Today. Genomic Imprinting & X-Inactivation

Transcription

1 Today 1. Quiz (~12 min) 2. Genomic imprinting in mammals 3. X-chromosome inactivation in mammals Note that readings on Dosage Compensation and Genomic Imprinting in Mammals are on our web site. Genomic Imprinting & X-Inactivation IMPRINTING A. What is it? B. Evidence for it? 1) Non-random inactivation of X chromosomes 2) Disomic mice 3) Transplantation of nuclei 4) Aborted human triploids 5) Genetic Diseases 6) Imprinted Genes C. Why does it exist? D. How might it work? X-INACTIVATION 1

2 Genomic Imprinting -What is it? Definition: differential modification of maternal and paternal genetic contributions to zygote - resulting in differential expression of parental alleles during development and/or in adult.! Genomic Imprinting -Evidence for it? 1. Non-random inactivation of X chromosomes Preferential inactivation of paternal X chromosome in marsupials; Preferential inactivation of X P in extra-embryonic tissue in mouse 2. Maternal and paternal contributions to zygote not interchangeable a) Effects of disomy can depend on parental origin! e.g. Mat. disomy for chrom. 11 leads to small progeny; Paternal disomy for chrom. 11 leads to large progeny. b) Similarly, effects of deletions can depend on parental origin! e.g. Mat. Inheritance of Hairpin-tail chrom. 17Δ gives big embryos that die, while paternal transmission give normal mice. 2

3 Evidence for imprinting, continued: c) Nuclear transplantations: neither gynogenones nor androgenones develop past blastocyst stage gynogenones dev. to 25 somite stage but show meager extraembryonic tissue; tropoblast of androgenones develops OK but embryo is retarded. d) Aborted human triploids Phenotypes differ depending on source of extra genome: 2 male, 1 female ("diandry") results in large placenta 2 female, 1 male results in small underdeveloped placenta e) Genetic diseases E.g. Prader-Willi syndrome when Δ on chrom. 15 comes from father; Angelman syndrome results when Δ on chrom. 15 inherited from mother. Not uncommon for loss of chromosomes in cancer cells to be nonrandom: E.g. In Wilms tumors maternal copy of chrom. 11 usually lost; In many BWS patients, both copies of chrom. 11 come from father. Evidence for imprinting, continued: f) Many transgenes show evidence of imprinting Interesting example with complex transgene: RNAse protection analysis If inherited from the male, transgene was expressed in heart. If inherited from the female parent, it was not expressed at all. 3

4 Imprinting of transgene (Swain et al, 1987), continued: Pattern of expression correlated with DNA methylation status of transgene Methylation of the transgene acquired by passage through the female parent and eliminated during gametogenesis in the male. Authors suggested that their findings provide a plausible mechanism for understanding parental imprinting that may be relevant to the failure of parthenogenesis in mammals, the apparent non-mendelian behavior of some autosomal genes, and the role of methylation in gene regulation. Evidence for imprinting, continued: g) Native imprinted genes (first found in 1991). Key early examples: Igf2r (maternally expressed; accounts for Hairpin-tail mouse) Igf2 (paternally expressed growth hormone gene) H19 (maternally expressed large ncrna; near Igf2) paternal Igf2 H19 maternal DNA methylation Igf2 H19 Is DNA methylation cause or effect of imprinting? Biallelic expression in Dnmt1 mutant: paternal Igf2 H19 maternal Igf2 H19 4

5 Imprinted genes, continued: ~200 imprinted genes in mammals such as mouse and humans Many, but not all, are imprinted in both humans and rodents Most are in clusters (16 clusters on 8 mouse chromosomes; 3 solo) Why does imprinting exist? CLUES? Most imprinted genes affect growth of placenta, fetus or neonate In this top group: ~half paternally expressed growth promoters ~half maternally expressed growth repressors Consistent with two theories for purpose of imprinting: a. parental conflict b. trophoblast defense 5

