Development of a recombinant live attenuated vaccine to protect channel catfish against motile Aeromonas septicemia

Aeromonas hydrophila causes motile Aeromonas septicemia (MAS), which affects farm-raised catfish. Since 2009, outbreaks caused by a clonal genetic group of A. hydrophila (virulent A. hydrophila, or vAh) have caused significant losses of market-size catfish in Alabama and Mississippi. Previously, our group reported that recombinant vAh surface proteins are effective in protecting catfish against MAS. Furthermore, live attenuated Edwardsiella ictaluri vaccine strain ESC-NDKL1 is an efficacious vaccine for enteric septicemia of catfish (ESC). In the current study, we investigated ESC-NDKL1 as a potential vector for expressing vAh antigens. Genes encoding vAh surface antigens were integrated into the ESC-NDKL1 chromosome by conjugation and homologous recombination. Six recombinant ESC-NDKL1 strains expressing one vAh antigen (ESC-NDKL1:: atpase, ESC-NDKL1:: fim, ESC-NDKL1:: fimMrfG, ESC-NDKL1:: fimA, ESC-NDKL1::tdr, and ESC-NDKL1:: ompA) showed significant protection against MAS with relative percent survival (RPS) of 52.10%, 25.67%, 39.99%, 38.93%, 47.83%, and 34.07%, respectively. Eleven recombinant ESC-NDKL1 strains expressing two vAh antigens showed significant protection against MAS and improved protection compared to recombinant ESC-NDKL1 strains expressing one vAh antigen. In particular, four recombinant ESC-NDKL1 strains expressing two vAh antigens (ESC-NDKL1:: atpase:: fimMrfG, ESC-NDKL1:: fim:: fimMrfG, ESC-NDKL1:: tdr::f imMrfG, and ESC-NDKL1:: fim:: ompA) showed significant protection (p < 0.05) compared to non-recombinant ESC-NDKL1 with RPS of 55.72%, 60.18%, 61.74%, and 54.81%, respectively. We identified four promising vaccine candidate strains to control MAS in the catfish industry.

Aeromonas hydrophila is a Gram-negative pathogen that causes motile Aeromonas septicemia (MAS) in many species of freshwater fish (Janda and Abbott, 2010; Fernández-Bravo and Figueras, 2020). It affects farm-raised catfish, the largest aquaculture industry in the U.S. Although it is historically considered a secondary pathogen, outbreaks caused by A. hydrophila as a primary pathogen began impacting U.S. catfish aquaculture in 2009 (Tuttle et al., 2023). Mortality rates in ponds ranged up to 50-60%, and predominantly market-sized fish were impacted (Janda and Abbott, 2010; Fernández-Bravo and Figueras, 2020).

Genomic analysis revealed that the A. hydrophila strains causing MAS in the U.S. catfish industry were in a clonal pathotype categorized as sequence type 251 (ST251), and they were designated virulent A. hydrophila (vAh) (Hossain et al., 2014). The U.S. vAh pathotype was further differentiated into two subclades: one found only in Mississippi catfish farms (MS subclade represented by strain S14-452) and one found in both Mississippi and Alabama catfish farms (AL/MS subclade represented by ML09-119) (Rasmussen-Ivey et al., 2016).

In affected farms, disease caused by vAh in catfish is characterized by acute onset of anorexia followed by high mortality rates. Affected fish have clinical signs typical of a systemic bacterial infection, but the disease also involves hemorrhages in muscles and visceral organs, which makes it readily distinguishable from enteric septicemia of catfish caused by E. ictaluri. In addition, MAS cases often include pan-ophthalmitis and orbital cellulitis, and the brain may be hyperemic with mild acute hemorrhage. Diseased fish often have bloody ascites (Baumgartner et al., 2017).

Edwardsiella ictaluri, a Gram-negative, facultative intracellular bacterium, is the causative agent of enteric septicemia of catfish (ESC), a major disease affecting farmed channel catfish in the United States. Our research team developed a novel live attenuated E. ictaluri vaccine strain, ESC-NDKL1, which has demonstrated effective protection against ESC in catfish fingerlings (Nho et al., 2017). The ESC-NDKL1 strain (ΔgcvPΔsdhCΔfrdA) has in-frame deletions of the gcvP gene (encoding glycine dehydrogenase) and the sdhC and frdA genes (encoding enzymes in the tricarboxylic acid cycle) (Dahal et al., 2014a; Nho et al., 2017).

Options for treatment of MAS and ESC are limited. Only one feed additive antibiotic, oxytetracycline (Terramycin ®), is currently labeled by the U.S. Food and Drug Administration for MAS (Plumb et al., 1995). Potential problems associated with any medicated feed include the cost, variable dosage (due to varying feed consumption), and the potential development of antimicrobial resistance (Okocha et al., 2018; Serwecińska, 2020). Improper or continuous antibiotic use has the potential to increase antimicrobial resistance in A. hydrophila (Fair and Tor, 2014).

