SUMMARYAS an attempt to elucidate further the pathogenesis of human cytomegalovirus (HCMV) infection the replication of HCMV in primary human bone marrow cells (BMC) has been investigated. It was found that BMC held in culture in general were susceptible to HCMV infection. Compared to human embryonic lung cells, however, the replicative cycle of HCMV AD169 in BMC as determined by the analysis of viral protein and DNA synthesis was delayed and productive virus infection was restricted to a subset of BMC not exceeding 21% of the total cell population. Both of these phenomena may explain the short-term persistence of HCMV in BMC cultures which was observed over 3 months. By experiments with specifically enriched and depleted cell populations and by indirect double immunofluorescence experiments we found that both bone marrow fibroblasts and a subset of bone marrow stem cells supported productive virus infection. The finding that HCMV replicates in early stem cells of the human bone marrow may explain important aspects of the pathogenesis of HCMV infection including the presence of HCMV in peripheral blood leukocytes.
INTRODUCTIONHuman cytomegalovirus (HCMV) has been frequently isolated from human leukocytes and their subsets as reported from different laboratories throughout the last 10 years (Harnden et al., 1967;Fiala et at., 1975;Rinaldo et aL, 1977;Macher et al., 1983; summarized by Kirchner, 1983). Positive results of virus isolation were most frequently achieved by co-cultivation of human granulocytes with indicator cells, suggesting that the granulocyte cell fraction is one of the major virus reservoirs (Fiala et al., 1975 ;Rinaldo et aL, 1980). Attempts to cultivate HCMV in vitro in primary human polymorphonuclear or mononuclear cells, however, were not successful and even infection of progenitor cell lines such as myeloblastoid lines did not lead to any appreciable virus replication (Rinaldo et al., 1978 ; Einhorn & Ost, 1984), The recent finding of HCMV RNA in lymphocytes of naturally infected individuals (Schrier et al., 1985) would suggest the possibility that HCMV does replicate in lymphocytes in vivo. These results, however, would not exclude the possibility that virus replication can also occur in early stem cells of the human bone marrow (Hirsch, 1984). The latter theory is supported by the finding that HCMV patients frequently show alterations of bone marrow cells (BMC) especially thrombocytopenia (Verdonck et al., 1985), and that in generalized infections the presence of HCMV can be demonstrated in the bone marrow (Brunning, 1981 ;Bodem et al., 1983). Furthermore, HCMV infection is the most frequent disease observed after bone marrow transplantation (Neimann et al., 1977) and the reported T cell subset alterations after HCMV infection (Carney et al., 1981) may well be explained by replication of the virus in early stem cells. Moreover, several reports on the replication of HCMV in different erythroleukaemic and lymphoblastoid cell lines in vitro suggest a susceptibility to HCMV of immature cells ...