A paper by Kis and colleagues demonstrates that B cells latently infected with EBV can modulate their transcription program in response to host-derived signals as the cells transit the germinal center (GC) into memory.
Epstein-Barr virus (EBV) is associated with important human cancers and is known to exhibit very disparate biologies. Infection of B lymphocytes in vitro invariably leads to their transformation into proliferating, activated lymphoblasts through the expression of 9 latent proteins (the growth transcription program or latency III). In contrast, the virus is quiescent in vivo, persisting in resting, peripheral memory B cells that express none of the latent proteins (the latency program). The reconciliation of these behaviors came with the understanding that EBV uses normal B-cell biology to establish and maintain persistent infection.1 Specifically, EBV uses the growth program in vivo to activate newly infected B cells so they can differentiate through a GC to become resting, latently infected memory B cells (see figure). However, the growth program does not allow differentiation, so, once activated, the latently infected cell in the GC must switch its transcription program to a more restricted form of latency (the default program or latency II) where only 3 latent proteins are expressed. The crucial players in this program are LMP1 and LMP2a, which are able to give signals necessary for the latently infected cells to enter, survive, and leave the GC. We know this is a bona fide program of the virus because it is also found in Hodgkin disease, which is believed to arise from a latently infected GC cell. What we do not know is how EBV turns on the default program. In the growth program, expression of LMP1 and LMP2a is strictly controlled by the viral transcription factor EBNA2, but EBNA2 is not expressed in the default program. Kis and colleagues now show that LMP1 expression can be turned on in the absence of EBNA2 through signals provided by the cytokine environment present in GCs. Specifically, IL-10 is able to mediate this response. This is particularly apropos because IL-10 is known to drive IgA class switching.2 This raises the possibility that the interplay between IL-10 and the latently infected cells in the GCs may extend beyond turning on the default program to ensuring a predisposition for homing to the mucosal epithelium where EBV ultimately replicates and is shed into saliva.
Crucially, the authors do not address LMP2a, the other key signaling component of the default program. Is it also turned on or is it regulated independently? There is reason to believe that LMP2A is expressed before LMP1 in vivo since LMP2a drives GC development,3 whereas LMP1 turns off bcl-6,4 a crucial step for exiting the GC. Curiously, Kis et al did not see down-regulation of bcl-6 by LMP1—a conflict that needs to be resolved.
Many questions remain unanswered about how EBV switches from growth to default to latency transcription programs in its progress from new infection to persistence in memory cells. This paper demonstrates that this progress involves an interplay between virus- and host-derived signals. ▪FIG1
Comment on Kis et al, page 2928