Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press

COMMENTARY Genomic imprinting Genomic imprinting is the differential modification of sine content in the DNA of the sperm and the egg, and the maternal and paternal genetic contributions to the this has been proposed as a possible molecular basis for zygote, resulting in the differential expression of pa­ imprinting (Monk 1987; Monk et al. 1987). Normally, it rental alleles during development and in the adult. It has is not possible to distinguish the levels of methylation of been known for some time that imprinting plays a role different parental , or specific alleles, in the in­ in a diversity of biological phenomena, including de­ dividual. Therefore, transgenic marker sequences, inte­ termination and germ cell differentiation in certain in­ grated at various sites in the and inherited from sects (Brown and Nur 1964; Scarbrough et al. 1984), pref­ one or the other parent, have been used to show differ­ erential inactivation of the paternal in ential methylation. The frequency of differential meth­ (for review, see VandeBurgh et al. 1987) and ylation of different transgenes has been reported in 4 out in the extraembryonic membranes of rodents (Takagi of 5 transgene loci studied by Sapienza et al. (1987), 1 out and Sasaki 1975), uniparental inheritance of chloroplast of 7 studied by Reik et al. (1987), and 1 out of 10 studied DNA (Sager and Kitchin 1975), and mating-type inter- by Swain et al. (1987). In most cases, the transgene in­ conversion in fission yeast (Klar 1987). Lately, the phe­ herited from the father is less methylated than if it were nomenon of imprinting is exciting considerable interest. inherited from the . So far, for any of these trans­ This is due to the recent discoveries in the mouse that loci showing differential imprinting in the fetus or the differential modification of the maternal and pa­ adult, it is not known whether the difference also ex­ ternal genomes is essential for successful development isted earlier between the sperm and the ^gg (because of of the conceptus (McGrath and Solter 1984; Surani et al. the difficulty in obtaining enough oocyte DNA for anal­ 1984) and survival and normal of the adult ysis); but in three cases examined by the workers men­ (Cattanach and Kirk 1985). In addition, recent reports tioned above, undermethylation of the paternally inher­ that transgenic DNA marker sequences are methylated ited loci in the offspring is reflected by undermethyla­ differently, depending on the parent of origin, suggest tion of these loci in the testes. This would seem to be at that methylation may be involved in the mechanism of variance with the fact that sperm DNA is more methyl­ imprinting (Reik et al. 1987; Sapienza et al. 1987; Swain ated than oocyte DNA overall. However, irrespective of et al. 1987). The regions of chromatin involved are now overall genomic methylation differences, imprinting accessible to further analysis. may be due to differential modulation of methylation The investigations that have shown a requirement for along particular domains of sperm and oocyte DNA. A the differentially modified maternal and paternal genetic precedent for differential modulation of methylation is contributions for successful development and normal seen, e.g., between the active and inactive X chromo­ phenotype in the mouse have utilized duplications of somes in somatic tissues; clustered CpG sequences 5' to maternal or paternal genomes or chromosome regions. certain are more methylated the compared to in­ Mouse containing either two female pronuclei active X chromosome, whereas other CpG sequences in or two male pronuclei were created by exchange of pron­ the body of the gene are less methylated on the inactive uclei between eggs (Barton et al. 1984; Mann and Loveil- X chromosome. The inactive X chromosome also re­ Badge 1984; McGrath and Solter 1984; Surani et al. minds us that there are more ways to inactivate a chro­ 1984). In embryos with two female genetic comple­ mosome or chromosomal domain than methylation ments, fetal development was good but development of alone; chromosome configuration must also play a part the extraembryonic membranes and was poor; in maintaining inactivation in segmental fashion along in embryos with two male genetic complements, the re­ the inactive X chromosome (for review, see Monk 1986). verse was true—the extraembryonic tissues developed Imprinting must be established during (or before) ga- well, and the fetus developed poorly. The reason for the metogenesis, persist stably throughout DNA replication uneven contribution to the different parts of the con­ and cell division in the soma, and be erased in the germ ceptus is not yet understood. Duplications of maternal line to be differentially established once more in the or paternal specific chromosome regions may be sperm and tgg genomes. Stable, heritable, differential achieved by matings between mice with different chro­ modification of chromatin is required. Differential pat­ mosome translocations (Searle and Beechey 1978) or car­ terns of methylation are stable, heritable (HoUiday and rying different metacentric chromosomes (Cattanach Pugh 1975), and, moreover, have been implicated in the and Kirk 1985). The combinations may be lethal or re­ regulation of , chromatin configuration, sult in different , suggesting differential and X-chromosome expression. The configurational dif­ functioning of maternal and paternal gene loci within ferences between repressed and derepressed chromo­ the regions concerned. Again, the loci involved and the some domains are also heritable (Weintraub 1985) and causes of the anomalies are unknown. correlated with methylation (Keshet et al. 1986). Sperm In the mouse, there is a difference in the methyl cyto- and oocyte are methylated differently; they are also sub-

