Activator protein-1 transactivation of the major immediate early locus is a determinant of cytomegalovirus reactivation from latency

Abstract

Human cytomegalovirus (HCMV) is a ubiquitous pathogen that latently infects hematopoietic cells and has the ability to reactivate when triggered by immunological stress. This reactivation causes significant morbidity and mortality in immune-deficient patients, who are unable to control viral dissemination. While a competent immune system helps prevent clinically detectable viremia, a portrait of the factors that induce reactivation following the proper cues remains incomplete. Our understanding of the complex molecular mechanisms underlying latency and reactivation continue to evolve. We previously showed the HCMV-encoded G-protein coupled receptor US28 is expressed during latency and facilitates latent infection by attenuating the activator protein-1 (AP-1) transcription factor subunit, c-fos, expression and activity. We now show AP-1 is a critical component for HCMV reactivation. Pharmacological inhibition of c-fos significantly attenuates viral reactivation. In agreement, infection with a virus in which we disrupted the proximal AP-1 binding site in the major immediate early (MIE) enhancer results in inefficient reactivation compared to wild type. Concomitantly, AP-1 recruitment to the MIE enhancer is significantly decreased following reactivation of the mutant virus. Further, AP-1 is critical for de-repression of MIE-driven transcripts and downstream early and late genes, while immediate early genes from other loci remain unaffected. Our data also reveal MIE transcripts driven from the MIE promoter, the distal promoter, and the internal promoter, iP2, are dependent upon AP-1 recruitment, while iP1-driven transcripts are AP-1-independent. Collectively, our data demonstrate AP-1 binding to and activation of the MIE enhancer is a key molecular process controlling reactivation from latency.Significance StatementHuman cytomegalovirus (HCMV) is a common pathogen that infects the majority of the population for life. This infection poses little threat in immunologically healthy individuals, but can be fatal in people with compromised immune systems. Our understanding of the mechanisms underlying latency and reactivation remains incomplete. Here, we show the cellular transcription factor, AP-1, is a key to regulating HCMV reactivation. Our findings reveal AP-1 binding to the major immediate early enhancer/promoter is critical for switching this locus from one that is repressed during latency to one that is highly active following reactivation. Our work provides a novel mechanism HCMV exploits to reactivate, highlighting AP-1 as a potential target to prevent HCMV reactivation.

Regulation of the MIE Locus During HCMV Latency and Reactivation

“…As iP1 and iP2 are located within intron A, each of their transcripts lack the noncoding exon 1 [ 42 ]. In addition to functioning during lytic infection, these promoters are active in myeloid cells under conditions that favor latency, and these promoters play a role in reactivation in myeloid cells [ 42 , 81 , 82 , 83 ]. That these alternative promoters were only recently identified suggests our understanding of the MIE locus is far from complete.…”

“…The AP-1 consensus site is between positions −174 and −168, relative to the transcription start site of the core promoter at position +1 [ 214 ], while the non-consensus site (5′-TGACTAA-3′) is located between positions −239 and −233 [ 159 , 160 , 215 ]. While neither site alone or in combination is required for efficient lytic replication in fibroblast or epithelial cells [ 83 , 160 ], AP-1 binding to its consensus sequence in the MIE enhancer is critical for successful reactivation from latency, as it is necessary for de-repression of MIE-driven transcripts, but not other IE genes [ 83 ]. Additionally, AP-1 recruitment to its consensus site in the MIE enhancer was necessary for the expression of MIEP-, iP2-, and dP-derived transcripts, while iP1-driven transcription was not altered with disruption of the AP-1 canonical binding site [ 83 ], suggesting the MIE canonical and alternative promoters differentially respond to transcription factor binding.…”

“…While neither site alone or in combination is required for efficient lytic replication in fibroblast or epithelial cells [ 83 , 160 ], AP-1 binding to its consensus sequence in the MIE enhancer is critical for successful reactivation from latency, as it is necessary for de-repression of MIE-driven transcripts, but not other IE genes [ 83 ]. Additionally, AP-1 recruitment to its consensus site in the MIE enhancer was necessary for the expression of MIEP-, iP2-, and dP-derived transcripts, while iP1-driven transcription was not altered with disruption of the AP-1 canonical binding site [ 83 ], suggesting the MIE canonical and alternative promoters differentially respond to transcription factor binding. In agreement with this, Hale et al recently identified forkhead box (FOX) binding sites in MIE intron A, capable of supporting FOXO1 and FOXO3a binding [ 161 ].…”

“…Thus, US28-mediated downregulation of c-fos prevents AP-1 binding to the MIE enhancer, suppressing transcription from this region [ 272 ]. As mentioned above, AP-1 is critical to the balance between latency and reactivation, as it is required for successful viral reactivation [ 83 ]. Thus, US28 likely prevents premature AP-1 binding to the MIE enhancer during latent infection in the absence of reactivation cues.…”