The caspase gene family in Micropterus salmoides: genome-wide characterization, bioinformatics, and expression profiling analysis

Abstract

Background

Apoptosis, a programmed cell death process mediated by caspases, is crucial for immunity and development. Caspases are categorized into initiator and effector types, regulating apoptosis via intrinsic and extrinsic pathways, and are also involved in pyroptosis and other cellular processes. In fish, diverse caspase genes have been identified and shown to participate in embryonic development and immune responses against bacterial infection. However, their role in largemouth bass against Aeromonas hydrophila—a pathogen causing severe intestinal inflammation and septicemia—remains unclear. This study identifies caspase genes in largemouth bass and investigates their expression and evolution post-infection, providing insights into their immune functions.

Results

This study presents the first comprehensive genomic analysis of caspases in largemouth bass, identifying 19 caspase genes: 3 inflammatory (caspase-1 orthologs), 6 effector (caspase-3/6/7), 8 initiator (caspase-2/8/9/10/20), and 2 unclassified (caspase-17) caspases. Functional and evolutionary analyses demonstrated a high level of homology between largemouth bass caspases and those of other teleosts. The sequence diversity analysis revealed that caspase genes in largemouth bass are subject to extensive mutational changes, predominantly through nonsynonymous single nucleotide variants, which may play a role in immune adaptation and resistance to disease. Alternative splicing analysis revealed multiple splicing sites and events in caspases of largemouth bass. Differential alternative splicing analysis showed significant induction of MsCASP8_1 variants after infection with Aeromonas veronii. Gene structure and sequence alignment analyses demonstrated that these isoforms result from alternative splicing events involving the 5th exon. Transcriptome profiling and qRT-PCR analyses revealed that most caspase genes in largemouth bass are responsive to Aeromonas hydrophila infection. This study systematically analyzed the characteristics of the caspase gene family in largemouth bass and identified associative patterns consistent with potential roles in innate immune responses, providing testable hypotheses for future functional validation.

Conclusions

This study comprehensively identified the caspase gene family in largemouth bass, analyzing their genomic organization, evolutionary relationships, sequence diversity, alternative splicing patterns, and expression responses to Aeromonas hydrophila infection. The findings provide valuable insights for further investigation into caspase-mediated apoptosis and immune mechanisms in teleost fish under Aeromonas hydrophila stress.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-026-12644-8.

Introduction

Apoptosis, an essential gene-regulated process of cell death, plays a crucial role in maintaining homeostasis in living organisms [1]. This process not only eliminates individual cells but also significantly contributes to immune function, particularly by clearing autoreactive lymphocytes to support tissue development and homeostasis [2]. Central to this process are cysteine-aspartic acid proteases (caspases), which are vital mediators of both apoptosis and inflammation [3, 4].

Caspases are initially synthesized as inactive zymogens that require either oligomerization or proteolytic cleavage at specific aspartic acid residues for activation. They consist of an N-terminal prodomain and a catalytic CASc domain, which is divided into large P20 and small P10 subunits [3, 4]. Based on their roles, caspases can be categorized into two major types: those involved in inflammation (e.g., caspase-1, -4, -5, -11, -12) and those that execute or initiate apoptotic processes [5]. Initiator caspases (caspase-2, -8, -9, -10) feature extended N-terminal domains containing CARD or DED interaction motifs, while effector caspases (caspase-3, -6, -7) possess truncated N-termini lacking these domains [5]. Initiator caspases activate apoptosis through either the intrinsic mitochondrial pathway or the extrinsic death receptor pathway [4]. In the intrinsic pathway, stress signals trigger the formation of the apoptosome complex, activating initiator caspases like caspase-9. The extrinsic pathway begins when ligands bind to death receptors, leading to the assembly of the death-inducing signaling complex (DISC), which activates caspases-8 and − 10 [6–8]. Effector caspases, including caspase-3, -6, and − 7, are then activated by initiator caspases, driving the final stages of apoptosis [9]. Beyond apoptosis, certain caspases (caspase-1, -4, -5, -11) mediate pyroptosis via gasdermin D cleavage or inflammasome-mediated caspase-1 activation [10]. Recent studies have expanded the role of caspases beyond apoptosis and pyroptosis, showing that they also regulate cellular processes such as proliferation, differentiation, migration, and tissue regeneration at non-lethal levels, demonstrating their broad functional diversity [8, 11].

Research into caspases in non-mammalian species, particularly fish, has revealed notable findings. Caspases-1, -2, -3, -6, -7, -8, -9, and − 10 have been identified in black rockfish (Sebastes schlegelii) and turbot (Scophthalmus maximus) [12, 13]. Novel caspase family members such as caspase-17, -19, -20, -21, -22, and − 23 have been discovered in zebrafish [14]. Furthermore, 18 complete caspase genes have been reported in rainbow trout, including caspase-1, -2, -3, -6, -7, -8, -9, -10, -17, and − 20 members [15], providing new insights into the caspase family in fish. Studies have demonstrated that caspases regulate apoptosis during zebrafish embryonic development, with distinct expression patterns observed from the cleavage stage to the larval phase, highlighting their specialized and redundant roles in development [14]. Additionally, bacterial infections in fish species such as black rockfish, turbot, spotted sea bass (Lateolabrax maculatus), and orange-spotted grouper (Epinephelus coioides) have been shown to significantly alter caspase gene expression, indicating their critical involvement in immune responses [12, 13, 16, 17]. Elevated expression of caspase genes in mucosal tissues, including the intestine, skin, and gills of euryhaline fish, further underscores their role in mucosal immunity [18].

The largemouth bass (Micropterus salmoides, M. salmoides,), is a freshwater species indigenous to North America. Due to its rapid growth, ease of capture, environmental resilience, and palatable flesh, it has been introduced globally since the 19th century [19–22]. In the 1980s, the species was introduced to China, where it underwent domestication and adapted to artificial feed, allowing for large-scale farming [23]. However, the largemouth bass aquaculture industry has recently faced substantial economic challenges, primarily due to recurring bacterial intestinal inflammation and septicemia caused by Aeromonas hydrophila (A. hydrophila), leading to significant losses [24]. A. hydrophila is a Gram-negative bacterium with a broad pathogenic spectrum. In fish, it induces diseases such as bacterial enteritis, ascites, and motile aeromonas septicemia, which can result in high mortality rates [25]. The bacterium can cause conditions ranging from mild enteritis to potentially lethal septicemia [26]. Despite its known pathogenicity, the role of caspase genes in the immune response of largemouth bass to A. hydrophila remains unclear. This study identified 19 caspase family members in largemouth bass and examined their expression patterns following A. hydrophila stimulation. Furthermore, functional, evolutionary, sequence diversity, alternative splicing (AS༉and differential alternative splicing (DAS) analysis were also performed to explore the evolutionary history and potential immunological mechanisms of these caspases genes. Our findings offer valuable insights into the functions of caspase genes in M. salmoides and enhance our understanding of their role in apoptosis mechanisms.

Materials and methods

Identification and classification of caspases

Genomic data were retrieved to identify caspase sequences in M. salmoides (Accession: GCF_014851395.1, https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_014851395.1/) and Micropterus dolomieu (M. dolomieu) (Accession: GCF_021292245.1, https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_021292245.1/) from the NCBI database (https://www.ncbi.nlm.nih.gov/). Amino acid sequences from Danio rerio (zebrafish), L. maculatus, Rattus norvegicus, and Homo sapiens were also retrieved for use as query sequences in BLASTP searches (E-value < 1×e^− 5). Conserved domain HMM profiles (CARD: PF00619, DED: PF01335, Peptidase_C14: PF00656) were retrieved from InterPro and utilized for HMMER searches (E-value < 1×e^− 5). Candidate caspases were verified through domain architecture analysis using InterProScan, followed by MW and pI calculation with TBtools.

