Introduction

Kaposi sarcoma-associated herpesvirus (KSHV), which is also known as human gammaherpesvirus 8 (HHV-8), is a member of the rhadinovirus genus, and was first identified in 1994 [1]. The highly-conserved, circular, 165-kb double stranded DNA genome of KSHV has a 140-kb unique region encoding approximately 90 genes flanked by 20-30 kb of terminal repeat sequences [2]. The viral genome is maintained as an episome in infected cells and persists in a latent state during which it expresses a latency-associated nuclear antigen (LANA, ORF73), Kaposin (K12), v-FLIP (K13), v-Cyclin (ORF72), and 12 microRNAs (miRNAs) [3]. Induction of viral replication and lytic gene expression, often by inflammation, promotes the expression of the Replication Transactivation Activator (RTA, ORF50), and a resulting cascade of secondary and tertiary viral proteins that make the virus capsid and DNA synthesis enzymes.

The etiological agent of Kaposi sarcoma (KS), KSHV, exists in at least 5 subtypes [4] and is endemic in sub-Saharan Africa, parts of Eastern Europe, the Mediterranean, and parts of China, where rates can range from 30-90% [5]. In the U.S. and many high-resource nations the prevalence of KSHV infection is low in the general population, but substantially elevated in high-risk groups such as HIV-1 infected individuals who have sex with men and in immunosuppressed subjects. Saliva is the major route of KSHV transmission and in endemic regions of the world, most infections occur within the first 5 years of life [6]. KSHV is a class I carcinogen and about 1% of human tumors are associated with KSHV infection [7]. Fewer than 1% of immunocompetent KSHV infected individuals develop disease, however most immunosuppressed individuals infected with the virus manifest one (or more) disorders [8] including KS, multicentric Castleman disease (MCD), KSHV inflammatory cytokine (KICS), immune reconstitution syndromes (IRIS), and primary effusion lymphoma (PEL).

KS is an incurable disease originally described as a blood vessel tumor in 1872 by Hungarian Moritz Kaposi, and is now understood to be a highly vascularized solid tumor of endothelial origin, characterized by KSHV-positive “spindle cells”, cellular pleomorphism, an inflammatory infiltrate of lymphocytes and plasma cells, sinuous vascular spaces, extravasated erythrocytes, and fibrosis [9]. Forms of KS include classic KS (cKS), iatrogenic immunosuppression-associated KS (iKS), endemic KS (enKS), and epidemic HIV-1/AIDS-associated KS (epKS) [10]. The enKS and cKS are often indolent, whereas epKS can have widespread mucocutaneous, nodal, and visceral involvement [9]. Although the majority of infected cells in a KS lesion manifest KSHV in latency, lytic reactivation is a critical step in oncogenesis [11]. KSHV infection of endothelial cells or hematopoietic progenitors leads to changes in their morphology, glucose metabolism, proliferation, lifespan, and gene expression. KSHV oncogenicity is reflected by numerous pro-angiogenic molecules that are induced, including members of the vascular endothelial growth factor (VEGF)-VEGF receptor and angiopoietin families, Interleukin 6 (IL6) and IL8, and platelet-derived growth factor, through the activities of lytic proteins K1, K15, and viral G protein coupled receptor (v-GPCR) [12]. Latency proteins contribute to tumorigenesis through repression of apoptosis (LANA, v-FLIP), and activation of cyclin-dependent kinase (v-Cyclin). KSHV also elaborates an array of mediators of immune evasion [13].

Although recent transcriptomic studies have provided novel insights into latent and lytic KSHV genomic expression in tumors [14–18], questions remain that can only be addressed by a single-cell approach to transcriptomics. Given the cellular complexity, rarity of infected cells, and variable clinical course of KSHV associated disorders, single-cell analyses provide a unique opportunity to explore key interactions of lytic and latent infected tumor cells with the tumor microenvironment.

KS therapy is focused on disease palliation to improve quality of life and survival, but it is not curative [19]. A key management component is to minimize immune suppression, whether by reducing immunosuppressive medications for iKS, or optimizing antiretroviral therapy for epKS. For indolent localized KS with minimal cosmetic or functional disturbance, topical or localized therapies may be indicated. For aggressive or visceral disease, or lesions with moderate-severe cosmetic or functional disturbance, systemic therapy is indicated. This may include FDA-approved chemotherapies such as liposomal doxorubicin or taxanes, or the immune or cereblon-modulator drug (IMiD, cel-mod) pomalidomide. The mechanism of action of these drugs remains unclear but they are known to alter angiogenesis, cytokine production, and T-cell activation [20]. Other chemotherapeutic, anti-angiogenic, proteasome inhibitor, and immune checkpoint inhibitor drugs showed preliminary activity. However, response rates of 30-60% are seen with most approaches, and biomarkers of activity remain to be defined. In rare cases, exacerbation of KS-associated inflammatory disorders were seen [21].

