Cell-homologous genes in the Kaposi's sarcoma-associated rhadinovirus human herpesvirus 8: determinants of its pathogenicity?

Abstract

The epidemiology of Kaposi's sarcoma (KS) among patients with AIDS has suggested that a sexually transmitted infectious agent other than human immunodeficiency virus (HIV) must be involved in its pathogenesis. KS is about 20,000 times more common in patients with AIDS than in the general population of the United States, and other immunosuppressed groups develop KS approximately 600 times more frequently than the healthy population. During the first decade of the AIDS epidemic , about 20% of homosexual and bisexual patients developed KS, in contrast to about 1% of men with hemophilia. Women were more likely to have KS if their partners were bisexual rather than parenterally infected men (4). This led to a broad search by PCR in patients with KS for the presence of viruses known to affect humans, such as cytomegalovirus (53, 57), human herpesvirus 6 (HHV-6) (29), and BK virus (43). Some of these viruses were found frequently in KS biopsy specimens, but none of them was consistently present (25, 30). A new era of KS research began when Y. Chang, P. S. Moore, and their colleagues (14) detected by representational difference analysis (35) two short DNA fragments from a herpesvi-rus that was distinct from all previously known herpesviruses. Remarkably, more than 90% of KS tissues obtained from patients with AIDS contained the virus. The viral sequences were not present in biopsy specimens from patients without AIDS but were found in 15% of non-KS tissue DNA from patients with AIDS. The new virus, tentatively termed KS-associated herpesvirus or HHV-8, was soon found to be common in all epidemiological forms of KS (24). Viral DNA is consistently present in AIDS-associated KS lesions (1, 36) and in the vast majority of classical European-Mediterranean KS lesions (1, 18), while it is far less frequent in uninvolved skin of patients with KS and in the various biopsy specimens from Caucasian patients without KS and HIV. AIDS-associated African KS specimens (92%) and non-AIDS-associated KS lesions in Uganda (85%) had the virus (16). Based on PCR, the prevalence of HHV-8 appeared to be high in the general population in Uganda, while searches in non-KS tumors and in normal tissues showed that the virus is rarely detectable in Caucasians. While some lymphomas carry other herpesviruses, such as Ep-stein-Barr virus (EBV) and HHV-6 (22), those specimens did not contain HHV-8 DNA. However, one type of lymphoid tumor, the AIDS-associated body cavity-based lymphoma (BCBL), was positive for HHV-8 DNA by PCR and Southern blotting (11, 12). Multifocal Castleman's disease (MCD) is a rare lymphoproliferative disorder; it occurs more frequently in association with KS. HHV-8 DNA was always found in patients with AIDS-associated MCD, including the cases without detectable KS, and it was also seen in the MCD cases of HIV-negative patients (60). Thus, there are now three distinct disease conditions for which PCR epidemiology has indicated that HHV-8 nearly always persists, leaving the question whether the few HHV-8-negative cases of KS are simply due to occasional technical problems in sample collection, DNA extraction , or PCR unanswered. Although most PCR-based studies indicated that HHV-8 is rare in the healthy general population, a few studies were contradictory. The frequent detection of HHV-8 DNA in skin lesions of transplant patients (49) was questioned by others (8). Some studies frequently found viral transcripts in prostate tissues (61) or DNA in semen and pros-tate tissues of patients without AIDS and KS at a frequency between 20 and 90% (34, 42), while other studies found the virus only in semen and prostate tissues of patients at risk for KS (17, 36, 64). The availability of B-lymphoid cell lines from BCBL harboring the virus allowed the performance of the first seroepide-miology studies. Antibodies against nucleic antigens of HHV-8 were seen in 70 to 80% of patients with KS, but no more than 1% of normal healthy subjects had antibodies detectable by this assay (20, 28, 41). Similarly, enzyme-linked immunosor-bent assays using a procaryotically expressed small capsid protein indicated that there was a high level of seroprevalence in KS cases but not in the general population (58). This was in contrast to an immunofluorescence-based serological study which determined that 25% of healthy adults had antibodies against a phorbol ester-induced BCBL lymphoid cell line (21, 32, 51).