Identification of blood mononuclear cells permissive of cytomegalovirus infection in vitro.

We conclude that both the experimental strain Ad169 and a CMV isolate obtained from a patient could infect PBMCs in vitro. The identity of infected cells was established as CD14+ monocytes and a small population of CD3/CD8+ large granular lymphocytes. No evidence of sensitivity to infection was obtained in small lymphocytes, in either the B or T cell population. Furthermore, flow cytometric analysis of cells expressing CMV-encoded antigens is a sensitive assay, as is the demonstration of viral RNA using reversed PCR and nested primer pairs. Since three different types of analyses all indicated active production of structural CMV antigens in infected cells, we conclude that monocytes and some CD8+ large lymphocytes might serve as reservoirs for latent CMV infection and might be responsible for the transfer of CMV infection. Presently, it cannot be determined whether PBMCs are indeed sensitive to a primary infection. Direct sequencing of virus isolates from in vitro-infected cells will have to be carried out to settle this issue. However, we favor the explanation that CMV causes a primary in vitro infection, since, as discussed above, various activating substances can induce the expression of CMV-encoded antigens in blood cells from seropositive donors. Finally, we found that the presence of the CD13 marker was a common denominator of all cells sensitive to CMV infection. We are presently attempting to elucidate whether the CD13 molecule is instrumental in the infection of cells by CMV.