Danny Reinberg studies the dynamic processes shaping chromatin structure and channeling transcriptional outcome: key to fostering a cellular identity (Figure 1).
The intricate processes establishing the regulated expression of genes in mammalian cells give rise to the distinctive patterns of proteins produced and the distinguishing traits and functions of the various tissues. This specific patterning of gene expression must be conveyed to daughter cells upon cell division to ensure their identical properties. Our strategy in investigating these astounding phenomena involves intensive mechanistic analyses performed in vitro and with this information, exploring the integration and functioning of these mechanisms in vivo. We steadily incorporated increasingly complex biological contexts into our studies over the years and we are now advancing our investigations into the fascinating process of chromatin compartmentalization and its role in gene regulation from embryonic stem cells (ESC) to cells of a defined lineage. As well, our goal of studying epigenetics in the context of a whole organism is now coming to fruition based on our extensive studies of the eusocial insect, ants.
The expression or repression of genes at the transcriptional level boils down to the structural state of chromatin⏤whether it is made accessible or inaccessible to the transcriptional machinery. “Epi”genetic features of the nucleosome-packaged DNA (chromatin) foster a particular chromatin structure and importantly, are distinct from the DNA itself and yet, are inherited during cellular division. This inheritance is paramount to maintaining a cellular identity.
Below, I describe our recent findings that in several cases arose from distinct systems devised to capture the dynamics of a particular process in action.
Polycomb group (PcG) of epigenetic regulators
Key to maintaining genes in a repressed state are the Polycomb group (PcG) of proteins that function ultimately to compact chromatin. PcG proteins exist in two major multi-subunit Polycomb Repressive Complexes (PRC) in the cell: PRC1 and PRC2. Our isolation and characterization of mammalian PRC2 revealed that its Ezh2 component catalyzes di- and tri-methylation of histone H3 at lysine 27 (H3K27me2/3). The presence of H3K27me2/3 in chromatin provides the platform for chromatin compaction and thus, gene repression. Notably, PRC2 can efficiently generate expansive regions of H3K27me3 that foster extensive repressed chromatin domains due to its “write-and-read” mechanism, a self-propagating phenomenon: its Eed component binds to the PRC2 product, H3K27me3, resulting in a conformational change in PRC2 (allosteric activation) that stimulates its catalytic activity towards naive nucleosomes (Figures 2 and 3). This property is likely pivotal to the restoration of repressed chromatin domains during cell division (see Inheriting chromatin domains, below).
This property is likely pivotal to the restoration of repressed chromatin domains during cell division (see Inheriting chromatin domains, below).
PRC2 recruitment to chromatin The vast majority of PRC2 target genes are those that are developmentally regulated. We sought to expose the salient features of repressive chromatin domain formation as it occurs in vivo by tracking PRC2 recruitment de novo to chromatin and its subsequent formation of repressive domains. After completely depleting ESC of PRC2 and its catalytic products, an intact PRC2 was re-constructed and its interaction with chromatin and the emergence of its catalytic products, H3K27me2/me3, was followed in real time (Figure 4). Amongst our findings, we identified the preferred nucleation sites (10 or more repeats of GA or GCN) for PRC2 recruitment to CpG islands (CGI) within the promoters of developmental genes. We are following-up on the role of long-range interactions that form “PRC2 hubs” in the nucleus in fostering PRC2 interaction with weaker sites (<10 CGI repeats). Importantly, by re-constructing PRC2 with an EED mutant that cannot interact with H3K27me3, PRC2 is confined to the strong nucleation sites and cannot spread (Figure 4), underscoring the importance of its allosteric activation in forming extensive, repressive chromatin domains. The basic scheme of this system is applicable to other processes for which appropriate assays can follow the evolving events.
How a naturally occurring mutant in the PRC2 substrate thwarts PRC2 The presence of a naturally occurring mutation in the substrate of PRC2, H3K27M (methionine substitution of lysine 27 of histone H3), gives rise to a severely defective profile of PRC2-mediated repressive chromatin domains in an untreatable pediatric cancer, Diffuse Intrinsic Pontine Glioma (DIPG). We approached this anomaly by tracking over time the repercussions to PRC2 chromatin occupation and repressive domain formation when H3K27M is expressed de novo in unexposed cells. PRC2 interacts with H3K27M-containing chromatin, but only transiently. Surprisingly, PRC2 is released in a permanently impaired state, suggesting that an irreversible conformation switch is induced by the oncohistone H3K27M, poisoning PRC2 activity. By following the cascade of events that culminate in this highly defective gene expression profile, we found that active chromatin domains comprising H3K36me2 inappropriately invade the domains normally repressed through PRC2. We are pursuing both of these signatures, PRC2 conformation and proteins interacting with H3K36me2, which may serve as a novel vulnerability for therapeutic intervention.
