Fig 1.
Schematic representation of sample collection, processing, and sequencing.
Liq. N2 –Liquid nitrogen.
Table 1.
Characteristics of the study subjects.
Table 2.
HLA types and super-types for paired lesion and control.
Fig 2.
Comparison of cellular gene expression between healthy donors’ skin and control skin from KS patients in a context of differences versus KS lesion.
(A) Scatter plot of Lesion/control (mean lesion/mean control from the same patient) and Lesion/normal (mean lesion of all the patients/mean healthy skin samples) expression log2-ratios. Regression line is shown as a green dotted line. Gray dots indicate genes whose expression was not significantly different (p>0.005) between healthy skin from donors and control skin from KS patients, squares show genes whose expression was significantly different between healthy skin from donors and control skin from KS patients at P<0.005 (N = 184) with red indicating up- and blue–down-regulation in control as compared to normal skin. Circle marks gene with expression significantly different between healthy skin from donors and control skin from KS patients at FDR<5% (N = 1). FDR–False discovery rate.
Fig 3.
Comparison of transcriptome profiles (KS lesion vs control) between EpKS and EnKS patients.
(A) A scatter plot comparing transcriptome profiles between EpKS and EnKS patients. Green dotted line denotes actual regression line; orange dotted line denotes expected relationship without any difference between groups. Grey dots denote genes whose expression was not significantly different between EpKS and EnKS patients. Black dots denote genes whose expression was nominally significantly different between EpKS and EnKS patients at P = 0.005 (N = 1112). Filled circles denote genes whose expression was significantly different between EpKS and EnKS patients at FDR<5% (N = 428). FDR–False discovery rate. (B) Heatmap of top coding genes that are significantly different in KS lesion vs controls and significantly more dysregulated in EnKS than EpKS patients. (C) Comparison of KS lesion vs control differences for 428 genes that are significantly different between EnKS and EpKS indicates that majority of genes are not uniquely dysregulated in endemic KS, but that dysregulation is more extreme in EnKS compared to EpKS. White filled circles highlight 29 genes that had at least 1.5-fold KS lesion/control fold change in one group, but less than 50% of the fold in another group. EnKS–Endemic Kaposi’s sarcoma, EpKS–Epidemic Kaposi’s sarcoma.
Fig 4.
Comparison of transcriptome profile in KS patients by gender and by ART-use.
(A) A scatter plot comparing transcriptome profiles between male and female EpKS patients. Green dotted line denotes regression line. Grey dots denote genes whose expression was not significantly different between male and female EpKS patients, black dots denote genes whose expression was significantly different between male and female EpKS patients at P = 0.005 (N = 34). (B) A scatter plot comparing transcriptome profiles between ART treated and ART naïve KS patients. Green dotted line denotes regression line. Gray dots denote genes whose expression was not significantly different between ART treated and ART naïve KS patients, black dots denote genes whose expression was significantly different between ART treated and ART naïve KS patients at P = 0.005 (N = 46) and unfilled circles denote genes whose expression was significantly different between ART treated and ART naïve KS patients at FDR<5% (N = 2). FDR–False discovery rate. EnKS–Endemic Kaposi’s sarcoma, EpKS–Epidemic Kaposi’s sarcoma.
Fig 5.
Hierarchical clustering of KSHV genes and samples.
ART–Antiretroviral therapy, KSHV–Kaposi’s sarcoma (KS)-associated herpesvirus, en–endemic KS, ep–epidemic KS, M–Male, F–Female, A–ART treated, H–ART naïve, n–HIV-1 negative.
Fig 6.
Correlation between KSHV RNA load/reads and cellular gene expression.
(A) Principal component analysis (PCA) showing relationship between first and second principal components (PC1, PC2) and KSHV RNA load/RNA in KS lesion vs control. (B) Heatmap for top 50 genes correlated with KSHV RNA load/read that were also up- or down-regulated in KS lesions vs controls. KSHV–Kaposi’s sarcoma (KS)-associated herpesvirus.
Fig 7.
Correlation of cellular gene expression in KS lesion/control between previous and current study.
(A) Heatmap of lesion/control difference in expression from 21 new sample pairs for 3589 up- or down-regulated cellular genes reported in the previous study. (B) Correlation of cellular genes differences between KS lesion vs control reported in the previous study versus current study. Top changes for 30 up- and down-regulated genes (>4-fold change) unique to the current study are highlighted in red and blue, respectively. Correlation line is shown by dotted orange line. FDR–False discovery rate. KSHV–Kaposi’s sarcoma (KS)-associated herpesvirus.
Fig 8.
Expression heatmap for cellular genes.
(A) Overlapped with previous study and (B) uniquely found in the current study. FDR–False discovery rate. KSHV–Kaposi’s sarcoma (KS)-associated herpesvirus.