X chromosome reactivation in mouse embryonal carcinoma cells

The influence of DNA methylation on monoallelic expression

Monoallelic gene expression occurs in diploid cells when only one of the two alleles of a gene is active. There are three main classes of genes that display monoallelic expression in mammalian genomes: (1) imprinted genes that are monoallelically expressed in a parent-of-origin dependent manner; (2) X-linked genes that undergo random X-chromosome inactivation in female cells; (3) random monoallelically expressed single and clustered genes located on autosomes. The heritability of monoallelic expression patterns during cell divisions implies that epigenetic mechanisms are involved in the cellular memory of these expression states. Among these, methylation of CpG sites on DNA is one of the best described modification to explain somatic inheritance. Here, we discuss the relevance of DNA methylation for the establishment and maintenance of monoallelic expression patterns among these three groups of genes, and how this is intrinsically linked to development and cellular states.

Reexpression of a cluster of silenced transgenes is associated with their rearrangement

Irreversible inactivation or silencing of tumor suppressor genes occurs frequently in the development of cancer. A similar process of silencing can occur after the integration of transfected or microinjected genes into the genomes of recipient cells. The inactivation of transfected genes seems particularly efficient in cells with stem cell characteristics. We have been studying the inactivation of genes transfected into cultured P19 embryonal carcinoma cells and found that the CpG-rich sequence comprising the coding region of the lacZ reporter gene becomes extensively methylated after integration into the genome. 5-Aza-2'-deoxycytidine (5AdC), an inhibitor of DNA methylation, induced the reexpression of silent transgenes in one clone of P19 cells studied in detail. However, the reexpressed genes remained heavily methylated over the lacZ coding sequence. We used pulsed-field gel electrophoresis to analyze the structure of the transgenic locus in the parental and in 5AdC-treated cells and found that, in each of the cells reexpressing the transgene, the cluster of transgenes had been rearranged. Each clone had undergone a different rearrangement that appeared to involve recombination within the tandemly repeated copies of the transgene. Our data seem consistent with the idea that 5AdC induces efficient DNA recombination between tandemly repeated genes and that the reexpression of silenced genes induced by 5AdC might be triggered by the chromatin reorganization at the site of DNA recombination.

DNA methylation pattern of a tandemly repeated LacZ transgene indicates that most copies are silent

The Pgk-1,2-lacZ transgene consists of the ubiquitously-expressed Pgk-1 promoter driving expression of the E. coli lacZ reporter gene. We studied the expression of this transgene in a mouse strain carrying 8-9 tandem copies of this construct. When inherited through the male germ line, the transgene was expressed in all tissues examined but when inherited through the female germ line, the transgene became irreversibly inactivated. The lacZ region is a CpG-rich island that was essentially entirely methylated in all copies of the silent, maternally-inherited transgene. At the active transgenic locus, all but one of the copies were entirely methylated. This one unmethylated copy was adjacent to the cellular DNA and was presumed to be the expressed transgene copy. These results suggest that the tandem repeats of transgenes become silenced by a mechanism associated with DNAmethylation and that proximity to the cellular genome may be important in maintaining expression against the spread of inactivation from the adjacent silent transgenes.

Mouse embryonal carcinoma cell-somatic cell hybrids as experimental tools for the study of cell differentiation and X chromosome activity

Remarkable properties of murine embryonal carcinoma cells, stem cells of teratocarcinomas, manifest themselves when they are fused with differentiated somatic cells. Among others the capacity to reprogram genomes from these somatic cells and restore genetic pluripotency is most striking.

