Genomic imprinting and Igf2 influence liver tumorigenesis and loss of heterozygosity in SV40 T antigen transgenic mice

Maternal-specific loss of heterozygosity (LOH) and allelic imbalances [i.e., partial LOH (pLOH)] observed in SV40 T/t antigen-induced liver tumors suggests that an imprinted gene on chromosome 7 is involved in liver tumorigenesis. Maternal-specific LOH/pLOH may reflect the loss of a maternally expressed tumor suppressor gene or the acquisition of paternally active alleles of a growth promoter. In addition, two oppositely imprinted genes on distal chromosome 7, Igf2 and H19, are re-expressed in most liver tumors from an SV40 T/t antigen transgenic line (M11T-G). Igf2 is a paternally expressed growth promoter, and H19 is a maternally expressed gene that can suppress growth in some tumor cell lines. We studied the role of Igf2 during liver tumorigenesis by creating Igf2 (+/-) M11T-G mice. These mice are essentially null for Igf2 expression because imprinting normally precludes maternal Igf2 expression. M11T-G, Igf2 (+/-) males exhibit a 15-fold reduction in the frequency of large tumors. Igf2 (+/-) tumors do not express maternal Igf2, indicating rigid imprinting control in the liver. LOH/pLOH analysis was performed on the tumors and indicates that acquisition of paternally active Igf2 alleles is a major selective event for M11T-G liver tumorigenesis. This also implies the existence of an imprinted, maternally expressed tumor suppressor gene on chromosome 7 that is unlikely to be H19.     

Chromatin conformation of the H19 epigenetic mark

Genomic imprinting in mammals is an epigenetic process that results in differential expression of the two parental alleles. The tightly linked murine H19 and Igf2 genes are reciprocally imprinted: H19 is expressed from the maternal chromosome while Igf2 is expressed from the paternal chromosome. A single regulatory region in the 5' flank of the H19 gene has been implicated in silencing both genes. On the paternal chromosome, this region is heavily methylated at CpG residues, leading to repression of the H19 gene. The mechanism by which the same region in an unmethylated state on the maternal chromosome silences Igf2 is less well understood. We have probed the chromatin structure of the region by assessing its sensitivity to nuclease digestion. Two regions of nuclease hypersensitivity that are specific to the maternal chromosome were identified. These coincide with the region that is most heavily methylated on the paternal chromosome. As is the case with paternal methylation, hypersensitivity is present in all tissues surveyed, irrespective of H19 expression. We suggest that the chromatin structure of the maternal 5' flank of the H19 gene may represent an epigenetic mark involved in the silencing of Igf2.     

Location of enhancers is essential for the imprinting of H19 and Igf2 genes

Genomic imprinting is the process in mammals by which gamete-specific epigenetic modifications establish the differential expression of the two alleles of a gene. The tightly linked H19 and Igf2 genes are expressed in tissues of endodermal and mesodermal origin, with H19 expressed from the maternal chromosome and Igf2 expressed from the paternal chromosome. A model has been proposed to explain the reciprocal imprinting of these genes; in this model, expression of the genes is governed by competition between their promoters for a common set of enhancers. An extra set of enhancers might be predicted to relieve the competition, thereby eliminating imprinting. Here we tested this prediction by generating mice with a duplication of the endoderm-specific enhancers. The normally silent Igf2 gene on the maternal chromosome was expressed in liver, consistent with relief from competition. We then generated a maternal chromosome containing a single set of enhancers located equidistant from 1gf2 and H19; the direction of the imprint was reversed. Thus, the location of the enhancers determines the outcome of competition in liver, and the strength of the H19 promoter is not sufficient to silence Igf2.     

Deletion of the H19 differentially methylated domain results in loss of imprinted expression of H19 and Igf2

Differentially methylated sequences associated with imprinted genes are proposed to control genomic imprinting. A 2-kb region located 5' to the imprinted mouse H19 gene is hypermethylated on the inactive paternal allele throughout development. To determine whether this differentially methylated domain (DMD) is required for imprinted expression at the endogenous locus, we have generated mice harboring a 1.6-kb targeted deletion of the DMD and assayed for allelic expression of H19 and the linked, oppositely imprinted Igf2 gene. H19 is activated and Igf2 expression is reduced when the DMD deletion is paternally inherited; conversely, upon maternal transmission of the mutation, H19 expression is reduced and Igf2 is activated. Consistent with the DMD's hypothesized role of setting up the methylation imprint, the mutation also perturbs allele-specific methylation of the remaining H19 sequences. In conclusion, these experiments show that the H19 hypermethylated 5' flanking sequences are required to silence paternally derived H19. Additionally, these experiments demonstrate a novel role for the DMD on the maternal chromosome where it is required for the maximal expression of H19 and the silencing of Igf2. Thus, the H19 differentially methylated sequences are required for both H19 and Igf2 imprinting.     

