Embryonic development starts from a single cell known as a zygote. Through multiple rounds of proliferation and differentiation, this initial cell generates the vast array of specialized cells that make up the body. But each time a cell proliferates and goes through mitosis—the process of cell division that results in two genetically identical daughter cells—it faces an identity crisis. It must decide whether to keep its original identity through self-renewal or give rise to a new one through differentiation.
“Mitosis poses a temporal challenge to cell identity,” says Effie Apostolou, Medicine, Weill Cornell Medicine. “During mitosis there are massive structural and functional changes in the cell. Chromatin [the DNA and protein that make up a chromosome] becomes very compacted and loses its typical, cell-type-specific, 3D organization, while transcription is globally shut down and important transcription factors are disassociated from the chromatin. In the end, you have something that doesn’t necessarily know what it is anymore because everything that originally defined that cell is perturbed. So how does that cell remember what it was, if it decides to self-renew, or how does it choose a new identity?”
Next-Gen Technology and Pluripotent Stem Cells
Apostolou’s research focuses on the complicated mechanics behind these cell-fate decisions. To peer into this dynamic world of the nucleus, the Apostolou lab uses a broad array of genome-wide, next-generation sequencing technologies: from PRO-seq (which provides insight into the transcriptional activity of a cell) to ChIP-seq (which analyzes protein interactions with DNA) to Hi-C (which looks at genome-wide, three-dimensional chromatin organization), and many more.
“[The cell] doesn’t necessarily know what it is anymore [after mitosis] because everything that originally defined that cell is perturbed.”
Apostolou and her colleagues apply these methods primarily to mouse pluripotent stem cells (cells able to give rise to almost any cell type in the body). The researchers either isolate the cells from early embryos or reprogram somatic cells (any cell other than egg or sperm cells) by pushing them all the way back to an early embryonic state where they become pluripotent. Known as induced pluripotent stem cells (iPSCs), they are identical molecularly and functionally to embryonic stem cells.
“PSCs divide very rapidly and can go through multiple cell divisions while maintaining their immature state and ability to differentiate into any cell type,” Apostolou explains. “They constitute a great system to study the mechanisms of cell-fate decisions.”
Mechanisms of Cell Self-Renewal
In a recent study looking at how a cell resets its identity after the temporal mitotic crisis, the researchers took PSCs and tracked them through the cell cycle. They wanted to discover how transcription (the first step of gene expression from DNA to RNA) and three-dimensional genome architecture (the hierarchical manner in which a cell’s two-meter-long DNA is packed into its tiny 10-micron nucleus) are reset, since they are both lost during mitosis.
“We isolated cells in mitosis and their daughter cells at consecutive stages after they exited mitosis and entered a new cell cycle,” Apostolou says. “Then we generated a four-dimensional map that allowed us to capture the dynamic, 3D chromatin reorganization and transcriptional reactivation on a genome-wide scale and over time.”
At the transcriptional level, the researchers discovered that all the genes that are important for stem cell identity are the first ones to reactivate. By contrast, genes critical for differentiation, which are normally inactive in pluripotent stem cells, turn on transiently after mitosis. “We believe these differentiational genes are probing for developmental signals that will tell them to become a different cell,” Apostolou says. “If they don’t receive those signals, they switch off again, and the cell goes back to self-renewal.”
How Do Cells Reset 3D Organization?
In parallel with the transcription study, the Apostolou lab explored the hierarchical three-dimensional organization in the nucleus. Normally the continuous, threadlike DNA coils in sections around cores of histone proteins to form nucleosomes. The chain of nucleosomes then forms loops in order to fit inside the nucleus. As a result of the looping, disparate sections of the linear genome end up next to each other, which allows for DNA regulatory elements—called enhancers—to control target genes over large distances. The loops are further organized into small neighborhoods known as contact domains, which are usually separated by boundaries. Finally, the whole thing is arranged into two compartments in which regions of the genome that are more active and gene-rich tend to cluster together, sequestered from the inactive, gene-poor regions.
Apostolou and her colleagues found that the boundaries of contact domains are reset very early in the post-mitosis cycle. The chromatin loops, however, are established more gradually and mostly after transcription. “This is one piece of evidence that the two processes of transcription and 3D organization are at least partially uncoupled, at least in the context of mitotic exit,” Apostolou explains.
“This is significant and partly surprising, since there is a large number of publications from us, and from others, which support very tight and potentially causal links between 3D organization and transcription during various cell-fate decisions,” she continues. “My group is excited to discover and learn more about the intricacies and interconnections between these two processes.”
Bookmarking for Gene Reactivation
One of the researchers’ most important findings involved the mechanism behind the asynchrony of gene activation after mitosis. “We stumbled on a very intriguing concept called bookmarking,” Apostolou says. “These are select histone marks, or proteins, on the chromatin that are retained during mitosis only in certain parts of the genome. They are thought to promote a quick activation of the bookmarked genes after mitosis. Our study supports this as it shows that mitotic retention of the histone mark called H3K27 acetylation can strongly predict the rapid resetting of stem cell genes and enhancers, suggesting a bookmarking function.”
Once they had established this, Apostolou and her collaborators depleted the mark specifically during mitosis and found that while transcriptional reactivation was perturbed, there was no effect on three-dimensional organization. “This further indicates that, in this context, while transcription and 3D organization support each other, they are not necessarily driven by the same mechanism,” she says.
Cell-Fate Decisions and Human Disease
After years of working with mouse pluripotent stem cells and reprogramming, Apostolou is now branching out to disease-relevant, cell-fate decisions in humans. Together with new interdisciplinary collaborators from Weill Cornell Medicine and Memorial Sloan Kettering Cancer Center, she is planning to dissect how enhancer-gene interactions in the three-dimensional genome regulate both normal pancreatic differentiation and diabetes. She will also explore the tumorigenic properties in such cancers as glioblastoma and lymphoma.
In addition, Apostolou has just begun an intercampus collaboration with John T. Lis and Elizabeth H. Kellogg, both Molecular Biology and Genetics; Haiyuan Yu, Computational Biology; and Hening Lin, Chemistry and Chemical Biology. The researchers will work together to identify the properties and composition of large, macromolecular complexes during the re-establishment of a cell’s identity after mitosis, among other things. “I feel incredibly lucky to be part of such an interdisciplinary team,” Apostolou says. “I think we’ll soon get very exciting results as we exchange tools and expertise.”