Several recombinant surface and extracellular proteins from A. hydrophila have vaccine efficacy. Examples include Omp38, which improved the efficacy of the inactivated A. hydrophila vaccine in Chinese bream (Wang et al., 2013); Aha1 and OmpW, which provided significant protection against A. hydrophila in common carp (Maiti et al., 2012); an extracellular protease, which significantly improved the efficacy of an inactivated A. hydrophila vaccine in mice (Wu et al., 2012); and Omp48, which provided significant protection against A. hydrophila infection in rohu (Labeo rohita) (Khushiramani et al., 2012). Recombinant flagellar protein FlgK also reacted strongly with channel catfish A. hydrophila immune serum, suggesting potential as a vaccine (Yeh and Klesius, 2011). Another recombinant A. hydrophila protein, Omp-G, provided significant protection against A. hydrophila in eels (Guan et al., 2011).

Comparative genomics showed that vAh strains (ST251) possess several surface and extracellular proteins that are unique to this clonal lineage, including major outer membrane protein (OmpA), P pilus assembly protein (Fim), fimbrial protein MrfG (FimMrfG), fimbrial protein (FimA), TonB dependent receptor (Tdr), and ATPase (Abdelhamed et al., 2016, Abdelhamed et al., 2019).

In the current study, we hypothesized that expression of vAh surface proteins in a live attenuated E. ictaluri vaccine (ESC-NDKL1) (Nho et al., 2017) will provide significant protection against MAS. ESC-NDKL1 is known to efficiently colonize mucosal tissues following immersion vaccination, generating strong protective immunity against ESC. By engineering ESC-NDKL1 to express vAh antigens, the vaccine could serve as a biological delivery platform to generate immunity toward two important Gram-negative pathogens simultaneously, enabling dual protection. To stably incorporate vAh protein coding genes in the E. ictaluri vaccine strain, we used a homologous recombination procedure (Heap et al., 2012). A total of 18 recombinant attenuated E. ictaluri strains were constructed, each with chromosomal integration of one or two genes encoding vAh protective antigens under control of the T7lac promoter, and their efficacy to provide protection against MAS was evaluated.

Ethics statement

Experimental infections and vaccine trials using specific pathogen free (SPF) channel catfish fingerlings were performed at Mississippi State University under a protocol approved by the MSU Institutional Animal Care and Use Committee (IACUC-22-338).

Bacterial strains and plasmids

Bacterial strains and plasmids are listed in Table 1. Live attenuated Edwardsiella ictaluri vaccine ESC-NDKL1 and recombinant strains were cultured in brain–heart infusion (BHI) agar and broth (Becton Dickinson, Sparks, MD, USA) and incubated at 30 °C. Virulent Aeromonas hydrophila ML09–119 was cultured in brain–heart infusion (BHI) agar and broth (Becton Dickinson, Sparks, MD, USA). Escherichia coli strains were cultured on lysogeny agar or lysogeny broth (LB) (Becton Dickinson) and incubated at 37 °C. Escherichia coli CC118λ pir was used for propagation of pMEG-375 and derivative plasmids. E. coli BW19851λ pir was used for plasmid transfer into ESC-NDKL1 by conjugation. Ampicillin was used at 100 μg/ml to maintain pMEG-375 and its derivatives, and colistin was used at 12.5 μg/ml for counter-selection against E. coli donor strains following conjugation.

Integration plasmid construction

We constructed two plasmid derivatives from pMEG-375 (Table 1) that allowed for integration of vAh antigen-encoding genes (fim, fimA, fimMrfG, ompA, atpase, and tdr) into the in-frame deleted site of Δ frdA and Δ gcvP genes in the ESC-NDKL1 chromosome. PCR amplicons of approximately 1000 base pairs upstream and downstream of frdA and gcvP genes were amplified using E. ictaluri 93–146 genomic DNA and primers pairs shown in Table 2. PCR reactions were performed in a 50 μL volume containing 1x PCR reaction buffer, 200 μM of dNTP, 0.1 μM of each primer, 1 unit of Taq DNA polymerase (0.3 μL of X U/μL stock), ~100 ng of genomic DNA, and Nuclease-free water to volume. Amplification was carried out using the following conditions: initial denaturation at 95 °C for 5 min; followed by 30 cycles of denaturation at 95 °C for 30 s, annealing at 56 °C for 1 min, and extension at 72 °C for 1 min; with a final extension at 72 °C for 10 min. The amplified upstream fragments of the gcvP and frdA genes were purified and digested with specific restriction endonucleases, and the size of the digested fragments were analyzed on 1% agarose gel using electrophoresis. Each processed gcvP or frdA upstream DNA fragment was ligated into pMEG-375. Ligation mixtures were transformed into chemically competent E. coli CC118λ pir as described (Tang et al., 1994). Positive clones were selected on LB agar plates with 100 µg/ml ampicillin. Candidate plasmids with appropriate cloned fragments were sequenced using primers designed to confirm correct orientation of the insert.

