Immune Response Analysis of Head Kidney in Large Yellow Croaker (Larimichthys crocea) Following Nocardia seriolae Infection

Abstract

The large yellow croaker (Larimichthys crocea), one of the most economically valuable marine fish species in China, suffers significant economic losses in aquaculture due to infectious diseases caused by marine pathogens, such as Nocardia seriolae. The pathogenic mechanisms underlying N. seriolae infection in L. crocea and the host immune responses remain inadequately characterized. To investigate the molecular mechanisms of this infection, we conducted transcriptome sequencing on the head kidney tissues of L. crocea at 1, 3, 7, and 14 days post-infection with N. seriolae. In total, 421, 1052, 3215, and 2459 upregulated genes, along with 1853, 1777, 3718, and 3134 downregulated genes were identified, respectively. KEGG enrichment analysis revealed that differentially expressed genes were predominantly associated with immune and metabolic pathways. Notably, pathways involved in Toll-like receptor signaling, ECM–receptor interaction, cytokine–cytokine receptor interaction, and focal adhesion were significantly enriched, highlighting an immune response to N. seriolae infection in L. crocea. In addition, significant enrichment of the citrate cycle (TCA cycle) and oxidative phosphorylation pathways in metabolic processes suggests an upregulated ATP synthesis to meet the heightened energy demand associated with the immune response to infection. These findings contribute to a deeper understanding of the immune defense mechanisms in the head kidney of L. crocea against N. seriolae infection and elucidate aspects of N. seriolae pathogenicity.

1. Introduction

The large yellow croaker (Larimichthys crocea), a member of the order Perciformes, family Sciaenidae, and genus Larimichthys [1], is a warm-water species predominantly distributed across the southern Yellow Sea, East China Sea, and northern South China Sea. This species is of considerable economic importance in marine aquaculture [2]. Overfishing in the 1970s led to a sharp decline in L. crocea populations. However, with governmental support beginning in the 1980s, aquaculture efforts expanded significantly. By the 1990s, advancements in artificial breeding techniques established a robust industry, making L. crocea the most extensively cultured marine fish species in China. It currently holds the highest global production in fry, farming scale, and adult yield, positioning it among eight major advantageous export aquaculture products in China [3,4]. Despite industry growth, the increasing scale of L. crocea aquaculture has been accompanied by outbreaks of diseases caused by pathogens such as Vibrio harveyi, Pseudomonas plecoglossicida, and Nocardia seriolae, resulting in substantial economic losses. Among these, N. seriolae represents a significant threat due to a limited understanding of its pathogenic mechanisms and challenges in effective prevention and control.

N. seriolae is a Gram-positive, facultative intracellular pathogen that infects a range of marine fish species, including L. crocea, Trachinotus ovatusand, and Lateolabrax japonicas, as well as freshwater species such as Channa argus and Micropterus salmoides [5,6]. This opportunistic pathogen is prone to outbreak in high-density aquaculture environments, often triggered by poor water quality and reduced host immunity. Infection typically occurs through compromised skin or gill tissues, resulting in clinical manifestations, including multiple white nodules in internal organs and skin ulcers [7,8]. Characterized by a prolonged incubation period, N. seriolae infection presents as a chronic disease that is often asymptomatic in early stages, complicating detection and control efforts, with conventional antibiotic therapies proving largely ineffective.

RNA sequencing technology enables the identification of gene expression levels under various treatment conditions, offering particular value for non-model organisms. In recent years, RNA sequencing has become widely applied in studying diseases in aquatic animals [9]. Given the complex and protracted nature of bacterial infections in fish, transcriptome sequencing provides a comprehensive transcriptomic dataset, facilitating the in-depth analysis of differentially expressed genes and their functions. This approach offers valuable insights into pathogenesis and immune response mechanisms following pathogen infection.

The head kidney, as the primary immune organ in teleost fish, plays an essential role in the development, differentiation, and proliferation of immune cells [10]. To elucidate the pathogenic mechanisms of N. seriolae infection in L. crocea and its immune responses, we collected head kidney tissues from both uninfected and N. seriolae-infected L. crocea at 1, 3, 7, and 14 days post-infection (dpi) for transcriptome sequencing. This analysis allowed us to assess gene expression changes across infection stages, investigating the interaction mechanisms between N. seriolae and L. crocea and providing a reference framework for further research on N. seriolae control strategies.