Inactivation of H19, an imprinted and putative tumor repressor gene, is a preneoplastic event during Wilms' tumorigenesis

Genetic evidence shows that the parent of origin-dependent expression patterns of the Igf2 and H19 genes is coordinated in mouse, such that H19 controls the activity of Igf2 in cis. Equally compelling evidence for a similar situation in humans is absent, although the frequently observed activation of the maternal IGF2 allele (ie., loss of imprinting) in Wilms' tumors has been attributed to the silencing of the maternal H19 locus. We show here that loss of H19 activity is generally a preneoplastic event, which may be linked with an overgrowth lesion that has been proposed to be permissive for tumor formation. Although our results document one instance in which a postneoplastic loss of H19 activity correlates with loss of IGF2 imprinting at the cellular level, it appears that inactivation of H19 is more generally independent of loss of imprinting of IGF2, at least in our specimens. Our results imply that inactivation of H19 correlates with blastema overgrowth and can be independent of a regulatory role with respect to IGF2 imprinting status in cis.     

The structural H19 gene is required for transgene imprinting

The product of the H19 gene is an untranslated RNA that is expressed exclusively from the maternal chromosome during mammalian development. The H19 gene and its 5'-flanking sequence are required for the genomic imprinting of two paternally expressed genes, Ins-2 (encodes insulin-2) and Igf-2 (encodes insulin-like growth factor-2), that lie 90 and 115 kb 5' to the H19 gene, respectively. In this report, the role of the H19 gene in its own imprinting is investigated by introducing a Mus spretus H19 gene into heterologous locations in the mouse genome. Multiple copies of the transgene were sufficient for its paternal silencing and DNA methylation. Replacing the H19 structural gene with a luciferase reporter gene resulted in loss of imprinting of the transgene. That is, high expression and low levels of DNA methylation were observed upon both paternal and maternal inheritance. The removal of 701 bp at the 5' end of the structural gene resulted in a similar loss of paternal-specific DNA methylation, arguing that those sequences are required for both the establishment and maintenance of the sperm-specific gametic mark. The M. spretus H19 transgene could not rescue the loss of Igf-2 imprinting in trans in H19 deletion mice, implying a cis requirement for the H19 gene. In contrast to a previous report in which overexpression of a marked H19 gene was a prenatal lethal, expression of the M. spretus transgene had no deleterious effect, leading to the conclusion that the 20-base insertion in the marked gene created a neomorphic mutation.     

Igf2r and Igf2 gene expression in androgenetic, gynogenetic, and parthenogenetic preimplantation mouse embryos: absence of regulation by genomic imprinting

Genomic imprinting in mammals is believed to result from modifications to chromosomes during gametogenesis that inactivate the paternal or maternal allele. The genes encoding the insulin-like growth factor type 2 (Igf2) and its receptor (Igf2r) are reciprocally imprinted and expressed from the paternal and maternal genomes, respectively, in the fetal and adult mouse. We find that both genes are expressed in androgenetic, gynogenetic, and parthenogenetic preimplantation mouse embryos. These results indicate that inactivation of imprinted genes occurs postfertilization (most likely postimplantation) and that genomic imprinting and gene inactivation are separate processes. We propose that imprinting marks the chromosome so that regulatory factors expressed in cells at later times can recognize the imprint and selectively inactivate the maternal or paternal allele. For these genes, this finding invalidates models of genomic imprinting that require them to be inactive from the time of fertilization.     

