Epigenetic Landscape of Kaposi's Sarcoma-Associated Herpesvirus Genome in Classic Kaposi's Sarcoma Tissues

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV) is etiologically related to Kaposi's sarcoma (KS), primary effusion lymphoma (PEL) and multicentric Castleman's disease (MCD). It typically displays two different phases in its life cycle, the default latency and occasional lytic replication. The epigenetic modifications are thought to determine the fate of KSHV infection. Previous studies elegantly depicted epigenetic landscape of latent viral genome in in vitro cell culture systems. However, the physiologically relevant scenario in clinical KS tissue samples is unclear. In the present study, we established a protocol of ChIP-Seq for clinical KS tissue samples and mapped out the epigenetic landscape of KSHV genome in classic KS tissues. We examined AcH3 and H3K27me3 histone modifications on KSHV genome, as well as the genome-wide binding sites of latency associated nuclear antigen (LANA). Our results demonstrated that the enriched AcH3 was mainly restricted at latent locus while H3K27me3 was widespread on KSHV genome in classic KS tissues. The epigenetic landscape at the region of vIRF3 gene confirmed its silenced state in KS tissues. Meanwhile, the abundant enrichment of LANA at the terminal repeat (TR) region was also validated in the classic KS tissues, however, different LANA binding sites were observed on the host genome. Furthermore, we verified the histone modifications by ChIP-qPCR and found the dominant repressive H3K27me3 at the promoter region of replication and transcription activator (RTA) in classic KS tissues. Intriguingly, we found that the TR region in classic KS tissues was lacking in AcH3 histone modifications. These data now established the epigenetic landscape of KSHV genome in classic KS tissues, which provides new insights for understanding KSHV epigenetics and pathogenesis.

Fig 1. Flowchart of ChIP in KS tissues.

Main steps in the protocol include: (1) Pre-treat; (2) Cross-link; (3) Cell acquisition; (4) Normal ChIP procedures.

Fig 2. Activating AcH3 and repressive H3K27me3 histone modifications on the KSHV genome in classic KS tissues.

The general maps of AcH3 and H3K27 histone modifications on KSHV genome are illustrated. Sequence reads for AcH3, H3K27me3 and Input samples were aligned to the KSHV genome (HQ404500+35TR) and visualized in IGV software. Values shown on the y axis represent the relative enrichment of ChIP-Seq signals and has been normalized according to the calculated normalization factors. The epigenetic maps illustrated in the figure contain information from two cases of classic KS tissues. 1#: the first case. 2#: the second case. The signals of AcH3 and H3K27 histone modifications were overlaid with the signals of Input (baseline) respectively. Red: AcH3 or H3K27. Blue: Input. *: the promoter region of vIRF3 gene. **: the latent locus.

Fig 3. AcH3 histone modification on KSHV genome in classic KS tissues.

The enlarged map of AcH3 histone modification on KSHV genome at higher resolution is illustrated. The epigenetic maps illustrated in the figure contain information from two cases of classic KS tissues. 1#: the first case. 2#: the second case. The signals of AcH3 histone modification are overlaid with the signals of Input (baseline). Red: AcH3. Blue: Input.

Fig 4. H3K27me3 histone modification on the KSHV genome in classic KS tissues.

The enlarged map of H3K27me3 histone modification on KSHV genome at higher resolution is illustrated. The epigenetic maps illustrated in the figure contain information from two cases of classic KS tissues. 1#: the first case. 2#: the second case. The signals of H3K27me3 histone modification were overlaid with the signals of Input (baseline). Red: H3K27me3. Blue: Input.

Fig 5. Genome-wide LANA binding sites in classic KS tissues.