Isolation and cultivation of blastocyst-derived stem cell lines from American mink (Mustela vison)

Ten embryonic stem (ES) cell lines from mink blastocysts were isolated and characterized. All the lines had a normal diploid karyotype; of the ten lines studied, five had the XX and five had the XY constitution. Testing of the pluripotency of the ES-like cells demonstrated that 1) among four lines of genotype XX, and X was late-replicating in three; both Xs were active in about one-third of cells of line MES8, and analysis of glucose-6-phosphate dehydrogenase revealed no dosage compensation for the X-linked gene; 2) when cultured in suspension, the majority of lines were capable of forming "simple" embryoid bodies (EB), and two only showed the capacity for forming "cystic" multilayer EBs. However, formation of ectoderm or foci of yolk sac hematopoiesis, a feature of mouse ES cells, was not observed in the "cystic" EB; 3) when cultured as a monolayer without feeder, the ES cells differentiated into either vimentin-positive fibroblast-like cells or cytokeratin-positive epithelial-like cells (less frequently); neural cells appeared in two lines; 4) when injected into athymic mice, only one of the four tested lines gave rise to tumors. These were fibrosarcomas composed of fibroblast-like cells, with an admixture of smooth muscular elements and stray islets of epithelial tissue; (5) when the ES cells of line MES1 were injected into 102 blastocyst cavities and subsequently transplanted into foster mothers, we obtained 30 offspring. Analysis of the biochemical markers and coat color did not demonstrate the presence of chimaeras among offspring. Thus the cell lines derived from mink blastocysts are true ES cells. However, their pluripotential capacities are restricted.

DNA methylation of two X chromosome genes in female somatic and embryonal carcinoma cells

The extent of methylation of DNA sequences upstream and within the two X-linked genes, Pgk-1 and Hprt, was analyzed in male and female somatic cells and in female embryonal carcinoma cells carrying either two active X chromosomes (Xa) or one active and one inactive X chromosome (Xi). Sites upstream and within the first intron of both Pgk-1 and Hprt were heavily methylated on the Xi in somatic cells and in embryonal carcinoma cells with an Xi. Reactivation of this Xi was accompanied by extensive demethylation of these sites. In female embryonal carcinoma cells with two active X chromosomes, one X inactivates during differentiation in culture; however, methylation did not occur during differentiation, consistent with the idea that DNA methylation does not play a role in the initiation of X inactivation but may be involved in maintaining inactivation of those genes on the Xi.

The search for the mouse X-chromosome inactivation centre

The phenomenon of X-chromosome inactivation in female mammals, whereby one of the two X chromosome present in each cell of the female embryo is inactivated early in development, was first described by Mary Lyon in 1961. Nearly 30 years later, the mechanism of X-chromosome inactivation remains unknown. Strong evidence has accumulated over the years, however, for the involvement of a major switch or inactivation centre on the mouse X chromosome. Identification of the inactivation centre at the molecular level would be an important step in understanding the mechanism of X-inactivation. In this paper we review the evidence for the existence and location of the X-inactivation centre on the mouse X-chromosome, present data on the molecular genetic mapping of this region, and describe ongoing strategies we are using to attempt to identify the inactivation centre at the molecular level.

Pharmacodynamic and DNA methylation studies of high-dose 1-beta-D-arabinofuranosyl cytosine before and after in vivo 5-azacytidine treatment in pediatric patients with refractory acute lymphocytic leukemia