Previous studies suggested that KSHV latent and lytic gene expression occurs in KS, and disruption of either program results in tumor regression [22]. Animal models for KS are lacking, and there is a dearth of genomic studies on primary KS tissue due to the admixture of multiple cell types, the small proportion of KSHV positive cells, and the complexity of fibrotic skin tumors. Here we used a scRNAseq multiomic approach to characterize the cellular and viral KS transcriptome at a single cell level in primary tissue. These findings may have applications for discovery of prognostic and predictive biomarkers and therapeutic insights for the design of safe and effective therapies for KS. Application of these technologies to understand primary KS pathogenesis and therapeutic responses may be applied to understanding oncogenic virus biology, as well as defining the evaluation and treatment of other infection-associated cancers.

Results

Sixteen participants contributed twenty-five samples for scRNAseq analysis, comprising nineteen KS skin biopsies, one non-KS skin biopsy, and five PBMC samples (Table 1). Eleven HIV+ participants had AIDS-associated KS (epKS), two HIV- participants had classic KS (cKS), one HIV- participant had iatrogenic KS (iKS), one HIV+ participant who also had a renal transplant had KS (ep/iKS), and one HIV+ participant had neither KS nor AIDS. Three participants contributed samples before and after therapy, one treated with nivolumab and ipilimumab, one treated with antiretroviral therapy, and one treated with pomalidomide. The participants with classic KS were the only females in the cohort. The sample set is small, diverse, and representative of the patient population presenting for clinical care in a non-endemic region. The current study is intended to demonstrate the feasibility and reproducibility of the methodology, generate hypotheses for future studies, and offer novel insights into potential biomarkers, pathways, and therapeutic targets.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Kaposi Sarcoma Primary Patient Samples.

https://doi.org/10.1371/journal.ppat.1012233.t001

Cells utilized for scRNAseq were obtained from viably frozen single-cell suspensions prepared from fresh primary skin and blood samples and submitted in three batches for 10X Genomics 5’ or 3’ gene expression with multiomic TCR profiling (S1 Fig). KSHV transcripts could be detected in all KS skin tumors. HIV-1 and EBV infected cells were not detected in KS lesions. The number of cells analyzed by scRNAseq and the percent KSHV positive cells for each sample (S1C Fig) are indicated in Table 1. Comparisons to scRNAseq of normal skin utilized published data. (S2 Fig) [23].

The rarity of KSHV transcript reads in infected cells and the rarity of infected cells in tumors resulted in an analysis challenge in which stringent quality control measures typically applied to filter out noise (cells that expressed less than 100 features; genes that were expressed in less than 10 cells) also filtered out most KSHV-infected cells. Filters for dying cells (>20% mitochondrial genes), low-quality reads (proportion of UMI > 93^rd percentile), and doublets removed between 15 and 25% of cells from each skin sample and an average of 32% of cells from PBMC samples. For identifying KSHV+ cells in skin samples, most suspected false-positives resulted from 1 KSHV read per cell. Evaluating cells with more than 1 KSHV read per specified gene retained 70% of suspected true positives and removed ~90% of suspected false positives (S3 Fig).

The landscape of primary KS

To observe the cellular landscape of primary KS, 248,741 cells from twenty-three samples (5 peripheral blood mononuclear cell, PBMC, samples and 18 skin samples submitted in three batches; S1 Fig) were merged into a single dataset, forming 25 unsupervised clusters (Fig 1A). Results for each of the individual samples are highlighted in Fig 1B. Cells from the peripheral blood (n=73,269) were clearly distinct from those in the skin samples (n=175,472) (Fig 1C). In addition to human transcripts, expressed KSHV genes were also detected in 3,118 cells (1.8% of total skin tumor cells), exclusively from skin samples (Fig 1D). In the merged plot, 95% of KSHV-expressing cells clustered together in cluster 13 and migrated very closely to, though separate from, non-infected endothelial cells in clusters 2 and 19.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Landscape of Primary KS.

Single cell suspensions from viably frozen primary KS blood and enzymatically separated skin tumor samples processed for scRNAseq resulted in A) an unsupervised UMAP clustering projection of the merged data set containing 248,741 cells from 23 samples representing a landscape of 25 clusters. Between 15 and 25% of cells from each sample were removed when filtering out dying cells (>20% mitochondrial genes), low quality cells (proportion of UMI > 93^rd percentile), and doublets (identified by DoubletFinder package). Seurat’s default settings were used for normalization and scaling. Principal component (PC) analysis was based on 22 high variable genes with clustering of cells set at a resolution of 0.6. The box around clusters 2, 13, and 19 indicates they are endothelial cells in skin tumors with 95% of KSHV-infected cells mapping to cluster 13. B) UMAP clustering projection color coded based on individual sample distribution across the merged dataset. Samples were submitted in three batches indicated by the color of the sample name in the legend (batch 1, red; batch 2, blue; batch 3, black) and the percent of KSHV+ cells contributed by each sample and the percent of KSHV+ cells present in each sample is shown in S1C Fig. C) UMAP clustering projection color-coded based on tissue source of each sample demonstrated that cells obtained from skin (n=175,472) and peripheral blood mononuclear cell (PBMC) (n=73,269) form clearly delineated clusters. D) UMAP clustering projection color-coded based on reads corresponded to the KSHV genome in which 3,118 KSHV+ cells (with 2 or more reads of viral transcripts) in the merged object corresponding to an average of 1.8% of total tumor cells. 2,970 (95%) of KSHV+ cells were in cluster 13. KSHV+ cells were not detected in PBMC samples. E) UMAP clustering projection color-coded were based on cell annotation using SingleR and the Human Primary Cell Atlas (HPCA) reference dataset. The inset focuses on vascular (red) vs. lymphatic endothelial cells (green) within clusters 2, 13, and 19 identifying cluster 13 as predominantly lymphatic endothelial cells.

