KDM1B as a Link Between Histone Modification and DNA Methylation of Gene Imprinting During Gametogenesis

Transcription

1 KDM1B as a Link Between Histone Modification and DNA Methylation of Gene Imprinting During Gametogenesis BI 424 Advanced Molecular Genetics, Winter /14/11 Abstract Genomic imprinting is a highly complex, dynamic process that has yet to be fully characterized but is known to occur during gametogenesis. 1 A key component of this process, the germ cell regulatory factor DNMT3L, has been shown to act as a bridge between histone modification and DNA methylation by acting as an activator for the related de novo DNA methyltransferase DNMT3A, as well as, by binding preferentially to unmethylated H3K4. 2,3 Recently a new histone demethylase known as KDM1B (also LSD2 or Aof1) has been found to specifically demethylate me2h3k4 and has been found to be present during gametogenesis. 4,5 Therefore the purpose of this proposal is to examine the hypothesis that KDM1B may act as a recruiter of DNMT3L to H3K4 that has been unmethylated. This will be done by experimentally determining whether these proteins co-localize, interact and how, and whether KDM1B interacts with other DNMT s if there is no interaction with DNMT3L. Background Gene imprinting is an epigenetic process that occurs during gametogenesis, in which parental-specific genes are targeted for expression. 1 This leads to parental-specific expression of these genes during embryogenesis and is a key

2 genetic regulatory process, since loss of imprinting can lead to lethality. 6 This process is modulated by the DNA methyltransferase family (DNMT) of proteins, particularly the de novo DNA methyltransferases DNMT3A and DNMT3B. 6 One member of the DNMT3 family, DNMT3-like (DNMT3L) has been implicated as a crucial member of this pathway. Although it shares sequence similarity to the de novo DNMT s, it lacks the catalytic domain necessary for DNA methylation. 7 However it has been shown to be crucial in maternal genetic imprinting, since a knock-out of this gene shows a maternal-effect lethal mutation in the next generation. 7,8 An attempt to explain this phenotype was made by Ooi et. al. in their research on DNMT3L in which they discovered that this protein bound preferentially to unmethylated H3K4 through its amino-terminus PHD (plant homeo domain). 3 Since H3K4 methylation is often the mark of active chromatin, then removal of this mark is necessary to lead to expression changes of these genes. This protein has also been shown to bind to DNMT3A, which gives DNMT3L an interesting role as a link between histone modification and DNA methylation. This role was characterized further by Jia et. al. in their crystallographic structure of DNMT3L-DNMT3A complex. This showed that DNMT3L binds with its carboxy-terminus in the catalytic domain of DNMT3A, thereby acting as an activator of the DNA methyltransferase activity. 2 A newer area of research in this field has been the discovery of histone

3 demethylases. Before their discovery, it was often believed that histone methylation was permanent, but it has since been shown to be highly dynamic. One such histone demethylase known as KDM1B (also LSD2 or Aof1) was recently discovered. This histone demethylase belongs to a family known as flavin-dependent amine oxidases. These proteins are dependent on a flavin cofactor, in this case FAD. This histone demethylase is one of only two in mammals, the other being the more characterized LSD1 (also KDM1 or Aof2). 4 LSD1 has been shown to demethylate H3K4 and is also known to interact with chromatin remodeling complexes through its tower domain. 4,9 KDM1B on the other hand does not share this tower domain but instead has a CW-zinc finger type domain, in which it may interact with other proteins or complexes; however that is yet to be determined. Besides that it shares the SWIRM domain, which binds to histones, and the amino-oxidase domain, which carries out the demethylation. 4 It has also been shown that KDM1B is expressed during gametogenesis and that a knock-out of this gene leads to loss of imprinting on a subset of maternally imprinted genes. 5 Given this information, this paper serves to propose that KDM1B may serve to recruit the cofactor DNMT3L to recently demethylated H3K4, which would then lead to DNA methylation of the gene via DNMT3A. This will be done through experimentation testing the following specific aims:

4 1. Determine whether KDM1B and DNMT3L co-localize in nuclei of mouse embryonic stem cells. 2. Determine whether there is a direct interaction between KDM1B and DNMT3L. If there is a real interaction between KDM1B and DNMT3L: 3. Determine the regions of each protein that interact with each other. And if there is no real interaction between KDM1B and DNMT3L: 4. Determine whether there is a direct interaction between KDM1B and DNMT3A. Specific Aims and Experiments Specific Aim 1 Proposal: KDM1B and DNMT3L co-localize to the same regions of the nucleus/chromosomes in mouse embryonic stem cells. Experiments: To determine whether KDM1B and DNMT3L co-localize within the nucleus, I will follow the procedure laid out in Ciccone et. al. using the same antibodies for DNMT3L but I will attempt to have antibodies made specifically for KDM1B, since they used an anti-flag antibody for Flag-tagged KDM1B. 5 I will do this in either mouse embryonic stem cells that naturally contain higher levels of DNMT3L and are transfected with KDM1B or harvested mouse oocytes, which naturally contain higher levels of KDM1B and DNMT3L. 5,7

5 This is in contrast to the NIH 3T3 cells used in the paper. As a control, I will also confirm whether DNMT3L and DNMT3A co-localize again using the antibodies from the Ciccone et. al. paper. 5 Specific Aim 2 Proposal: KDM1B serves to recruit DNMT3L to the newly unmethylated H3K4 via direct protein:protein interactions. Experiments: To determine whether there is a real interaction between KDM1B and DNMT3L, I will perform a co-immunoprecipitation (CoIP) as described in Ooi et. al. using Flag-tagged KDM1B transfected into mouse embryonic stem cells and Flag-tagged and KDM1B (from specific aim 1) transfected into mouse embryonic stem cells. 3 I will then submit the results from both pulldowns to mass spectrometric analysis to determine the identities of all associated proteins. Since this assay tends to identify strong interactions, I will also do a yeast two-hybrid assay as described in Fields and Song in Saccharomyces cerevisiae. 10 I will do this assay in duplicate, one with KDM1B as the bait bound to the Gal4 binding domain and DNMT3L as the prey bound to the Gal4 activation domain and the one as vice versa. These will function at the upstream activating sequence (UAS) for the lacz reporter gene. I will determine the strength of the interaction by monitoring the amount of β-galactosidase produced (blue color) from the cells.

