Research Article: A Viral microRNA Cluster Regulates the Expression of PTEN, p27 and of a bcl-2 Homolog

Date Published: January 22, 2016

Publisher: Public Library of Science

Author(s): Katharina Bernhardt, Janina Haar, Ming-Han Tsai, Remy Poirey, Regina Feederle, Henri-Jacques Delecluse, Erik K Flemington.


The Epstein-Barr virus (EBV) infects and transforms B-lymphocytes with high efficiency. This process requires expression of the viral latent proteins and of the 3 miR-BHRF1 microRNAs. Here we show that B-cells infected by a virus that lacks these non-coding RNAs (Δ123) grew more slowly between day 5 and day 20, relative to wild type controls. This effect could be ascribed to a reduced S phase entry combined with a moderately increased apoptosis rate. Whilst the first phenotypic trait was consistent with an enhanced PTEN expression in B-cells infected with Δ123, the second could be explained by very low BHRF1 protein and RNA levels in the same cells. Indeed, B-cells infected either by a recombinant virus that lacks the BHRF1 protein, a viral bcl-2 homolog, or by Δ123 underwent a similar degree of apoptosis, whereas knockouts of both BHRF1 microRNAs and protein proved transformation-incompetent. We find that that the miR-BHRF1-3 seed regions, and to a lesser extent those of miR-BHRF1-2 mediate these stimulatory effects. After this critical period, B-cells infected with the Δ123 mutant recovered a normal growth rate and became more resistant to provoked apoptosis. This resulted from an enhanced BHRF1 protein expression relative to cells infected with wild type viruses and correlated with decreased p27 expression, two pro-oncogenic events. The upregulation of BHRF1 can be explained by the observation that large BHRF1 mRNAs are the source of BHRF1 protein but are destroyed following BHRF1 microRNA processing, in particular of miR-BHRF1-2. The BHRF1 microRNAs are unlikely to directly target p27 but their absence may facilitate the selection of B-cells that express low levels of this protein. Thus, the BHRF1 microRNAs allowed a time-restricted expression of the BHRF1 protein to innocuously expand the virus B-cell reservoir during the first weeks post-infection without increasing long-term immune pressure.

Partial Text

The Epstein-Barr virus (EBV) is the first discovered tumor human virus and is etiologically associated with approximately 2% of all tumors worldwide [1, 2]. These tumors are largely diverse in terms of lineage and include multiple types of lymphomas and carcinomas [3]. Immune deficiency, e.g. caused by immunosuppressive regimen is a strong risk factor for the development of EBV-associated lymphomas [2]. These tumors are thought, at least to some extent, to reflect EBV’s ability to transform primary B-cells [2]. This process can be easily observed in vitro as it leads to the establishment of lymphoblastoid cell lines (LCLs) and requires the simultaneous expression of some members of the viral latent gene family [2]. In recent years, it has become clear that the BHRF1 microRNAs (miRNAs) encoded by the virus markedly potentiate this process. Recombinant viruses that lack one or several of these three miR-BHRF1s are less transforming than their wild type counterparts and the effect is cumulative [4–6]. One study has ascribed this property to the ability of the BHRF1 miRNAs to prevent massive apoptosis in the first days of infection [4]. Furthermore, viruses that lack the three BHRF1 miRNAs (Δ123) grow more slowly and display abnormalities of the cell cycle [4, 5]. Humanized NSG mice infected by Δ123 eventually develop B-cell proliferations that are indistinguishable from those caused by wild type infection, but cell growth induced by the mutant is delayed by several weeks, confirming that the BHRF1 miRNAs are particularly required in the early phases of infection [7]. The BHRF1 protein, around which the BHRF1 microRNAs are located, has also been implicated in EBV-mediated B-cell transformation, although its role appears to be more difficult to define. BHRF1 is a bcl-2 homolog that shares its anti-apoptotic properties [8, 9]. Although its expression level is hardly detectable in LCLs, it is strongly expressed in the Wp-restricted Burkitt’s lymphoma (BL) cells, a subset of Burkitt’s lymphomas that are infected by EBVs that carry a deletion of the EBNA2 gene and whose latent genes are driven by the Wp promoter [10, 11]. A recombinant virus that lacks the BHRF1 protein retains full transformation abilities, suggesting that this protein is dispensable for transformation [12]. However, its enhanced expression in Wp-restricted BLs leads to a markedly enhanced resistance to apoptosis induced by ionomycin [11, 13]. Thus, both the BHRF1 protein and the BHRF1 miRNAs have been implicated in the regulation of apoptosis.

The clarification of the molecular mechanisms fine tuned by miRNAs is frequently a difficult undertaking and the BHRF1 miRNAs are no exception to that rule. CLIP-based strategies have identified potential targets of these miRNAs but the identity of the crucial proteins that they modulate remains enshrouded in mystery [17]. We have evaluated in detail the role served by the BHRF1 miRNAs during the first weeks of EBV-mediated B-cell transformation and found that cells infected with a virus that lacks the three BHRF1 miRNAs undergo on average twice as much apoptosis than cells infected with wild type controls between day 5 and day 20. Our results are only partially concordant with an earlier report that described a massive apoptosis in cells infected with a virus devoid of the BHRF1 miRNAs during the first five days of infection, followed by a quick recovery of cell numbers 10 days after infection [4]. These data are difficult to reconcile with the fact that the BHRF1 miRNAs are fully produced only 5 to 8 days after the onset of infection [21, 22] and we indeed did not see any differences between cells infected with the Δ123 mutant and cells infected with wild type controls during the first 5 days of infection. One difference between the studies lies in that we investigated the infected B-cells at the single cell level and used a larger panel of markers to characterize apoptotic cells. Indeed, positive staining for Annexin-V used in the study by Seto et al. is not specific for apoptosis and is found in many other modes of cell death [23].




0 0 vote
Article Rating
Notify of
Inline Feedbacks
View all comments