The primary development of clinical resistance to 1-beta-D-arabinofuranosyl cytosine (ara-C) in leukemic blast cells is expressed as decreased cellular concentrations of its active anabolite. Correlations exist between the cellular concentrations of 1-beta-D-arabinofuranosyl cytosine 5'-triphosphate (ara-CTP) in leukemic blast cells and inhibition of DNA synthetic capacity with the clinical response to high-dose cytosine arabinoside (HDara-C). 5-Azacytidine (5-Aza-C) and its congeners are potent DNA hypomethylating agents, an action closely associated with the reexpression of certain genes such as that for deoxycytidine kinase (dCk) in ara-C-resistant mouse and human leukemic cells. Reexpression of dCk could increase the cellular ara-CTP concentrations and the sensitivity to ara-C. A total of 17 pediatric patients with refractory acute lymphocytic leukemia (ALL) received a continuous infusion of 5-Aza-C at 150 mg/m2 daily for 5 days after not responding to (13/17) or relapsing from (4/17) an HDara-C regimen (3 g/m2 over 3 h, every 12 h, x 8 doses). Approximately 3 days after the end of the 5-Aza-C infusion, the HDara-C regimen was given again with the idea that the induced DNA hypomethylation in the leukemic cells may have increased the dCk activity and that a reversal of the tumor drug resistance to ara-C could have occurred. Deoxycytidine kinase (expressed as cellular ara-CTP concentrations) in untreated blasts, DNA synthetic capacity (DSC), and the percentage of DNA methylcytidine levels were determined before and after 5-Aza-C administration. Cellular ara-CTP was enhanced to varying degrees in 15 of 16 patients after 5-Aza-C treatment. The average cellular concentration of ara-CTP determined in vitro by the sensitivity test was 314 +/- 390 microM, 2.3-fold higher than the average value before 5-Aza-C treatment. In 12 patients in whom the DNA methylation studies were completed before and after 5-Aza-C treatment, the average DNA hypomethylation level was 55.6% + 15.8% of pretreatment values (n = 13; mean +/- SD). DSC showed a profound decline in 2/9 evaluable patients who achieved a complete response (CR) after this regimen. The data suggest that treatment with a cytostatic but DNA-modulatory regimen of 5-Aza-C causes DNA hypomethylation in vivo, which is associated with dCk reexpression in the patients' leukemic blasts. The partial reversal of drug resistance to ara-C by 5-Aza-C yielded two CRs in this poor-prognosis, multiply relapsed patient population with refractory ALL.(ABSTRACT TRUNCATED AT 400 WORDS)

Reactivation of hprt on the inactive X chromosome with DNA demethylating agents

The C86 line of female embryonal carcinoma cells contains one active and one inactive X chromosome. Following methylnitrosourea mutagenesis, a clone called C86AGM2 was isolated that carries a mutated hprt gene on the active X chromosome. This hprtm allele encodes an HPRT enzyme that has less than 1% normal enzyme activity, is thermolabile, and has an altered isoelectric point. Following treatment with drugs that demethylate DNA, the hprt+ gene from the inactive X chromosome in C86AGM2 cells became active as determined by the appearance of HPRT activity with the thermodenaturation and electrofocusing characteristics of the normal enzyme. No expression of this hprt+ gene occurred if C86AGM2 cells were induced to differentiate prior to DNA demethylation. Stable lines of C86AGM2 cells expressing both the hprtm and hprt+ genes did not inactivate either gene following differentiation.

Methylation is co-ordinated on the putative replication origins of Physarum ribosomal DNA

In Physarum polycephalum, the ribosomal DNA is found as 60,000 base-pair palindromes. Each rDNA has four symmetrically arranged replication origins flanked by ribosomal RNA genes. A particular sequence, the putative replication origin, is repeated at the approximate position of each origin and nowhere else in the molecule. On a typical rDNA molecule, only one origin is active per replication cycle. We show that both the level and co-ordination of methylation result in asymmetrically methylated rDNA molecules that are particularly hypomethylated at one of their four putative replication origins. This pattern of methylation on a typical rDNA molecule is consistent with a model where hypomethylation is a determinant of origin activity.

High-frequency reactivation of X-linked genes in Chinese hamster X human hybrid cells

Three genes on the human inactive X chromosome retained in the Chinese hamster X human hybrid cell line X8/6T2 have been reactivated using the demethylating agent, 5-azacytidine (5-aza-CR). Pulse-labeling and histochemical methods permitted detection and measurement of reactivation rates of the hypoxanthine phosphoribosyltransferase (Hpt) and glucose-6-phosphate dehydrogenase (G6pd) genes within 48 h of treatment. About 50% of the cells became active for these genes, which represents a reactivation rate some 30-fold greater than previously reported in similar systems. The phosphoglycerate kinase (Pgk) gene was not reactivated as frequently as the Hpt or G6pd genes. Segregation analysis of progeny of treated cells showed that enzyme-positive and enzyme-negative cells were produced in proportions supporting the notion that 5-aza-CR causes demethylation by replicative loss and that demethylation leads to reactivation.

