DOT1L interaction partner AF10 controls patterning of H3K79 methylation and RNA polymerase II to maintain cell identity.

Histone 3 lysine 79 methylation (H3K79me) is enriched on gene bodies proportional to gene expression levels and serves as a strong barrier for the reprogramming of somatic cells to induced pluripotent stem cells (iPSCs). DOT1L is the sole histone methyltransferase that deposits all three orders-mono (me1), di (me2), and tri (me3) methylation-at H3K79. Here, we leverage genetic and chemical approaches to parse the specific functions of orders of H3K79me in maintaining cell identity. DOT1L interacts with AF10 (Mllt10), which recognizes unmodified H3K27 and boosts H3K79me2/3 methylation. AF10 deletion evicts H3K79me2/3 and reorganizes H3K79me1 to the transcription start site to facilitate iPSC formation in the absence of steady-state transcriptional changes. Instead, AF10 loss redistributes RNA polymerase II to a uniquely pluripotent pattern at highly expressed, rapidly transcribed housekeeping genes. Taken together, we reveal a specific mechanism for H3K79me2/3 located at the gene body in reinforcing cell identity.

DOT1L is a barrier to histone acetylation during reprogramming to pluripotency.

Embryonic stem cells (ESCs) have transcriptionally permissive chromatin enriched for gene activation-associated histone modifications. A striking exception is DOT1L-mediated H3K79 dimethylation (H3K79me2) that is considered a positive regulator of transcription. We find that ESCs are depleted for H3K79me2 at shared locations of enrichment with somatic cells, which are highly and ubiquitously expressed housekeeping genes, and have lower RNA polymerase II (RNAPII) at the transcription start site (TSS) despite greater nascent transcription. Inhibiting DOT1L increases the efficiency of reprogramming of somatic to induced pluripotent stem cells, enables an ESC-like RNAPII pattern at the TSS, and functionally compensates for enforced RNAPII pausing. DOT1L inhibition increases H3K27 methylation and RNAPII elongation-enhancing histone acetylation without changing the expression of the causal histone-modifying enzymes. Only the maintenance of elevated histone acetylation is essential for enhanced reprogramming and occurs at loci that are depleted for H3K79me2. Thus, DOT1L inhibition promotes the hyperacetylation and hypertranscription pluripotent properties.

Connecting the DOTs on Cell Identity.

DOT1-Like (DOT1L) is the sole methyltransferase of histone H3K79, a modification enriched mainly on the bodies of actively transcribing genes. DOT1L has been extensively studied in leukemia were some of the most frequent onco-fusion proteins contain portions of DOT1L associated factors that mislocalize H3K79 methylation and drive oncogenesis. However, the role of DOT1L in non-transformed, developmental contexts is less clear. Here we assess the known functional roles of DOT1L both in vitro cell culture and in vivo models of mammalian development. DOT1L is evicted during the 2-cell stage when cells are totipotent and massive epigenetic and transcriptional alterations occur. Embryonic stem cell lines that are derived from the blastocyst tolerate the loss of DOT1L, while the reduction of DOT1L protein levels or its catalytic activity greatly enhances somatic cell reprogramming to induced pluripotent stem cells. DOT1L knockout mice are embryonically lethal when organogenesis commences. We catalog the rapidly increasing studies of total and lineage specific knockout model systems that show that DOT1L is broadly required for differentiation. Reduced DOT1L activity is concomitant with increased developmental potential. Contrary to what would be expected of a modification that is associated with active transcription, loss of DOT1L activity results in more upregulated than downregulated genes. DOT1L also participates in various epigenetic networks that are both cell type and developmental stage specific. Taken together, the functions of DOT1L during development are pleiotropic and involve gene regulation at the locus specific and global levels.

DOT1L inhibition enhances pluripotency beyond acquisition of epithelial identity and without immediate suppression of the somatic transcriptome.

Inhibiting the histone 3 lysine 79 (H3K79) methyltransferase, disruptor of telomeric silencing 1-like (DOT1L), increases the efficiency of reprogramming somatic cells to induced pluripotent stem cells (iPSCs). Here, we find that, despite the enrichment of H3K79 methylation on thousands of actively transcribed genes in somatic cells, DOT1L inhibition (DOT1Li) does not immediately cause the shutdown of the somatic transcriptional profile to enable transition to pluripotency. Contrary to the prevalent view, DOT1Li promotes iPSC generation beyond the mesenchymal to epithelial transition and even from already epithelial cell types. DOT1Li is most potent at the midpoint of reprogramming in part by repressing Nfix that persists at late stages of reprogramming. Importantly, regulation of single genes cannot substitute for DOT1Li, demonstrating that H3K79 methylation has pleiotropic effects in maintaining cell identity.

Methyl-Metabolite Depletion Elicits Adaptive Responses to Support Heterochromatin Stability and Epigenetic Persistence.

