The P. multocida strain used in our experiments was strain HN01, serotype D (GenBank: CP037861.1), originally isolated from lung tissue obtained from a Hainan black goat and preserved in our laboratory. The bacterial culture was grown in tryptic soy broth containing 5% newborn calf serum (Sijiqing, China).

Goat bronchial epithelial cells (Hefei Wanwu Biotechnology Co., Ltd., China) were cultured in a specialized medium containing 6% fetal bovine serum and 1% penicillin–streptomycin and maintained under standard conditions at 37 °C in a humidified 5% CO₂ incubator.

For lentiviral packaging, two FCGBP-shRNA lentiviral vectors (pLV3tltr-ZsGreen-Puro-U6-FCGBP-i1 and pLV3tltr-ZsGreen-Puro-U6-FCGBP-i2) and a negative control vector (pLV3tltr-ZsGreen-Puro-U6-NC) were co-transduced with two packaging plasmids into 293T cells using Lipo293™ transduction reagent (Beyotime Biotechnology, China). After 48 h of transduction, the viral supernatant was harvested and sterile-filtered through a 0.45-μm membrane and concentrated with PEG to obtain high-titer viral stocks. The viral titer was determined using the serial dilution method and calculated based on the number of cells positive for green fluorescent protein (GFP) observed under a fluorescence microscope. Calculations revealed a final titer of 5.3 × 10⁸ TU/mL for FCGBP-shRNA-i1, 5.5 × 10⁸ TU/mL for FCGBP-shRNA-i2, and 5.1 × 10⁸ TU/mL for NC-shRNA. To determine the optimal multiplicity of infection (MOI), goat bronchial epithelial cells were distributed into 96-well plates at 1.5 × 10⁴ cells per well. The cells were divided into a nontransfection control group and an NC-shRNA lentivirus transfection group. After 24 h, the NC-shRNA lentivirus transfection group was incubated in antibiotic-free medium with NC-shRNA lentivirus at MOI values of 2, 5, 10, 20, 50, 100, and 200, along with 6 μg/mL polybrene. All experiments were performed in duplicate. After 8 h of incubation, the culture medium was replaced with fresh complete medium supplemented with antibiotics. Transfection efficiency was assessed by observing GFP fluorescence 24 h after transfection.

To verify whether lentiviral transfection has cytotoxic effects, the CCK-8 assay was employed to assess cell viability after the lentiviral transfection. Cells were plated in 6-well plates at a density of 3 × 10⁵ cells per well and divided into four groups: Ctrl (untreated control), NC (NC-shRNA transfection), F (FCGBP-shRNA-i2 transfection), and blank (blank control). After 10 h, the NC and F groups were transfected with the corresponding lentivirus at an MOI of 200. Following a 10 h incubation, the lentivirus-containing medium was removed, and cells were washed once with phosphate-buffered saline (PBS) before adding fresh complete medium. After 24 h, cells were trypsinized and replated into 96-well plates at 1 × 10⁵ cells per well.

After 12 h, the culture medium was replaced with 90 μL of serum-free medium, and 10 μL of CCK-8 reagent (CoWin Biosciences Co., Ltd., China) was added to each well. Absorbance at 450 nm (OD450) was measured using a microplate reader at 2, 4, 6, 24, and 48 h after the addition of CCK-8 reagent. Each experimental group included six replicate wells, and the entire procedure was performed in three independent experiments (n = 3).

To select higher efficiency interference fragments, we compared the interference efficiency of two different lentiviral interference fragments. Cells were plated into 24-well plates at a density of 5.5 × 10⁴ cells per well and assigned to the following groups (with two replicates per group): control (no transfection), F1 (transfected with FCGBP-shRNA-i1), and F2 (transfected with FCGBP-shRNA-i2). At 24 h after seeding, the F1 and F2 groups were transfected with their respective lentiviruses at MOI 100. To enhance transfection efficiency, the viral supernatant was supplemented with 3 μL of polybrene (6 μg/mL). After 8 h of incubation, the viral supernatant was removed, and the cells were subjected to two gentle washes with PBS before adding fresh complete medium containing 1% penicillin–streptomycin.

