Tumor Necrosis Factor Alpha Induces Reactivation of Human Cytomegalovirus Independently of Myeloid Cell Differentiation following Posttranscriptional Establishment of Latency.

We used the Kasumi-3 model to study human cytomegalovirus (HCMV) latency and reactivation in myeloid progenitor cells. Kasumi-3 cells were infected with HCMV strain TB40/E wt -GFP, flow sorted for green fluorescent protein-positive (GFP+ ) cells, and cultured for various times to monitor establishment of latency, as judged by repression of viral gene expression (RNA/DNA ratio) and loss of virus production. We found that, in the vast majority of cells, latency was established posttranscriptionally in the GFP+ infected cells: transcription was initially turned on and then turned off. We also found that some of the GFP- cells were infected, suggesting that latency might be established in these cells at the outset of infection. We were not able to test this hypothesis because some GFP- cells expressed lytic genes and thus it was not possible to separate them from GFP- quiescent cells. In addition, we found that the pattern of expression of lytic genes that have been associated with latency, including UL138, US28, and RNA2.7, was the same as that of other lytic genes, indicating that there was no preferential expression of these genes once latency was established. We confirmed previous studies showing that tumor necrosis factor alpha (TNF-α) induced reactivation of infectious virus, and by analyzing expression of the progenitor cell marker CD34 as well as myeloid cell differentiation markers in IE+ cells after treatment with TNF-α, we showed that TNF-α induced transcriptional reactivation of IE gene expression independently of differentiation. TNF-α-mediated reactivation in Kasumi-3 cells was correlated with activation of NF-κB, KAP-1, and ATM. IMPORTANCE HCMV is an important human pathogen that establishes lifelong latent infection in myeloid progenitor cells and reactivates frequently to cause significant disease in immunocompromised people. Our observation that viral gene expression is first turned on and then turned off to establish latency suggests that there is a host defense, which may be myeloid cell specific, responsible for transcriptional silencing of viral gene expression. Our observation that TNF-α induces reactivation independently of differentiation provides insight into molecular mechanisms that control reactivation.