Sequence analysis and transcription factor binding sites (TFBS) prediction of caspases

To characterize the structural features of caspase family members, conserved motifs were identified using MEME Suite (https://meme-suite.org/meme/tools/meme). In addition, conserved domains were predicted through InterProScan. The resulting conserved protein domains, motifs, and gene structures were visualized with TBtools. For transcriptional regulatory analysis, 2-kb promoter regions upstream of the translation start site (ATG) were extracted for TFBS prediction via AnimalTFDB4 (https://guolab.wchscu.cn/AnimalTFDB4/#/TFBS_Predict). Visualization of the TFBS prediction results was performed using ggplot2 (v3.5.1) within the R environment.

Phylogenetic, collinearity, and syntenic analysis

To elucidate the evolutionary patterns of caspase genes, we conducted comparative sequence analysis using caspase orthologs from diverse fish species (M. salmoides, M. dolomieu, D. rerio, L. maculatus, Dicentrarchus labrax, Siniperca chuatsi, Plectropomus leopardus, Xiphias gladius, Sander lucioperca) and representative vertebrates (R. norvegicus, H. sapiens). Multiple sequence alignments were performed using Clustal W, followed by phylogenetic analysis with the maximum likelihood (ML) algorithm and the VT + F+R6 model in iqtree (v2.0.4), with bootstrap support assessed over 1000 replicates. The phylogenetic tree were visualized and annotated using iTOL (v7). Collinearity analysis was conducted with MCScanX (v1.0.0) across six species — M. dolomieu, M. salmoides, Oncorhynchus mykiss (rainbow trout, Accession: GCF_013265735.2), Cyprinus carpio (Accession: GCF_018340385.1), D. rerio (Accession: GCF_000002035.6), and Salmo salar (Accession: GCF_905237065.1). The identified caspase collinear gene pairs were visualized using TBtools. To validate the genomic context of caspase genes, syntenic analyses were also performed for selected caspase genes (CASP10, CASP2, and CASP7) across the same six species used in the collinearity analysis via TBtools. Genome annotations were sourced from the NCBI database, allowing for the determination of transcriptional orientations, annotations, and positional relationships of adjacent genes based on BLAST comparison results.

Quantitative expression, sequence diversity, differential expression, AS and DAS analysis

Transcriptome data for M. salmoides (Accession: PRJNA1113509, https://www.ncbi.nlm.nih.gov/sra?linkname=bioproject_sra_all&from_uid=1113509 and Accession: PRJNA933545, https://www.ncbi.nlm.nih.gov/sra?linkname=bioproject_sra_all&from_uid=933545), collected under normal conditions from various tissues, were obtained from the NCBI database to investigate sequence diversity [27, 28]. Transcriptome datasets from NCBI (PRJNA697259: spleen 24 h post-live/inactivated A. hydrophila infection, https://www.ncbi.nlm.nih.gov/sra/?term=PRJNA697259; PRJNA748458: spleen 6/12 h post- Aeromonas veronii (A. veronii) infection, https://www.ncbi.nlm.nih.gov/sra/?term=PRJNA748458) were analyzed for AS, differential expression and DAS [29, 30]. Raw sequencing reads were processed through Trimmomatic (v0.39) for adapter removal and quality filtering, generating high-quality paired-end reads. These processed reads were subsequently mapped to the M. salmoides reference genome using HISAT2 (v2.2.1). Normalization and quantitative expression analyses of caspase genes were conducted using StringTie (v2.1.7) and featureCounts (v2.0.1). DESeq2 (v1.30.1) was used for differential expression analysis with a threshold of |log₂FC| > 1 and padj < 0.05. To investigate the response mechanisms of caspase family members following A. hydrophila infection, focusing on their role in the transcriptional upregulation of apoptosis-related genes in M. salmoides, we submitted the protein sequence file of M. salmoides to the EggNOG-mapper website (http://eggnog-mapper.embl.de/) to perform GO and KEGG annotations. Subsequently, TBtools was used to map gene expression levels onto the transcriptional upregulation of genes in the apoptosis pathway. Analysis of single nucleotide polymorphisms (SNPs) identified from RNA-seq data was performed using bcftools (v1.8), followed by simple filtering of VCF files based on QUAL > = 20 && DP > = 10 && FORMAT/DP > = 5.Then, with subsequent filtering via bcftools filter to apply quality thresholds such as minor Allele Frequency (MAF ≥ 0.05), genotype quality (GQ ≥ 20) to retain high-confidence variants.

The effect of identified SNPs from RNA-seq data and indels on missense mutations was assessed by annotating caspase genes with ANNOVAR (v2020Apr16), evaluating their impact on protein sequences. AS and DAS (FDR ≤ 0.05) events were analyzed using rMATS (v4.1.1) with an event-based approach, and visualization was achieved using rmats2sashimiplot (v3.0.0).

Tissue distribution of caspase genes in largemouth bass and their response to A. hydrophila infection

Healthy juvenile largemouth bass, averaging 8.51 ± 1.92 g (mean ± SD, n = 90) in weight and 89.06 ± 6.80 mm in total length, were obtained from Guangdong Heshi Aquatic Products Co., Ltd. (Foshan, China). Following one week of acclimatization, the fish were randomly allocated into two experimental groups: a control group (n = 42) receiving intraperitoneal injections of 50 µL phosphate-buffered saline (PBS) at 20 °C, and an A. hydrophila-inoculated group (n = 42) that received the same volume of a bacterial suspension at a concentration of 5.4 × 10⁸ CFU/mL. Using a microsyringe, the solutions were administered intraperitoneally, marking the injection time as the 0-hour point. Prior to sampling, individuals were first anaesthetized by immersion in a buffered solution of tricaine methanesulfonate (MS-222) at a concentration of 100 mg/L until loss of equilibrium and opercular movement ceased (typically within 3–5 min), indicating a state of profound anesthesia and unconsciousness. Following verification of anesthetic depth, euthanasia was completed by rapid cervical transection (spinal cord severance). Samples of intestine, spleen, and head kidney were collected from each fish at 6, 12, 24, 48, 72, 96, and 120 h post-injection, rapidly frozen in liquid nitrogen, and maintained at -80 °C until RNA extraction. Six untreated fish were also sampled at baseline (0 h) to evaluate caspase gene expression patterns in multiple tissues: intestine, spleen, head kidney, liver, gill, blood, brain, muscle, and skin. At different sampling time points post-injection, the intestine, spleen, and head kidney were collected, each sample was set up with four replicates.

Expression profiling of caspase genes via qRT-PCR and statistical analysis

Specific primers targeting individual caspase genes were designed using Primer5 software, with sequences listed in Supplementary Table S1. cDNA samples were diluted to an optimal concentration based on standard curves and amplified with the LineGene 9600 System. The Actin Beta gene (ACTB, accession number XM_038695351.1) was selected as an internal reference for normalization based on its stable expression across the tissues and experimental conditions examined in this study, as supported by preliminary stability analysis and common practice in teleost gene expression studies [31], and all reactions were conducted in triplicate.The qPCR cycling conditions consisted of an initial denaturation step at 95 °C for 20 s, followed by 40 cycles of denaturation at 95 °C for 15 s, annealing at 60 °C for 20 s, and extension at 72 °C for 10 s. Relative expression levels of the caspase genes were quantified using the 2^{− ΔΔCT} method.