Transfection efficiency was preliminarily assessed by observing GFP expression under a fluorescence microscope after 48 h of transfection. Then, RNA was extracted from the cells using the TRIzol method (1 mL per well), and the levels of FCGBP mRNA expression were quantified by qRT-PCR to identify the most interference-efficient shRNA fragment.

To further evaluate the efficiency of FCGBP protein interference, we assessed the expression levels of FCGBP protein via Western blotting. Cells were plated in 6-well plates at a density of 3 × 10⁵ and assigned to Ctrl (control) and F (FCGBP-i2-shRNA transfection) groups. After 10 h of lentiviral transfection, the medium was replaced with fresh complete medium. Following 48 h, cells were lysed using IP lysis buffer (Servicebio Technology Co., Ltd., China). Supernatants were collected after centrifugation, supplemented with 5× SDS-PAGE Protein Loading Buffer (Servicebio Technology Co., Ltd., China), and boiled in a metal bath at 96 °C for 10 min. Samples were separated on a 15% Bis–Tris gel and transferred to a polyvinylidene difluoride membrane. Western blotting was performed using primary antibodies against FCGBP (Cloud-Clone PAP389Hu01) and GAPDH (Proteintech 1E6D9). After incubation with the appropriate secondary antibodies, protein bands were visualized using the Special Supersensitive ECL Chemiluminescence Kit (Beyotime P0018AS). Each group included three technical replicates.

To select the optimal infection MOI and the optimal infection time, we conducted a selection of infection conditions. Cells were distributed into 24-well plates at a density of 5 × 10⁴ cells per well. After 24 h of incubation, the culture medium was changed to antibiotic-free medium. Based on the preliminary work conducted in our laboratory, we selected the optimal infection conditions by testing a range of MOIs (10, 50, 100, and 200) for P. multocida. With two replicate wells per MOI, followed by co-incubation. Cellular morphology was monitored at 2, 4, and 6 h post-infection (hpi) to define the optimal conditions for infection, i.e., the minimal MOI and earliest time point inducing a significant CPE.

After determining the optimal infection conditions, we conducted a full-scale infection experiment. Cells were plated into 6-well plates at a density of 3 × 10⁵ cells per well and assigned to the following six experimental groups (with three replicates per group): Ctrl (untreated control), CP (P. multocida infection), NC (NC-shRNA transfection), NCP (NC-shRNA transfection + P. multocida infection), F (FCGBP-shRNA-i2 transfection), and FP (FCGBP-shRNA-i2 transfection + P. multocida infection). The CP, NCP, and FP groups were infected with P. multocida at an MOI of 100 for 4 h following a 24 h period. After infection, the supernatant from each well was collected for subsequent analysis via ELISA. The cells were then washed three times with PBS and lysed directly in the wells using TRIzol reagent. The resulting lysates were snap-frozen in liquid nitrogen and maintained at −80 °C until subsequent RNA extraction.

Each experimental group included three technical repetitions. A total of 18 cell samples from different treatment groups were sent to Guangzhou GENEDENOVO Biotechnology Co., Ltd. (Guangzhou, China) for RNA sequencing. Total RNA was isolated using TRIzol reagent. After extraction, RNA integrity was assessed by agarose gel electrophoresis and using an Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The Hieff NGS® Ultima Dual-mode mRNA Library Prep Kit was employed to prepare sequencing libraries, which were then subjected to quality control using the DNA 1000 Assay Kit (Agilent Technologies, 5067-1504, USA). The libraries with sufficient quality were subsequently sequenced on an Illumina NovaSeq X Plus platform for transcriptome analysis.

