- William B. Graham Professor of Human Genetics; Department of Human Genetics; Department of Molecular Genetics and Cell Biology; Howard Hughes Medical Institute; Committee on Cancer Biology; Committee on Neurobiology; Committee on Evolutionary Biology; Committee on Genetics, Genomics, and Systems Biology
Ph.D., Massachusetts Institute of Technology, 1998
The University of Chicago
929 East 57th Street
Chicago, Illinois 60637
Phone: (773) 834-4393
Website (Dept. of Human Genetics)
Mammalian genetics, development and function of the brain, and comparative and evolutionary genomics
Our main research interest lies at the interface of two disciplines: stem cell biology and epigenetics. Specifically, we wish to understand the epigenetic mechanisms of how cell fate becomes progressively restricted when pluripotent stem cells become progressively differentiated during development. While our study of cell fate restriction utilizes approaches in stem cell biology and epigenetics, it also intersects with many other fields, including developmental biology, genomics and systems biology, cancer biology, neurobiology, and microRNA biology. Our other research interests include the study of human evolution – especially evolution of the brain – using genetic, genomic and bioinformatic approaches, and the development of new technologies in genetics and stem cell biology. Below is a brief summary of our work on cell fate restriction.
The hallmark of multicellular life is the presence of diverse cell types in a single organism that exhibit disparate physiological functions despite bearing an identical genome. The multitude of cell types in an organism is created during development through the process of lineage differentiation – the progressive transformation of stem cells into a wide range of terminally differentiated, specialized cell types.
In mammals, as in many other taxa, lineage differentiation is characterized by two complementary aspects that can perhaps be thought of as the Yin and Yang of the process. The Yang (or the positive aspect) is cell fate determination, whereby differentiating cells progressively acquire phenotypic identities of specialized cell types. The Yin (or negative aspect) is cell fate restriction, whereby differentiating cells progressively lose their potential for all but the lineage that they are committed to. Cell fate restriction is manifested in two ways. First, cells in advanced stages of differentiation do not normally revert back to less differentiated states under physiological conditions (i.e., cells do not typically dedifferentiate). Second, cells that have differentiated down one developmental lineage do not normally switch into cell types of different lineages (i.e., cells do not typically transdifferentiate).
Of the Yin and Yang of lineage differentiation, the Yang – cell fate determination – has been a central focus of developmental biology for many decades, and a great deal of insight is now available regarding its mechanistic details. By contrast, the Yin – cell fate restriction – is scantly studied and its mechanism remains essentially unknown.
We have recently proposed a simple model, termed “occlusis”, to account for the mechanism of cell fate restriction. We coined this word by appending the suffix “sis” (meaning the process of) to the word “occlude”. Occlusis literally means “the process of occluding”, and it refers to the conjectured process of gene occlusion during lineage differentiation as detailed below.
Before delving into the model, it is useful to briefly recall a few current ideas about how transcription is regulated inside a eukaryotic cell. Broadly speaking, the transcriptional output of a gene in a cell is determined by two inputs. One is the trans-acting milieu of the cell, defined as all the diffusible factors that collectively impinge on the gene’s regulatory sequences to promote or repress its expression. The other is the cis-acting chromatin state of the gene itself, defined as the full complement of chromatin marks around the gene such as DNA methylation, histone modifications, and the binding of chromatin remodeling proteins, which in combination influence how the gene responds to the trans-acting milieu. It is also important to recognize that trans-acting milieu can, in many cases, directly influence cis-acting chromatin state of a gene, such that the trans and cis regulation of gene expression cannot always be considered as independent of each other.
Given the above discussion, it is conceivable that different cis-acting chromatin states may be able to lead to different expression levels of a gene even when the gene is present in the exact same trans-acting milieu. Should this be the case, it can be said that different chromatin states have conferred different transcriptional potential to the gene. This idea is central to the occlusis model. In its stripped-down version, the occlusis model consists of three components. The first postulates that the transcriptional potential of a gene can assume one of two states: either competent or occluded. In the competent state, a gene is capable of responding to trans-acting factors in the cellular milieu such that it is active when appropriate transcriptional activators are present, though it can also be silent when activators are absent or repressors are present. In the occluded state, by contrast, a gene is blocked by its cis-acting chromatin state from responding to trans-acting factors in the cellular milieu, such that it remains silent even if transcriptional activators for that gene are present in the cell. Importantly, this means that not only is it the case that an occluded gene can no longer be activated directly by trans-acting factors, it is also the case that trans-acting factors cannot turn on the gene indirectly by erasing the cis-acting chromatin state responsible for the occlusion of the gene. The second component of the model posits that, during development, an increasing number of genes shift from competent to occluded state as lineage differentiation progresses, and that this shift is essentially irreversible. In other words, a gene can transition from competent to occluded state during lineage differentiation, but once a gene is occluded, it cannot typically revert back to the competent state under physiological conditions. The third component postulates that the erasure of occlusion takes place during reproduction, such that all or nearly all genes are reset to the competent state at the very start of organismal development. Such erasure (which can be referred to as “deocclusion”) could occur in pluripotent stem cells post fertilization or in germ cells prior to fertilization.
Taken together, the occlusis model basically asserts that the progressive restriction of cell fate during lineage differentiation is achieved via the progressive occlusion of an increasing number of lineage-inappropriate genes, and furthermore, cells regain pluripotency during reproduction via genome-wide deocclusion. Here, lineage-inappropriate genes refer to genes in a given cell type that, if aberrantly expressed, would promote cell fate outside of the lineage that the cell is committed to.
Our lab is currently testing the occlusis model using a range of experimental approaches. Several pieces of evidence have emerged that support this model. The most critical of which is the development of an assay that led to the identification of occluded genes. The assay involves fusing two disparate cell types, and searching for genes in fused cells that are silent in the genome of one fusion partner but active in the genome of the other fusion partner. Here, the active copies of these genes serve as a positive control for the presence of a transcriptionally supportive milieu. With this positive control, the silent copies of these same genes in the fused cells can be ascertained as being occluded. Interestingly, we found that occluded genes in a given cell type tend to include master triggers of alternative cell fates. For example, in fibroblasts, occluded genes include master triggers of myogenic differentiation (e.g.,Myf5 and Myod1), neurogenic differentiation (e.g., Neurod1, Neurog2, Pax6 and Rbpj), hepatogenic differentiation (Hnf1a, Hnf3b, Hnf4a and Cebpa), and osteogenic differentiation (e.g., Runx2). This is consistent with the notion that the occlusion of key lineage-inappropriate genes in a given cell type serves the purpose of safeguarding the identity of that cell type. Our preliminary data also show that, consistent with the occlusis model, pluripotent embryonic stem cells are characterized by a dearth of occluded genes. Additionally, we also have preliminary evidence supporting the idea that occlusion is irreversible during lineage differentiation.
Currently, we are studying the occlusis model from several angles. First, we are continuing to ascertain the validity of the model by testing several key predictions of the model. Second, we are investigating the biochemical basis of gene occlusion, focusing on specific chromatin marks such as DNA methylation and covalent histone modifications. Third, we are exploring the implications of gene occlusion in a variety of biological processes including stem cell differentiation, cancer, aging, and microRNA regulation.