2. Materials and Methods

2.1. L. crocea and N. seriolae

Healthy L. crocea were obtained from Tongzhao Village, Ningbo City, Zhejiang Province, with an average weight of 50–60 g and a length of 15–18 cm. The fish were acclimated for 7 days in a recirculating aquaculture system (RAS) with continuous aeration and were fed twice daily at 2% of their total body weight, with feeding halted 24 h prior to infection experiments. N. seriolae cultures were maintained in our laboratory, initially stored at −80 °C. Bacteria were streaked onto trypticase soy agar (TSA) plates and incubated at 28 °C for 5–7 days. The resulting bacterial culture was homogenized in sterile PBS to prepare a bacterial suspension.

2.2. Infection Experiment

Healthy L. crocea were randomly assigned to two groups: an infection group and a control group, with 60 fish per group. The bacterial suspension was prepared at an LD₅₀ concentration of 5.99 × 10⁶ CFU/mL in sterile PBS. The infection group received 500 μL of the bacterial suspension via intraperitoneal injection, while the control group was administered an equivalent volume of sterile PBS using the same route. Throughout the 14-day observation period, water temperature was maintained at 27 ± 1 °C, pH at 7.5–8.0, dissolved oxygen levels above 6.0 mg/L, and salinity at 25–30‰. Fish were fed twice daily at 2% of their total body weight, with a 50% water exchange performed daily.

2.3. Sample Collection

Before sampling, L. crocea were anesthetized with MS-222 (100 mg/L, Sigma-Aldrich, St. Louis, MO, USA) until reaching a state of immobility, characterized by reduced opercular movement (~3 min). To minimize contamination, both the fish surface and dissection instruments were sterilized with 75% ethanol prior to tissue collection. Head kidney tissues were collected from both the control and infection groups at 1, 3, 7, and 14 dpi and immediately frozen in liquid nitrogen. Additionally, tissues from the liver, spleen, head kidney, and trunk kidney were fixed in 4% paraformaldehyde for 24 h for subsequent histopathological examination.

2.4. Total RNA Extraction, Library Construction, and Sequencing

Total RNA was extracted from head kidney tissues using Trizol reagent (Invitrogen, Waltham, MA, USA). RNA concentration and purity were evaluated with a NanoDrop spectrophotometer, and integrity was assessed by agarose gel electrophoresis and an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Following quality confirmation, mRNA was enriched from total RNA using Oligo dT magnetic beads. Fragmentation was followed by synthesis of first-strand cDNA using random hexamer primers, with second-strand cDNA synthesized subsequently. Library preparation included end repair, A-tailing, adapter ligation, fragment selection, amplification, and purification steps. The library concentration was initially quantified with a Qubit 2.0 Fluorometer, diluted to 1.5 ng/μL, and insert size was confirmed using an Agilent 2100 Bioanalyzer. Quantitative real-time PCR (qRT-PCR) was conducted to determine the effective concentration of the library (>1.5 nM) to ensure library quality. Libraries passing quality control were pooled according to effective concentration and sequencing specifications for Illumina sequencing.

2.5. RNA Sequencing Data Processing and Analysis

The raw sequencing data contained a minor proportion of reads with sequencing adapters or low quality. To ensure data integrity and reliability, the raw data were filtered to remove reads containing adapters, ambiguous bases (N), and low-quality bases (Qphred ≤ 5 across more than 50% of the read length). Clean reads were subsequently aligned to the reference genome using HISAT2 software (version 2.0.5) to obtain positional information. Transcript assembly was conducted with StringTie software (version 2.2.0), which utilizes network flow algorithms and supports de novo assembly when necessary. Newly assembled transcripts were annotated using databases such as Pfam, SUPERFAMILY, GO, and KEGG. To confirm experimental reliability and validate sample selection, correlation analysis of gene expression levels and principal component analysis (PCA) were performed. Correlation analysis, based on FPKM values of all genes across samples, was visualized as a heatmap to illustrate inter-group differences and consistency within replicates. PCA, applied to evaluate both inter-group differences and within-group replicates, used linear algebra methods for dimensionality reduction, extracting principal components from the extensive set of gene variables. The raw data have been deposited at the NCBI Sequence Read Archive under the accession number SUB14871298.