https://doi.org/10.1371/journal.ppat.1012233.g001

In the peripheral blood, separable clusters of cells were found for monocytes, B and T lymphocytes, and natural killer cells (Fig 1E). Within skin biopsies, several clusters of endothelial cells and fibroblasts were detected as well as keratinocytes and epithelial cells. Clusters of B and T cells, macrophages, and dendritic cells were also detected in skin preparations and, interestingly, formed distinct clusters from the same cell types obtained from peripheral blood, though they migrate closely.

Two populations of KSHV infected cells in primary KS skin lesions.

Although in the composite object KSHV infected cells were clustered in a single cluster (Figs 1,2A,S1D), in individual KS skin tumors, cells in which KSHV genes were expressed typically formed two distinct clusters (Figs 2B,2C,3,S4 and S5B-D). Both clusters contained cells expressing lytic and latent KSHV genes (Figs 2C,S5B and S5D). Both clusters expressed endothelial markers PECAM1 (CD31), FLT4 (VEGFR3), and CD36; lacked ITGA2B, PDGFRA, and ACKR4; and expressed the vascular-to-lymphatic endothelial determinant prospero homeobox protein 1 (PROX1) (S5A Fig) [24]. However, one of the two KSHV-infected clusters (cluster A) was highly transcriptionally active (S5B,S5C,S5E and S5G Figs). Thousands of host genes were differentially expressed in this cluster, including CD34 and, unexpectedly, housekeeping genes including GAPDH (Figs 3C, S2B,S2D,S4 and S5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Latent and Lytic KSHV in Primary KS.

A “latent” cell is defined as any cell expressing at least two reads from any of the following KSHV latency transcripts: LANA (ORF73, gp81), Kaposin (K12, gp79), vFLIP (ORF71, gp80), K15 (ORF75, gp85), vOX-2 (K14, gp83), or vIRF-2 (gp65) AND expressing no other KSHV gene. A “lytic” cell is defined as any cell expressing at least one read from a latency transcript AND at least one read from any other viral gene. When all samples are combined in a single data set as in Fig 1, KSHV infected cells populate in a single cluster (cluster 13). A) Lytic vs. Latent: Within the UMAP clustering projection described in Fig 1, 2,970 KSHV+ cells are in cluster 13; of which the virus is in latency in 1,257 cells (42%) and lytic replication in 1,713 cells (58%). However, when samples are evaluated individually, unsupervised clustering divides KSHV infected cells into two distinct clusters. B) Two clusters: 10X Cell Ranger, unsupervised, graph-based t-SNE plots of KS6B, KS8, and KS9 skin tumors with KSHV+ (red) clusters (A and B) highlighted. Remaining cluster colors are not coordinated between samples. C) An exploded diagram of the t-SNE plot of sample KS6B in which clusters A (brown) and B (purple) are magnified and color coded based on Log2 KSHV gene expression and lytic vs latent cell identity demonstrate that viral gene expression and lytic vs latent replication are not the defining features that distinguish the cluster A from cluster B. Samples shown in panels B and C were those samples with >2% KSHV-positive cells (S1C Fig).

https://doi.org/10.1371/journal.ppat.1012233.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Two Populations of KSHV Infected Cells in Primary KS.

To characterize infected tumor cells, a new unsupervised analysis of the 2,970 KSHV+ cells within cluster 13 was performed (with number of principle components = 10 and resolution = 0.3) in which cells were annotated using SingleR and Human Primary Cell Atlas (HPCA) dataset resulting in A) a UMAP projection of 10 KSHV+ clusters (numbered 13-0 through 13-9). Although cells in cluster 13 were identified as lymphatic endothelial, small subclusters of KSHV+ cells also express keratinocyte markers (cluster 13-8, n=19 of 135), fibroblast markers (subcluster 13-5, n=23 of 204), vascular endothelial markers (subcluster 13-6, n=40 of 182), or markers of mitosis, including MKI-67, RRM2, and DIAPH3 (subcluster 13-9, n=76). B) The remaining subclusters can be divided into two groups with subclusters 13-1, 13-2, and 13-3 representing cluster B from Fig 2 and subclusters 13-0, 13-4, 13-7 representing cluster A. A UMAP projection color-coded based on sample reveals that subclusters 13-3 and 13-4 are exclusively from sample KS6B and subclusters 13-2 and 13-7 are exclusively from KS8. C) A FeaturePlot of Log2 expression values representing normalized read counts for GAPDH, CD34, EP300 and CREBBP identify cluster A (subclusters 13-0,4,7) as CD34⁺GAPDH^HI and cluster B (subclusters 1,2,3) as CD34^-GAPDH^LO.