Statistical analysis and graphical representation were per formed using R packages and SPSS 20.0. Error bars represented the standard error calculated from four biological replicates, and data were subjected to analysis of variance (ANOVA). Then Duncan’s multiple range test was used to do post hoc test. All statistical analyses were performed using SPSS statistical software package (SPSS 20.0, IBM Company, USA). The significance of treatments was tested at the level of P < 0.05.

Results

Identification and classification of caspase

Genomic analysis revealed 19 caspase genes in M. salmoides genome and 17 in that of M. dolomieu (Table 1). The composition and distribution of caspase family members differed between the two species. In M. salmoides, the identified caspase genes included 3 copies of caspase-1, 1 of caspase-2, 2 of caspase-3, 3 of caspase-6, 1 of caspase-7, 2 of caspase-8, 3 of caspase-9, 1 of caspase-10, 2 of caspase-17, and 1 of caspase-20. In contrast, M. dolomieu was found to contain 3 copies in caspase-1, 1 in caspase-2, 2 in caspase-3, 3 in caspase-6, 1 in caspase-7, 2 in caspase-8, 1 in caspase-9, 1 in caspase-10, 2 in caspase-17, 1 in caspase-20. The amino acid sequences in M. salmoides ranged from 229 to 558 residues in length, with molecular weights between 25,235.78 and 62,395.09 Da, and pI ranging from 5.08 to 8.28. Summary data for M. dolomieu are also provided in Table 1.

Table 1.

Comparative analysis of caspase family members in M. salmoides and M. dolomieu

Sequence ID	Gene ID	Amino Acid(aa)	Molecular Weight(Da)	Theoretical pI
MsCASP1_1	119,891,577	386	43262.37	5.56
MsCASP1_2	119,883,616	397	44773.04	6.54
MsCASP1_3	119,883,617	391	43957.21	6.84
MsCASP2	119,897,458	506	57011.51	6.12
MsCASP3_1	119,888,653	294	32737.9	5.48
MsCASP3_2	119,898,808	283	31236.38	5.89
MsCASP6_1	119,896,761	301	34010.37	5.6
MsCASP6_2	119,898,911	303	34289.76	5.38
MsCASP6_3	119,898,913	302	33988.52	8.28
MsCASP7	119,916,041	313	35443.87	5.09
MsCASP8_1	119,902,446	481	55010.39	5.12
MsCASP8_2	119,902,447	440	50061.72	6.13
MsCASP9_1	119,905,914	435	48911.15	5.59
MsCASP9_2	119,915,377	513	56915.76	5.08
MsCASP9_3	119,915,381	435	48911.15	5.59
MsCASP10	119,904,125	558	62395.09	5.91
MsCASP17_1	119,914,744	229	25235.78	5.63
MsCASP17_2	119,919,358	229	25250.76	7.1
MsCASP20	119,902,448	385	43118.19	6.18
MdCASP1_1	123,960,194	386	43198.31	5.92
MdCASP1_2	123,985,597	397	44740.02	6.3
MdCASP1_3	123,985,598	392	44129.26	6.29
MdCASP2	123,968,476	504	56794.25	6.12
MdCASP3_1	123,957,505	295	32863.11	5.63
MdCASP3_2	123,970,040	283	31339.51	6.09
MdCASP6_1	123,969,782	301	34023.41	5.73
MdCASP6_2	123,969,783	301	34083.46	5.47
MdCASP6_3	123,969,955	253	28689.5	7.03
MdCASP7	123,982,834	322	36506.19	5.34
MdCASP8_1	123,972,363	481	54987.44	5.23
MdCASP8_2	123,972,364	441	50430.26	6.13
MdCASP9	123,966,227	433	48646.92	5.77
MdCASP10	123,973,633	558	62395.08	6.03
MdCASP17_1	123,981,800	253	27860.84	5.82
MdCASP17_2	123,982,350	274	30315.51	7.58
MdCASP20	123,972,365	385	42908.88	6.42

Note: Gene ID listed in Table 1 refer to the publicly available reference genome assemblies from NCBI’s Genome database (Accession: GCF_014851395.1, and accession: GCF_021292245.1), which were used for gene identification and analysis in this study

Sequence analysis and prediction of transcription factor binding sites of caspase

To investigate the sequence architecture of caspases in M. salmoides and M. dolomieu, a comprehensive analysis was conducted focusing on conserved domains, motifs, and gene structures for each family member (Fig. 1). InterProScan validation confirmed that all identified caspase sequences contained the CASc domain. However, structural differences were observed among various caspase subtypes. For instance, caspase-1, -2, -9, and − 20 contained C-terminal CARD domains, while caspase-8 and − 10 possessed dual C-terminal DED domains (Fig. 1A). Ten conserved motifs were detected among caspase family members (Fig. 1B), including motif 2 which featured the characteristic QACNG sequence (Fig. 1D). The gene structures of the caspase family members were largely similar, especially among orthologs. For example, caspase-2 had 14 exons in both M. salmoides and M. dolomieu, caspase-10 had 15 exons, and caspase-20 had 8 exons (Fig. 1C). Genes with multiple copies exhibited variations in exon-intron organization. For instance, MsCASP8_1 and MdCASP8_1 each contained 14 exons, whereas MsCASP8_2 and MdCASP8_2 had only 11 exons. Figure 2 displayed the distribution of various transcription factor binding motifs across a set of caspase genes. Several transcription factors, including Fork_head, P53, ETS, MYB, and AP-2, play roles in diverse developmental processes and are associated with the regulation of genes involved in cell cycle regulation and apoptosis. ZBTB, RFX, and TF_bZIP are linked to immune responses and cell cycle control. Meanwhile, T-box and Homeobox transcription factors are essential for tissue differentiation, morphogenesis, and organogenesis. These transcription factors may regulate caspases, which in turn affect cellular processes including apoptosis, development, and stress responses.

Fig. 1.

Protein domain, conserved motifs, and gene structure analysis of the caspase family in M. dolomieu and M. salmoides. (A) Protein domains; (B) Conserved motifs; (C) Gene structures; (D) Motif 2 contains the highly conserved amino acid sequence QACNG (QACNG, N represents G for caspase-9; Q for caspase-8, -10, and − 20; and R for the other caspase genes)

Fig. 2.