Following RNA-seq, the raw data were subjected to quality control and filtered using fastp software.¹ The filtered reads were then mapped to the reference genome in HISAT2 (17). Gene expression levels were estimated using RSEM (18), and differential expression analysis was performed using DESeq2 (19). The OmicShare tool suite² was used for data visualization and functional analyses, which specifically included principal component analysis (PCA), heatmap generation, and Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Finally, protein–protein interaction (PPI) networks were generated using the STRING database³ and visualized in Cytoscape (v3.10.2).

To validate the dysregulated immune pathways identified by RNA-seq, the levels of secreted cytokines. The levels of key proinflammatory cytokines (IFN-α, TNF-α, CCL5, IL-17, and IL-6) were measured by ELISA kits (Xiamen Lun Changshuo Biotechnology Co., Ltd., China) according to the manufacturer’s instructions. Optical density was promptly measured at 450 nm for each well to calculate sample concentrations.

To validate the results of RNA-seq, genes ISG15, RSAD2, TRANK1, IFIH1, DDX58, IMP3, RRP9, FADD, PIK3CA, and COL1A2 were selected for analysis via qRT-PCR. Total RNA was first reverse-transcribed into cDNA using the 5× FastKing-RT SuperMix (Tiangen Biotech, Beijing, China). Then, qRT-PCR was conducted on a Langji Q2000B real-time system (Langji, Hangzhou, China) using the SYBR Green Pro Taq HS Premixed qPCR Kit (Hunan Aikrui Bioengineering, China). Gene-specific primers (Supplementary Table S1) were designed using the NCBI Primer-BLAST tool.⁴ The gene encoding glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal reference, with relative gene expression determined via the comparative 2^−ΔΔCT method. The data obtained from qRT-PCR are presented as mean ± standard deviation (SD) of three independent biological replicates, each biological replicates measured in 3 technical replicates.

Data are presented as mean ± SD from at least three biological replicates (n = 3). Differences between groups were analyzed using one-way ANOVA followed by multiple comparisons tests in GraphPad Prism (v10.1.2). Statistical significance was set at p < 0.05.

Goat bronchial epithelial cells exhibited an irregular, cobblestone-like shape under bright-field microscopy (Figures 1A,B). To establish optimal conditions for transfection, cells were transfected with lentiviral particles at MOI values of 2, 5, 10, 20, 50, 100, and 200. Fluorescence imaging at 24 h post-transfection (Figures 1C–I) demonstrated a positive correlation between MOI and transfection efficiency. To ensure that subsequent phenotypic observations were not influenced by cytotoxicity, we first evaluated whether the cell viability was affected by lentiviral transfection. We assessed the impact of lentiviral transfection on cell viability using CCK-8 assays. Results showed that at 48 h, the NC group exhibited a cell viability rate of 111%, while the F group demonstrated a 97% cell viability rate (Figure 1J). No significant cytotoxicity or morphological changes were induced by the lentivirus across the MOI range tested. As the maximal fluorescence efficiency was observed at MOI 200, this value was selected for all subsequent experiments.

To identify the most effective FCGBP-targeting shRNA fragment, goat bronchial epithelial cells were transfected at MOI 200 with one of two distinct lentiviral constructs: group F1 (FCGBP-shRNA-i1) and group F2 (FCGBP-shRNA-i2). As shown in Figures 2A–D, qRT-PCR analysis confirmed that FCGBP-shRNA-i2 achieved a superior knockdown efficiency compared to FCGBP-shRNA-i1. Therefore, the FCGBP-shRNA-i2 construct at MOI 200 was selected for all subsequent experiments (Figure 2E). Western blotting analysis showed that FCGBP expression was significantly reduced in the F2 (FCGBP-i2-shRNA) group compared with that in the Ctrl group (Figure 2F).

The goat bronchial epithelial cells in the Ctrl group maintained an irregular cobblestone-like shape throughout the observation period (6 h) (Figures 3A–C). The changes in cell morphology following infection with P. multocida at various MOIs (10, 50, 100, and 200) over time are summarized in Figures 3D–O. The onset of cell death was observed across all infected groups as early as 2 hpi, though without notable morphological changes, and within 4 hpi, cell death became more pronounced. Concurrently, cells in the group at 200 MOI began to exhibit a slightly round shape. This progressive rounding was evident in both the MOI 100 and MOI 200 groups by 6 hpi. Based on these observations, a robust infection model at MOI 100 with an infection duration of 6 h was established for the subsequent experiments involving the CP, NCP, and FP groups.

