Nucleotide sequence of the Kaposi sarcoma-associated herpesvirus (HHV8)

Abstract

The genome of the Kaposi sarcoma-associated herpesvirus (KSHV or HHV8) was mapped with cosmid and phage genomic libraries from the BC-1 cell line. Its nucleotide sequence was determined except for a 3-kb region at the right end of the genome that was refractory to cloning. The BC-1 KSHV genome consists of a 140.5-kb-long unique coding region flanked by multiple G + C-rich 801-bp terminal repeat sequences. A genomic duplication that apparently arose in the parental tumor is present in this cell culture-derived strain. At least 81 ORFs, including 66 with homology to herpesvirus saimiri ORFs, and 5 internal repeat regions are present in the long unique region. The virus encodes homologs to complement-binding proteins, three cytokines (two macrophage inflammatory proteins and interleukin 6), dihydrofolate reductase, bcl-2, interferon regulatory factors, interleukin 8 receptor, neural cell adhesion molecule-like adhesin, and a D-type cyclin, as well as viral structural and metabolic proteins. Terminal repeat analysis of virus DNA from a KS lesion suggests a monoclonal expansion of KSHV in the KS tumor.

Figure 1

Annotated LUR and TR of the KSHV genome. The orientation of identified ORFs in the LUR are denoted by the direction of arrows, with HVS homologous ORFs in dark blue and nonhomologous ORFs in light blue. Seven blocks (numbered) of conserved herpesvirus genes with nonconserved interblock (IB) regions (lettered) are shown under the kilobase marker; the block numbering scheme differs from the original description by Chee et al. (30). The overlapping cosmid (Z prefix) and lambda (L prefix) clones used to map the KSHV genome are compared with the KS5 lambda phage clone from a KS lesion and shown below. Features and putative coding regions not specifically designated are shown above the ORF map. Repeat regions are shown as white lines (frnk, vnct, waka/jwka, zppa, moi, mdsk). Putative coding regions and other features (see text) not designated as ORFs are shown as solid lines.

Figure 2

(A) Sequence of TR unit demonstrating its high G+C content. Sequences highly homologous to conserved herpesvirus pac1 sites are underlined, with lesser homology sites to specific pac1 and pac2 sequences italicized. (B) Southern blot of DNA from BC-1 (lane 1), BCP-1 (lane 2) and a KS lesion (lane 3) digested with NdeII which cuts once in the TR sequence and probed with a plasmid containing the TR sequence. The intense hybridization band at 0.8 kb represents multiple copies of the NdeII-digested single unit TR (C). A schematic representation (C) of genome structures of KSHV in BCP-1 and BC-1 cell lines consistent with the data presented in B and D. Taq I (T) sites flank the TR regions and NdeII (N) sites are within the TRs. Lowercase tr refers to the deleted truncated TR unit at the left end of the unique region. DR represents the duplicated region of the LUR buried within the TR. (D) Southern blot hybridization with TR probe of DNA from BC-1 (lane 1), BCP-1 (lane 2), a KS lesion (lane 3), and HBL-6 (lane 4) digested with Taq I, which does not cut in the TR. Taq I-digested DNA from both BC-1 (lane 1) and HBL-6 (lane 4) show similar TR hybridization patterns, suggesting identical insertion of a unique sequence into the TR region, which sequencing studies demonstrate is a duplicated portion of the LUR (see text). BCP-1 TR hybridization (lane 2) shows laddering consistent with a virus population having variable TR region lengths within this cell line due to lytic replication. The absence of TR laddering in KS lesion DNA (lane 3) suggests that a clonal virus population is present in the tumor.