6 How does imprinting work? Apparently a cis-acting silencing mechanism Presumably imprint occurs when two genomes are separate (during gamate formation) Imprint must be erased since new germ cells arise from embryonic diploid cells In principle, imprint could be any epigenetic mark How have critical marks been found? Gametic DMRs (Differentially Methylation Regions) identified; tested if ICE (Imprint Control Element) NG=non-imprinted gene; IG=imprinted gene; IG-NC=imprinted noncoding gene How does imprinting work? (continued) Clusters of imprinted genes typically show M or P DMR & lncrna 6

7 Since these DMR s are required for lncrnas (which repress other genes), it makes sense that deletion of unmethylated DMR matters: (unmethylated) ncrnas typically antisense and long (~ kb); mechanism? Mechanism unclear but truncation of ncrna prevents function Direct transcriptional interference at IG promoters? (NG=non-imprinted gene; IG=imprinted gene; IG-NC=imprinted long noncoding RNA gene) A different mechanism controls imprinting in Igf2r region Deletion of intergenic DMR abolished imprinting (thus ICE) The ICE binds methylation-sensitive insulator, CTCF CTCF-binding sites also near some other imprinted genes 7

8 Genomic Imprinting - Key Features cis-acting mechanism Affects both male & female (consequence of inheritance, not sex) Imprints are epigenetic modifications acquired by one gamate DNA methylation is the only imprint identified to date Imprinted genes are mostly clustered together with a noncoding RNA Imprints can act over multiple genes over great distance Imprinted genes play roles in mammalian development Imprinting may balance interest of sexes and protect mother X-inactivation/Dosage compensation Mammals: Y chromosome has <50 functional genes, X has ~1500 Without dosage compensation, XX individuals would have 2x more of gene products, relative to XY individuals. Mammals switch off one X in female, flies double expression of the X in males, C. elegans reduces expression of both Xs. Lyon suggested (1961) that Barr body is inactivated X Clones from females heterozygous for G6PD express only one allele XXX & XXXX females inactivate all but one X ( n-1 rule ) Both X s active early; X inactivation coincident with cellular differentiation Except in marsupials, random inactivation of X in embryo (but imprinted initially?) 8

9 Cycle of X-Inactivation and Reactivation Inactivation; Reactivation X-Inactivation, continued ~15% of genes on human X escape inactivation (PAR & elsewhere) Escapees lack the hallmarks of inactivation: Late replication Greater DNA methylation of CpG islands Hypoacetylation of histones (all of them) Heterochromatic histone methylation (high K27m; low K4m) Presence of H2A variant ( Macro H2A ) H2A K119 Ubiquitylation coating of inactive X with Xist (X inactive specific transcript) Ability to inactivate X maps to ~80 kb region, XIC (X-inactivation center) XIC encodes 17b kb Xist and its antisense Tsix Lyon noticed that X has lots of LINEs; serve as booster elements? Inactivation can spread from XIC into translocated autosomal sequences though not as well as on X sequences (and Xist absent) 9

10 How does X-inactivation occur? The process is considered in steps, which may be connected: 1. counting (inactivate all but one) 2. choice (random? If not, which?) 3. initiation 4. spreading 5. maintenance Blocking factor is attractive but not identified Imprinted inactivation may reflect a mark established in gamate that affects binding of blocking factor. In ES cells deletions 3 of Xist cause that allele to express Xist (BF &/or Tsix affected) Blocking Factor Model Kinetics of X-inactivation process provides clues Xist is uniquely expressed from Xi, its level dramatically increases in pre-implantation embryos at time of inactivation and it coats Xi. Over-expression of Xist on autosome induces heterochromatin in cis! Xist is transcribed from Xs in undiff. ES cells (XX & XY) but is unstable, probably because of fold excess of Tsix. 10

11 Particular steps may play critical roles in Xi but then may become redundant as more stable silencing is established Reactivation of Xi occurs in primordial germ cells at same time as genome-wide demethylation and erasure of parental imprints. Lots of unknowns remain such as: Xist necessary & sufficient for initiation but is not needed for maintenance Similarly, K27m only needed early; macroh2a & Ub-H2A depend on Xist DNA methylation is not required for establishment of Xi but serves as lock. Mechanisms of counting and choice (blocking factor? imprint?) Mechanistic interrelationship of various features of Xi Mechanism of spreading (and involvement of way stations?) 11