Prediction of transcription factor binding sites (TFBS) in the caspase of M. dolomieu and M. salmoides. A Distribution of TFBS along the 5’ to 3’ direction of the caspase gene sequences; B Functional classification of TF associated with the predicted binding sites

Phylogenetic, collinearity, and syntenic analysis

As shown in Fig. 3, the caspases from M. salmoides clustered with their orthologs, forming ten distinct clades. These clades were grouped into three subfamilies: effector, inflammatory, and regulatory caspases, which align well with their functional annotations. Specifically, the apoptosis-executing caspases—caspase-3, -6, and − 7—formed a cluster. The effector caspases occupied the innermost branches of the tree, while the inflammatory and regulatory caspases grouped with their corresponding teleost homologs and occupied the outer branches. Overall, this phylogenetic analysis supports the accuracy of caspase gene annotation in M. salmoides, suggesting that these genes have been largely conserved throughout evolution. Synteny analysis was carried out among caspase gene pairs from M. dolomieu, M. salmoides, O. mykiss, C. carpio, S. salar, and D. rerio using MCScanX and BLASTP. The results revealed several collinearity gene pairs: 5 between D. rerio and M. dolomieu, 5 between D. rerio and M. salmoides, 13 between C. carpio and M. dolomieu, 12 between C. carpio and M. salmoides, 13 between O. mykiss and M. dolomieu, 10 between O. mykiss and M. salmoides, 14 between S. salar and M. dolomieu, 13 between S. salar and M. salmoides, 10 between M. dolomieu and M. salmoides (Fig. 4). Furthermore, syntenic relationships for CASP10, CASP2, and CASP7 was examined. To further validate their identities, the genomic regions surrounding these caspase genes and their neighboring genes in M. dolomieu, M. salmoides, and other teleost species were displayed in Fig. 5. Notably, the CASP10 locus showed high conservation among the analyzed species, with AGPS, pde11a, Cyt C, osbpl6, prkra, fam237, ADAM23, lmtk2, and pvalb5 consistently found in proximity. This suggests that these genes may have co-evolved with CASP10 and potentially participate in shared cellular processes such as apoptosis or inflammation. Similarly, the region around CASP2 also exhibited strong synteny, with fam131bb, clcn1b, rap1gapl, GSTK1, styk1b, phc1, and m6pr commonly found nearby, indicating potential functional interactions. Likewise, the CASP7 locus was highly conserved, with adrb1, nhlrc2, dclre1a, TRIM39, NRAP, dnmbp, cpn1, and itgb3b frequently observed in its vicinity, implying possible co-regulation or shared biological roles. Despite this overall conservation, some variations in gene order and presence were detected. For instance, in M. dolomieu, lmtk2, pvalb5, and PA were present downstream of MdCASP10, and styk1b was found near MdCASP2, but these genes were absent in other species, possibly reflecting species-specific adaptations. In contrast, dnmbp was missing downstream of MsCASP7 in M. salmoides, suggesting gene loss or structural rearrangement.

Fig. 3.

Phylogenetic analysis of the caspase gene family. Distinct clades are represented by different colors

Fig. 4.

Caspase gene collinearity analysis. Red lines indicate collinearity gene pairs among species (DR: D. rerio, MD: M. dolomieu, MS:M. salmoides, CC: C. carpio, OM: O. mykiss, SS: S. salar)

Fig. 5.

Syntenic analysis of caspase genes (CASP10, CASP2, and CASP7) in M. dolomieu, M. salmoides, O. mykiss, C. carpio, S. salar and D. rerio. agps: alkyldihydroxyacetonephosphate synthase, peroxisomal; pde11a: phosphodiesterase 11a; Cyt C: cytochrome c-like; osbpl6: oxysterol binding protein-like 6; prkra: protein kinase, interferon-inducible double stranded RNA dependent activator; fam237: protein FAM237A-like; ADAM23: disintegrin and metalloproteinase domain-containing protein 23; lmtk2: lemur tyrosine kinase 2; pvalb5: fuscoguttatus parvalbumin 5; PA: parvalbumin beta-like; fam131bb: Family with sequence similarity 131 member B; un: uncharacterized protein; chrna2b: neuronal acetylcholine receptor subunit alpha-2; ptk2bb: protein tyrosine kinase 2 beta, b; clcn1b: chloride channel protein 1; rap1gapl: RAP1 GTPase activating protein-like; GSTK1: glutathione S-transferase kappa 1; styk1b: serine/threonine/tyrosine kinase 1b; phc1: polyhomeotic homolog 1; m6pr: cation-dependent mannose-6-phosphate receptor; adrb1: adrenoceptor beta 1; nhlrc2: NHL repeat containing 2; dclre1a: DNA cross-link repair 1 A protein; PLEKHS1: pleckstrin homology domain-containing family S member 1-like; TRIM39: E3 ubiquitin-protein ligase TRIM39; NRAP: nebulin-related anchoring protein; dnmbp: dynamin binding protein; cpn1: carboxypeptidase N, polypeptide 1; itgb3b: integrin beta 3b

Quantitative expression, sequence diversity, differential expression, AS and DAS analysis

To examine the expression dynamics of caspase genes in response to Aeromonas infection, we aligned spleen transcriptome data collected at multiple time points after infection with two Aeromonas strains (A. hydrophila and A. veronii) to the reference genome for responsive gene quantification. Analysis demonstrated significant expression changes in multiple caspase genes upon bacterial challenge (Figs. 6A and 7A). Differential expression analysis indicated that eight caspase genes (MsCASP3_1, MsCASP6_2, MsCASP6_3, MsCASP7, MsCASP8_2, MsCASP9_2, MsCASP9_3, MsCASP20) were up-regulated, while one gene (MsCASP1_2) was down-regulated at 24 h post-immunization with inactivated A. hydrophila in the spleen of M. salmoides (Fig. 6B). Furthermore, two genes (MsCASP3_1, MsCASP8_2) were up-regulated and one (MsCASP1_2) was down-regulated at 24 h after live A. hydrophila infection (Fig. 6C). Notably, one caspase gene (MsCASP7) was up-regulated, while two others (MsCASP3_2, MsCASP8_1) were down-regulated at 12 h post-A. veronii infection in the spleen of M. salmoides (Fig. 7B). To explore the response mechanisms of caspase family members following A. hydrophila infection, focusing on their role in the transcriptional upregulation of apoptosis-related genes in M. salmoides, we analyzed differentially expressed genes related to this process. Six hours after infection, key components, including caspase family members (caspases-2, -3, -6, -7, -8, -9, and − 10) as well as pro-apoptotic genes (Bak, Bid, and Bad), showed no significant changes (Fig. 8A, yellow squares), while some anti-apoptotic genes (Bcl-2 and Bcl-XL) potentially still elevated (Fig. 8A, red squares). By twelve hours post-infection, a greater number of genes showed significant changes, further activating the transcriptional upregulation of genes in the apoptosis pathway. Notably, several anti-apoptotic genes began to be down-regulated (Bcl-2 and Bcl-XL). Additionally, critical components, including caspases-3, -7, and − 8, started to show responses. At twenty-four hours post-immunization with inactivated A. hydrophila (Fig. 8C), the transcriptional upregulation of genes in the apoptosis pathway exhibited modest changes compared to that observed with live A. hydrophila (Fig. 8D), although notable gene expression variations were still apparent. The extrinsic apoptotic pathway initiates through death receptor (e.g., Fas and TNFR1) engagement with their ligands, inducing FADD-mediated DISC assembly. This complex promotes the auto-proteolytic activation of caspases-8/-10, which subsequently either directly activate effector caspases (-3, -6, -7) or indirectly stimulate the intrinsic pathway via Bid cleavage. This complex facilitates the self-cleavage and activation of pro-caspases-8 and − 10. Activated caspase-8 can directly trigger downstream effector caspases, such as caspase-3, -6, and − 7, or can indirectly activate the intrinsic apoptotic pathway by cleaving the pro-apoptotic protein Bid. In Fig. 8C, alterations were noted in essential components of the extrinsic pathway, including death receptors (e.g., Fas and TNFR1), the adaptor protein FADD, caspase-8, and its downstream effectors (caspases-3, -6, and − 7). However, no significant changes were observed in Bid—the principal activator of the intrinsic apoptotic pathway. It is proposed that by twenty-four hours after immunization with inactivated A. hydrophila, the extrinsic apoptotic pathway was activated, while the intrinsic pathway was not. Conversely, at twenty-four hours post-infection with live A. hydrophila, the transcriptional upregulation of genes in the apoptosis pathway displayed more extensive gene expression changes. As depicted in Fig. 8D, numerous red and green squares were evident, including death receptors (e.g., Fas and TNFR1), the adaptor protein FADD, caspase-8, and its downstream effectors (caspase-3), along with pro-apoptotic (Bid) and anti-apoptotic genes (Bcl-2 and Bcl-XL). Key components associated with the caspase cascade and DNA damage repair (e.g., p53) may have been significantly activated, resulting in a comprehensive apoptosis program. Sequence diversity analysis identified frameshift insertions and nonsynonymous single nucleotide variants (SNVs)—across in three tissues (spleen, kidney, and intestine) of M. salmoides (Fig. 9). Among these, nonsynonymous SNVs were the most common mutation type, indicating their widespread presence. The presence of numerous nonsynonymous SNVs in immune-related tissues is consistent with a potential role in genetic diversification, with associative patterns that may relate to immune adaptation (requires further functional validation). However, direct functional validation is required to confirm their impact on protein function and disease resistance. We will also cite this as a key direction for future research in the Discussion.

