization and influence metabolism(s) in E. piscicida. In E. coli, the cAMP-CRP complex is required for malT expression, the transcriptional activator of five transcriptional units related to the maltose regulon [32]. The regulation mechanism between Crp and maltose operon in E. piscicida still needs further investigation. Thirdly, the Δcrp mutant also loses motility which is conferred by flagella of bacterial pathogens [33]. Motility is critical for cell invasion and is associated with chemotaxis and secretion of effector proteins that act as virulence factors [34,35].

Discussion
cAMP receptor protein (CRP; also known as catabolite activator protein, CAP) is a regulatory protein in bacteria. The binding with cAMP causes a conformation change that allows CRP to bind tightly to the DNA site in the promoters of the genes it controls and interact with RNA polymerase [26,27]. Crp is widely distributed among the bacteria and relatively conserved, indicating its regulation of similar gene families related to carbohydrate metabolism [10,28], competence, growth, and virulence determinants [20,[29][30][31].
E. piscicida is a facultative aerobic pathogen belonging to the Enterobacteriaceae family, with a wide ecological niche and host range including various species of fish [1]. In this study, a crp gene deletion mutant (∆crp) and a crp promoter mutant (∆P crp ) of E. piscicida were constructed in order to investigate the role of the crp gene in E. piscicida physiology, virulence and ability to confer immune protection to fish hosts. Replacement of the promoter of crp gene by the arabinose-regulated araC P BAD promoter, yielding ∆P crp in which crp transcription was arabinose dependent. In this study, ∆P crp was used as complement strain when the arabinose was added. Firstly, the ∆crp mutant exhibited impaired growth in vitro. The growth delay in crp mutants seem common in diverse bacteria species, relating to its role as a global regulator [9]. Secondly, lacking of Crp, E. piscicida could not utilize maltose, which means that Crp does regulate the genes related to carbohydrate utilization and influence metabolism(s) in E. piscicida. In E. coli, the cAMP-CRP complex is required for malT expression, the transcriptional activator of five transcriptional units related to the maltose regulon [32]. The regulation mechanism between Crp and maltose operon in E. piscicida still needs further investigation. Thirdly, the ∆crp mutant also loses motility which is conferred by flagella of bacterial pathogens [33]. Motility is critical for cell invasion and is associated with chemotaxis and secretion of effector proteins that act as virulence factors [34,35].
Considering this, we conducted a competitive infection assay and a fish serum suvival assay to compare systemic infection capacity of ∆crp and the parent strains. The ∆crp mutant showed poor colonization in different tissues and a weakened survival rate against catfish serum, indicating its attenuated evasion capacity against host immune clearance. In further virulence assessment assays, the LD 50 s of the ∆crp mutant in zebrafish and channel catfish increased by 40-fold and 10-fold, respectively, comparing to the parent strain, and ∆crp mutant infection group exhibited a delayed death (in the zebrafish) and mild symptoms (in the channel catfish). This was consistent with previous study that E. piscicida CK216, a ∆crp mutant strain exhibits 1000-fold attenuation comparing to wild type strain in gold fish [36]. Indeed, recent unpublished work in our laboratory indicates that ∆crp mutant could evoke host controllable inflammation which could effectively contribute to bacterial clearance, while J118 strain infection evoked excessive inflammatory cytokine production in the susceptible organ liver and trunk kidney, leading to severe tissue injury. Therefore, the above results support that Crp contributes to the virulence of E. piscicida and that the ∆crp mutant is attenuated.
Being a mucosal facultative intracellular pathogen, E. piscicida is an ideal candidate to develop a live attenuated vaccine for the aquaculture industry, which could elicit both humoral and cellular immune responses [37]. We further evaluated the immune protection mediated by the ∆crp mutant in zebrafish and catfish fingerlings. And as expected, the ∆crp mutant group had a significantly higher survival rate than the PBS control group. We determined that the ∆crp mutant could elicit host immune responses and offer moderate protection in zebrafish and catfish. Thence, the introduction of a ∆crp mutation could be considered as part of genetic plan to construct an E. piscicida live attenuated vaccine in further studies.

Conclusions
In summary, we conclude that Crp is involved in the regulation of maltose utilization and is required for flagella synthesis in E. piscicida. More importantly, Crp is also required for E. piscicida resistance against host immune elimination and that it contributes to the systemic dissemination and virulence of this bacterium. Finally, the ∆crp mutant could be a promising vaccine candidate for further genetic modification to control the E. piscicida infections.