After confirmation, the amplified downstream fragments of the gcvP and frdA genes were digested and ligated to the respective plasmids (pMEG-375:: gcvP and pMEG-375:: frdA). Positive clones were selected on LB agar plates with 100 µg/ml ampicillin. Candidate plasmids with appropriate fragments were sequenced to confirm correct orientation of the insert. The constructed plasmids were designated pMEG frdA and pMEG gcvP.

Cloning of vAh genes with a T7 promotor

The strategy for construction of recombinant strains is shown in Figure 1. PCR amplicons for six vAh genes (fim, fimA, fimMrfG, ompA1, tdr, and atpase) were amplified from pET28a:: fimA, pET28a:: fim, pET28a:: fimMrfG, pET28a:: ompA1, pET28a:: tdr, and pET28a:: atpase plasmids using the T7 promoter and T7 terminator primer pairs. These plasmids were constructed previously by our lab (Abdelhamed et al., 2016, Abdelhamed et al., 2017, Abdelhamed et al., 2019). Each of the amplicons contained a vAh gene under the control of T7lac promoter (Table 2). Because E. ictaluri lacks the lacI repressor gene, the T7lac promoter expresses constitutively in this species (Karsi and Lawrence, 2007). Restriction endonuclease sites were incorporated at the 5’ end of each primer. The six amplified products were purified and digested with appropriate restriction endonucleases (New England Biolabs, Ipswich, MA, USA), and the size of each digested fragment was analyzed on 1% agarose gel using electrophoresis. Each DNA fragment was ligated in both pMEG frdA and pMEG gcvP using T4 ligase (New England Biolabs). In this step, amplicons were ligated between the gcvP or frdA upstream and downstream fragments. Ligation mixtures were transformed into chemically competent E. coli CC118λ pir as described (Tang et al., 1994). Positive clones were selected on LB agar plates with 100 µg/ml ampicillin. Candidate plasmids with appropriately sized inserts were purified using Hi-Speed Mini Plasmid Kit (IBI Scientific) and sent for sequencing to Eurofins genomics using T7 promoter and T7 terminator primers to confirm correct orientation of the insert.

At the end of this step, twelve recombinant plasmids were constructed: pMEG gcvP:: fim, pMEG gcvP:: fimA, pMEG gcvP:: fimMrfG, pMEG gcvP:: ompA1, pMEG gcvP:: tdr, pMEG gcvP:: atpase, pMEG frdA:: fim, pMEG frdA:: fimA, pMEG frdA:: fimMrfG, pMEG frdA:: ompA1, pMEG frdA:: tdr, and pMEG frdA:: atpase. Plasmids were then transformed into chemically competent conjugative E. coli BW19851λ pir as described (Tang et al., 1994). Positive clones were selected on LB agar plates with 100 µg/ml ampicillin.

Construction of recombinant ESC- NDKL1 strains expressing one vAh antigen

For incorporation in the chromosome of ESC-NDKL1, vAh genes were inserted in the strain’s frdA deletion site. Plasmids in conjugative E. coli BW19851λ pir (pMEG frdA:: fim, pMEG frdA:: fimA, pMEG frdA:: fimMrfG, pMEG frdA:: ompA1, pMEG frdA:: tdr, and pMEG frdA:: atpase) were introduced by conjugation into ESC-NDKL1, and positive colonies were selected on BHI plus 100 µg/ml ampicillin and 12.5 µg/ml colistin.

Selection of ESC-NDKL1 recombinant strains was conducted in two steps. In the first step, colonies with the entire plasmid inserted by a single crossover in frdA were selected on BHI agar plates with ampicillin and colistin. In the second step, the single crossover mutants were propagated on LB agar with 5% sucrose and 0.35% mannitol to allow loss of the suicide vector by a second homologous recombination (Abdelhamed et al., 2013). After this, mutant colonies were picked and grown in BHI broth with colistin. At each step, colony PCR was conducted to check for the inserted gene using primer pairs specific to the inserted gene (Table 3) (Abdelhamed et al., 2013). Candidate positive colonies were tested for ampicillin sensitivity to ensure loss of the plasmid from the ESC-NDKL1 chromosome. PCR products amplified from positive clones were sequenced at Eurofins genomics after treatment with ExoSAP-IT (Affymetrix, Santa Clara, CA) using a Big Dye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA) (Abdelhamed et al., 2013). The strains were designated ESC-NDKL1:: atpase, ESC-NDKL1:: fim, ESC-NDKL1:: fimMrfG, ESC-NDKL1:: fimA, ESC-NDKL1::tdr, and ESC-NDKL1:: ompA1. Bacterial stocks were prepared in 20% glycerol and stored at − 80 °C.