GENES & DEVELOPMENT 2:921-925 © 1988 by Cold Spring Harbor Laboratory ISSN 0890-9369/88 $1.00 921 Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press

Monk

ject to different configurational constraints in the pack­ }— ^^ sperm aging of their genomes. Therefore, I would like to pro­ pose a model of establishment, propagation, and erasure of imprinting that works on at least two different levels ijlTr —differential methylation and differential chromatin 1 1 8 cell configuration. Three important aspects of this model CLEAVAGE follow: (1) Initial imprinting differences between the ga­ metes influence the timing of onset of parental allele ex­ pression. Thereafter, 'function determines form' in the 1 1 ICM sense that recruitment of specific maternal or paternal 1 bla 1 domains of chromatin into active gene expression during BLASTOCYST development leads to 'open' chromatin configuration in these regions and undermethylation. Conversely, the failure of a specific maternal or paternal chromosome 1 1 ICM 1 bla 1 domain to enter active expression during certain in­ IMPLANTATION terval of time in development may lead to condensation and subsequent methylation of that domain. (2) The Lyt germ line escapes extensive de novo DNA methylation, 1 6''i epj 1 6^a tr\i PRE-GA5TRULATI0N and this hypomethylation is a prerequisite or a precon­ 1 dition for erasure of the configurational constraints of imprinting and for of the germ line to developmental totipotency; this process may be con­ 1 /~V^ 7^2 emb 1 7^ cho nected to meiosis. (3) Some apparent differential im­ GASTRULATION printing phenomena (preferential allele expression, ab­ normal phenotype, or nonreciprocal lethality) may be a consequence of DNA sequence differences between the 1 GC^ \ 1 GCc/ maternal and paternal chromosome regions concerned. GERM CELLS Further considerations related to this model are out­ lined below. We have studied the temporal and regional changes in Figure 1. Densitometer tracings of DNA isolated from dif­ overall DNA methylation in the embryonic, extraem­ ferent tissues of the mouse conceptus and digested with Hpflll. bryonic, and germ cell lineages during development of Fragments resulting from Hpflll digestion were end-labeled, the mouse (Monk et al. 1987; Sanford et al, 1987). A electrophoresed, blotted to a nitrocellulose filter, and autora- summary of the results is given in Figure 1 in the form of diographed. The discrete band in the oocyte track is one of the densitometer tracings of the distributions of the larger mitochondrial Hpflll fragments. (ICM) Irmer cell mass; (bla) fragment sizes resulting from Hp^II digestion of DNA blastocyst; (epi) epiblast; (end) primary endoderm; (emb) em­ bryo; (cho) chorion; (GC) germ cell; 6.5 and 7.5, days of gesta­ from eggs, sperm, eight-cell embryos, blastocysts and tion. (For details, see Monk et al. 1987.) implanting blastocysts (and their isolated inner cell masses), embryonic and extraembryonic regions of pos- timplantation embryos (pre- and postgastrulation), and premeiotic germ cells of male and female embryos. In­ ating or confirming differential programming of the de­ creased genomic methylation (e.g., sperm and gastru- finitive germ layers. We proposed that much of the de lating ; Fig. 1) is correlated with higher- novo methylation observed in somatic tissues acts to molecular-weight Hpall fragments to the left of the dis­ stabilize and reinforce prior events regulating the ac­ tributions (top of the lanes on the gel) and decreased tivity of specific genes, chromosome domains, or the X genomic methylation, with a skewing of the distribution chromosome (in females). A similar conclusion—that away from the top of the gel (e.g., oocyte, blastocysts, methylation may be a secondary event involved in fetal germ cell DNAs; Fig. 1). It is clear that the Qgg maintenance of the inactive state—was reached by Lock genome is strikingly undermethylated (the peak fraction et al. (1987), who showed that methylation of gene se­ is a Hpall fragment of mitochrondrial DNA) and the quences on the inactive X chromosome occurred after sperm genome is relatively methylated. There is a loss of X-chromosome inactivation in development. methylation during preimplantation development These observations must be taken into account when (perhaps due to a delay in synthesis of the embryo-coded considering the propagation of differential methylation methylase) and the blastocyst is markedly undermethy­ as a mechanism for the propagation of imprinted infor­ lated. Thereafter, embryonic and extraembryonic mation in development (see Fig. 2). If there is a delay in genomes are methylated progressively de novo, indepen­ the onset of expression of the gene for maintenance dently (and therefore potentially differently), and to dif­ methylase, we expect that each cell doubling will expo­ ferent final extents. De novo methylation also continues nentially decrease the methylated DNA strands re­ postgastrulation and, hence, could be a mechanism initi- maining. If the methylation difference between the ga-