https://doi.org/10.1371/journal.ppat.1012233.g003

Expression of CD34 and lymphatic vessel endothelial hyaluron receptor 1 (LYVE1) in this cluster, along with ribosomes, exosomes, and markers like LGALS1, IGFBP7, CD63, and BSG was consistent with genes expressed in high endothelial venules (HEVs) (S5E Fig) [25]. However, while cell types like telocytes can be excluded (due to expression of CD31, PDPN, and LYVE1, and the absence of PDGFRA), the expression of markers for fenestrated capillaries (PLVAP) and endothelial tip cells (NID2, ITGB1, CAV1, GMFG, IGFBP3, TGFBI, KLF2, PLAUR, SOX17) makes the identity of the CD34+ cells unclear [25–28]. Interestingly, the CD34+ KSHV+ cluster expressed transcripts associated with spliceosome and electron transport machinery, and also exhibited high expression of lymphocyte antigen 6 complex, locus H (LY6H), which has been described by Moorad et al. as associated with “inflammatory” KS lesions (S5F Fig) [15]. Although CD34+ and CD34- endothelial cells were also seen in normal skin samples, there were no significant differences in GAPDH expression (S2 Fig).

In the CD34- cluster, housekeeping genes were suppressed while cellular proliferation factors were enriched including EP300 and CREBBP (S4 and S5D Fig), along with factors like DTX1, DTX4, HEY1, and CTNNB1 that regulate the NOTCH and WNT signaling pathways. Expressing biomarkers consistent with a lymphatic endothelial lineage (S5A Fig), the CD34- cluster is likely representative of proliferating KS cells. Despite the significant differences in viral, cellular, and specific biomarker gene expression, the two infected clusters were grouped together in our merged UMAP plot, supporting similarities between these 2 clusters that could be due to a common endothelial cell lineage origin. KSHV+ cells with characteristics of both clusters could be detected in skin tumors from all KS samples with >50 KSHV+ cells (S4A-O Fig), both may be involved in tumor growth, and each cluster represents a largely uncharacterized and rare sub-population of cells within KS tumors (Figs 1D and S1C) that are readily distinguishable by expression of housekeeping genes.

Novel biomarkers of KSHV infected cells in primary KS lesions.

In addition to viral genes, several cellular genes were commonly expressed in the KSHV positive cells including PROX1, mannose receptor C-type 1 (MRC1; CD206), fms-related tyrosine kinase 4 (FLT4), and Kir2.1 inward-rectifier potassium channel (KCNJ2) (S5A Fig). These markers were previously described in KS lesions and provide confirmation for the sensitivity and reproducibility of scRNAseq data obtained from primary skin lesions [29–31]. Differential expression analysis revealed 1,118 significantly enriched genes in KSHV+ cells including several voltage gated sodium channel (VGSC) genes with SCN9A as a top biomarker candidate along with the KSHV Latency cluster (Figs 4, S1D, S6, S7 and S8). The expression of the VGSC gene SCN9A has not been previously described in KS, was tightly associated with both clusters of KSHV infected cells (S7A-C Fig), was minimally expressed in uninfected endothelial or other stromal cells, and was not expressed in the KS-negative skin samples (S2 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Differential Expression Analysis of KSHV

+ cells. A) Differential expression analysis of genes expressed in KSHV+ cells included 1,118 genes with a p value less than 1e-300. When graphed as the average Log2 fold change of gene expression in KSHV+ vs. KSHV negative cells against the ratio of percent of KSHV+ cells expressing that gene (Pct.1) to the percent of KSHV- cells expressing that gene (Pct. 2) the KSHV latency cluster downstream of the LTd promoter (designated KSHV LAT, including viral genes v-FLIP, v-CYC, and K12) along with the lytic transcript K5 were the top viral biomarkers. The voltage-gated sodium channel SCN9A (Nav1.7) emerged as the top, non-KSHV biomarker among a number of previously described factors including FLT4, PROX1, STC1 and CD36.

https://doi.org/10.1371/journal.ppat.1012233.g004

The number of KSHV+ cells is inversely proportional to immune cell number in primary KS.

In most samples, including two HIV-negative samples, KSHV+ cells were very rare, comprising less than 2% of the total cells in the sample (Figs 5A and S1C). In three samples, all of which were HIV-associated, KSHV+ cells were notably more abundant, representing 3-9% of the total cells in the sample. In samples in which KSHV+ cells were rare, macrophages (S9 Fig) and CD8+ T lymphocytes were more abundant than in samples in which KSHV+ cells were prevalent (Fig 5). In these samples there was an inverse correlation between the number of macrophages and CD8 T cells and the number of KSHV+ cells in the primary tumor (Fig 5D) suggesting their role in KS immunity. HIV-negative samples had significantly more CD8 T cells and macrophages than samples obtained from AIDS-associated lesions (Fig 5C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Immunity in Primary KS.