2.6. Identification and Analysis of Differentially Expressed Genes (DEGs)

Following the assessment of gene expression levels to identify DEGs, statistical analyses were conducted to isolate genes with significant expression changes across conditions and to link these to specific gene functions. Differential expression analysis was performed using DESeq1.20.0, applying thresholds of log₂ (FoldChange) ≥ 1 and padj ≤ 0.05 to identify significant DEGs. Functional enrichment analysis was then used to categorize hundreds to thousands of genes, proteins, or other molecules into specific pathways, simplifying the complexity of data interpretation. Enrichment analysis annotated DEGs using the GO and KEGG databases, with subsequent GO functional enrichment and KEGG pathway enrichment analyses conducted using clusterProfiler 3.8.1.

2.7. Quantitative Real-Time PCR (qPCR) Analysis

To assess the reliability and accuracy of the transcriptome data, eight immune-related DEGs were selected for qPCR validation. Gene-specific primers were designed using the NCBI primer-BLAST tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast, accessed on 10 August 2024), with primer sequences detailed in

Table A1. Primers used for qPCR verification of differently expressed genes.

Primer Name	Forward Primer	Reverse Primer
β-actin	GTGATGGTTGGTATGGGCCA	TCGATGGGGTACTTCAGGGT
CDHE	GTGTGGATAAGGAGACGGGC	ACGTTCCTCAATGGTGGGTG
CDH5	TGCAAACCTGCGTTTTCCTG	GCTGTGACCGCTCCTATGTT
TRX1	ACTGTATGCCCACGTTCCAG	CTATGACGTGCCAGCTTTTCTT
LAMP	ACGATCAACACCACCTACCG	TGGAGTTACGGTTGTAGCGG
CXCL12	GTCAACAACATCCCACGAAGC	GCTTGGCAATCACTTGGAAGG
IL7R	TGGCAACACACAGGACAACT	TAGGATGAGGGATGCTTGGC
CXCR3	TCTCTTGGGCACCGTACAAC	GGTTCAGAGCACAATGCGAC
CARD9	GGACGGAGTGTGTGAGGATG	TGTCGGAGGTATGGGGTGAT
MKK4	GATAAGCCTGTCTGGAGTGTCC	GACGACTCGATGCTGTGTGT

. Primer validation was performed by running PCR followed by agarose gel electrophoresis to assess primer dimer formation. Serial dilutions of cDNA were used to generate melt curve plots, and only primers with an efficiency greater than 98% were used for subsequent qPCR experiments. QPCR assays were conducted on a Roche LightCycler 480 system using 2× Universal Blue SYBR Green qPCR SuperMix (Servicebio, Wuhan, China). The qPCR cycling program consisted of an initial denaturation at 95 °C for 30 s, followed by 40 cycles of denaturation at 95 °C for 15 s, annealing at 57 °C for 10 s, and extension at 72 °C for 30 s, concluding with a melt curve analysis. Each reaction was performed in triplicate, using β-actin as the reference gene. Relative expression levels were calculated using the 2^−ΔΔCt method, and statistical analysis was conducted using GraphPad Prism 10.0.

3. Results

3.1. Clinical Symptoms and Histopathological Observations

Infected L. crocea initially displayed non-specific symptoms, including lethargy and reduced appetite. As the infection advanced, external bleeding and ulcerations developed, accompanied by hemorrhaging in the fins (Figure 1a) and abdominal swelling (Figure 1b). Post-mortem analysis revealed numerous white nodules in the liver, spleen, and kidney tissues (Figure 1c). Histopathological analysis revealed progressive lesion development in the liver, spleen, head kidney, and trunk kidney over the infection course. At 1 and 3 dpi, hepatocyte deformation with indistinct boundaries was observed, spleen tissues displayed hemorrhage and hemosiderin deposition, and kidney cells showed a decrease in number with associated bleeding. At 7 dpi, granulomatous nodules began forming in these organs, characterized by a fibrous connective tissue layer and a small number of lymphocytes. At 14 dpi, these nodules became larger, well demarcated, and more abundant, encapsulated by an outer layer of lymphocytes and fibrous connective tissue (Figure 2).

