Virus-Specific Defense Responses in Sweetpotato: Transcriptomic Insights into Resistance and Susceptibility to SPFMV, SPCSV, and SPVD

Abstract

Sweetpotato (Ipomoea batatas L. Lam) production is threatened by complex viral diseases, notably sweet potato virus disease (SPVD) worldwide, which results from co-infection by sweet potato feathery mottle virus (SPFMV) and sweet potato chlorotic stunt virus (SPCSV). This study provides virus-specific transcriptomic insights into the immune responses of three sweetpotato cultivars, 'Beauregard', 'Tanzania', and 'New Kawogo', to SPFMV, SPCSV, and SPVD. Using RNA-seq profiling across three timepoints post-infection at 3, 6, and 12 weeks, we identified distinct virus- and genotype-specific gene expression responses. 'New Kawogo' activated early and sustained immune pathways involving redox regulation, transcriptional control, and hormonal signaling in response to both SPCSV and SPFMV, while showing minimal transcriptional disruption under SPVD, reflecting robust tolerance. 'Beauregard' exhibited early suppression of immune and metabolic genes, with delayed and disorganized recovery efforts, particularly under SPVD. Defense-related pathways including NBS-LRR signaling, RNA silencing, and hormonal regulation were consistently upregulated in 'New Kawogo' and to a lesser extent in 'Tanzania', but remained inactive in 'Beauregard'. This study highlights candidate resistance and susceptibility genes for each virus, providing a molecular basis for developing virus-resilient sweetpotato cultivars through functional genomics and marker-assisted breeding. These findings elucidate the molecular basis of virus resistance in sweetpotato and identify candidate genes for marker-assisted breeding, despite limitations arising from the use of a diploid reference genome and discrete sampling intervals.

Keywords: sweet potato chlorotic stunt virus (SPCSV); sweet potato feathery mottle virus (SPFMV); sweet potato virus disease (SPVD); sweetpotato; transcriptome profiling.

Figure 1

Symptoms observed on leaves of plants co-infected with SPCSV and SPFMV at different timepoints. (A) Graft inoculation using I. setosa. (B) ‘Tanzania’ at 6 WPI showing mild symptoms of veinal chlorosis. (C) ‘Beauregard’ at 6 WPI showing vein clearing, mosaic, and roll down. (D) ‘New Kawogo’ at 6 WPI showing no symptoms.

Figure 2

Bar charts illustrate the mean percentage of virus-derived small interfering RNA (vsiRNA) reads mapped to SPFMV and SPCSV genomes in (A) ‘New Kawogo’ and (B) ‘Tanzania’ across mock, SPFMV, SPCSV, and SPVD treatments at 3, 6, and 12 WPI.

Figure 3

Principal component analysis of sweet potato transcriptional responses to viral infection. (A) Cultivar and temporal clustering. Samples are colored by timepoint (red: 3 weeks; green: 6 weeks; blue: 12 weeks) and shaped by cultivar (circle: ‘Beauregard’; triangle: ‘Tanzania’; square: ‘New Kawogo’). (B) Treatment and temporal clustering. Samples are colored by timepoint and shaped by treatment (circle: mock; triangle: SPCSV; square: SPFMV; cross: SPVD). Principal components explain 44% of total variance (PC1: 23%; PC2: 21%).

Figure 4

Expression patterns of top 50 differentially expressed genes. Heatmap shows row-scaled log₂ fold change values for the 50 most significantly differentially expressed genes (lowest adjusted p-values) across treatment comparisons. Columns represent individual comparisons organized by cultivar (‘Beauregard’, ‘Tanzania’, ‘New Kawogo’), treatment (SPCSV, SPFMV, SPVD), and timepoint (3, 6, 12 weeks). Hierarchical clustering was performed for both genes (rows) and comparisons (columns). Values represent z-scored log₂ fold changes relative to control conditions.

Figure 5

UpSet plot of DEG intersections across treatment conditions. (Bottom panel) Set sizes (number of DEGs) for each condition labeled as Virus_Cultivar_Timepoint. (Top panel) Intersection sizes between condition combinations, with connected dots indicating which sets contribute to each intersection. Sets are ordered by frequency, with intersections showing both shared and condition-specific differential expression signatures across viral treatments, cultivars, and timepoints.

Figure 6

Temporal dynamics of differential gene expression across virus–cultivar combinations. Line plots show log10-transformed DEG counts over time (3, 6, 12 weeks) for each virus–cultivar combination. Red lines indicate upregulated genes (log₂FC ≥ 0.5); blue lines indicate downregulated genes (log₂FC ≤ −0.5). Panels are organized by virus treatment (rows) and cultivar (columns). DEGs were filtered using padj < 0.05 and |log₂FC| ≥ 0.5 thresholds.

Figure 7

Functional enrichment analysis of differentially expressed genes across treatment conditions. Dot plots show significantly enriched gene ontology driver terms (p < 0.05, g:SCS correction) for downregulated (left) and upregulated (right) genes within each condition. Driver terms represent the most statistically robust and biologically relevant enrichments. Terms are organized by GO domain: biological process, molecular function, and cellular component. Dot size represents DEG count contributing to enrichment; dot color represents −log10(p-value). Conditions are labeled as Cultivar_Virus_Timepoint.

Figure 8

Defense pathway differential expression across virus–cultivar–timepoint combinations. Heatmap shows unscaled log₂ fold change values for defense pathway gene sets relative to mock controls. Red indicates upregulation; blue indicates downregulation. Pathways include cell wall modification, NBS-LRR resistance, protein quality control/HSPs, RNA silencing, oxidative stress response, hormone signaling (SA and JA/ET), RNA decay, miRNA regulation, pattern recognition, and secondary metabolism. Conditions are labeled as Cultivar_Virus_Timepoint and organized by cultivar (‘Beauregard’: red; ‘New Kawogo’: orange; ‘Tanzania’: blue), treatment (SPCSV: red; SPFMV: blue; SPVD: green), and timepoint (3, 6, 12 weeks).