A) t-SNE plots of primary KS tumor samples in two groups, three samples with greater than (>) 2% KSHV positive cells (KS6B, KS8, KS9) and three samples with less than (<) 2% KSHV positive cells (KS7A, KS10A, KS12) color- coded to highlight Macrophages (Black dots; defined as all cells with elevated Log2 sum expression >2 for CD14, CD80, CD86, TLR2, MARCO, IL10, IL1B, IL1A, and ITGAX (CD11C) also see S9 Fig), CD8+T cells (Blue dots; defined as all cells with elevated Log2 sum expression >2 for CD2, CD3E, CD8A, NKG7, LAG3, GZMH, PTPRCAP, and negative for CD4), and KSHV+ tumor cells (Red dots, defined as all cells expressing KSHV genes). All samples with >2% KSHV-positive cells are shown, and samples shown with <2% KSHV-positive cells are representative of the entire group of samples with that characteristic. B) Graph representing the ratio of Macrophages to KSHV+ cells (black bars), and CD8+T cells to KSHV positive cells (blue bars) in samples are shown in panel A. C) Percent Macrophages, Percent CD8+ T cells and Percent KSHV+ cells in samples from HIV+ vs. HIV- donors with p values shown for comparisons of the results from each group. D) Dot plot of the ratio (Fold) of Macrophages/ Total; CD8+ T cells/ Total; and KSHV+ cells/ Total, for each skin sample. Correlation coefficient for Macrophage vs. KSHV = -0.273 (1 tailed p value = 0.14); CD8+T cell vs. KSHV = -0.356 (one tailed p value = 0.074); Macrophage vs. CD8+ T cell = 0.69 (one tailed p value = 0.00007).

https://doi.org/10.1371/journal.ppat.1012233.g005

The peripheral blood of KS subjects

Whether truly absent or simply below the level of detection for scRNAseq, there were few EBV+, HIV-1+, or KSHV+ cells in the peripheral blood mononuclear cells (PBMCs) (S10 Fig). However, the data did reveal unique characteristics of T cells in the peripheral blood of KS patients.

Low CD4:CD8 ratio in peripheral blood in HIV+ KS subjects.

The CD4:CD8 T cell ratio is an important biomarker of pathogenesis. In healthy subjects the CD4:CD8 T cell ratio in whole blood is usually greater than 1.0, indicating that CD4+ T cells are typically present in greater abundance than CD8+ T cells [32]. In our PBMC samples, the CD4:CD8 T cell ratio was less than 1 (Fig 6), with all four of these KS patients being immune-suppressed due to AIDS or immunosuppressant medication. For the HIV+ subjects, the low ratio CD4:CD8 T cell ratio resulted from both low representation of CD4 T cells (6.3-12.5% of total PBMC sample) and notably high representation of CD8 T cells (67.5-80.4% of total PBMC cells). While PBMC counts of CD4+ and CD8+ T cells based on transcript counts are not comparable to whole blood reference values based on cell surface protein expression, these proportions could be reflective of these subjects’ persistent HIV infection [32]. The KS12 blood sample from the HIV negative iKS participant had the highest CD4:CD8 T cell ratio, albeit still <1, due to a low proportion of both CD4+ T cells (22.9% of total cells) and CD8+ T cells (42.3% of total cells). Similar results were described in Ravishankar et al (Fig 6C) [33]. In calculating these ratios, the value of combining scRNAseq with TCR sequencing was apparent, clearly distinguishing TCR-CD4+ monocytes (S9B Fig) from TCR+CD4+ T cells, and TCR-CD8+ NK cells from TCR+CD8+ T cells [33–35].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Low CD4:CD8 Ratio in KS Blood.

A) UMAP plots for 4 primary peripheral blood mononuclear cell (PBMC) samples in which multiomic scRNAseq and T-cell receptor (TCR) sequencing was performed (for batch 1 samples including KS6A PBMC, TCRseq was not performed). The combination of TCR reads with gene expression for CD4 and CD8A clearly defined clusters representing the following cell types: TCR+CD4+ (Blue; CD4+ T cells), TCR-CD4+ (Orange; monocytes) also see S9 Fig, TCR+CD8- (Green; CD8+ T cells), TCR-CD8+ (Red; NK), TCR+DP (Purple; double-positive T cells), TCR+DN (Brown; double-negative T cells). The ratio of CD4+ T cells to CD8+ T cells is indicated, as well as the percentage of CD4+ and CD8+ T cells out of total TCR+ cells. B) Quantitation of the percent CD4+ TCR+ and percent CD8+ TCR+ cells in KS peripheral blood mononuclear cell (PBMC) samples (three AIDS-KS and one HIV-negative iatrogenic KS sample are compared to publically available PBMC data from a normal reference data set from ScaleBio [80]. C) Data reported in S4 Fig of Ravishankar et al. [33] in which the CD4:CD8 ratio in AIDS-KS is similar to the samples shown in panel B.

https://doi.org/10.1371/journal.ppat.1012233.g006

Expansion of T cell clones in peripheral blood of KS subjects.