Construction of recombinant ESC- NDKL1 strains expressing two vAh antigens

To construct ESC-NDKL1 strains expressing combinations of two vAh antigens, vAh genes were inserted in the gcvP deletion site of ESC-NDKL1:: atpase, ESC-NDKL1:: fim, ESC-NDKL1:: fimMrfG, ESC-NDKL1:: fimA, ESC-NDKL1:: tdr, or ESC-NDKL1:: ompA1. Plasmids in E. coli BW19851λ pir (pMEG gcvP:: fimA, pMEG gcvP:: fim, pMEG gcvP:: fimMrfG, pMEG gcvP:: ompA1, pMEG gcvP:: tdr, and pMEG gcvP:: atpase) were transferred by conjugation using the combinations in Table 2. Positive colonies were selected on BHI plus 100 µg/ml ampicillin and 12.5 µg/ml colistin. The second recombination event was selected as described (Nho et al., 2017). At each step, colony PCR was conducted to check for the inserted gene using primer pairs specific to the inserted gene, as shown in Table 3. Candidate recombinant colonies were tested for ampicillin sensitivity to ensure loss of the plasmid from the ESC-NDKL1 chromosome and confirmed by sequencing.

Confirmation of vAh genes expression from the chromosome of the single and double recombinant ESC-NDKL1 strains

Expression of the inserted vAh genes from the ESC-NDKL1 chromosome was confirmed using qPCR. Recombinant ESC-NDKL1 strains and non-recombinant ESC-NDKL1 were grown at 30 0 C for 24 h. Bacterial pellets were collected, and RNA was extracted using The FastRNA™ SPIN Kit for Microbes with FastPrep-24™ according to the manufacturer’s protocol (MP Biomedicals, Santa Ana, CA). Each column was treated with RNase-Free DNase (QIAGEN, Hilden, Germany) to remove potential genomic DNA contamination of the total RNA. A Nanodrop ND-1000 spectrophotometer (Thermo Scientific, USA) was used for estimating RNA concentration and evaluating the quality (Kordon et al., 2019).

A Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Scientific, USA) was used to transcribe total RNA into cDNA according to the manufacturer’s instructions. Synthesized cDNA reactions were stored at −80 °C. For qPCR, primers were designed by IDT primer design software based on locus tag sequences listed in Table 4 and synthesized commercially (MWG Eurofins Genomics). E. ictaluri 16s ribosomal RNA primers were used for normalization (Kordon et al., 2018). Real-time qPCR was conducted as described (Kordon et al., 2019, Kordon et al., 2021). Six technical replicates were performed for each qPCR assay (two technical replicates for each of three biological replicates). Relative gene expression levels were calculated using the 2^(-ΔΔCt) method. Briefly, the cycle threshold (Ct) values of target genes were normalized to the housekeeping gene (e.g., 16S rRNA), and expression differences between treatment groups were expressed as fold changes relative to the control group.

Western blot

Western blotting was performed to detect expression of the vAh antigens in recombinant ESC-NDKL1 strains. Polyclonal rabbit antibodies specific to the recombinant proteins were produced by Fortis Life Sciences (Boston, MA) using synthetic peptides corresponding to the recombinant proteins. To prepare protein lysates from recombinant strains, bacterial culture was grown to mid-log phase and harvested by centrifugation. Bacteria were lysed using a cell lysis buffer consisting of 50 mM Tris–HCl (pH 8.0), 5% glycerol, 0.5% Triton X-100, 2 mM PMSF, and 1.5 mM EDTA. The resulting suspensions were incubated on ice for 30 min followed by sonication. DNase I (20 μL/mL) was added and followed by incubation at 4 °C for 1 h. The lysate was centrifuged, and the supernatant containing soluble proteins was collected.

Protein samples were mixed with 4× Laemmli sample buffer containing β-mercaptoethanol and heated at 100 °C for 10 min. Prepared samples were then subjected to SDS-PAGE and Western blotting. Total protein concentration was determined using the BCA assay (Thermo Fisher Scientific). Equal amounts of protein (20–30 µg) were loaded per lane and separated on a 12% SDS-PAGE gel. Proteins were transferred to a PVDF membrane using a semi-dry transfer system. Membranes were blocked in 5% non-fat dry milk in TBST (Tris-buffered saline with 0.1% Tween-20) for 1 hour at room temperature and incubated overnight at 4 °C with primary polyclonal antibodies (1:1000 dilution). After washing, membranes were incubated with horse radish peroxidase-conjugated secondary antibodies (1:5000 dilution) for 1 hour at room temperature. Signal detection was performed using enhanced chemiluminescence reagents (Bio-Rad), and images were captured using a ChemiDoc Imaging System (Bio-Rad).