Fig. 6.

Expression profiles and DEGs analysis of caspase genes in the spleen of M. salmoides 24 h post-infection with live A. hydrophila and post-immunization with inactivated A. hydrophila. (A) Heatmap of caspase genes expression in spleen (FPKM-based quantification); (B) Volcano plot of caspase DEGs between control and inactivated A. hydrophila-immunized groups; (C) Volcano plot of caspase DEGs between control and live A. hydrophila-infected groups

Fig. 7.

Expression profiles and DEGs analysis of caspase genes in the spleen of M. salmoides at 6 h and 12 h post-infection with A. veronii. (A) Heatmap of caspase genes expression in spleen (FPKM-based quantification); (B) Volcano plot of caspase DEGs between control and 12 h post-A. veronii infection groups

Fig. 8.

Mapping of differentially expressed genes (DEGs) to the genes involved in the Apoptosis pathway. A-B. DEGs from the spleen of M. salmoides at 6 h and 12 h post-A. veronii infection were mapped to the Apoptosis pathway. C-D. DEGs from the spleen at 24 h following immunization with inactivated A. hydrophila and challenge with live A. hydrophila were similarly analyzed. The up-regulated expression was shown in red, the down-regulated expression was shown in green, and yellow indicated that the component did not show significant changes, undetected genes are shown by white coloring in the current analysis

Fig. 9.

Bubble plot depicting patterns of sequence diversity in the caspase gene across different tissues of M. salmoides. Each bubble represents a specific mutation type in the caspase gene, with its size corresponding to the number of observed missense variants

The intestine showed the highest overall mutation count (60), suggesting it may be under greater selective pressure or environmental exposure compared to other tissues. Notably, MsCASP1_1 and MsCASP1_3 exhibited frequent nonsynonymous SNVs, highlighting their potential involvement in immune regulation. Figure 10 illustrated the number and location of missense variants across MsCASP1_1 and MsCASP1_3. Some genes, such as MsCASP8_1, were exclusively mutated in the intestine, implying tissue-specific immune functions. A single frameshift insertion was detected in MsCASP8_2 in the kidney, which could lead to significant functional alterations or protein dysfunction. Overall, the results indicate that caspase genes in M. salmoides experience substantial genetic variation, potentially playing a role in immune adaptation and resistance to disease. AS events were identified in caspase-1, caspase-2, caspase-3, caspase-7, caspase-8, and caspase-10 across the spleen in M. salmoides (Table 2). These caspases exhibited multiple splicing sites, with three distinct AS events observed. Notably, caspase-8 (MsCASP8_1) showed seven skipped exon splice sites and one mutually exclusive exon (MXE) splice site. In contrast, MsCASP8_2 displayed only three SE splice sites. Caspase-2 (MsCASP2) contained six SE splice sites, while caspase-3 (MsCASP3_1) had two SE splice sites. Additionally, a single SE splice site was identified in caspase-1 (MsCASP1_3), caspase-7 (MsCASP7), and caspase-10 (MsCASP10) respectively. Through DAS analysis, we observed DAS events in MsCASP8_1 of M. salmoides following A. veronii infection in the spleen. As illustrated in Fig. 11, a skipped exon event was identified in the 5th exon of MsCASP8_1 in the spleen of M. salmoides at 6 h and 12 h post-infection with A. veronii (IncLevel < 1, FDR ≤ 0.05). Specifically, the sequence highlighted in the red box in Fig. 11C represents the nucleotide sequence of the excised 5th exon. definitive evidence of functional divergence event involving the skipped exon in the 5th exon of MsCASP8_1, its precise functional consequences require experimental validation. This splice isoform may act as a regulatory switch, fine-tuning the intensity of apoptotic signaling by generating a functionally impaired protein variant, which could serve to limit potential immunopathological damage, a hypothesis that warrants experimental validation. This finding provides new insights into how fish utilize post-transcriptional regulation to balance effective pathogen clearance with host self-protection.

Fig. 10.

Mapping of missense variants in MsCASP1_1 and MsCASP1_3 transcripts from the kidney, spleen, and intestine of M. salmoides. The schematic shows the gene structure of each transcript, with green boxes indicating exons, dashed lines representing introns, and colored boxes highlighting variant positions (blue: non-mutated; pink: mutated). Transcript length is indicated by the scale bar at the top. (A) MsCASP1_1; (B) MsCASP1_3

Table 2.

The number of AS events sites in caspases of M. salmoides

AS type	A5SS	MXE	SE
MsCASP1_3	0	0	1
MsCASP2	1	0	6
MsCASP3_1	0	0	2
MsCASP7	0	0	1
MsCASP8_1	0	1	7
MsCASP8_2	0	0	3
MsCASP10	0	0	1

Fig. 11.

DAS of MsCASP8_1 in response to A. veronii infection. A, B. SE AS events in the 5th exon of MsCASP8_1 at 6 h and 12 h post-A. veronii infection. Red bars represent exons, and black lines indicate introns. The black rectangles below each panel depict CDS. Junction read counts are shown above the curved splice junctions. The inclusion level (IncLevel) for each splicing event is displayed in the upper right corner; IncLevel represents the normalized read count for each exon, with a value of 1.00 indicating complete exon inclusion (i.e., no exon skipping). C Nucleotide sequence of the 5th exon of MsCASP8_1 in M. salmoides

Tissue distribution of caspase genes in largemouth bass and their response to A. hydrophila infection

The findings demonstrated tissue-specific expression patterns among ten caspase genes in M. salmoides (Fig. 12, P < 0.05). Several genes—including MsCASP1_2 (2.76-fold), MsCASP2 (1.72-fold), MsCASP3_2 (15.57-fold), MsCASP7 (4.76-fold), and MsCASP8_1 (20.83-fold)—showed high expression in the blood, suggesting possible involvement in circulatory or immune-related processes. Conversely, MsCASP6_2 (211.88-fold) and MsCASP17_1 (20.69-fold) were highly expressed in intestine, indicating possible involvement in intestine structure, contraction, or repair mechanisms. MsCASP9_2 (6.11-fold) was highly expressed in brain. Notably, MsCASP10 (5.29-fold) and MsCASP20 (3.00-fold) exhibited high expression in spleen and gill respectively, implying functions related to immune processes in these organs. Additionally, MsCASP17_1 (13.67-fold) was predominantly expressed in the skin, pointing toward functions related to skin integrity, barrier maintenance, or wound healing processes.

