Crucian carps (~14.2 g) were obtained from Hunan Yuelu Mountain Science and Technology Co., Ltd.: Changsha, China, which carries out aquatic breeding in Changsha. During the experiment, fish were housed pre-experimentally under controlled conditions (25.25 ± 0.58 °C, pH 7.05 ± 0.12, >7.0 mg/L O₂). Pre-experimental screening confirmed that the fish were disease-free [18]. Tissues were acquired following ethical guidelines (GB/T 35892-2018 [19]). All animal experiments were approved by the Hunan Normal University’s Animal Care Committee (No. 2023109).

To generate the A. hydrophila ΔAmoG mutant, a 300 bp in-frame deletion of AmoG (aa 661–760) was engineered via an overlap extension PCR. The primers AmoG-F1/AmoG-R1 and AmoG-F2/AmoG-R2 were used sequentially for overlap PCRs, followed by a fusion PCR using AmoG-F1/AmoG-R2 (Table S1). The resultant amplicon was ligated into pDM4 at the BglII site, yielding pDMAmoG. Escherichia coli S17-1 λpir harboring pDMAmoG was conjugated with A. hydrophila CCL1. Following coculture and selection on polymyxin B/chloramphenicol plates, sucrose-resistant yet chloramphenicol-sensitive colonies were isolated. Confirmation of in-frame deletion in these isolates was achieved through PCR and sequencing, leading to the identification of the ΔAmoG mutant.

For the creation of the AmoG complement strain, designated ΔAmoG-C, the gene encoding AmoG, along with its native promoter, was PCR-amplified using the primers AmoG-F3 and AmoG-R3 (Table S1). This amplicon was then cloned into the pACYC184 plasmid at its EcoRV restriction site, forming pACYCAmoG-C. Subsequently, this construct was introduced into E. coli S17-1 λpir, which was subsequently mated with the ΔAmoG mutant. Transconjugants were identified on selective LB agar plates containing polymyxin B and chloramphenicol. One selected colony, designated ΔAmoG-C, was further validated by PCR and sequence confirmation.

The strains CCL1, ΔAmoG, and ΔAmoG-C were cultured in LB broth at 28 °C for 24 h. Scanning electron microscopy was employed to assess cell morphology. Growth dynamics were monitored by cultivating the strains in LB with or without 100 μM 2, 2′-dipyridyl (Dp), starting at an A600 of 0.01; samples were taken at various intervals to measure A600. Motility assays and biofilm formation tests were conducted according to Reference [16]. The Chrome Azurol S (CAS) plate assay and Arnow’s test were used to analyze siderophore production according to Reference [20]. Each assay was replicated three times.

For in vivo infection, fish were injected intramuscularly (i.m.) with 100 µL suspensions containing 1 × 10⁶ CFU/mL of CCL1, ΔAmoG, ΔAmoG-C, or an equivalent volume of PBS (control). At 24 h post infection (hpi), distal gut, kidney, spleen, and blood from CCL1-, ΔAmoG-, ΔAmoG-C-, or PBS (control)-treated fish were carefully sampled under sterile conditions for subsequent experimental procedures. (i) Bacterial dissemination in the kidney and spleen and AB-PAS staining were carried out as reported previously [21]. (ii) A qRT-PCR was performed to analyze the genes’ expression in the distal gut. The genes and PCR primers used are listed in Table S1. Expression levels were measured through the 2^−ΔΔCT method, referencing beta-actin, with PCR efficiency (E) and correlation (R²) checked as previously detailed [22]. (iii) Plasma was derived from the blood of each group. D-lactic acid concentrations were measured using a D-lactic acid ELISA assay kit procured from Nanjing Jiancheng Bioengineering Institute, Nanjing, China. Additionally, Limulus Amebocyte Lysate (LAL) QCL-1000 kits from Lonza were employed to quantify lipopolysaccharide (LPS) levels within the plasma. (iv) An immunohistochemistry (IHC) assay was used to analyze the expression of Occludin in the distal gut as reported previously [16]. Mouse anti-rOccludin antibody (prepared in [16]) and Cy3-conjugated goat anti-mouse secondary antibody (Servicebio, Wuhan, China) were used as the primary and secondary antibodies, respectively. (v) Iron concentrations in the serum and distal gut were determined using an iron-unsaturated iron binding capacity kit (Nanjing Jiancheng) according to the manufacturer’s instruction. The DAB-enhanced Prussian blue iron staining kit (Servicebio, Wuhan, China) was utilized to detect iron deposits within liver samples according to the manufacturer’s instructions. (vi) To conduct an apoptosis assessment targeting the distal guts, an apoptosis assay was carried out using the TUNEL Apoptosis Detection Kit (Servicebio). (vii) For assessing survival rates, twenty crucian carps per group were challenged with CCL1, ΔAmoG, ΔAmoG-C, or received a PBS control, following the same inoculation procedure as described above. Over a period of two weeks, fish mortality was meticulously documented post infection as reported previously [21]. Each assay was replicated three times, with three fish used each time.

Transcriptome analysis was performed as outlined in previous work [16]. At 24 hpi, as above, distal guts from fish infected with CCL1, ΔAmoG, ΔAmoG-C, or exposed to PBS (serving as control) were collected under sterile conditions (with four fish in each group). Gut specimens from each group were frozen at −80 °C prior to RNA extraction for subsequent RNA sequencing (RNA-Seq). RNA extraction, library construction, and data analysis were performed by Novogene (Beijing, China).

For a comprehensive analysis of the microbiome shifts in the guts of fish exposed to CCL1, ΔAmoG, ΔAmoG-C, or treated with PBS (control), as above, distal guts from each group were taken and forwarded to Novogene for expert processing and sequencing (with four fish in each group). Following CTAB/SDS DNA extraction, libraries were constructed and sequenced using the 515F and 806R primer pair on an Illumina NovaSeq6000. Rigorous data analysis was performed through QIIME2 as reported previously [16].