Growth kinetics of recombinant ESC-NDKL1 strains

Overnight cultures of E. ictaluri ESC-NDKL1 and recombinant strains were normalized based on optical density at 600 nm (OD600) and used to inoculate fresh BHI broth at 1:100 dilution in a 96-well plate. Cytation 5 Cell Imaging Multi-Mode Reader (BioTek, Winooski, VT, United States) was used to monitor the growth by measuring OD600 every hour for 24 h. The growth assay was conducted with four technical replicates per strain.

Efficacy of recombinant ESC-NDKL1 strains expressing one vAh antigen to protect catfish fingerlings against MAS

SPF channel catfish fingerlings (11.9 ± 1.4 cm, 34.9 ± 2.85 g) were transferred 20/tank into 20 L tanks supplied with flow-through dechlorinated municipal water. Five replicate tanks were used per recombinant ESC-NDKL1 strain (ESC-NDKL1:: atpase, ESC-NDKL1:: fim, ESC-NDKL1:: fimMrfG, ESC-NDKL1:: fimA, ESC-NDKL1::tdr, and ESC-NDKL1:: ompA1). One group was sham vaccinated, and one group was vaccinated with the non-recombinant ESC-NDKL1 strain. Catfish were vaccinated by immersion in water containing approximately 1 x 10 7 CFU/ml for 1 h at 26°C, followed by gradual removal of bacteria, as described (Dahal et al., 2014b). Sham-vaccinated control fish were immersed in sterile BHI for 1 hour at 26 °C.

Three weeks following vaccination, the water temperature was gradually increased to 30 °C. A total of 10 fish from each group were removed, and serum was obtained to determine antibody response against vAh whole cell lysate. The remaining fish were intraperitoneally infected with vAh strain ML09-119 (approximately 1 × 10 5 CFU/fish) at 30 °C. This challenge dose is routinely confirmed in preliminary trials before each experiment to verify that it produces the expected level of mortality appropriate for the specific fish size and water temperature, ensuring reproducibility and suitability of the challenge conditions. Mortalities were monitored daily until no mortalities were observed for 72 h (Dahal et al., 2014a; Nho et al., 2017).

For statistical analysis, the mean percent mortality was calculated for each treatment and arcsine-transformed. Transformed mean percent mortality data were then analyzed by Student’s t -test to compare treatments. Relative percent survival (RPS) was calculated according to the following formula: RPS = [1 − (% mortality of vaccinated fish/% mortality of non-vaccinated fish)] × 100 (Amend, 1981).

Efficacy of recombinant ESC-NDKL1 strains expressing two vAh antigens to protect catfish fingerlings against MAS

SPF channel catfish fingerlings (12.34 ± 2.17 cm, 38.69 ± 3.51 g) were transferred 20/tank into 20 L tanks supplied with flow-through dechlorinated municipal water. Five replicate tanks were used per recombinant ESC-NDKL1 strain (ESC-NDKL1:: atpase:: ompA1, ESC-NDKL1:: fim:: ompA1, ESC-NDKL1:: fimMrfG::ompA1, ESC-NDKL1:: fimA:: ompA1, ESC-NDKL1::tdr::ompA1, ESC-NDKL1:: atpase:: tdr, ESC-NDKL1:: fim:: tdr, ESC-NDKL1:: fimA:: fimMrfG, ESC-NDKL1:: fim:: tdr, ESC-NDKL1:: atpase:: fimMrfG, ESC-NDKL1:: fim:: mrfG, and ESC-NDKL1:: tdr:: fimMrfG). One group was sham vaccinated, and one group was vaccinated with non-recombinant ESC-NDKL1 strain. Catfish were vaccinated by immersion in water containing approximately 1 x 10 7 CFU/ml of water for 1 h at 26°C, followed by gradual removal of bacteria. Three weeks following vaccination, water temperature was gradually increased to 30 °C. A total of 10 fish from each group was removed, and serum was obtained to determine antibody response against vAh whole cell lysate. The remaining fish were intraperitoneally injected with vAh strain ML09-119 (approximately 1 × 10 5 CFU/fish) at 30 °C. Mortalities were monitored daily until no mortalities were observed for 72 h (Dahal et al., 2014a; Nho et al., 2017) and RPS was calculated (Amend, 1981).