S-adenosylmethionine (SAM) is the methyl-donor substrate for DNA and histone methyltransferases that regulate epigenetic states and subsequent gene expression. This metabolism-epigenome link sensitizes chromatin methylation to altered SAM abundance, yet the mechanisms that allow organisms to adapt and protect epigenetic information during life-experienced fluctuations in SAM availability are unknown. We identified a robust response to SAM depletion that is highlighted by preferential cytoplasmic and nuclear mono-methylation of H3 Lys 9 (H3K9) at the expense of broad losses in histone di- and tri-methylation. Under SAM-depleted conditions, H3K9 mono-methylation preserves heterochromatin stability and supports global epigenetic persistence upon metabolic recovery. This unique chromatin response was robust across the mouse lifespan and correlated with improved metabolic health, supporting a significant role for epigenetic adaptation to SAM depletion in vivo. Together, these studies provide evidence for an adaptive response that enables epigenetic persistence to metabolic stress.

Restricted TET2 Expression in Germinal Center Type B Cells Promotes Stringent Epstein-Barr Virus Latency.

Epstein-Barr virus (EBV) latently infects normal B cells and contributes to the development of certain human lymphomas. Newly infected B cells support a highly transforming form (type III) of viral latency; however, long-term EBV infection in immunocompetent hosts is limited to B cells with a more restricted form of latency (type I) in which most viral gene expression is silenced by promoter DNA methylation. How EBV converts latency type is unclear, although it is known that type I latency is associated with a germinal center (GC) B cell phenotype, and type III latency with an activated B cell (ABC) phenotype. In this study, we have examined whether expression of TET2, a cellular enzyme that initiates DNA demethylation by converting 5-methylcytosine (5mC) into 5-hydroxymethylcytosine (5hmC), regulates EBV latency type in B cells. We found that TET2 expression is inhibited in normal GC cells and GC type lymphomas. In contrast, TET2 is expressed in normal naive B cells and ABC type lymphomas. We also demonstrate that GC type cell lines have increased 5mC levels and reduced 5hmC levels in comparison to those of ABC type lines. Finally, we show that TET2 promotes the ability of the EBV transcription factor EBNA2 to convert EBV-infected cells from type I to type III latency. These findings demonstrate that TET2 expression is repressed in GC cells independent of EBV infection and suggest that TET2 promotes type III EBV latency in B cells with an ABC or naive phenotype by enhancing EBNA2 activation of methylated EBV promoters.IMPORTANCE EBV establishes several different types of viral latency in B cells. However, cellular factors that determine whether EBV enters the highly transforming type III latency, versus the more restricted type I latency, have not been well characterized. Here we show that TET2, a cellular enzyme that initiates DNA demethylation by converting 5-methylcytosine (5mC) into 5-hydroxymethylcytosine (5hmC), regulates EBV latency type in B cells by enhancing the ability of the viral transcription factor EBNA2 to activate methylated viral promoters that are expressed in type III (but not type I) latency. Furthermore, we demonstrate that (independent of EBV) TET2 is turned off in normal and malignant germinal center (GC) B cells but expressed in other B cell types. Thus, restricted TET2 expression in GC cells may promote type I EBV latency.

5-hydroxymethylation of the EBV genome regulates the latent to lytic switch.

Latent Epstein-Barr virus (EBV) infection and cellular hypermethylation are hallmarks of undifferentiated nasopharyngeal carcinoma (NPC). However, EBV infection of normal oral epithelial cells is confined to differentiated cells and is lytic. Here we demonstrate that the EBV genome can become 5-hydroxymethylated and that this DNA modification affects EBV lytic reactivation. We show that global 5-hydroxymethylcytosine (5hmC)-modified DNA accumulates during normal epithelial-cell differentiation, whereas EBV+ NPCs have little if any 5hmC-modified DNA. Furthermore, we find that increasing cellular ten-eleven translocation (TET) activity [which converts methylated cytosine (5mC) to 5hmC] decreases methylation, and increases 5hmC modification, of lytic EBV promoters in EBV-infected cell lines containing highly methylated viral genomes. Conversely, inhibition of endogenous TET activity increases lytic EBV promoter methylation in an EBV-infected telomerase-immortalized normal oral keratinocyte (NOKs) cell line where lytic viral promoters are largely unmethylated. We demonstrate that these cytosine modifications differentially affect the ability of the two EBV immediate-early proteins, BZLF1 (Z) and BRLF1 (R), to induce the lytic form of viral infection. Although methylation of lytic EBV promoters increases Z-mediated and inhibits R-mediated lytic reactivation, 5hmC modification of lytic EBV promoters has the opposite effect. We also identify a specific CpG-containing Z-binding site on the BRLF1 promoter that must be methylated for Z-mediated viral reactivation and show that TET-mediated 5hmC modification of this site in NOKs prevents Z-mediated viral reactivation. Decreased 5-hydroxymethylation of cellular and viral genes may contribute to NPC formation.

Differentiation-Dependent KLF4 Expression Promotes Lytic Epstein-Barr Virus Infection in Epithelial Cells.