Fig. 12.

Tissue-specific expression profiles of 10 caspase genes in M. salmoides (P < 0.05)

To investigate the potential involvement of M. salmoides caspase genes in innate immunity, the mRNA expression levels across three key immune tissues — the intestine, head kidney, and spleen — at seven time points following A. hydrophila infection (6, 12, 24, 48, 72, 96, and 120 h) were analyzed. Our results demonstrated that caspase gene expression profiles varied significantly over time and among tissues (Fig. 13, P < 0.05). MsCASP1_2 was significantly down-regulated in the intestine from 6 to 12 h and again from 96 to 120 h. In the spleen, it was down-regulated at 24 h but up-regulated at 96 h. In the head kidney, its expression remained significantly down-regulated from 72 to 120 h. MsCASP2 showed no significant changes in the intestine. However, it was down-regulated in the spleen from 6 to 12 h and up-regulated in the head kidney from 72 to 96 h. MsCASP3_2 was significantly up-regulated in the intestine at 24 h. In the spleen, it was down-regulated from 48 to 72 h but quickly returned to baseline. In the head kidney, it up-regulated from 12 to 24 h and down-regulated from 96 to 120 h. MsCASP6_2 was consistently down-regulated in the intestine from 24 to 120 h. In the spleen, it was up-regulated from 6 to 48 h but down-regulated at 96 h. In the head kidney, it was down-regulated from 48 to 120 h. MsCASP7 was significantly up-regulated in the intestine from 12 to 48 h. In the spleen, it up-regulated from 12 to 24 h but down-regulated at 96 h. In the head kidney, it was up-regulated from 12 to 72 h and normalized afterward. MsCASP8_1 no significant changes in expression in the intestine. In contrast, it was up-regulated in the spleen from 6 to 12 h and in the head kidney from 6 to 48 h, followed by down-regulation from 96 to 120 h. MsCASP9_2 was down-regulated in the intestine at 24 h and up-regulated in the spleen from 12 to 24 h. No significant changes were observed in the head kidney. MsCASP10 was down-regulated in the intestine at 48 h, up-regulated in the spleen from 24 to 72 h, and down-regulated in the head kidney at 96 h. MsCASP17_1 was down-regulated in the intestine from 72 to 96 h. In the spleen, it was up-regulated at 6 to 12 h and again at 96 h. In the head kidney, it was up-regulated from 6 to 24 h, then down-regulated from 72 to 120 h. MsCASP20 was significantly up-regulated in the intestine at 48 h. In the spleen, it was up-regulated from 12 to 48 h and down-regulated at 96 h. In the head kidney, it was up-regulated from 12 to 24 h and subsequently down-regulated at 96 h.

Fig. 13.

Expression of 10 caspase genes in the intestine, spleen, and head kidney of M. salmoides post-A. hydrophila inoculation. Bars labeled with an asterisk (*) denote significant up- or down-regulation relative to the control group (P < 0.05, LSD test)

Discussion

As established in vertebrate biology, caspases serve as critical mediators in extrinsic and intrinsic apoptotic pathways, functioning as pivotal regulators of death signal transduction. serve as critical mediators in extrinsic and intrinsic apoptotic pathways, functioning as pivotal regulators of death signal transduction. These proteases contribute significantly to diverse immune and physiological functions. Mammalian genomes encode 12 distinct caspases (caspase-1 to -12), with caspase-1, -4, -5, -11, and − 12 comprising the inflammatory subgroup. These inflammatory caspases facilitate cytokine maturation during innate immune responses and may induce pyroptosis. Notably, caspase-4, -5, and − 12 are human-specific, whereas caspase-11 is uniquely expressed in mice. In contrast, many inflammatory caspase orthologues identified in mammals are absent in teleosts, with only caspase-1 being detected in teleosts. This study identified 19 and 17 caspase genes in M. salmoides and M. dolomieu, respectively. Among these, both M. salmoides and M. dolomieu possess three caspase-1 orthologs, contrasting with one or two copies observed in black rockfish [12], turbot [13], rainbow trout [15], and zebrafish [14]. This suggests caspase-1 in fish may functionally compensate for multiple mammalian inflammatory caspases, resulting in a more streamlined inflammatory response mediated primarily through caspase-1. Interestingly, zebrafish uniquely harbor several non-mammalian caspases (caspase-17, -19, -20, -21, -22, and − 23) [14], with caspase-19 and − 23 representing teleost-specific inflammatory caspases [32]. However, these genes are absent in M. salmoides and M. dolomieu, mirroring the pattern in rainbow trout. Such findings indicate significant evolutionary divergence of inflammatory caspases not only between mammals and teleosts but also among teleost species [15]. In contrast, previous studies have shown that typical initiator and effector caspases are conserved across mammals and teleosts. We observed consistent orthologue numbers for effector caspases in M. salmoides and M. dolomieu, with two orthologues for caspase-3, three for caspase-6, and one for caspase-7 in both species, which aligns with observations in zebrafish [14]. Additionally, we found differing orthologue numbers for caspase-9 in M. salmoides and M. dolomieu. While M. dolomieu and zebrafish each possess one orthologue, M. salmoides contains three. For other initiator caspases—caspase-2 (1), caspase-8 (2), caspase-10 (1), and caspase-20 (1)—orthologue numbers were consistent among M. salmoides, M. dolomieu, and zebrafish [14]. Among several caspase members identified exclusively in teleosts, only caspase-17 and caspase-20 were found in a limited number of teleosts (zebrafish, rainbow trout, and bass in this study) [14, 15], while other caspases were reported only in zebrafish. The uncategorized caspase, caspase-17, is present as one orthologue in zebrafish and rainbow trout [14, 15], but as two orthologues in both M. salmoides and M. dolomieu. While the physiological roles of caspase-17 remain poorly characterized, structural studies indicate its distinct substrate specificity and enzymatic characteristics relative to other caspases [33]. In this study, a methionine replacement was observed at conserved positions in MsCASP17_1 (151), MsCASP17_2 (172), and MsCASP17_1 (175), and MsCASP17_2 (176) (Fig. S1), leading to disruption of the substrate binding site. This conserved amino acid substitution, observed in both zebrafish and rainbow trout, suggests evolutionary fragment loss in teleost caspase-17 lineages [14]. Notably, caspase-18 - an ancestral caspase-8 paralog present in birds but lost in placental mammals [34], may be functionally compensated by teleost-specific caspase-20, a product of whole genome duplication events [35–37]. In our research, we identified a caspase-20 subtype in M. salmoides and M. dolomieu respectively, but no caspase-18 was detected, which aligns with findings in zebrafish and rainbow trout [14, 15].

