Persistent foot-and-mouth disease virus infection of the bovine nasopharynx is associated with suppression of innate and cellular immunity

Fig 1. Tissue sample characteristics. (A) Ratio of FMDV RT-qPCR positive samples for DNP (green) and DSP (blue) samples. (B) Boxplot showing FMDV-specific RT-qPCR Cq values as correlate for viral load in DNP and DSP samples. Cq values for both tissue types were compared using the Wilcoxon rank-sum test. (C) Principal component analysis (PCA) of gene expression in selected DNP and DSP samples (n = 52) based on the normalized transcript abundancies of the 1000 transcripts with the highest variance within the dataset. The PCA illustrates the overall variance structure of the dataset, with PC1 and PC2 capturing the major axes of separation among samples. FMDV-specific RT-qPCR Cq values are indicated by color. A distinct cluster of highly FMDV-positive DSP samples is highlighted by a red area.

Fig 2. Proteomic analysis. (A) Principal component analysis (PCA) performed using normalized protein abundance values from the proteomic analysis of DSP samples. The PCA illustrates the overall variance structure of the dataset, with PC1 and PC2 capturing the major axes of separation among samples. Samples are grouped according to their FMDV Cq value: high FMDV-positive (Cq ≤ 30) in red, low FMDV-positive (Cq > 30) in orange and FMDV-negative (no Cq) in blue-gray. (B) Analysis of enriched Gene Ontology Biological Process (GOBP) terms derived from proteins differentially expressed between high FMDV-positive and FMDV-negative DSP samples. Each point represents a pathway term, plotted by log2 enrichment on the x-axis and –log10 adjusted p-value on the y-axis. Positive log2 enrichment indicates upregulation of pathways, negative log2 enrichment indicates downregulation. Red points denote pathways with statistically significant enrichment, whereas grey points represent nonsignificant terms.

Fig 3. Gene expression of follicle-associated epithelium marker genes in DSP samples. Samples are grouped by their FMDV Cq value into highly positive (Cq ≤ 30, red), weakly positive (Cq > 30, orange) and negative (no Cq, grey area). Gene counts were normalized using a regularized log transformation (rlog) and the result of a linear regression is shown. Correlation between normalized gene counts and FMDV genome load was assed using Spearman’s rank correlation coefficient (Rs). Rs ranges from −1 to 1 and values closer to 1 or −1 indicate a strong positive or negative monotonic relationship, respectively. Rs = 0 indicates no monotonic association. A correlation was considered statistically significant if the p-value was less than 0.05.

Fig 4. Gene expression of tight junction components, T-cell proliferation and markers for regulatory T cells in DSP samples. (A) Gene expression of the components of cellular tight junctions occludin (OCLN), claudin-1 (CLDN1) and the tight junction protein 1 (TJP1/ZO-1). (B) Gene expression of important immune checkpoints for regulating T-cell proliferation: PD-1, CTLA4 and LAG-3. (C) Gene expression of marker genes for Treg cells. Samples are grouped by their FMDV Cq value into highly positive (Cq ≤ 30, red), weakly positive (Cq > 30, orange) and negative (no Cq, grey area). Gene counts were normalized using a regularized log transformation (rlog) and the result of a linear regression is shown. Correlation between normalized gene counts and FMDV genome load was assed using Spearman’s rank correlation coefficient (Rs). Rs ranges from −1 to 1 and values closer to 1 or −1 indicate a strong positive or negative monotonic relationship, respectively. Rs = 0 indicates no monotonic association. A correlation was considered statistically significant if the p-value was less than 0.05.

Fig 5. Gene expression of selected genes of the innate immune response signalling cascade in DSP samples. (A) Genes of the TLR pathway including TLR7, TLR8, MYD88 and RNF152. (B) Genes involved in the detection of viral RNA: the RNA-sensing receptors RIG-I and MDA5, IFI16, which exerts a positive influence on RIG-I expression, as well as LRRC25, possibly upregulated by FMDV 3A and exerting a negative influence on RIG-I and MDA5 expression. (C) Downregulated genes downstream the TLR, RLR and NLR pathways including TRAF6 and MAPK13 expression. Samples are grouped by their FMDV Cq value into highly positive (Cq ≤ 30, red), weakly positive (Cq > 30, orange) and negative (no Cq, grey area). Gene counts were normalized using a regularized log transformation (rlog) and the result of a linear regression is shown. Correlation between normalized gene counts and FMDV genome load was assed using Spearman’s rank correlation coefficient (Rs). Rs ranges from −1 to 1 and values closer to 1 or −1 indicate a strong positive or negative monotonic relationship, respectively. Rs = 0 indicates no monotonic association. A correlation was considered statistically significant if the p-value was less than 0.05.