922 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press

Genomic imprinting metes is the sole source of imprinting information, how not in the germ line). Alternatively, or in addition, a is this information propagated if methylation is being second mechanism of propagating imprinted informa­ lost? It could be argued that certain 'key sites' remain tion may rely on heritable differential chromatin struc­ differentially methylated in the maternal and paternal ture. The process could be d3niamic (albeit circular) in genetic complements and that these key sites direct the that these initial imprinting mechanisms, both methyl­ subsequent patterns of de novo methylation (although ation and structural modification, may distinguish the

TWO LEVELS OF DIFFER­ ENTIAL IMPRINTING?

Fertilization PATERNAL IMPRINT MATERNAL IMPRINT / o o^ Male Female X"^X Pre - implantation \ i f LOSS OF METHYLATION Gametogenesis meiosis in female \ I I ERASURE OF IMPRINTING X'^'X"

Blastocyst

/ Germ line Implantation ESCAPES DE NOVO METHYLATION

DE NOVO METHYLATION. V\ /// \S< Pre-gastrulation ^^( ^conceptus'

Extra-embryonic tissues LOW METHYLATION Xp or XM Xp in female Figure 2. A diagrammatic model showing two levels of imprinting differences in the gametes, configurational (shown as different outlines of nuclei) and methylation (shown as dots within nuclei), the propagation of imprinting in the embryonic and extraembryonic regions of the conceptus, and erasure of imprinting in the germ line. The changes in overall genomic methylation during development are indicated. In the female conceptus, preferential inactivation of the paternal X chromosome (Xp) occurs in extraembryonic tissues, random X-chromosome inactivation (maternal, X^^ or Xp) occurs in embryonic tissue, and X-chromosome reactivation (X'^X" to X+X+) occurs at meiosis in the germ line.

GENES & DEVELOPMENT 923 Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press