6 Specific Aim 3 Proposal: If KDM1B and DNMT3L have a direct interaction, the particular regions that are required for binding will be elucidated by deletion and mutation studies. Experiments: To determine the exact regions of each protein that are necessary to accommodate an interaction, I will a series of deletion analyses with each protein. This will be done by gradual truncations of each protein. For example, I will truncate a portion of KDM1B while leaving DNMT3L complete. I will test the ability to bind by using both proteins/protein fragments in a yeast twohybrid assay (as in specific aim 2). Upon narrowing down a region, I will further determine the specific residues necessary for an interaction via targeted deletions within the identified regions. I will then repeat the yeast two-hybrid assay with only one or the other of the complete proteins containing a targeted deletion until the specific residues are identified. Specific Aim 4 Proposal: If KDM1B and DNMT3L do not have a direct interaction, any other DNMT s that may associate with KDM1B with be determined. Experiments: To determine which DNMT s may interact with KDM1B, I will review the results from specific aim 2 to determine whether an other DNMT s were recovered. If none are found or to confirm any interaction, I will perform a

7 yeast two-hybrid assay with KDM1B as the bait and other DNMT s as the prey. Discussion For my first specific aim, I would expect to find that KDM1B co-localizes with DNMT3L. I would also expect that my control would yield DNMT3L co-localizing with DNMT3A, since they are known to form a complex. 7 This would give positive feedback to continue on with the experiments. However there is the possibility that either I would not find co-localization or else I would not see enough of a signal due to low endogenous levels of protein. If the first case, then I would still continue on to my next aim since even a slight co-localization may be occurring that can not be spotted in the assay. Also if the second, then I would modifiy the experiment by ectopically expressing one or both proteins in the mouse embryonic stem cells and then continuing on with the experiment. For my second specific aim, I would hope to see a real, direct interaction in the co-immunoprecipitation assay and that this would be confirmed by the yeast two-hybrid assay. However as stated before no result in the CoIP assay would not discount a possible direct interaction between KDM1B and DNMT3L. There is still the possibility that this interaction could be weak and perhaps even transient. Therefore the yeast two-hybrid assay would be expected to give some type of result via blue color made by the expression of β-galactosidase. If any indication of an interaction were made, I would move on to my third

8 specific aim. However if no interaction was seen at all, then I would move on the my fourth specific aim. Assuming that some type of interaction was shown in the previous aim, I would expect to be able to narrow down the domains necessary for this interaction within both of the proteins. The truncation and mutational analysis should be able to narrow down to the exact residues that are important. This may serve to be more challenging if a whole region is required, as I would expect to be the case in KDM1B. This is because it contains a CW-zinc finger type motif that is likely to be involved in protein:protein interactions. 4 If no direct interaction is shown in specific aim two, I would expect that some other proteins had shown up in specific aim one which I would hope to include other DNMT s, particulary DNMT3A since it forms a complex with DNMT3L. 7 If these proteins were not found then any other proteins found would form a good future research proposal to study there interaction and affects with KDM1B.

9 References 1. Reik, W. & Walter, J. Genomic imprinting: parental influence on the genome. Nature Reviews Genetics 2, (2001). 2. Jai, D., Jurkowska, R.Z., Zhang, X., Jeltsch, A., & Cheng, X. Structure of DNMT3A bound to DNMT3L suggests a model for de novo DNA methylation. Nature 449, (2007). 3. Ooi, S.K.T., Qiu, C., Bernstein, E., Li, K., Jia, D., Yang, Z., Erdjument- Bromage, H., Tempst, P., Lin, S., Allis, C.D., Cheng, X., & Bestor, T.H. DNMT3L connects unmethylated lysine 4 of histone H3 to de novo methylation of DNA. Nature 448, (2007). 4. Karytinos, A., Forneris, F., Profumo, A., Ciossani, G., Battaglioli, E., Binda, C., & Mattevi, A. A novel mammalian flavin-dependent histone demethylase. Journal of Biological Chemistry 284, (2009). 5. Ciccone, D.N., Su, H., Hevi, S., Gay, F., Lei, H., Bajko, J., Xu, G., Li, E., & Chen, T. KDM1B is a histone H3K4 demethylase required to establish maternal genomic imprints. Nature 461, (2009). 6. Okana, M., Bell, D.W., Haber, D.A., & Li, E. DNA methyltransferases DNMT3A and DNMT3B are essential for de novo methylation and mammalian development. Cell 99, (1999). 7. Hata, K., Okano, M., Lei, H., & Li, E. DNMT3L cooperates with the DNMT3 family of de novo DNA methyltransferases to establish maternal imprints in mice. Development 129, (2002). 8. Bourc his, D., Xu, G., Lin, C., Bollman, B., & Bestor, T.H. DNMT3L and the establishment of maternal genomic imprints. Science 294, (2001). 9. Shi, Y., Lan, F., Matson, C., Mulligan, P., Whetstine, J.R., Cole, P.A., Casero, R.A., & Shi, Y. Histone demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell 119, (2004). 10. Fields, S. & Song, O. A novel genetic system to detect protein-protein interactions. Nature 340, (1989).