(A) The general maps of LANA binding sites on KSHV genome are illustrated. Sequence reads for LANA and Input samples were aligned to the KSHV genome (HQ404500+35TR) and visualized in IGV software. Values shown on the y axis represent the relative enrichment of ChIP-Seq signals and has been normalized according to the calculated normalization factors. The epigenetic maps illustrated in the figure contain information from two cases of classic KS tissues. 1#: the first case. 2#: the second case. The signals of LANA enrichment were overlaid with the signals of Input (baseline) respectively. Red: LANA. Blue: Input. (B) The distribution of peaks in relation to genes. The identified LANA peaks on the host genome were classified into different groups according to the annotation of peaks by PAVIS. (C) Distances from LANA peaks (-3000 bp to +3000 bp) to TSSs in bins of 100 bp. (D) The number of LANA peaks (-3000 bp to +3000 bp) identified from each case of KS tissues and the number of overlapped peaks. (E) LANA peaks at both the gene body and promoter region of MORC2 gene illustrated by IGV on the GRCh37/hg19. The promoter region of MORC2 gene is illustrated in the box. Bar graph on the right is the result of LANA ChIP–qPCR at the promoter region of MORC2. Data were normalized by the percent input method (signals obtained from ChIP were divided by signals obtained from an input sample) and were presented as mean±SD. The corresponding P value (0.0058) was calculated with Student’s t test.

Fig 6. Dominant repressive H3K27me3 histone modifications at the RTA promoter region in classic KS tissues.

(A) Illustration of AcH3 and H3K27me3 histone modifications at RTA promoter region (HQ404500: 68000–71000). The signals of AcH3 and H3K27 histone modifications were overlaid with the signals of Input (baseline) respectively. Red: AcH3 or H3K27. Blue: Input. *: the region of relative strong H3K27me3 enrichment. (B) Results of ChIP-qPCR at the region of RTA promoter and GAPDH gene (control) in classic KS tissues. Samples prepared for ChIP-Seq were divided and a small quantity (1/5) kept for ChIP-qPCR assay before library construction. ChIP–qPCR data were normalized by the percent input method (signals obtained from ChIP were divided by signals obtained from an input sample). Data are presented as mean±SD. N.A. represents no amplification.

Fig 7. Distinct epigenetic landscape at the TR region in classic KS tissues.

(A) Illustration of epigenetic landscape at the TR region (gi|139472801: 137169–137969). Data were presented as line plot in IGV. The signals of AcH3 and H3K27 histone modifications as well as LANA enrichment are overlaid with the signals of Input (baseline). Red: LANA. Grass Green: AcH3. Dark Blue: H3K27me3. Sky Blue: Input. (B) Results of ChIP-qPCR at TR region in classic KS tissues. Samples prepared for ChIP-Seq were kept small quantity (1/5) for ChIP-qPCR assay before library construction. ChIP–qPCR data were normalized by the percent input method (signals obtained from ChIP were divided by signals obtained from an input sample). Data are presented as mean±SD. (C) Results of ChIP-qPCR at TR region in in vitro cultured cell lines (iSLK.219, BCBL-1 and BC3).

Fig 8. Validation of histone modifications on the KSHV genome in new cases of classic KS tissue.

(A) New classic KS case1 and (B) New classic KS case2. Results of ChIP-qPCR at the region of TR, RTA promoter, miR-cluster, vIRF3 and GAPDH gene (control) in classic KS tissue. Samples prepared from new cases of classic KS tissue were used for ChIP-qPCR assay. ChIP–qPCR data were normalized by the percent input method (signals obtained from ChIP were divided by signals obtained from an input sample). Data are presented as mean±SD. N.A. represents no amplification.

Fig 9. Histone modifications on the KSHV genome in AIDS-related KS tissue.

Results of ChIP-qPCR at the region of TR, RTA promoter, miR-cluster, vIRF3 and GAPDH gene (control) in AIDS-related KS tissue. Samples prepared from AIDS-related KS tissue were used for ChIP-qPCR assay. ChIP–qPCR data were normalized by the percent input method (signals obtained from ChIP were divided by signals obtained from an input sample). Data are presented as mean±SD. N.A. represents no amplification.