Monk timing of onset of maternal or paternal gene activity, final aspect of the model to be considered. and differential gene activity itself may lead to further Some apparent imprinting phenomena observed will differences in methylation and structural modification be a consequence of DNA sequence divergence between of the maternal and paternal loci as development pro­ the maternal and patemal chromosome regions con- ceeds. cemed. This has been well established in interspecific We come now to a consideration of the erasure of the hybrids of fish, where differential timing of expression of biparental imprinting in the germ line and reestablish- the two parental loci, preferential allele expression, de­ ment of either the male or female uniparental imprint in velopmental abnormalities and lethality are correlated or , respectively. Here, a con­ with the degree of genetic divergence between the sideration of cycles of X-chromosome activation, inacti- parents (Whitt et al. 1973, 1977; Whitt 1981). The nonre- vation, and reactivation in different lineages of the fe­ ciprocal lethality of some crosses in fish is reminiscent male mouse embryo may be relevant (for review, see of the DDK strain of mice, where matings of DDK fe­ Monk 1981, 1986). The paternally inherited X chromo­ males with males of other strains of mice are sterile, some, previously inactive during spermatogenesis, be­ whereas reciprocal crosses are fertile. The DDK egg cy­ comes active during cleavage stages of development toplasm is rendered incapable of supporting develop­ (perhaps as a consequence of loss of methylation and/or ment by an interaction with 'foreign' sperm (Renard and of decondensation of paternal chromosomes) and is then Babinet 1986). A similar difference, perhaps in timing of preferentially inactivated in the extraembryonic lin­ sex-determining functions, may underlie the aberrant eages and randomly inactivated in the fetal and germ sexual differentiation involving a foreign Y chromosome cell lineages (see Fig. 2). Whatever the imprint is that on a C57B6 genetic background (Eicher and Washburn distinguishes the paternal X chromosome for inactiva- 1986). Restriction maps for a number of loci for C57B6 tion in the extraembryonic lineages, this imprint is ei­ DNA differ from that of other inbred strains of labora­ ther preferentially inherited by the extraembryonic lin­ tory mice (P. Martin and M. Monk, unpubl.). Hence, the eages (perhaps those cells retaining methylated DNA use of the C57B6 strain as one parent in investigations of 'sort out' into extraembryonic lineages) or it is lost or imprinting may show more modifications than would be 'not seen' when the time comes to inactivate an X-chro­ observed within an inbred strain. mosome in the fetal and germ line precursor cells. The In summary, the role of imprinting in a number of es­ reasons for preferential patemal X chromosome inacti- sential biological processes determining sex, germ cell vation in extraembryonic tissues of rodents and somatic differentiation, development, and normal adult pheno- tissues of marsupials is not known. Although the inac­ type is now well established. It is suggested that tissue- tive X chromosome is inactivated irreversibly in somatic and timing-specific functional haploidy in early develop­ cells, it is reactivated just prior to meiosis in the female ment occurs as a consequence of differential heritable germ cells (see Fig. 2). It has been suggested previously chromatin configuration imposed by packaging con­ (Monk 1981) that X-chromosome inactivation is linked straints in sperm, as well as differential DNA methyl­ to cell differentiation and, conversely, that X-chromo­ ation of sperm and &gg genomes. Later in development, some reactivation in the female germ line is linked to the configurational and methylation differences ob­ cell dedifferentiation or a reprogramming of the germ served in the fetus or in the adult may be correlated with line to developmental totipotency. It is tempting to the different patterns of gene expression occurring ear­ equate erasure of imprinting with this latter process (see lier. Some phenomena reminiscent of genomic im­ Fig. 2). If, as seems likely, the germ line of both female printing may occur as a consequence of genetic distance and male embryos escapes de novo methylation, erasure (DNA sequence differences) between the gametic of imprinting would require erasure of chromatin struc­ genomes, and the greater the distance is, the more pro­ ture constraints—a process that could conceivably be a nounced the effects. In this respect, differential expres­ part of the unraveling of chromosomes in preparation for sion of parental alleles may result in lethality and hence synapsis at meiosis. If erasure were not complete, im­ play a role in speciation, or 'protection against foreign printing could pass through more than one generation as DNA.' Finally, erasure of imprinting (and reactivation of grandparental memory or even a permanent imprint. In­ the inactive X chromosome) may be a function of deed, a permanent imprint has been observed (as a meth­ meiosis. ylation change) for one particular transgene marker fol­ lowing transmission through the female (Hadchouel et al. 1987) and proposed as a mechanism to explain the pattern of inheritance of the in References humans (Laird 1987). In other cases, what appears as a Barton, S.C., M.A.H. Surani, and M.L. Norris. 1984. Role of pa­ permanent imprint, or indeed as a , may arise temal and maternal genomes in mouse development. Na­ as a consequence of DNA sequence differences between ture 311: 374-376. the parental genomes in the chromatin regions con­ Brown, S.W. and U. Nur. 1964. Heterochromatic chromosomes cerned. This DNA sequence difference could lead to a in the coccids. Science 145: 130-136. difference in timing of activation of these regions during Cattanach, B.M. and M. Kirk. 1985. Differential activity of ma- temally and patemally derived chromosome regions in development. Inactivation would then be due to the fact mice. Nature 315: 496-498. that the region failed to be expressed. This leads to the Eicher, E.M. and L.L. Washbum. 1986. Genetic control of pri-