The caspases identified in M. salmoides and M. dolomieu possess characteristic N-terminal prodomains of varying lengths and conserved C-terminal CASc catalytic domains, mirroring the structural organization observed in other vertebrate orthologs (Fig. 1A). Domain architecture analysis demonstrated that caspase-1, -2, -9, and − 20 feature N-terminal CARD domains, whereas caspase-8 and − 10 contain dual DED domains. These structural features likely enable the proteins to interact with distinct adapter proteins, thereby initiating specific downstream signaling pathways. The effector caspases, including caspase-3, -6, and − 7, as well as the unclassified caspase-17, were characterized by a short prodomain. The motifs analysis indicated that motif 2 in the CASc catalytic domains of these sequences contains the active site (QACNG), as shown in Fig. 1D. Phylogenetic analysis, based on the identified caspase amino acid sequences from M. salmoides, M. dolomieu, and selected vertebrates, revealed that the caspase genes from M. salmoides and M. dolomieu cluster with their teleost counterparts (Fig. 3). This clustering supports their annotations and categorizes them into ten subfamilies, demonstrating the conservation of the caspase family during evolution. Furthermore, the identification of numerous collinearity gene pairs between M. salmoides, M. dolomieu, and various teleost species further corroborates this conclusion. Syntenic analysis also revealed highly conserved regions around the caspase-2, caspase-7, and caspase-10 genes in these species, suggesting that these genomic regions have undergone strong selective pressure. Genomic analysis revealed cytochrome c, a key regulator of the intrinsic apoptotic pathway that triggers caspase-9 and downstream effectors (caspase-3/7), located upstream of the caspase-10 gene in multiple teleost species. This genomic arrangement implies potential functional involvement of caspase-10 in teleost apoptosis. Finally, transcription factor binding site prediction identified multiple binding motifs in these caspase sequences associated with genes involved in cell cycle regulation and apoptosis, further validating the accuracy of the caspase gene annotations.

In this study, we identified numerous non-synonymous SNPs (nsSNPs) from RNA-seq data in the coding regions of caspase genes across three tissues of M. salmoides (Fig. 9). NsSNPs may substantially alter protein function by modifying encoded amino acids, thereby potentially compromising structural integrity and biological activity [38]. Caspases typically exist as inactive zymogens that undergo proteolytic cleavage for activation, playing pivotal roles in apoptotic protein processing [39]. SNPs may modulate caspase activity by creating novel cleavage sites or altering substrate cleavage efficiency, with evidence suggesting that SNPs near cleavage sites can significantly impact functionality [40]. However, it should be emphasized that the identification of nonsynonymous SNVs in this study is based on RNA-seq data and is subject to the aforementioned methodological limitations. Therefore, while the enrichment of these variants in immune tissues suggests a potential association with immune responses, it cannot be directly equated to positive selection or a causal link to adaptive phenotypes. These SNVs may be neutral, and their enrichment could reflect detection bias due to the high expression of these genes in immune tissues. Determining whether they have functional effects and contribute substantively to disease resistance requires further validation through DNA-level confirmation, population genetic analyses (such as calculations of selection pressure), and functional experiments (e.g., protein activity assays).

While comprehensive data on caspase proteolysis in M. salmoides remain scarce, the likely presence of functionally relevant mutations in its caspase genes warrants further investigation into their sequence diversity. AS expands proteomic diversity and enables context-dependent gene regulation in multiexon genes [41], which generates isoforms with cell-, tissue-, or condition-specific expression patterns. Several caspase splice variants, including caspase-2 [42], caspase-3 [43], caspase-8 [44], and caspase-9 [45], have been identified in mammals. AS generates caspase isoforms with antagonistic apoptotic functions, exemplified by pro-apoptotic caspase-2L/-9a versus anti-apoptotic caspase-2S/-9b [45, 46]. Additionally, AS has been reported in caspase-1 of D. labrax, leading to the formation of isoforms with distinct enzyme activities [47]. The presence of caspase-1 isoforms in D. labrax points to their evolutionary conservation and likely involvement in regulating the inflammatory response. In this study, RNA-seq analysis revealed AS events in caspase-1, caspase-2, caspase-3, caspase-7, caspase-8, and caspase-10 of M. salmoides. These splicing events lead to the production of more than one transcript for each gene. Notably, the splicing patterns and sites of the two caspase-8 genes, MsCASP8_1 and MsCASP8_2, showed significant differences. This also explains the distinct expression levels of these two genes at different time points in response to A.hydrophila infection. It also reveals that AS can regulate mRNA levels. Aquatic organisms utilize this post-transcriptional mechanism to respond to environmental stimuli, which plays a key role in regulating gene expression during stress responses [48, 49]. Moreover, RNA-seq analysis in this study revealed DAS of caspase-8 (CASP8_1) following A. veronii infection (Fig. 11), suggesting that AS may represent a key regulatory layer in the fish’s response to environmental stress.

It should be noted that the analysis of infection-induced DAS in this study is primarily based on transcriptome data from A.veronii infection. Although A.hydrophila and A. veronii belong to the same genus and are both important pathogens for largemouth bass, they may possess different virulence profiles and host interaction patterns. Therefore, while the specific DAS event observed in MsCASP8_1 provides a strong proof-of-concept for post-transcriptional regulation of caspase genes, its generalizability and specific patterns in A. hydrophila or other pathogen infections require confirmation through further targeted experiments (e.g., direct validation of AS using A. hydrophila-infected samples). Future studies should systematically compare how different pathogen species modulate the splicing programs of key host immune genes (such as the caspase family), which will contribute to a more refined understanding of the post-transcriptional regulatory mechanisms underlying fish immune responses.

In our study, the differential gene expression analysis of the transcriptome revealed that as the duration of Aeromonas infection lengthened, the activation of the transcriptional upregulation of genes in the apoptosis pathway progressed from initial subtle changes to full activation. Crucial components of this process include members of the caspase family, Bcl-2 family proteins, and p53, whose expression alterations significantly impact cell fate (Fig. 8). Lu discovered that apoptosis-related genes such as caspase-8, Fas, p53, and LRDD showed slight upregulation during the early stages of infection, with a marked increase in the later stages, indicating distinct temporal characteristics of the apoptotic response [50].Tissue-specific expression profiling of caspase genes in M. salmoides demonstrated differential expression across nine examined tissues, including intestine, spleen, head kidney, liver, gill, blood, brain, muscle, and skin. With the exception of caspase-9, all other caspases were predominantly expressed in immune-related organs or tissues. Caspase-1 had the highest expression in the blood, a pattern also observed in teleosts such as black rockfish and Cynoglossus semilaevis (tongue sole) [12, 51]. Caspase-2 was primarily found in the blood, head kidney, and muscle, with high expression levels in the blood of black rockfish [12], the head kidney of turbot [13], and the muscle of tongue sole [51]. Caspase-3/-8 showed blood-dominant expression, consistent with turbot and Paralichthys olivaceus (Japanese flounder) [13, 52]. Caspase-6 exhibited intestinal preference as in black rockfish and Japanese flounder [12, 52]. Caspase-7 displayed blood-predominant expression similar to turbot [13]. Caspase-9 was predominantly expressed in the brain, akin to tongue sole [51]. Caspase-10 had the highest expression in the spleen and intestine, resembling the expression pattern in black rockfish [12]. Additionally, this study also observed that caspase-17 was mainly expressed in the intestine and skin, while caspase-20 was most highly expressed in the gill. The above results indicate that multiple caspase paralogs exhibit transcriptional activity across different tissues, demonstrating divergent expression patterns. These findings support the hypothesis that these genes may have acquired distinct biological functions through subfunctionalization or neofunctionalization. However, definitive evidence of functional divergence (e.g., differences in substrate specificity or signaling pathway selectivity) awaits validation through future biochemical and cell biology experiments.