Epstein-Barr virus (EBV) is a human herpesvirus associated with B-cell and epithelial cell malignancies. EBV lytically infects normal differentiated oral epithelial cells, where it causes a tongue lesion known as oral hairy leukoplakia (OHL) in immunosuppressed patients. However, the cellular mechanism(s) that enable EBV to establish exclusively lytic infection in normal differentiated oral epithelial cells are not currently understood. Here we show that a cellular transcription factor known to promote epithelial cell differentiation, KLF4, induces differentiation-dependent lytic EBV infection by binding to and activating the two EBV immediate-early gene (BZLF1 and BRLF1) promoters. We demonstrate that latently EBV-infected, telomerase-immortalized normal oral keratinocyte (NOKs) cells undergo lytic viral reactivation confined to the more differentiated cell layers in organotypic raft culture. Furthermore, we show that endogenous KLF4 expression is required for efficient lytic viral reactivation in response to phorbol ester and sodium butyrate treatment in several different EBV-infected epithelial cell lines, and that the combination of KLF4 and another differentiation-dependent cellular transcription factor, BLIMP1, is highly synergistic for inducing lytic EBV infection. We confirm that both KLF4 and BLIMP1 are expressed in differentiated, but not undifferentiated, epithelial cells in normal tongue tissue, and show that KLF4 and BLIMP1 are both expressed in a patient-derived OHL lesion. In contrast, KLF4 protein is not detectably expressed in B cells, where EBV normally enters latent infection, although KLF4 over-expression is sufficient to induce lytic EBV reactivation in Burkitt lymphoma cells. Thus, KLF4, together with BLIMP1, plays a critical role in mediating lytic EBV reactivation in epithelial cells.

Viral genome methylation differentially affects the ability of BZLF1 versus BRLF1 to activate Epstein-Barr virus lytic gene expression and viral replication.

The Epstein-Barr virus (EBV) immediate-early proteins BZLF1 and BRLF1 can both induce lytic EBV reactivation when overexpressed in latently infected cells. Although EBV genome methylation is required for BZLF1-mediated activation of lytic gene expression, the effect of viral genome methylation on BRLF1-mediated viral reactivation has not been well studied. Here, we have compared the effect of viral DNA methylation on BZLF1- versus BRLF1-mediated activation of lytic EBV gene transcription and viral genome replication. We show that most early lytic viral promoters are preferentially activated by BZLF1 in the methylated form, while methylation decreases the ability of BRLF1 to activate most early lytic promoters, as well as the BLRF2 late viral promoter. Moreover, methylation of bacmid constructs containing the EBV genome enhances BZLF1-mediated, but decreases BRLF1-mediated, early lytic gene expression. Methylation of viral promoter DNA does not affect BRLF1 binding to a variety of different CpG-containing BRLF1 binding motifs (RREs) in vitro or in vivo. However, BRLF1 preferentially induces H3K9 histone acetylation of unmethylated promoters in vivo. The methylated and unmethylated forms of an oriLyt-containing plasmid replicate with similar efficiency when transfected into EBV-positive cells that express the essential viral replication proteins in trans. Most importantly, we demonstrate that lytic viral gene expression and replication can be induced by BRLF1, but not BZLF1, expression in an EBV-positive telomerase-immortalized epithelial cell line (NOKs-Akata) in which lytic viral gene promoters remain largely unmethylated. These results suggest that the unmethylated form of the EBV genome can undergo viral reactivation and replication in a BRLF1-dependent manner.

Cellular transcription factor Oct-1 interacts with the Epstein-Barr virus BRLF1 protein to promote disruption of viral latency.

The Epstein-Barr virus (EBV) latent-to-lytic switch is an essential part of the viral life cycle, but the cellular factors that promote viral reactivation are not well defined. In this report, we demonstrate that the cellular transcription factor Oct-1 cooperates with the EBV immediate-early protein BRLF1 (R, Rta) to induce lytic viral reactivation. We show that cotransfected Oct-1 enhances the ability of BRLF1 to activate lytic gene expression in 293 cells stably infected with a BRLF1-defective EBV mutant (BRLF1-stop) and that Oct-1 increases BRLF1-mediated activation of lytic EBV promoters in reporter gene assays. We find that Oct-1 interacts directly with BRLF1 in vitro and that a mutant BRLF1 protein (the M140A mutant) attenuated for the ability to interact with Oct-1 in vitro is also resistant to Oct-1-mediated transcriptional enhancement in 293 BRLF1-stop cells. Furthermore, we show that cotransfected Oct-1 augments BRLF1 binding to a variety of lytic EBV promoters in chromatin immunoprecipitation (ChIP) assays (including the BZLF1, BMRF1, and SM promoters) and that BRLF1 tethers Oct-1 to lytic EBV promoters. In addition, we demonstrate that an Oct-1 mutant defective in DNA binding (the S335D mutant) still retains the ability to enhance BRLF1 transcriptional effects. Finally, we show that knockdown of endogenous Oct-1 expression reduces the level of constitutive lytic EBV gene expression in both EBV-positive B-cell and EBV-positive epithelial cell lines. These results suggest that Oct-1 acts as a positive regulator of EBV lytic gene expression and that this effect is at least partially mediated through its interaction with the viral protein BRLF1.