T cells present in the peripheral blood exist as a unique oligoclonal pool of various T cell clones carrying a diverse array of antigen-specific T cell receptors (TCRs) [36]. An emerging challenge in tumor immunology is elucidating the role and identity of tumor or pathogen specific T cell clones. By combining single-cell TCR profiling with single-cell gene expression, multiplex scRNAseq provides a powerful tool to identify KS-specific CD8+ TCR clones [37]. In nine samples from four participants for whom TCR reads from both tumor and PBMC could be evaluated, medium to large clonal groups (with 0.1-10% abundance) were present in PBMCs (Fig 7A) and the distribution of VDJ recombination was non-random among the most abundant CD8+ clones in skin tumors (Fig 7C). One complementarity-determining region 3 (CDR3) sequence motif pattern was found in all 4 PBMC samples, 9 motifs were found in 3 of 4 PBMC samples, and 120 motifs were shared between two PBMC samples (Fig 7B). The most common motifs and the clones that endorse them are listed in S1 Table. Identical CD8 T cell clones were found in both the skin and the peripheral blood from the same patient and in two different skin samples taken before and after therapy from the same patient (Figs 7E and S11). Interestingly, in these cases, the most abundant T cell clone in the peripheral blood was not the most abundant clone in the tumor. The most abundant T cell clones among the KS samples carried a subset of frequent rearrangements, especially enriched in TCR β joining region 1-1 (TRBJ1-1) and TRBJ2-5, with recurring similarities among the CDR3 sequences of the most abundant clones in each patient. Interestingly, TRBJ1-1 and TRBJ2-5 were not the most frequently used TRBJ variants in the TCRβ repertoire of healthy controls [33,38,39], and the CDR3 sequences associated with TRBJ1-1 clones shared significant similarities with the CDR3 sequence (CASSILGLRNTEAFF) found in CD8+ T cell clones that react with the KSHV major capsid protein (ORF25) (Fig 7D and 7E) [40].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Characterization of T cell diversity in KS.

A) Diversity of T cell clonal expansion in peripheral blood mononuclear cell (PBMC) samples from 4 patients, characterized by proportion of space occupied by clones. Clonal expansion was ascribed to one of 4 relative abundance categories, based on a clone size score calculated by scRepertoire. Clonal size cut points are listed in the panel legend [81]. B) Number of samples expressing the most common complementarity-determining region 3 (CDR3) motifs found in GLIPH2 clustering (Grouping Lymphocyte Interactions by Paratope Hotspots). Among 289 motifs, only one was found among PBMC samples from all 4 subjects. C) The graph depicts the frequency of T-cell receptor β chain joining segment (TRBJ) usage in the 150 most abundant T-cell receptor (TCR) clones in primary KS samples with matched blood and skin samples (KS6B, KS10, KS11, KS12). D) Three sequence logos; top) 9 CDR3 sequences in abundant TRBJ1-1 clones in KS PBMC samples from 4 patients described in panel E; middle) 19 CDR3 sequences from TRB1-1 clones found by GLIPH2 clustering described in panel B; compared to bottom) the anti-open reading frame (ORF) 25 P88 sequence reported by Roshan et al. in [40]. The y-axis BITS indicates the height method used by the sequence logo generation package. E) Alignment of 9 CD3 TRBJ1-1 sequences from abundant T cell clones from 4 KS subjects to the sequence reported in [40]. ‘Max score’ refers to the highest bit score, or sequence similarity score, for the alignment. ‘Perc. Ident’ refers to the percent of total sequence amino acids that are identical to the Roshan et al. sequence [40]. ‘E value’ refers to the significance score. Similar to p values, E values closer to zero represent fewer chances of a result being due to chance, and hence higher significance. The sequence color code refers to the E values, with gray marking an alignment score of <40, blue marking a score of 40-50, and red marking identical residues (score >200) [82]. Samples shown in panels C-E are all samples for which TCR analysis was performed on both skin and PBMCs.

https://doi.org/10.1371/journal.ppat.1012233.g007

Discussion

KS, like other KSHV-associated diseases, exhibits widely different transcription programs [16,42,43]. Several recent studies have applied scRNAseq to acute virus infections [44–48], as well as latent infection with human cytomegalovirus and other herpesviruses [49–51]. In a previous scRNAseq study, Landis et al. found latency-associated transcripts in the majority of cells of two KSHV-infected PEL cell lines [17]. Fewer than 1% of cells expressed lytic viral RNAs, with a predominance of early over late lytic viral transcripts. Jung et al. performed scRNAseq on KSHV infected 3-dimensional organoid cultures. They noted high levels of lytic replication, a unique pattern of lytic K2-K5 gene expression, and marked changes in host gene expression in infected and uninfected cells [52].