924 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press

Genomic imprinting

mary sex determination in mice. Annu. Rev. Genet. Takagi, N. and M. Sasaki. 1975. Preferential expression of the 20: 327-360. paternally derived X chromosome in the extraembryonic Hadchouel, M., H. Farza, D. Simon, P. Tiollais, and C. Pourcel. membranes in the mouse. Nature 256: 640-642. 1987. Maternal inhibition of hepatitis B surface antigen gene VandeBerg, T.L., E.S. Robinson, P.B. Samollow, and P.G. expression in transgenic mice correlates with de novo meth- Johnson. 1987. X-linked gene expression and X-chromosome ylation. Nature 329: 454-456. inactivation: Marsupials, mouse, and man compared. Iso­ Holliday, R. and J.E. Pugh. 1975. DNA modification mecha­ zymes: Current Topics in Biological and Medical Research nisms and gene activity during development. Science (ed. C.L. Markert) vol. 15, pp. 225-253. Alan R. Liss, New 187: 226-232. York. Keshet, I., J. Lieman-Hurwitz, and H. Cedar. 1986. DNA meth- Weintraub, H. 1985. Assembly and propagation of repressed and ylation affects the formation of active chromatin. Cell derepressed chromosome states. Cell 42: 705-711. 44: 535-543. Whitt, G.S. 1981. Developmental genetics of fishes: isozymic Klar, A.J.S. 1987. Differentiated parental DNA strands confer analysis of differential gene expression. Am. Zool. 21: 549- developmental asymmetry on daughter cells in fission yeast. 572. Nature 326: 466-470. Whitt, G.S., W.F. Childers, and P.L. Cho. 1973. Allelic expres­ Laird, CD. 1987. Proposed mechanism of inheritance and ex­ sion at enzyme loci in an intertribal sunfish. /. Hered. pression of the human fragile X syndrome of mental retarda­ 64: 55-61. tion. Genetics 117: 587-599. Whitt, G.S., D.P. Phillip, and W.F. Childers. 1977. Aberrant Lock, L.F., N. Takagi, and G.R. Martin. 1987. Methylation of gene expression during the development of hybrid sunfishes the Hprt gene on the inactive X occurs after chromosome (Perciformes, Teleostei). Differentiation 9: 97-109. inactivation. Cell 48: 39-46. Mann, J.R. and R.H. Lovell-Badge. 1984. Inviability of parthen- Marilyn Monk ogenones is determined by pronuclei, not egg cytoplasm. Nature 310: 66-67. MRC Mammalian Development Unit McGrath, J. and D. Solter. 1984. Completion of mouse embryo- 4 Stephenson Way genesis requires both the maternal and patemal genomes. London NWl 2HE, UK Cell 37: 179-183. Monk, M. 1981. A stem line model for cellular and chromo­ somal differentiation in early mouse development. Differen­ tiation 19: 71-76. . 1986. Methylation and the X chromosome. BioEssays 4: 204-208. . 1987. Memories of mother and father. Nature 328: 203-204. Monk, M., M. Boubelik, and S. Lehnert. 1987. Temporal and regional changes in DNA methylation in the embryonic, ex­ traembryonic and germ cell lineages during mouse embryo development. Development 99: 371-382. Reik, W., A. ColUck, M.L. Norris, S.C. Barton, and M.A. Surani. 1987. Genomic imprinting determines methylation of pa­ rental alleles in transgenic mice. Nature 328: 248-251. Renard, J.P. and C. Babinet. 1986. Identification of a patemal development effect on the cytoplasm of one-cell-stage mouse embryos. Proc. Natl. Acad. Sci. 83: 6883-6886. Sager, R. and R. Kitchin. 1975. Selective silencing of eukaryotic DNA. Science 189: 426-433. Sanford, J.P., H.J. Clark, V.M. Chapman, and J. Rossant. 1987. Differences in DNA methylation during oogenesis and sper­ matogenesis and their persistance during early embryo- genesis in the mouse. Genes Dev. 1: 1039-1046. Sapienza, C, A.C. Peterson, J. Rossant, and R. Balling. 1987. Degree of methylation of transgenes is dependent on of origin. Nature 328: 251-254. Scarbrough, K., S. Hattman, and U. Nur. 1984. Relationship of DNA methylation level to the presence of in . Mol. Cell. Biol. 4: 599-603. Searle, A.G. and C.V. Beechey. 1978. Complementation studies with mouse translocations. Cytogenet. Cell Genet. 20: 282-303. Surani, M.A.H., S.C. Banton, and M.L. Norris. 1984. Develop­ ment of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature 308: 548-550. Swain, J.L., T.A. Stewart, and P. Leder. 1987. Parental legacy determines methylation and expression of an autosomal transgene: a molecular mechanisms for parental imprinting. Ceii 50: 718-727.

GENES & DEVELOPMENT 925 Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press

Genomic imprinting.

M Monk

Genes Dev. 1988, 2: Access the most recent version at doi:10.1101/gad.2.8.921

References This article cites 30 articles, 8 of which can be accessed free at: http://genesdev.cshlp.org/content/2/8/921.full.html#ref-list-1

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here.

Copyright © Cold Spring Harbor Laboratory Press