The majority of cytosine in the human genome is methylated on the 5-position (5mC) where it plays key roles in development and in disease, including cancer, through its potent gene repression activity. In vertebrates, 5mC is established by de novo enzymes (Dnmt3a/b-Dnmt3L) and maintained epigenetically in the context of symmetric sequence contexts (e.g. CpG) by the maintenance enzyme Dnmt1 and its cofactor Uhrf1. Because S. cerevisiae and S. pombe lack 5mC, facile yeast experimental systems have not been developed to investigate this important DNA mark. We have now developed such a system, namely the human fungal pathogen C. neoformans. This tractable, haploid budding yeast harbors symmetric CpG methylation at centromeres and subtelomeric regions. 5mC production is catalyzed by an enzyme called Dnmt5. In addition to a methyltransferase domain, Dnmt5 harbors a chromodomain that recognizes histone H3 modified on lysine 9 (H3K9me) and an ATPase domain. H3K9me promotes DNA methylation in vivo as does a homolog of the human hemimethyl-DNA-binding protein Uhrf1. Purified Dnmt5 is ATP-dependent and displays extraordinary specificity for hemimethylated DNA in vitro, with no detectable enzymatic activity on unmethyated substrates. In vivo, once 5mC is lost, it is not globally re-established, either in vegetatively-growing cells or during sexual reproduction. These data indicate that Dnmt5 is a maintenance-type enzyme. Surprisingly, Dnmt5 is the only cytosine DNA methyltransferase encoded by the C. neoformans genome. This finding raises the intriguing question of how 5mC can be established in an organism lacking a de novo enzyme. Through phylogenetic analysis of whole genome sequences, we determined that predicted Dnmt enzyme, DnmtX, was present in an ancestral species, but that its gene was lost at least 50 million years ago. To test whether DnmtX was a de novo enzyme, we introduced versions of DnmtX from existing species into a C. neoformans strain lacking 5mC but harboring Dnmt5. We find that 5mC is deposited by these enzymes de novo and is then maintained. Thus, DnmtX has de novo activity. Our results thus suggest a radical possibility, namely that 5mC has been maintained epigenetically by Dnmt5 since the DnmtX de novo methylase was lost. Such a model requires a combination of a high fidelity of epigenetic inheritance of 5mC together with some level of natural selection. We will test predictions of this model and use the power of a yeast system to investigate epigenetic inheritance fidelity and function. Specifically, we will measure and investigate 5mC inheritance fidelity, elucidate the evolution of 5mC patterns, investigate roles of 5mC in centromere function and genome defense, and identify determinants of 5mC maintenance and action. This work will probe the limits of epigenetic memory and illuminate how cytosine methylation of DNA is accurately maintained and how it functions.

Public Health Relevance

Cellular memory mediated by a DNA modification called cytosine methylation is critical for normal human health and goes awry in many diseases, including cancer. We have developed the yeast Cryptococcus neoformans as the first unicellular model organism to study this process. New knowledge gained from the application of the powerful tools available in this simple yeast system will illuminate how this important DNA modification is accurately maintained as cells divide and how it functions.

National Institute of Health (NIH)
National Institute of General Medical Sciences (NIGMS)
Research Project (R01)
Project #
Application #
Study Section
Molecular Genetics B Study Section (MGB)
Program Officer
Carter, Anthony D
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of California San Francisco
Schools of Medicine
San Francisco
United States
Zip Code
Harrigan, Patrick; Madhani, Hiten D; El-Samad, Hana (2018) Real-Time Genetic Compensation Defines the Dynamic Demands of Feedback Control. Cell 175:877-886.e10
Burke, Jordan E; Longhurst, Adam D; Merkurjev, Daria et al. (2018) Spliceosome Profiling Visualizes Operations of a Dynamic RNP at Nucleotide Resolution. Cell 173:1014-1030.e17
Mavor, David; Barlow, Kyle A; Asarnow, Daniel et al. (2018) Extending chemical perturbations of the ubiquitin fitness landscape in a classroom setting reveals new constraints on sequence tolerance. Biol Open 7:
Parsa, Jahan-Yar; Boudoukha, Selim; Burke, Jordan et al. (2018) Polymerase pausing induced by sequence-specific RNA-binding protein drives heterochromatin assembly. Genes Dev 32:953-964
Allshire, Robin C; Madhani, Hiten D (2018) Ten principles of heterochromatin formation and function. Nat Rev Mol Cell Biol 19:229-244
Roth, Robert; Madhani, Hiten D; Garcia, Jennifer F (2018) Total RNA Isolation and Quantification of Specific RNAs in Fission Yeast. Methods Mol Biol 1721:63-72
Al-Sady, Bassem; Greenstein, Rachel A; El-Samad, Hana J et al. (2016) Sensitive and Quantitative Three-Color Protein Imaging in Fission Yeast Using Spectrally Diverse, Recoded Fluorescent Proteins with Experimentally-Characterized In Vivo Maturation Kinetics. PLoS One 11:e0159292
Inada, Maki; Nichols, Robert J; Parsa, Jahan-Yar et al. (2016) Phospho-site mutants of the RNA Polymerase II C-terminal domain alter subtelomeric gene expression and chromatin modification state in fission yeast. Nucleic Acids Res 44:9180-9189
Garcia, Jennifer F; Al-Sady, Bassem; Madhani, Hiten D (2015) Intrinsic Toxicity of Unchecked Heterochromatin Spread Is Suppressed by Redundant Chromatin Boundary Functions in Schizosacchromyces pombe. G3 (Bethesda) 5:1453-61
Dumesic, Phillip A; Rosenblad, Magnus A; Samuelsson, Tore et al. (2015) Noncanoncial signal recognition particle RNAs in a major eukaryotic phylum revealed by purification of SRP from the human pathogen Cryptococcus neoformans. Nucleic Acids Res 43:9017-27

Showing the most recent 10 out of 32 publications