The Cell and Molecular Biology (CAMB) program spans a daunting breadth of biological inquiry, with students and investigators participating across six dynamic and complementary programs. This diversity is exemplified by the research of Ellen Elliott, recent alumna of the Developmental, Stem Cell and Regenerative Biology (DSRB) program and the Kaestner laboratory. Her work integrates the methods and concepts of regenerative biology, organismal development, physiology, and epigenetic regulation to investigate the role of DNA methylation in mammalian intestinal epithelia. In the January 2016 issue of eLife, Ellen and colleagues present that the DNA methyltransferase Dnmt3b can rescue detrimental DNA hypomethylation and stalled stem cell maturation caused by the loss of the methyltransferase Dnmt1 in adult mouse intestinal epithelia.
Ellen’s work is a product of the Kaestner lab’s efforts to understand the pathways of liver, pancreatic, and intestinal organogenesis using cell culture and mouse models. The lab uses these methods in part to study the contribution of DNA methylation to gene regulation. DNA methylation is a heritable covalent modification of eukaryotic DNA that some organisms use to differentially regulate maternally and
paternally derived chromosomes. This modification is essential for the timely, dosage-controlled expression of many genes, especially those genes involved in embryogenesis and development. These genes are methylated by two classes of methyltransferases, each with distinct targets and mechanisms of action. One class consists of the ‘maintenance’ methyltransferases Dnmt3a and Dnmt3b that faithfully copy the methylation patterns of parent DNA strands onto daughter DNA strands during semi-conservative DNA replication. The other class consists of the ‘de novo’ methyltransferase Dnmt1 that can lay down new methyl marks on DNA without a pre-existing template. The loss of either class of methyltransferases can lead to the loss of epigenetic information and can be fatal to actively-dividing cells, including embryonic stem cells.
Over the course of three publications, Ellen delves into the role of the Dnmt1 methyltransferase in the development and operation of the ‘crypt’ stem cells that replenish the mature cells constituting the intestinal villi. She proposes a novel relationship between the two methyltransferase classes by elegantly demonstrating that adult mouse crypt cells can overcome the loss of Dnmt1 by adapting and repurposing Dnmt3b. This compensation explains the foundational observations that Dnmt1 ablation only transiently affects the mature intestine, while being absolutely necessary for perinatal crypt maturation. Conditional Dnmt1-null adult intestines experience expansive hypomethylation at LINE1 and H19 elements and a global increase in DNA damage. However, these intestinal Dnmt1-null cells can survive and proliferate outwards, unlike in other actively-replicating tissues, where the Dnmt1 deletion is lethal. Contrary to expectations that Dnmt1 loss would arrest cell growth, the conditional knockout crypts even appear to proliferate outwards at an accelerated rate. The result is striking. The crypt cells amass without being able to differentiate into mature epithelial cells. Yet, the mice can survive the Dnmt1 deletion and recover wild-type crypt cell functionality. Within a week, Dnmt3b (though not Dnmt3a) mRNA and protein levels increase to compensate for the loss of Dnmt1, with a return to control levels of DNA methylation. Double mutants for Dnmt1 and Dnmt3b are not so fortunate. The double conditional knockout mice are severely morbid in 60% of cases, with decimated epithelial integrity, demethylated LINE1 and H19 loci, and significant DNA damage. The remaining 40% are only able to survive through the sporadic escape of crypt cells from Cre-recombination, repopulating the crypt zone with Dnmt3b-positive cells. Dnmt3b acquires added import and functionality as it compensates for Dnmt1. Control experiments show that Dnmt3b depletion alone does not affect crypt functionality, and that Dnmt3a cannot compensate for the Dnmt1/Dnmt3b null mice. Moreover, constitutive or inducible Dnmt3a or 3b loss in the intestinal epithelium by itself has no discernable phenotype or effect on viability.
The 'de novo' DNA methyltransferase Dnmt3b compensates Dnmt1-deficient intestinal epithelium.
Top Panel: Intestinal epithelium-specific ablation of the maintenance DNA methyltransferase Dnmt1 in adult mice leads to abnormalities in tissue morphology. Loss of Dnmt1 increases DNA double stranded breaks and cell apoptosis as determined by increases in gammaH2AX foci and TUNEL staining. Mutant crypt epithelial cells show decreased methylation at LINE1 retrotransposons and theH19 imprinting control.
Middle Panel: The intestinal epithelium can recover from Dnmt1 ablation. Within 8weeks Dnmt1mutant cells resemble healthy controls in morphology, levels of DNA double stranded breaks and apoptosis. Methylation at H19 but not LINE1 is also restored. Mutants have specifically increased expression of the DNAmethyltransferase Dnmt3b.
Bottom Panel: Tissue-specific deletion of both Dnmt1 and Dnmt3b is severely deleterious to the intestinal epithelium. Double mutants have grossly abnormal epithelium including increased crypt cell apoptosis, decreased DNA methylation at H19 and LINE1 loci and extensive DNA damage. The double mutant cells do not recover the control phenotype over time. (Illustration by Arwa Abbas)
Of course, Ellen’s work is not without unanswered questions. She acknowledges that her group does not understand the mechanism by which Dnmt3b can properly methylate the Dnmt1 targets – if Dnmt3b manages to remethylate hemimethylated DNA or if it waits until the DNA is entirely demethylated before restoring the wild-type methylation pattern through another mechanism. “It would be interesting to see where the methyltransferases are binding,” says Ellen. She hypothesizes that Dnmt3b might assume Dnmt1 binding sites and protein interactions as it compensates for Dnmt1 loss. Ellen also hopes that other specialists of intestinal cell biology will investigate the source and specific loci of the DNA damage occuring after Dnmt1 ablation. The answers to these questions might complement one another, as the damaged loci might be reconfigured targets for Dnmt3b.
This work by Ellen and colleagues demonstrates the fluid roles and relationships of epigenetic factors in developing and adult cells, clearing further ground for the research of intestinal cell biology and organogenesis on the whole. A better understanding of these processes and mechanisms has the potential to redefine current dogmas about DNA methylation and present novel strategies for the treatment of developmental and proliferative disorders in the intestine and other organs.
Ellen Elliott is currently studying the genetics of immune-cell lncRNAs as a post-doc at the Jackson Laboratory in Farmington, Connecticut.
A commentary on:
"The ‘de novo’ DNA methyltransferase Dnmt3b compensates the Dnmt1-deficient intestinal epithelium", Elliott et al., 2016.
Link to the PubMed page.