A functional analysis of imprinting in parthenogenetic embryonic stem cells

A detailed analysis of the developmental potential of parthenogenetic embryonic stem cells (PGES) was made in vivo and in vitro, and a comparison was made with the development of cells from parthenogenetic embryos (PG). In vivo, in chimeras with normal host cells (N), PGES cells showed a restricted tissue distribution consistent with that of PG cells, suggesting faithful imprinting in PGES cells with respect to genes involved in lineage allocation and differentiation. Restricted developmental potential was also observed in teratomas formed by ectopic transfer under the kidney capsule. In contrast, the classic phenotype of growth retardation normally observed in PG<==>N chimeras was not seen, suggesting aberrant regulation in PGES cells of genes involved in growth regulation. We also analysed the expression of known imprinted genes after ES cell differentiation. Igf2, H19 and Igf2r were all appropriately expressed in the PGES derived cells following induction of differentiation in vitro with all-trans retinoic acid or DMSO, when compared with control (D3) and androgenetic ES cells (AGES). Interestingly, H19 was found to be expressed at high levels following differentiation of the AGES cells. Due to the unexpected normal growth regulation of PGES<==>N chimeras we also examined Igf2 expression in PGES derived cells differentiated in vivo and found that this gene was still repressed. Our studies show that PGES cells provide a valuable in vitro model system to study the effects of imprinting on cell differentiation and they also provide invaluable material for extensive molecular studies on imprinted genes. In addition, the aberrant growth phenotype observed in chimeras has implications for mechanisms that regulate the somatic establishment and maintenance of some imprints. This is of particular interest as aberrant imprinting has recently been invoked in the etiology of some human diseases.     

Chromosome 11p15.5 regional imprinting: comparative analysis of KIP2 and H19 in human tissues and Wilms' tumors

The imprinted H19 gene is frequently inactivated in Wilms' tumors (WTs) either by chromosome 11p15.5 loss of heterozygosity (LOH) or by hypermethylation of the maternal allele and it is possible that there might be coordinate disruption of imprinting of multiple 11p15.5 genes in these tumors. To test this we have characterized total and allele-specific mRNA expression levels and DNA methylation of the 11p15.5 KIP2 gene in normal human tissues, WTs and embryonal rhabdomyosarcoma (RMS). Both KIP2 alleles are expressed but there is a bias with the maternal allele contributing 70-90% of mRNA. Tumors with LOH show moderate to marked reductions in KIP2 mRNA relative to control tissues and residual mRNA expression is from the imprinted paternal allele. Among WTs without LOH most cases with H19 inactivation also have reduced KIP2 expression and most cases with persistent H19 expression have high levels of KIP2 mRNA. In contrast to the extensive hypermethylation of the imprinted H19 allele, both KIP2 alleles are hypomethylated and WTs with biallelic H19 hypermethylation lack comparable hypermethylation of KIP2 DNA. 5-aza-2'-deoxycytidine (aza-C) increases H19 expression in RD RMS cells but does not activate KIP2 expression. These data indicate coordinately reduced expression of two linked paternally imprinted genes in most WTs and also suggest mechanistic differences in the maintenance of imprinting at these two loci.     

Purification and nucleic-acid-binding properties of a Saccharomyces cerevisiae protein involved in the control of ploidy

The mouse H19 gene is expressed exclusively from the maternal allele. The imprinted expression of the endogenous gene can be recapitulated in mice by using a 14-kb transgene encompassing 4 kb of 5'-flanking sequence, 8 kb of 3'-flanking sequence, which includes the two endoderm-specific enhancers, and an internally deleted structural gene. We have generated multiple transgenic lines with this 14-kb transgene and found that high-copy-number transgenes most closely follow the imprinted expression of the endogenous gene. To determine which sequences are important for imprinted expression, deletions were introduced into the transgene. Deletion of the 5' region, where a differentially methylated sequence proposed to be important in determining parental-specific expression is located, resulted in transgenes that were expressed and hypomethylated, regardless of parental origin. A 6-kb transgene, which contains most of the differentially methylated sequence but lacks the 8-kb 3' region, was not expressed and also not methylated. These results indicate that expression of either the H19 transgene or a 3' DNA sequence is key to establishing the differential methylation pattern observed at the endogenous locus. Finally, methylation analysis of transgenic sperm DNA from the lines that are not imprinted reveals that the transgenes are not capable of establishing and maintaining the paternal methylation pattern observed for imprinted transgenes and the endogenous paternal allele. Thus, the imprinting of the H19 gene requires a complex set of elements including the region of differential methylation and the 3'-flanking sequence.     