3.2. RNA Sequencing and Transcriptome Assembly

After filtering raw data, checking sequencing error rates, and examining GC content distribution, clean reads were obtained for further analyses. Quality assessment of sequencing data from the control and infection groups showed Q20 base percentages above 97.51%, Q30 base percentages above 93.16%, and GC content ranging from 45.81% to 48.6%, with sequencing error rates under 1%. Clean reads for the control group and head kidney tissues at 1, 3, 7, and 14 dpi totaled 41,182,671, 46,961,273, 45,076,950, 44,184,500, and 44,862,975, respectively (

Table A2. Summary of the L. crocea transcriptome assembly.

Samples	Raw Reads	Raw Bases	Clean Reads	Clean Bases	Error Rate	Q20	Q30	GC
N_HK_4	45,270,980	6.79G	43,075,476	6.46G	0.01	98.09	95.44	47.84
N_HK_5	46,974,800	7.05G	44,636,102	6.7G	0.01	98.14	95.57	47.98
N_HK_6	40,982,186	6.15G	38,848,096	5.83G	0.01	98.03	95.37	47.55
N_HK_7	40,274,004	6.04G	38,171,012	5.73G	0.01	98.02	95.28	48.5
P_HK_1d_5	47,345,312	7.1G	44,994,378	6.75G	0.01	98.08	95.48	48.22
P_HK_1d_6	45,743,588	6.86G	43,437,776	6.52G	0.01	98.02	95.3	47.81
P_HK_1d_7	47,512,608	7.13G	45,120,636	6.77G	0.01	98.03	95.33	47.91
P_HK_1d_8	56,976,132	8.55G	54,292,304	8.14G	0.01	97.99	95.32	46.91
P_HK_3d_6	48,648,122	7.3G	46,120,906	6.92G	0.01	97.96	95.2	47.53
P_HK_3d_7	46,225,332	6.93G	44,490,220	6.67G	0.01	98.07	95.47	45.84
P_HK_3d_8	50,153,988	7.52G	47,511,608	7.13G	0.01	97.98	95.27	45.81
P_HK_3d_9	44,439,958	6.67G	42,185,068	6.33G	0.01	98.12	95.55	48.21
P_HK_7d_6	46,642,454	7G	44,231,964	6.63G	0.01	98.01	95.22	48.68
P_HK_7d_7	44,570,078	6.69G	42,994,850	6.45G	0.01	97.78	94.6	47.08
P_HK_7d_8	43,698,142	6.55G	42,023,454	6.3G	0.01	97.87	94.78	47.93
P_HK_7d_9	48,596,398	7.29G	47,487,732	7.12G	0.01	97.51	93.16	46.78
P_HK_14d_6	41,531,408	6.23G	39,872,020	5.98G	0.01	97.69	94.5	47.71
P_HK_14d_7	45,741,040	6.86G	44,079,956	6.61G	0.01	97.86	94.78	48.03
P_HK_14d_8	48,603,658	7.29G	46,527,064	6.98G	0.01	97.71	94.53	46.74
P_HK_14d_9	50,941,480	7.64G	48,972,860	7.35G	0.01	97.88	94.86	48.18

). These metrics confirm that the sequencing data quality is appropriate for subsequent transcriptome analysis. Correlation coefficients based on FPKM values of all genes were calculated to evaluate gene expression correlations among samples, visualized in a heatmap (Figure 3a). PCA and differential gene clustering further highlighted substantial inter-group differences between the control and infection groups, with high reproducibility within each group (Figure 3b,c).

3.3. Identification of Differential Expression Gene

Transcriptome analysis of head kidney tissues infected with N. seriolae revealed a total of 17,629 DEGs across the four infection time points, with 7147 upregulated and 10,482 downregulated genes. In comparison to the control group (N), the P1 (1 dpi) group exhibited 421 upregulated and 1853 downregulated genes (Figure 4a), the P3 (3 dpi) group had 1052 upregulated and 1777 downregulated genes (Figure 4b), the P7 (7 dpi) group showed 3215 upregulated and 3718 downregulated genes (Figure 4c), and the P14 (14 dpi) group presented 2459 upregulated and 3134 downregulated genes (Figure 4d). The progressive increase in DEG numbers suggests a time-dependent immune activation in response to N. seriolae challenge.