Serum antibody response generated by recombinant ESC-NDKL1 strains against vAh

Analysis of antibody titer in immunized fish sera against vAh whole cell lysate was done using enzyme-linked immunosorbent assay (ELISA). Briefly, three weeks post-vaccination and prior to experimental infection with strain ML09-119, ten fish per group (two fish per tank) were removed for blood collection. Blood was collected from the caudal vein, and serum was obtained (Abdelhamed et al., 2019). Ninety-six well plates were coated overnight at 4 °C with 100 μl per well of heat-killed vAh (10 8 CFU/ml), then wells were washed and blocked with 100 μl/well of 5% nonfat dry milk (Bio-Rad) for 2 h (Abdelhamed et al., 2019). After blocking, plates were washed with PBS, and 100 µL of diluted immunized fish serum (1:100 dilution in PBS) was added to each well and incubated for 1 hour at 37 °C. Wells were washed three times with PBS containing 0.05% Tween-20 (PBS-T), followed by addition of 100 µL of monoclonal antibody 9E1 (anti-catfish Ig) at a 1:2,000 dilution and incubation for 1 hour at 37 °C. After washing, 100 µL of goat anti-mouse IgG-HRP conjugate (Fisher Scientific) at a 1:5,000 dilution was added and incubated for 1 hour at 37 °C (Lobb and Clem, 1982; Miller et al., 1987). Finally, 100 μl of p -nitrophenyl phosphate substrate (Sigma 104 phosphatase substrate) dissolved in 10% diethanolamine buffer was added to each well, and plates were incubated for 45 min at room temperature. Optical density of each well was measured at 405 nm in an ELISA Microplate Reader (CA, USA). Control groups included wells with PBS buffer (in place of serum) and wells with known positive A. hydrophila immune sera. To standardize, the average background absorbance for each plate from negative control was subtracted from treatments.

The effects of vaccine treatments on antibody concentrations (A 405) were analyzed using a mixed-model approach in PROC MIXED (SAS for Windows 9.4). A 405 values were log 10 -transformed prior to analysis to meet model assumptions. Treatment was included as a fixed effect, and block was treated as a random effect. When treatment effects were statistically significant, least squares mean were compared using Dunnett’s adjustment, with either the sham-vaccinated or non-recombinant vaccinated group serving as the reference. Residuals were examined to verify model fit and normality. A significance threshold of p < 0.05 was applied for all analyses.

Confirmation of vAh genes expression from the chromosome of the recombinant ESC-NDKL1 strains

Constructed recombinant ESC-NDKL1 strains are shown in Table 5. Relative gene expression of atpase, fim, fimMrfG, fimA, tdr, and ompA1 in recombinant ESC-NDKL1 strains when inserted in the frdA deletion site (compared to the non-recombinant ESC-NDKL1 strain) was significantly increased (p < 0.05) (Figure 2). Relative expression was also significantly increased when the genes were inserted in the gcvP deletion site (compared to the non-recombinant ESC-NDKL1 strain) (p < 0.05) (Figure 3).

Western blot analysis

Protein bands corresponding to the expected size were observed (Figure 4). P pilus assembly protein FimA (20.6 kDa), fimbrial protein Fim (20.1 kDa), fimbrial protein MrfG (19.6 kDa), major outer membrane protein OmpA1 (37.3 kDa), TonB-dependent receptor (78.6 kDa), and ATPase (81.5 kDa).

Growth kinetics of recombinant ESC-NDKL1 strains

The growth of the single and double recombinant ESC-NDKL1 strains did not differ significantly from non-recombinant ESC-NDKL1 (parent strain) when cultured in BHI broth (Figure 5).

Vaccine efficacy of recombinant live attenuated ESC-NDKL1 strains expressing one vAh antigen

Vaccination of catfish fingerlings with the recombinant live attenuated vaccine strains expressing one vAh antigen caused no mortalities by bath immersion. Vaccinated channel catfish groups using the constructed single recombinant ESC-NDKL1 strains (ESC-NDKL1:: atpase, ESC-NDKL1:: fim, ESC-NDKL1:: fimMrfG, ESC-NDKL1:: fimA, ESC-NDKL1::tdr, and ESC-NDKL1:: ompA1) and non-recombinant ESC-NDKL1 showed significant protection (p < 0.05) compared to sham-vaccinated catfish fingerlings at 21 d post-vaccination after vAh IP injection (Figure 6), with RPS 52.10%, 25.67%, 39.99%, 38.93%, 47.83%, 34.07%, and 34.28%, respectively. No significant difference was detected in protection from single recombinant strains compared to the non-recombinant ESC-NDKL1 strain.

Efficacy of recombinant live attenuated ESC-NDKL1 strains expressing two vAh antigens to protect catfish fingerlings against MAS caused by vAh

Vaccination of catfish fingerlings with the recombinant live attenuated vaccine strains expressing two vAh antigens caused no mortalities. Vaccinated channel catfish treatments using the double recombinant ESC-NDKL1 strains and non-recombinant ESC-NDKL1 showed significant protection against vAh compared to the sham-vaccinated group (p < 0.05) (Figures 7, 8).

Four double recombinant ESC-NDKL1 strains (ESC-NDKL1:: atpase:: fimMrfG, ESC-NDKL1:: fim:: fimMrfG, ESC-NDKL1:: tdr:: fimMrfG, and ESC-NDKL1:: fim:: ompA) showed significant protection (p < 0.05) compared to non-recombinant ESC-NDKL1. These four strains had RPS of 55.72%, 60.18%, 61.74%, and 54.81%, respectively.