In this study, after M. salmoides was infected with A. hydrophila, the initiator caspases—caspase-8, -9, -10, and − 20—were upregulated in the spleen, while caspase-2 was downregulated. Similarly, in the spleen of Vibrio harveyi-infected spotted sea bass, caspase-8, -9, and − 10 showed the same upregulation pattern [16]. A similar expression pattern of caspase-8, -9, and − 20 was observed in the transcriptome of A. hydrophila-infected M. salmoides spleen (as shown in Experiment Fig. 6). Additionally, caspase-6 and caspase-7 were upregulated in the spleen, while caspase-3 showed the opposite pattern. In V. anguillarum-infected black rockfish, caspase-6 was upregulated, but caspase-3 and − 7 were not expressed [12]. In the head kidney, the initiator caspases—caspase-2, -8, and − 20—were upregulated, while caspase-10 was downregulated in later stages, and caspase-9 showed no change. The spleen of V. harveyi-infected spotted sea bass showed a similar pattern for caspase-2 and − 8 [16]. In the present study, caspase-3 and − 7 were upregulated in the spleen, while caspase-6 displayed the opposite expression pattern. Similarly, in the head kidney of V. harveyi-infected spotted sea bass, caspase-3 and − 7 showed similar patterns [16]. Early in A. hydrophila infection of M. salmoides, caspase-1 was downregulated in the intestine and spleen, with similar downregulation observed in the intestine (12 h post-infection) of V. anguillarum-infected turbot [13], and in the spleen (24 h post-infection) of A. hydrophila-infected M. salmoides (as shown in Experiment Fig. 6). At 96 h, caspase-1 was significantly upregulated in the spleen, indicating a transient immune response. In later stages of infection, caspase-1 in the head kidney was downregulated. Caspase-1 is essential for the maturation of pro-inflammatory cytokines such as IL-1β and IL-18 [47]. Its downregulation suggests a reduction in the inflammatory response, though excessive inflammation can harm the host. Therefore, downregulation of caspase-1 may act as a protective mechanism against tissue damage. In the intestine of A. hydrophila-infected M. salmoides, effector caspases—caspase-3, -7, and initiator caspase-20—were transiently upregulated, while other caspase genes were either downregulated or unchanged. This pattern mirrored that of V. anguillarum-infected black rockfish, where caspase-3 and − 7 were transiently upregulated and other caspase genes were downregulated [12]. Unlike in the head kidney and spleen, the expression of caspase genes related to apoptosis initiation and execution was not induced in the intestine. This may be linked to the strong immune system of intestine [12]. In this experiment, caspase-17 in the intestine did not show significant early changes, but it was notably downregulated later. In contrast, it was significantly upregulated in the spleen and head kidney initially. The function of caspase-17 in teleosts remains under-explored, but our findings suggest it may play a key role in M. salmoides’ defense against A.hydrophila infections. Furthermore, we identified distinct different expression patterns for apoptotic initiator caspases and effector caspases at various time points across the intestine, spleen, and head kidney of A. hydrophila-infected M. salmoides, consistent with the coordinated transcriptional activation of apoptotic pathways during the immune response.

While this study provides a comprehensive genomic and transcriptomic profile of the caspase gene family in M. salmoides, several limitations highlight avenues for future research. First, the reported infection-induced AS events in caspase genes are based on computational predictions from RNA-seq data. Future work should prioritize experimental validation of these splice isoforms, using methods such as RT-PCR with isoform-specific primers followed by sanger sequencing, and ideally, detection at the protein level via western blotting with isoform-specific antibodies or targeted mass spectrometry. Second, the functional implications of the identified nonsynonymous SNVs and AS events remain speculative. Direct functional assays—such as expressing recombinant wild-type and variant/mutant caspase proteins and comparing their protease activity, substrate cleavage kinetics, or roles in apoptosis and immune signaling—are needed to establish causal relationships between sequence variation and biological function.Finally, expanding the study to include a larger cohort of individuals with known disease-resistance phenotypes would enable association studies to determine whether specific caspase alleles or expression isoforms are linked to enhanced resilience against A. hydrophila or other pathogens, thereby bridging the gap between molecular observations and practical aquaculture applications.

Conclusion

Genomic analysis identified 19 caspase genes in M. salmoides, with proteins and gene structure analysis, and phylogenetic comparisons revealing evolutionary patterns distinct from mammals but conserved with zebrafish and rainbow trout. The observed gene copy number variations of caspases in M. salmoides suggest potential functional divergence in apoptotic signaling pathways. AS analysis revealed numerous splicing events and sites in caspase-1, caspase-2, caspase-3, caspase-7, caspase-8, and caspase-10 of M. salmoides, suggesting that AS events may function to regulate mRNA levels. DAS analysis revealed that DAS in the 5th exon of MsCASP8_1 occurred at 6 and 12 h post- A.hydrophila infection highlighted stress-responsive regulatory mechanisms. Notably, multiple non-synonymous polymorphisms were detected in M. salmoides caspases. Tissue-specific expression profiling revealed predominant caspase gene expression in immune organs, with significant upregulation in intestine, spleen, and head kidney following A. hydrophila challenge, indicating apoptotic involvement in antimicrobial defense. These findings advance our understanding of apoptotic regulation in teleosts and explore the evolutionary adaptation of caspase-mediated immunity, providing valuable references for for studying resistance to A.hydrophila in teleosts.

Beyond elucidating the evolutionary and functional landscape of caspases in teleost immunity, our findings offer tangible avenues for application in largemouth bass aquaculture. Firstly, the identified infection-responsive caspase alleles and splice variants (e.g., the MsCASP8_1 isoform) serve as promising candidate molecular biomarkers. Their integration into selective breeding programs could enable the identification and propagation of genotypes with superior disease resistance, contributing to sustainable stock improvement. Secondly, the discovery of pathogen-induced AS highlights a previously underappreciated layer of immune regulation in fish, suggesting RNA-based regulation as a potential target for future therapeutic exploration. By bridging molecular mechanisms with industry needs, this study provides a foundation for developing novel strategies to mitigate losses caused by Aeromonas and other pathogens in intensive aquaculture systems.

Author contributions

Yuping Chen: Conceptualization, Methodology, Validation, Investigation, Writing - original draft & review& editing.Mengzhen Jiang, Yelin Zeng, Yaling Xian, Shengmei Tang, Jiayi Wen, Zecheng Zou: Methodology, Formal analysis, Investigation.Yongsheng Wang: Writing - Review & Editing, Supervision, Project administration.

Funding

This work was supported by the Scientific Research Foundation for Principle Investigator, Kunpeng Institute of Modern Agriculture at Foshan (KIMAQD2022002) and the Funding of Major Scientific Research Tasks, Kunpeng Institute of Modern Agriculture at Foshan (KIMAZDKY2022003).

Data availability

The authors confirm that no new genes or proteins were generated in this study and all analyses were based on existing data in the databases. The data underlying the findings of this study are presented in the article and its supplementary materials. The accession numbers of NCBI Reference Sequence or GenBank for the vertebrate caspase proteins used in this study are provided in Supplementary Material, while those for the M. salmoides caspase sequences id are listed in Table 1.The links to the data used for transcriptomics analysis in the manuscript, registered on NCBI, are as follows: PRJNA1113509: SRA Links for BioProject (Select 1113509) - SRA - NCBI）;PRJNA933545:SRA Links for BioProject (Select 933545) - SRA - NCBI）;PRJNA697259:PRJNA697259 - SRA - NCBI); PRJNA748458:PRJNA748458 - SRA - NCBI)

Declarations

Ethics approval and consent to participate

All animal experiments in this study were conducted in accordance with the guidelines and regulations for the care and use of laboratory animals. The experimental protocols were approved by the Animal Ethics Committee of Kunpeng Institute of Modern Agriculture at Foshan.

Not applicable. This study does not contain any individual person’s data.

Consent for publication

Not applicable. This study does not contain any individual person’s data.

Competing interests

The authors declare no competing interests.