3.4. Functional Annotation and Enrichment Analysis

To elucidate the molecular mechanisms underlying the immune response of L. crocea to N. seriolae infection, we performed Gene Ontology (GO) functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses on DEGs using the clusterProfiler package. This approach was employed to identify key biological pathways involved in the L. crocea’s response to N. seriolae infection, thereby elucidating the fundamental molecular mechanisms underlying this process.

GO analysis systematically categorizes gene functions into three ontologies: biological process (BP), cellular component (CC), and molecular function (MF). Using an adjusted p-value (padj) < 0.05, we visualized the 30 most significantly enriched GO terms (or all if fewer than 30) (Figure 5). Enrichment patterns varied temporally. At P1 (1 dpi), DEGs were predominantly enriched in biological adhesion (BP, 8 terms), extracellular region (CC, 4 terms), and receptor regulator activity (MF, 16 terms). By P3 (3 dpi), the enrichment profile shifted toward the oxidation–reduction process (BP, 9 terms), extracellular region (CC, 10 terms), and oxidoreductase activity (MF, 11 terms). At P7 (7 dpi), protein folding (BP, 9 terms), mitochondrion (CC, 27 terms), and unfolded protein binding (MF, 14 terms) showed significant enrichment. The P14 (14 dpi) time point demonstrated the most extensive changes, with DEGs primarily enriched in ribosome biogenesis (BP, 29 terms), mitochondrion (CC, 30 terms), and unfolded protein binding (MF, 15 terms).

KEGG pathway analysis (padj < 0.05) identified key immune-related and metabolic pathways activated during infection. The top 20 most significant pathways (or all if fewer than 20) were plotted (Figure 6). DEG annotation revealed the following: 2274 genes mapped to 150 pathways (10 significantly enriched) at P1 (1 dpi); 2829 genes to 160 pathways (3 enriched) at P3; 6933 genes to 164 pathways (41 enriched) at P7 (7 dpi); and 5593 genes to 165 pathways (37 enriched) at P14 (14 dpi). Notably, immune-related pathways showed dynamic temporal regulation. ECM–receptor interaction and focal adhesion were consistently enriched across all time points. Cytokine–cytokine receptor interaction, cell adhesion molecules, and MAPK signaling pathway were prominent at 1, 7, and 14 dpi, while the intestinal immune network for IgA production emerged at 7 and 14 dpi. Pattern recognition receptor pathways (NOD-like, C-type lectin, and RIG-I-like receptor signaling) were activated throughout infection. Metabolic pathways, including oxidative phosphorylation and citrate cycle (TCA cycle), became significantly enriched only at later stages (7 and 14 dpi).

The Toll-like receptor, NOD-like receptor, RIG-I-like receptor, and C-type lectin receptor signaling pathways facilitate pathogen recognition by detecting distinct molecular patterns, thereby initiating innate immune responses. ECM–receptor interactions and focal adhesion pathways dynamically regulate immune cell migration and tissue localization, while cytokine–receptor interactions orchestrate leukocyte recruitment and activation. Key immune-related genes enriched in these immune pathways include IL-1β and TLR5, which are critical for inflammatory regulation and bacterial detection, respectively. Additionally, CCL19, CCL20, and CXCR2 contribute significantly to immune modulation. These pathways and genes collectively form a coordinated defense network against N. seriolae infection. Furthermore, oxidative phosphorylation and the TCA cycle-central metabolic pathways support the immune response by enhancing ATP production through mitochondrial electron transport and substrate oxidation. The significant enrichment of these pathways suggests a close link between energy metabolism and immune defense in L. crocea.

3.5. qPCR Analysis

To validate the RNA-seq results, eight immune-related genes were selected from the four infection time points for qPCR verification. Three distinct expression patterns were identified (Figure 7): Relative to the N group, the expression levels of CDHE, TRX1, and LAMP progressively increased following N. seriolae exposure, reaching peak expression at 7 dpi. In contrast, CDH5 expression decreased at 1, 3, and 14 dpi, with an increase observed at 7 dpi. The expression levels of IL7R, CXCR3, CARD9, and MKK4 remained lower than pre-infection levels. These qPCR results align with the RNA-seq findings, supporting the reliability of the RNA-seq data.