The current study provides the first single cell transcriptomic analysis of primary KS tumors and PBMC samples from different subtypes of KS, including cKS, iKS, and epKS. In this study, KSHV transcripts were found in a minority of cells in each KS tumor but not in the peripheral blood. PCR assays have detected KSHV DNA in peripheral blood mononuclear cells in 52% of epKS subjects [53]. The lack of KSHV+ cells in PBMCs in this study could be due to low levels of expression, a limitation of the sensitivity of our current scRNAseq technology or analysis, or both. Future studies using CITE-Seq with a custom panel of KSHV antibodies could improve the sensitivity of this analysis.

A novel finding from our study was the consistent presence of two separate clusters of KSHV-infected cells in skin tumors, differentiated by the presence or absence of CD34 expression as well as ribosomal genes and housekeeping genes like GAPDH. The CD34- cluster was consistent with a lymphatic endothelial cell (LEC) lineage and was characterized by proliferative gene expression and voltage-gated ion channels [54]. The CD34+ cluster expresses many genes that correlate specifically with high endothelial venule cells which promote lymphocyte extravasation and maintain elevated levels of ribosomes and secretory vesicles [25]. However, this cluster is also enriched in genes associated with vascular endothelial “tip” cells, which drive angiogenic sprouting, are motile, and express long filopodia [27,28,55]. Cells in these clusters shared common biomarkers of endothelial cells, including PECAM1, PROX1, and CD36. However, in addition to marked differences in viral and cellular gene expression, we observed differential expression of several biomarkers, such as CD34, LYVE1, PLVAP, and PDPN. We conjecture that challenges in developing KS tissue culture models may be due to lack of cultivation conditions that include both the CD34- and CD34+ KSHV infected cell types. It is also possible that the characteristics of one or both of these subtypes of KSHV infected cells causes KS to be intrinsically sensitive to tumor immunity.

Interestingly, a study proposed a “transcriptional reprogramming” mechanism in which overexpressing PROX1, a known upregulated gene in KS, suppressed blood vascular endothelial cell (BVEC) genes and induced the LEC transcriptional program [56]. We found consistent upregulation of PROX1 in both subtypes of KSHV-infected cells. This finding supports a potential lineage relationship between the two infected clusters observed in our study, and suggests a potential PROX1 role in endothelial cell differentiation and/or malignant transformation in KSHV pathogenesis.

While latency-associated transcripts predominated, lytic transcripts were also identified in both KSHV-infected cell subtypes in our study. Although we cannot exclude some level of KSHV reactivation during sample processing, we consistently found lytic transcripts in all cases. The most abundantly expressed viral genes in our samples were the latency cluster genes downstream of the LTd promoter, a pathogenesis factor that encodes at least three proteins (v-CYC, v-FLIP and Kaposin); and K5, an early lytic ubiquitin ligase with immune evasion functions [57]. CD34- clusters showed higher average viral reads per cell than CD34+ clusters. One notable exception was K5, which was more highly expressed in CD34+ clusters while being predominant in both. A potential explanation for this over-representation could be related to genomic rearrangements, which were not fully investigated in this study but likely existed in one sample. One study showed K5 overexpression in almost a third of KS lesions, due to similar de novo KSHV genomic rearrangements of a 1.5kb section containing the K5 and K6 genes [41].

Differential expression of host genes, such as GAPDH and CD34, in KSHV infected cell clusters did not correlate with the latent or lytic viral programs. Pardamean et al described two lytic-associated mechanisms of KSHV host gene shutoff, involving v-SOX and ORF10 expression during early and late lytic replication, respectively [58]. We detected no v-SOX or ORF10 transcription in any of our samples, and we found notably low expression of GAPDH and other host genes in infected clusters with low lytic transcripts. It is unclear whether host gene shutoff is induced by v-SOX and/or ORF10 expression below the level of detection for our assay, or other host gene inhibition mechanisms are at play.

Our study identified several potential cellular biomarkers of KSHV infection, including some that were previously described [29–31]. One of these markers, FLT4, has previously been linked to malignancy [59] and discussed as a potential therapeutic target [60]. Another marker, PROX1, has been found to co-localize with CD34 in KS lesions [29] and other cancers [61], and may be involved in regulation of endothelial to mesenchymal transition. An interesting finding was the expression pattern of LYVE1, a known lymphatic endothelial marker, which was present in both KSHV-infected clusters and particularly overexpressed in CD34+, transcriptionally active cells. LYVE1 has been described in close correlation with PROX1 in “budding endothelial cells”, a pattern thought to be a hallmark of lymphatic phenotype induction in budding vascular endothelium [24,54]. LYVE1 and PROX1 colocalization was also observed in tumors and tumor-adjacent lymphatic vessels [24]. Our findings support a role of endothelial phenotype transition in KSHV-induced host cell changes and oncogenesis.