Murine Stim1 maps to distal chromosome 7 and is not imprinted

STIM1 (GOK) maps to a region of human Chromosome (Chr) 11p15.5 that is implicated in several embryonal tumors, and some evidence indicates that STIM1 may have a growth suppressor role in rhabdomyosarcoma. In this study we have mapped the murine homolog, Stim1, to the same position as Hbb on distal mouse Chr 7. This region is separated by 20 cM from the region of distal Chr 7 that contains Igf2, H19, and other imprinted genes. Using strain-specific polymorphisms, we have shown that Stim1 is expressed from both parental alleles in fetal and neonatal mouse tissues. Similar analyses of human Wilms' tumor and normal kidney tissues demonstrated biallelic expression of STIM1 in the majority of samples. These data demonstrate that Stim1 expression is not regulated by genomic imprinting in either mouse or human tissues. Thus, if STIM1 is a tumor suppressor at 11p15.5, loss of expression is not due to imprinting effects.     

Why is there no diploid overdose effect in Prader-Willi syndrome due to uniparental disomy?

Due to DNA technology, it is now apparent that the mechanisms of genetic disease are more complex than the model of a gene with biallelic expression in the diploid state. If a gene is imprinted, monoallelic expression is the norm when the chromosomes of a pair are inherited normally from each parent. Uniparental disomy (UPD) is the abnormal situation where both chromosomes of a pair come from the same parent. When the chromosome contains an imprinted gene, UPD may result in nullisomy or disomy for a functional copy of that gene. If there are two imprinted loci on the same chromosome, UPD for that chromosome results in nullisomy for one imprinted gene but functional disomy for the other a "diploid overdose" (DO). This situation has been well demonstrated in the Prader-Willi syndrome (PWS) which is the nullisomic phenotype for the PWS gene(s) on chromosome 15q11-13. Chromosome 15q11-13 also contains the gene for Angelman syndrome (AS) which has a phenotype distinct from PWS. Both loci are subject to imprinting--in PWS, the imprint is on the maternal chromosome 15, in AS it is on the paternal chromosome 15. All individuals with PWS due to maternal UPD, while functionally nullisomic for the PWS locus, are functionally disomic for the AS locus--a DO situation. Assuming that biallelic expression of an imprinted gene is harmful, one would expect DO for an imprinted gene to produce a phenotypic effect. Cases of PWS due to UPD do not appear to differ from those due to deletion (hypopigmentation in deletional cases can be explained by loss of D15S12 downstream from the critical region). There is no good evidence of DO for the AS locus in PWS due to UPD. Why then was it 'necessary' in evolutionary terms to imprint the AS locus and maintain the imprint faithfully for life. A similar situation of two imprinted genes on the same chromosome occurs with IGF2 and H19 on chromosome 11p15. Maternal imprinting for IGF2 and paternal imprinting for H19 is the norm. Paternal UPD in this situation does lead to a DO effect, namely Beckwith-Wiedemann syndrome. The possibility of a DO effect needs to be considered when assessing the phenotypic spectrum of UPD for other chromosomes currently under investigation.     

LCR-dependent gene expression in beta-globin YAC transgenics: detailed structural studies validate functional analysis even in the presence of fragmented YACs

Yeast artificial chromosome (YAC) transgenesis is associated with a high frequency of deletions in the integrated transgenes. To determine the impact of these rearrangements on the ability to derive structure-function relationships using YACs, transgenic mice were generated with 248 or 155 kb beta-globin locus YACs. The transgenics were examined for structural integrity of the YAC using an approach of structural analysis that unambiguously demonstrates intactness of YAC transgene copies. Globin gene expression per copy of each integrated transgene and the profiles of globin gene expression during development were determined. Diverse deletion patterns were observed in one or more integrated YACs in all the 248 and most of the 155 kb transgenic lines we analyzed. However, when the structure of the major regulatory element of the beta-globin locus, the locus control region, was preserved, the genes of the beta-globin locus functioned normally and globin transgenes of both the 248 and 155 kb beta-YACs were expressed in a position-independent, copy number-dependent manner. Furthermore, the globin genes of both beta-YACs displayed normal developmental regulation. We conclude that YACs can be used for analysis of structure-function relationships of large genes or multigene loci in spite of the tendency for rearrangements and deletions of the integrated transgenes. However, detailed structural evidence for integrity and continuity of locus sequences is required for correct interpretation of functional data.