4. Discussion

N. seriolae poses a major threat to aquaculture, causing significant economic losses, yet its pathogenic mechanisms remain incompletely understood [11]. Since pathogenic manifestations directly reflect host–pathogen interactions [12], we observed the clinical and histopathological features of L. crocea infected with N. seriolae. Notably, yellow-white granulomatous nodules were observed in visceral tissues and the abdominal cavity, showing remarkable consistency with previous reports in Micropterus salmoides [13], Salmo salar [14], Pampus argenteus [15], and Channa argus [16]. The disease progression exhibited distinct temporal patterns: while the initial infection stages (1 and 3 dpi) were asymptomatic, prominent granuloma formation occurred in multiple organs during later stages (7 and 14 dpi), accompanied by progressive tissue damage and hemorrhagic manifestations. Histopathological examination revealed distinct granulomatous nodules in the liver, spleen, head kidney, and trunk kidney during advanced infection stages (7 and 14 dpi), with enlarged nodules featuring defined boundaries and darkened centers. These nodules were encapsulated by a layer of lymphocytes and fibrous connective tissue. Previous studies indicate that N. seriolae can evade phagocytic killing, disseminate within the host, and infect multiple tissues [17]. In response to N. seriolae infection and spread, hosts accumulate immune and epithelial cells around the lesions, leading to granuloma formation [11]. These findings suggest that in the late stages of infection (7 and 14 dpi), L. crocea mounts an immune response by forming granulomatous nodules to counteract N. seriolae.

To further investigate the immune mechanisms of L. crocea against N. seriolae, we performed RNA-Seq analysis on the head kidney of both uninfected and infected fish. KEGG pathway analysis identified significant enrichment in both immune-related (Toll-like receptor signaling, cytokine–cytokine receptor interaction) and metabolic pathways (TCA cycle, oxidative phosphorylation). Notably, key immune genes, including IL-1β, TLR5, CCL19, CCL20, and CXCR2, were significantly upregulated. These findings provide a foundation for subsequent mechanistic discussions regarding host–pathogen interactions. We systematically analyzed and discussed these pathways and immune-related genes to elucidate their functional significance in the host defense mechanism.

Innate immune responses serve as the first line of host defense against pathogenic infections [18]. Upon pathogen entry, pattern recognition receptors (PRRs) detect pathogen-associated molecular patterns (PAMPs) through specific signaling pathways, thereby resisting pathogen invasion and initiating immune responses. This process forms the core of cellular innate immunity [19]. Our KEGG enrichment analysis identified the involvement of four key receptor signaling pathways: C-type lectin receptor (CLR), Toll-like receptor (TLR), NOD-like receptor (NLR), and RIG-I-like receptor (RLR) pathways. This multi-PRR activation pattern is consistent with observations in N. seriolae-infected L. maculatus [20]. It is common for multiple PRRs to be activated simultaneously in response to pathogen invasion [21]. Notably, IL-1β expression was significantly upregulated at all post-infection time points within these four receptor pathways. As a pivotal early-response cytokine, IL-1β orchestrates immune defense by activating monocytes, macrophages, and neutrophils, thereby regulating inflammatory processes [22]. These results demonstrate that L. crocea employs coordinated PRR signaling and IL-1β-mediated inflammation to combat N. seriolae infection.

The Toll-like receptor (TLR) family represents the primary and most extensively studied category of PRRs [23,24]. In fish, the TLR signaling pathway plays a pivotal role in defense against both bacterial and viral infections [25]. Our transcriptomic analysis revealed significant upregulation of TLR5 in the head kidney at 7 and 14 dpi, consistent with observations in silver pomfret infected with N. seriolae [15]. Mechanistically, surface-expressed TLR5 detects bacterial flagellin [26,27], triggering downstream signaling cascades that orchestrate comprehensive immune activation through precise PAMP recognition [28]. These findings establish TLR5 as a key gene in L. crocea’s immune recognition and defense against N. seriolae.