Quality control and filtering of the raw RNA-seq reads in fastp software involved removing reads containing adapters, those with >10% N bases, low-quality reads (where >50% of bases had a Phred score ≤20), and poly-A reads (Supplementary Table S2). The clean reads were then aligned to the reference genome (Assembly GCA_001704415.2), achieving a high mapping rate of over 93% across all samples (Supplementary Table S3).

Using the high-quality sequencing data, we first evaluated the overall transcriptomic differences among the samples. PCA of the transcriptomic data revealed a clear segregation of the experimental groups (Figure 4A). The first two principal components (PC1 and PC2) explained 70.9 and 18.1% of the total variance, respectively. Notably, the Ctrl and CP groups formed distinct clusters, separate from the NC and F groups as well as from the FP and NCP groups.

The correlation heatmap generated for gene expression profiles further supported these findings (Figure 4B). As expected, biological replicates within the same group clustered together, demonstrating high correlation. In contrast, the Ctrl and CP groups showed low levels of correlation with the other treatment groups.

Following quality control, DEGs were identified from the RNA-seq data. For the NCP versus FP comparison, significance was defined as p < 0.05 and |log2FC| > 1, whereas for all other comparisons, an FDR < 0.05 and |log2FC| > 1 were applied. Comparative transcriptomic analysis revealed a profound impact of FCGBP knockdown on the host response to infection. Infection alone induced 50 DEGs for Ctrl versus CP (49 upregulated and 1 downregulated) and 695 DEGs for NC versus NCP (376 upregulated and 319 downregulated), FCGBP-knockdown cells exhibited a dramatically amplified response with 802 DEGs for F versus FP (373 upregulated and 429 downregulated), 1,515 DEGs for CP versus FP (1,000 upregulated and 515 downregulated) (Figure 5A), and 70 DEGs for NCP versus FP (47 upregulated and 23 downregulated) (Supplementary Figure S1A), with the CP versus FP group comparison exhibiting the most substantial transcriptional alterations. We annotated the top five significantly upregulated and downregulated DEGs for each comparison group (Figures 5B–E, Supplementary Figure S1B).

To functionally characterize the DEGs, GO enrichment analysis was conducted for the top 20 terms with the smallest Q-values (Figure 6). In groups infected with P. multocida DEGs were enriched in the type I interferon signaling pathway and defense response pathways (Figures 6A,C,D), whereas the Ctrl versus F group showed no enrichment in these immune-related pathways, confirming successful immune activation in goat bronchial epithelial cells upon P. multocida infection. We observed from the GO pathway enrichment analysis of the Ctrl versus F comparison group (Figure 6B) that under non-infected conditions, enriched pathways predominantly clustered around Cell part, Cell, and Organelle pathways. This indicates that under normal physiological conditions, the primary function of the FCGBP is associated with maintaining the fundamental structural integrity of cells. Interestingly, we observed, in the comparison between the Ctrl and F groups at the transcriptomic level, that IL-6 gene expression was significantly downregulated, indicating that FCGBP deficiency leads to reduced expression of certain fundamental immune factors. This finding was confirmed by ELISA (Figure 6F, p = 0.0153), indicating that FCGBP deficiency compromises baseline immune signaling even in the absence of pathogen challenge. Further supporting its role in structural integrity, in the Ctrl vs. F group comparison, DEGs were also associated with pathways like “biological adhesion” and “cell junction” (Supplementary Figure S3), reinforcing the notion that FCGBP contributes to epithelial barrier maintenance. In infected cells, the CP versus FP comparison revealed DEGs enriched in “cell surface receptor signaling”, “response to lipopolysaccharide,” and “response to molecules of bacterial origin” (Figure 6D). Similarly, NCP versus FP comparisons showed DEGs enriched in the cell periphery and plasma membrane, further suggesting that FCGBP maintains the structural integrity of the cell membrane during infection (Supplementary Figure S2). This pivotal finding indicated that the FCGBP is closely related to the cell barrier and cell surface receptor signaling pathways and directly participates in the host cell’s perception and defense response against bacteria.