The expression of VGSCs in both clusters of KSHV-infected cells was a novel observation. VGSCs are abnormally expressed in many types of cancers, and their level of expression and activity are related to the aggressiveness of the disease. Their tumor effects may be mediated through sodium ion effects, membrane potential modulation, or other pathways [62]. Of particular note was SCN9A (Nav1.7), a tetrodotoxin-sensitive VGSC with known roles in angiogenesis and regulation of chemotaxis [63]. SCN9A expression and activity were shown to steadily increase in herpes simples virus type 1 latent infection [64]. SCN9A is normally expressed in skin vasculature [65], and it has been associated with several cancer types, including endometrial, gastric, and prostate carcinomas [66–68]. In this study, SCN9A was tightly correlated with viral gene expression. Whether SCN9A could potentially be involved in KSHV-induced angiogenesis and blood vessel migration remains to be determined. Nonetheless, SCN9A emerges as a top RNA biomarker candidate of KSHV infection within primary KS lesions. Notably, another cation ion channel, despite relative low levels of expression, may be a therapeutic target in Kaposi sarcoma [69].

A recent bulk RNAseq study identified two types of KS lesions, inflammatory and proliferative, distinct by host gene transcription patterns [15]. Of the transcripts associated with the inflammatory subtype, we detected cell migration-inducing and hyaluronan-binding protein (KIAA1199; CEMIP) expression in all KS tumor samples, in both CD34+ and CD34- clusters. CEMIP was among the most commonly expressed genes among KSHV infected cells. Vesicle amine transport protein 1 homolog (Trixis californica)-like (VATL1) was also expressed in both infected clusters, to a lesser level and not in all samples; and LY6H was enriched in CD34+ clusters. Additionally, the proliferative subtype marker annexin A8-like 1 (ANXA8L1) was detected in non-infected epithelial cells and keratinocytes in all of our samples. While the inflammatory and proliferative distinction was not readily apparent in our analysis, quantification using single cell technology may further elucidate the transcriptional patterns of these newly identified markers.

Among the KSHV-negative cell population in KS tumors were T cells. KSHV-specific CD8 T cells are known to recognize and kill cells expressing KSHV lytic antigens [70], and they are more frequent in KSHV-seropositive individuals without KS than those with active disease, suggesting an anti-tumor role of these cells [71]. Perhaps reflective of these anti-tumor and anti-viral functions, the number of KSHV+ cells in our sample tumors were inversely proportional to immune cell number. One caveat to this relationship is that the proportion of immune cells in each sample also positively correlated with the total number of cells in the sample, though there was no correlation between percent of KSHV positive cells and total number of cells. In distinction to previous reports [72,73], we noted clonally expanded CD8+ T cells in tumors in paired tumor and blood samples, as well as longitudinal tumor samples. TCR Vβ CDR3 sequences of T cell clones in the peripheral blood were similar, but not identical, to a clone known to be directed against the major KSHV capsid protein [40]. Another KS PBMC single cell study [33] also reported common TCR sequence motifs of unknown antigen specificity, postulating a potential poly-specific T cell response to KSHV. Further investigation is warranted to elucidate whether T cell clonal expansion is indeed directed against KSHV, cellular tumor-specific antigens, HIV-1, or other persistent viral infections in these individuals.

To evaluate the potential feasibility of scRNAseq to monitor KS therapy, we examined samples obtained longitudinally from three individuals. In one individual each, we assessed the effects of antiretroviral therapy, immune checkpoint inhibition, and pomalidomide therapy. With each therapy we noted an increase in the abundance of stromal endothelial cells, distinct from both KSHV+ endothelial cell subtypes in unsupervised cluster formation, expression of KSHV genes, and KSHV infection markers in host cells. This change correlated with elevated expression of genes associated with tumor infiltrating CD8+ T cells. Studies of additional individuals are warranted to comprehensively assess effects of immunotherapy on the KS tumor microenvironment.

Our study has several limitations. The sensitivity of our scRNAseq assay may play a role in our consistently low KSHV detection in primary skin tumors compared to other assays [74,75], as well as the rarity of EBV, KSHV or HIV-1 detection in PBMC samples. Whether assay sensitivity prevented KSHV detection in non-endothelial tumor cell types, such as epithelial and immune cells, is unknown. Another limitation is that we did not examine micro- or long non-coding RNAs [76,77]. Our subjects constitute a small, heterogeneous subject pool of different age ranges and gender, who had different KS subtypes. While some demographic traits, such as gender, have not been found to significantly influence KS transcriptomics (34), the inter-individual differences in scRNAseq between and within tumors remain to be characterized. In addition, subjects with epKS had received differing durations of antiretroviral therapy. Nevertheless, the current study demonstrates the feasibility and utility of scRNAseq for interrogating the KS biology.

Taken together, these studies demonstrate the feasibility of single cell, multiomic analyses to characterize the malignant and stromal composition of primary KS blood and tumor tissue, quantitate viral and host gene expression, identify novel prognostic and predictive biomarkers and potential therapeutic targets, and evaluate the efficacy of therapeutic interventions.

Materials and methods

Supporting information

Acknowledgments

We thank Dr. Feng Gao (Washington University, Department of Surgery) with assistance with statistical analysis. We thank the Alvin J. Siteman Cancer Center at Washington University School of Medicine and Barnes-Jewish Hospital in St. Louis, MO. and the Institute of Clinical and Translational Sciences (ICTS) at Washington University in St. Louis, for the use of the Genome Technology Access Center, which provided scRNAseq services and support.

Financial statement

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.