The extracellular matrix (ECM) serves as a critical modulator of post-infection immune responses, mediating complex pathogen–host–immune cell interactions [29]. Our transcriptomic analysis revealed consistent enrichment of ECM–receptor interaction pathways throughout infection, accompanied by the downregulation of most ECM-related genes (laminin, integrin, collagen, and fibronectin), indicating infection-induced ECM remodeling. Such ECM degradation may promote host defense by facilitating immune cell trafficking and releasing bioactive factors [30]. These findings suggest the active participation of ECM dynamics in L. crocea’s immune defense against N. seriolae.

Cytokines are a family of secreted proteins integral to immune regulation and inflammation, including chemokines (CKs), interleukins (ILs), interferons (IFNs), colony-stimulating factors (CSFs), and tumor necrosis factors (TNFs) [31]. Transcriptional profiling revealed the temporal regulation of key cytokine genes: Antimicrobial CCL19 and CCL20 were significantly upregulated at 7 dpi, while CXCR2 and IL1R2 showed sustained induction at 3, 7, and 14 dpi. These chemokines coordinate leukocyte recruitment and differentiation, bridging innate and adaptive immunity [32]. CXCR2 upregulation in L. crocea mirrors previous observations during bacterial and viral challenges [33], reinforcing its central role in antimicrobial defense. The lymphoid chemokine CCL19, which orchestrates antigen-presenting cell migration [34], exhibited infection-responsive expression patterns consistent with reports in A. hydrophila-infected goldfish [35] and V. harveyi-challenged S. ocellatus [36]. Similarly, IL1R2 induction paralleled findings in V. anguillarum-infected S. salar and A. hydrophila-exposed fish [37,38]. Collectively, these results demonstrate that L. crocea mobilizes a conserved cytokine network (CXCR2, IL1R2, CCL19, CCL20) to combat N. seriolae infection through enhanced immune activation.

Focal adhesions (FAs) are essential for cell adhesion, migration, and antimicrobial immunity [39]. Previous studies have shown that the FA signaling pathway plays an active role in the skin immune response of Cynoglossus semilaevis following V. vulnificus infection [40]. Our results demonstrate the sustained downregulation of FA pathway components in N. seriolae-challenged L. crocea. This suppression coincided with visible skin lesions and hemorrhages, likely resulting from both infection-induced tissue damage and husbandry-related trauma. Given the skin’s vital role as a primary physical and immunological barrier [41], the observed FA pathway inhibition suggests compromised barrier function, which may facilitate secondary infections during pathogenesis.

Host metabolic regulation plays a critical role during pathogen infection by influencing immune responses [42]. Transcriptomic analysis revealed a significant activation of oxidative phosphorylation and TCA cycle pathways at 7 and 14 dpi in N. seriolae-infected L. crocea, reflecting heightened energy demands for immune defense [43]. This metabolic shift, characterized by coordinated upregulation of pathway genes, mirrors observations in P. plecoglossicida-infected L. crocea [44], suggesting a conserved metabolic strategy for antibacterial immunity. The elevated energy metabolism likely supports resource-intensive immune processes, including leukocyte proliferation and effector molecule production, during late-stage infection.

5. Conclusions

This study provides a comprehensive transcriptomic analysis of the immune response in the L. crocea head kidney following N. seriolae infection. Our findings reveal (1) the robust activation of immune defense mechanisms, as evidenced by the enrichment of Toll-like receptor signaling, ECM–receptor interaction, cytokine–cytokine receptor interaction, and focal adhesion pathways, alongside the upregulation of key immune mediators (IL-1β, TLR5, CCL19/20, CXCR2); (2) metabolic reprogramming favoring energy production through enhanced TCA cycle activity and oxidative phosphorylation; and (3) histopathological confirmation of granuloma formation as a host defense strategy.

While this study provides valuable insights into molecular immune responses in the head kidney, its scope is limited in terms of organ coverage and mechanistic depth. Future investigations should employ multi-omics approaches (proteomics, metabolomics, and single-cell RNA sequencing) across key immune organs (e.g., spleen, liver). Additionally, in situ hybridization and immunohistochemical analyses would provide further validation and spatial resolution of gene expression patterns. These advancements will deepen our understanding of host–pathogen interactions and contribute to the development of effective disease control strategies.