Serum antibody response generated by recombinant ESC-NDKL1 strains against vAh

Using plates coated with ML09–119 whole cell lysate, serum antibody titers were higher in fish vaccinated with recombinant ESC-NDKL1 strains expressing vAh antigens compared to the non-recombinant ESC-NDKL1 strain (Figure 9). ESC-NDKL1:: fimMrfG generated the strongest antibody response among the single recombinant strains, and the two double recombinant strains that generated the highest antibody response both included fimMrfG. In general, double recombinant strains generate higher antibody titers than single recombinant strains, but not in all cases.

This study aimed to develop a stable recombinant live attenuated ESC-NDKL1 that acts as a dual vaccine to provide protection against MAS caused by vAh and ESC caused by Edwardsiella ictaluri. This strategy was shown to be effective with an Edwardsiella piscicida strain expressing an Ichthyophthirius multifiliis antigen (Swain et al., 2022). MAS is a devastating bacterial disease that affects catfish aquaculture in the U.S., and there is urgent need to develop a prevention method to control this disease.

Several researchers have identified candidate DNA or recombinant protein vaccines for MAS (Hossain et al., 2014; Abdelhamed et al., 2016, Abdelhamed et al., 2019; Wise et al., 2021; Miryala and Swain, 2025), but to date, there is no commercial vaccine available against MAS caused by vAh. Previously, our group identified three fimbrial proteins (Fim, FimA, and FimMrfG) and three outer membrane proteins (major outer membrane protein OmpA1, TonB-dependent receptor Tdr, and ATPase protein) that stimulate protective immunity in channel catfish fingerlings against vAh infection (Abdelhamed et al., 2016, Abdelhamed et al., 2017, Abdelhamed et al., 2019).

Our group developed live attenuated E. ictaluri vaccine ESC-NDKL1 by in-frame deletion of two genes encoding enzymes in the tricarboxylic acid cycle (sdhC and frdA) and one gene encoding a one-carbon metabolism enzyme (gcvP) (Nho et al., 2017) from wild type E. ictaluri strain 93-146. Laboratory and field challenge study revealed that ESC-NDKL1 is attenuated in catfish fingerlings and provides significant protection from ESC (Nho et al., 2017). We hypothesized that ESC-NDKL1 would be an effective carrier to express vAh antigens and provide protection against MAS.

In the present study, integration of one or two vAh genes encoding Fim, FimA, FimMrfG, ATPase, Tdr, and OmpA1 under the control of the T7lac promoter into the chromosome of live attenuated vaccine ESC-NDKL1 at frdA and gcvP deletion sites was accomplished. The ΔfrdA and ΔgcvP sites were intentionally selected because these loci are already deleted in the ESC-NDKL1 attenuated vaccine (Δ frdAΔgcvPΔsdhC). These genes were previously shown to contribute to attenuation without impairing in vitro growth (Nho et al., 2017). Insertion of antigen cassettes into these nonfunctional deletion sites avoids disruption of other essential genes.

A total of 18 recombinant live attenuated ESC-NDKL1 strains were constructed: six expressing a single recombinant vAh protein by integration of one vAh gene into the in-frame deletion site of frdA, and twelve expressing two vAh antigens constructed by integration of two vAh genes into deletion sites of frdA and gcvP. qPCR and Western blots confirmed expression of vAh genes from the T7 promoter in ESC-NDKL1. The growth kinetics of all single and double recombinants were unchanged compared to parent ESC-NDKL1 (Figure 5), supporting that integration did not affect physiological fitness.

Immersion vaccination of catfish fingerlings with single recombinant live attenuated ESC-NDKL1 (ESC-NDKL1:: atpase, ESC-NDKL1:: fim, ESC-NDKL1:: fimMrfG, ESC-NDKL1:: fimA, ESC-NDKL1:: tdr, and ESC-NDKL1:: ompA1) and non-recombinant ESC-NDKL1 showed significant protection (p < 0.05) against vAh infection compared to sham-vaccinated fish. Interestingly, non-recombinant ESC-NDKL1 provided significant protection against vAh infection, and there was no significant difference in protection comparing single recombinant ESC-NDKL1 strains to the non-recombinant ESC-NDKL1. Several studies found that protection of catfish against bacterial infection appears to rely heavily on both innate and acquired mechanisms of immunity because lymphoid and hematopoietic organs lack the complexity equated to their mammalian counterparts (Akira et al., 2006; Magnadóttir, 2006). Cross-protection between fish pathogens has been documented; for example, vaccination of channel catfish and blue catfish hybrids with live attenuated E. ictaluri gives protection against Edwardsiella piscicida (López-Porras et al., 2022). Live attenuated cold-water disease vaccine cross-protects against an array of Flavobacterium psychrophilum strains through immersion vaccination (Ma et al., 2019). E. piscicida basS and basR mutants induce significant cross-protection against E. ictaluri (Sayed et al., 2022). Cross-protection between these species may be due to the conserved innate immune response of catfish against Gram-negative bacteria, coupled with epitope similarity (López-Porras et al., 2022).