5-aza-C-induced changes in the time of replication of the X chromosomes of Microtus agrestis are followed by non-random reversion to a late pattern of replication

Treatment with 5-azacytidine (5-aza-C) causes an advance in the time of replication and enhances the DNase-I sensitivity of the inactive X chromosome in Gerbillus gerbilllus fibroblasts. We found that these changes were not stably inherited and upon removal of the drug the cells reverted to the original state of one active and one inactive X chromosome. In order to determine whether this reversion was random, we used a cell line of female Microtus agrestis fibroblasts in which the two X chromosomes are morphologically distinguishable. In this work we show that the reversion to a late pattern of replication is not random, and the originally late replicating X chromosome is preferentially "reinactivated", suggesting an imprinting-like marking of one or both X chromosomes. The changes in the replication pattern of the X chromosome were associated with changes in total DNA methylation. Double treatment of cells with 5-aza-C did not alter this pattern of euchromatin activation and reinactivation. A dramatic advance in the time of replication of the entire X linked constitutive heterochromatin (XCH) region was however, observed in the doubly treated cells. This change in the replication timing of the XCH occurred in both X chromosomes and was independent of the changes observed in the euchromatic region. These observations suggest the existence of at least two independent regulatory sites which control the timing of replication of two large chromosomal regions.

Methylation of the Hprt gene on the inactive X occurs after chromosome inactivation

DNA sequences have previously been identified in the first intron of the mouse Hprt gene that are methylated on the inactive but not the active X chromosome. The temporal relationship between methylation of these sequences and X-inactivation was studied in teratocarcinoma cells and postimplantation mouse embryos: the sequences are unmethylated prior to X-inactivation and do not become methylated on the inactive X in most fetal cells until several days postinactivation. Such inactive X-specific methylation occurs in a significantly smaller proportion of the cells in the extra-embryonic tissues, yolk sac mesoderm and endoderm, than in the fetus. These data suggest that the inactive X-specific methylation of sequences such as those in the first intron of the Hprt gene does not play any role in the primary events of X-inactivation, but may function as part of a secondary, tissue-specific mechanism for maintaining the inactive state.

Cloning and expression of the mouse pgk-1 gene and the nucleotide sequence of its promoter

We report the cloning of the mouse pgk-1 gene encoding the somatic cell isoform of the enzyme phosphoglycerate kinase. The gene is contained within a 16-kb region of the X chromosome and is interrupted by at least ten introns. The promoter region of the pgk-1 gene is rich in G and C nucleotides and contains five copies of the hexadeoxynucleotide, GGGCGG, the potential binding site for the Sp 1 transcription factor, a CCAAT sequence, but no TATA box. This promoter functions following DNA-mediated transfection into mammalian cells. The promoter of the mouse pgk-1 gene is homologous to the human pgk-1 promoter. A number of conserved motifs in the promoter may indicate a significant role for these sequences in expression of the pgk-1 gene.

Reactivation of X-linked genes in human fibroblasts transformed by origin-defective SV40

To determine if expression of genes on the inactive X is inducible in human cells, we looked for reactivation events in a clone of fibroblasts transformed with origin-defective SV40. The karyotype of these cells was grossly heteroploid so that the aneuploidy associated with SV40 transformation occurs even in the absence of viral replication. This transformed clone, heterozygous for hypoxanthine phosphoribosyltransferase (HPRT), lacks HPRT activity, as the mutant allele is on the active X and the normal allele on the inactive X. Reactivation of the HPRT+ allele on the inactive X was observed at a frequency of 6 X 10(-5) per cell and increased approximately eightfold following treatment with the cytidine analogs 5-azacytidine (5azaC) and 5-azadeoxycytidine. The fact that spontaneous reactivation is detectable in some clones, but not all, suggests that the environment of the SV40-transformed cell, although not sufficient to induce generalized derepression, increases the frequency of rare reactivation events. The methylation pattern at the HPRT locus revealed transformation-associated alterations that may have predisposed these cells to reactivation events, spontaneous as well as 5azaC-induced.