This role was further confirmed by KEGG pathway analysis (top 20 pathways based on the smallest Q-value; Figure 7). While the DEGs identified in the Ctrl versus CP group comparison were enriched in core inflammatory pathways (e.g., NLR, TNF, and cytokine–cytokine receptor interaction pathways; Figure 7A), those detected in both the NC versus NCP and F versus FP group comparisons showed additional enrichment in the C-type lectin receptor signaling pathway and Fc gamma R-mediated phagocytosis (Figures 7B,C), underscoring the broad involvement of the FCGBP in pulmonary immunity. Most significantly, the CP versus FP group comparison exhibited the most pronounced inflammatory signature, with DEGs highly enriched in the NLR, TNF, TLR, NF-kappa B, and cytokine–cytokine receptor interaction pathways (Figure 7D). This demonstrated that FCGBP deficiency exacerbates the inflammatory response to P. multocida infection. To identify a consensus immune signature, the overlap of DEGs across the four group comparisons was then analyzed, revealing 28 common DEGs, as shown in the Venn diagram in Figure 8A. PPI network analysis of these genes highlighted a core set of hub genes, with the top 10 ordered as follows based on the degree value: ISG15, OAS1, IFIH1, RSAD2, MX1, IFI44, DDX58, RTP4, MX2, XAF1, and CMPK2 (Figure 8B). The central position of these genes in the network suggests they play an essential regulatory role in the host response to P. multocida infection.

To further elucidate the functional relationships among dysregulated genes identified in the preceding pathway analyses, protein–protein interaction (PPI) networks were constructed. PPI analysis reinforced the transcriptomic findings, revealing two major dysregulated modules in FCGBP-deficient infected cells (Figure 9). First, a core cluster of interferon-stimulated genes (e.g., ISG15, MX1, RSAD2, OAS1, and IFIH1) was hyperactivated, quantitatively confirming the amplified interferon response. Concurrently, a module comprising genes essential for cell adhesion and junction integrity (e.g., ITGAE, NGFR, GDNF and AQP2) was suppressed. These findings indicate that FCGBP deficiency may compromise barrier integrity and disrupt fundamental cellular processes, potentially leading to a stressed and dysfunctional state in infected cells.

To validate the transcriptional changes at the protein level, we quantified the concentrations of secreted cytokines. Quantification of the concentrations of IFN-α, TNF-α, CCL5, IL-17 across different treatment groups via ELISA revealed that the levels of all five cytokines in the CP, NCP, and FP groups were significantly upregulated compared to those in their respective uninfected controls (Ctrl, NC, and F groups). Notably, FCGBP knockdown exerted distinct, cytokine-specific effects. Compared to the CP group, the FP group exhibited significantly lower levels of IL-17 and CCL5, whereas IFN-α concentration was markedly elevated (Figure 10). These protein levels are fully consistent with the transcriptomic profiles.

To experimentally validate the transcriptomic data, the expression levels of the 10 above-mentioned hub genes consistently identified in the PPI networks were quantified by qRT-PCR. As shown in Figure 11, a strong concordance between the results of RNA-seq and those of qRT-PCR was observed for 9 out of 10 genes (ISG15, RSAD2, TRANK1, IFIH1, DDX58, IMP3, RRP9, PIK3CA, and COL1A2), thereby confirming the overall accuracy of our sequencing analysis. An exception was the FADD gene, which showed a divergent expression pattern. This discrepancy is likely attributable to differences in transcript variants or post-translational modifications, which can lead to inconsistent results between RNA-seq and qRT-PCR assays.