Vaccination of catfish fingerlings with recombinant ESC-NDKL1 expressing two vAh antigens showed significant protection (p < 0.05) compared to the sham-vaccinated group. Four recombinant ESC-NDKL1 strains expressing two vAh antigens (ESC-NDKL1:: atpase:: fimMrfG, ESC-NDKL1:: fim:: fimMrfG, ESC-NDKL1:: tdr:: fimMrfG and ESC-NDKL1:: fim:: ompA) showed significant protection compared to non-recombinant ESC-NDKL1. Thus, the strategy of including more than one vAh antigen in the vaccine carrier strain resulted in improved efficacy. The four antigens expressed in these double recombinant vaccine strains include FimMrfG, Fim, Atpase, and Tdr. Our group has previously determined potential of these recombinant proteins as vaccine antigens and found that catfish vaccinated with FimMrfG, Fim, ATPase, and Tdr had RPS of 85.77%, 95.41%, 89.16%, and 95.59% compared to sham-vaccinated catfish (Abdelhamed et al., 2017).

Vaccine combinations that included fimMrfG showed the best protection, and three of the most effective double recombinant vaccines included this antigen. FimMrfG is a fimbrial protein, which is often an important virulence factor in a wide range of pathogenic bacteria (Sauer et al., 2004). Fimbriae contribute to invasion of host tissue, phage binding, DNA transfer, biofilm formation, cell aggregation, and twitching motility (Doughty et al., 2000; Kline et al., 2009; Proft, 2009). Because they are located on the bacterial surface, they are often antigenic and are important targets for vaccine development against several bacterial diseases.

Interestingly, another fimbrial protein encoded by the fim gene was also one of the more effective antigens in the double recombinants, and it was in two of the most effective double recombinant strains. Often pathogenic bacteria express more than one fimbrial type (Mobley et al., 1993; Snyder et al., 2005). The specific roles of these two fimbrial types in vAh are not known, but it is interesting to note that all four of the most effective double recombinant vaccine candidates included at least one fimbrial antigen, and one of them included both FimMrfG and Fim.

TonB-dependent receptor has been used as an immunogen in other bacterial vaccines. For example, Tdr provided protection in Japanese flounder (Paralichthys olivaceous) against Pseudomonas fluorescens infection with 80.6% RPS (Hu et al., 2012). Tdr was shown to induce bactericidal antibodies in mice against Neisseria meningitides (Stork et al., 2010). The function of the vAh ATPase protein used as an antigen in the current study is not known, but it is predicted to be an outer membrane protein.

OmpA proteins are among the most immunodominant antigens in the outer membrane family (Singh and Williams, 2003). In common carp, recombinant OmpA protein shows high antibody production against Edwardsiella tarda infection (Maiti et al., 2011). Additionally, rainbow trout immunized with OmpA purified from Flavobacterium psychrophilum revealed a high antibody titer. Studies suggest that OmpA proteins could be used as immunogens to protect common carp against E. tarda with higher survival rate (60%) compared to non-vaccinated fish (Maiti et al., 2011). Vaccination of catfish with recombinant vAh OmpA protein showed significant protection against MAS with RPS of 98.59% (Abdelhamed et al., 2017).

The serum antibody response generated by the recombinant ESC-NDKL1 strains varied, but it is noteworthy that expression of one or two vAh antigens generated a serum antibody response that was detectable using plates coated with whole bacterial cell lysate. It is also noteworthy that two antigens generally resulted in a higher catfish antibody response. Antibody titers do not always correlate with protection, including in fish (Ellis, 1999), but our results clearly indicate that expression of vAh antigens in the ESC-NDKL1 carrier generated antibody response. It is likely that antibody response contributes to the protective effect of the constructed recombinant vaccines, but cell-mediated immunity and innate immune components (complement, lysozyme, antimicrobial peptides, or acute phase proteins) are also probably important contributors to protective immunity against vAh (Peatman et al., 2007; Dash et al., 2014; Luo et al., 2016).

In conclusion, we identified four recombinant double live attenuated ESC-NDKL1 strains expressing vAh antigens that demonstrate strong potential as dual-purpose vaccine candidates capable of protecting channel catfish against both motile Aeromonas septicemia (MAS) and enteric septicemia of catfish (ESC). Developing a single E. ictaluri –based live attenuated vaccine that simultaneously delivers vAh antigens offers clear practical advantages: it reduces handling stress, streamlines vaccination procedures, and decreases labor and production costs for farmers who routinely face both pathogens within the same production cycle. This dual-protection strategy is therefore not only biologically effective but also economically attractive for commercial aquaculture. Continued research aimed at optimizing vaccine delivery, including the evaluation of oral administration, will further enhance the feasibility and impact of this approach, supporting industry adoption.