Inheriting chromatin domains The essence of epigenetics entails those features of chromatin other than DNA that are inherited during cell division, preserving the integrity of the gene expression profile inherent to a specific cell type. Yet, which, if any, of the many histone post-translational modifications (hPTMs) within chromatin are epigenetic was unknown. To address this fundamental question, we devised a powerful, new method to temporally track in vivo the fate of parental nucleosomes during DNA replication at single gene resolution to rigorously assess nucleosome segregation as a function of the transcription status. This procedure incorporated several stringent criteria to ensure that endogenous histones H3.1 and H3.2 were “marked” only during a brief period in the G1 phase of the cell-cycle.
Fighting (dueling) in the wild type ants
A fight between two female workers in a queenless colony. Two female workers (without wing) are fighting for the queen throne with antennae twitching behavior. Smaller ants with wings and lighter colors are male ants.
Wandering behavior in the Orco mutant
Wandering phenotypes displayed by Orco mutants. Three Orco mutant workers (ORR: Orange-Red-Red, OBR:Orange-Blue-Red and OWO: Orange-White-Orange) leave the nest and wander outside the nest.
What Can Ants Teach Us About Epigenetics?
What better way to probe changes in the epigenetic profile as a result of environmental cues than in a whole organism? In actuality, there is not much known about epigenetic changes that may underlie behavior. But which organism constitutes an experimentally-tenable resource of epigenetic processes? It took a chance encounter with my scientific colleague and friend, Shelley Berger, to identify one. At a scientific meeting, Shelley mentioned her recent fascination with ant behavior after observing leaf cutter ants in Costa Rica. She suggested that ants might be a perfect organism to address the big question of an epigenetic basis underlying behavior.
Epigenetics includes chromatin constituents that impact directly or indirectly the chromatin structure and thus, the transcription status of genes. Importantly, these constituents are independent of the DNA sequence and yet, are also inherited during cell division. Epigenetic changes foster specific gene expression profiles during development and must be maintained to safeguard cellular identity. Ants seemed to present an incredibly exciting, yet challenging opportunity to develop a model system-exciting because members of an ant colony exhibit well-defined behavioral and morphological differences that might arise from epigenetic phenomena, and challenging as we would need to develop genetics in ants to fully explore the processes involved.
In an invigorating collaborative effort funded by the Howard Hughes Medical Institute, we set out to establish ants as a model system to explore the epigenetic input underlying differences in longevity, social behavior and aging among the queen and other castes of an ant colony. Shelley and I brought several top-notch scientists of different expertise onboard. For example, Juergen Liebig at Arizona State University is one of the top experts in studying how division of labor is maintained in insect societies. For years, Juergen had been planning to expand his research to study differential gene expression underlying ant behavior and the differences in aging in his established model ant systems. From Juergen, Shelley and I learned that ants exhibit a range of interesting behaviors, including high sociability and well-defined division of labor within their colonies. For example, the ants within a colony can assume either reproductive or non-reproductive roles. The queen is the sole egg-layer and does not have to leave the nest and hence, the queen’s brain does not function in complex tasks and is accordingly smaller. In contrast, non-reproductive nestmates maintain the colony, raise the brood and forage for food, requiring complex brain function and hence a larger brain. The different reproductive roles also have a strong impact on the longevity of a queen and workers: a queen lives up to 10 times longer than worker ants. Even though these two types of ants have a high degree of genetic relatedness and begin life similarly, their individual experiences sculpt their brains and behaviors in vastly different ways.
One of the first goals of this extensive collaborative program was to fully sequence, assemble and begin annotating the genomes of two ant species: Camponotus floridanus (Cflo) and Harpegnathos saltatory (Hsal), which we successfully achieved. Some of our subsequent studies regarding the epigenetic basis of reproduction and aging in Hsal are described in Current Research, above. In the case of Cflo, the colony comprises one queen and two sub-castes: “majors” that guard the ant colony and “minors” that forage for food. Our evidence from studies with Cflo point straight to epigenetic regulation of caste-specific behavior: altering the balance between acetylated and de-acetylated versions of lysine 27 of histone H3 in the brain of “majors” leads to their conversion into “minors”-a bona fide epigenetic phenomenon whereby changes in histone post-translational modifications that impact chromatin structure gives rise to altered behavior. Because we succeeded in establishing genetics in Hsal using CRISPR technology, we can now fully study the function of genes involved in the processes described here and in Current Research and their effect on the epigenetic program-the overarching goal of this intensive endeavor.