Incomplete X chromosome dosage compensation in chorionic villi of human placenta.

Studies of glucose-6-phosphate dehydrogenase (G6PD) in heterozygous cells from chorionic villi of five fetal and one newborn placenta show that the locus on the allocyclic X is expressed in many cells of this trophectoderm derivative. Heterodimers were present in clonal populations of cells with normal diploid karyotype and a late replicating X chromosome. The expression of the two X chromosomes was unequal, based on ratios of homodimers and heterodimers in clones. Studies of DNA, digested with Hpa II and probed with cloned genomic G6PD sequences, indicate that expression of the locus in chorionic villi is associated with hypomethylation of 3' CpG clusters. These findings suggest that dosage compensation, at least at the G6PD locus, has not been well established or maintained (or both) in placental tissue. Furthermore, the active X chromosome in these human cells of trophoblastic origin can be either the paternal or maternal one; therefore, paternal X inactivation in extraembryonic lineages is not an essential feature of mammalian X dosage compensation.     

Isolation of the human chromosomal band Xq28 within somatic cell hybrids by fragile X site breakage.

The chromosomal fragile-site mapping to Xq27.3 is associated with a frequent form of mental retardation and is prone to breakage after induced deoxyribonucleotide pool perturbation. The human hypoxanthine phosphoribosyltransferase (HPRT) and glucose-6-phosphate dehydrogenase (G6PD) genes flank the fragile X chromosome site and can be used to monitor integrity of the site in human-hamster somatic cell hybrids deficient in the rodent forms of these activities. After induction of the fragile X site, negative selection for HPRT and positive enrichment for G6PD resulted in 31 independent colonies of HPRT-,G6PD+ phenotype. Southern blot analysis demonstrated the loss of all tested markers proximal to the fragile X site with retention of all tested human Xq28 loci in a majority of the hybrids. In situ hybridization with a human-specific probe demonstrated the translocation of a small amount of human DNA to rodent chromosomes in these hybrids, suggesting chromosome breakage at the fragile X site and the subsequent translocation of Xq28. Southern blot hybridization of hybrid-cell DNA, resolved by pulsed-field gel electrophoresis, for human-specific repetitive sequences revealed abundant CpG-islands within Xq28, consistent with its known gene density. The electrophoretic banding patterns of human DNA among the hybrids were remarkably consistent, suggesting that fragile X site breakage is limited to a relatively small region in Xq27-28. These somatic cell hybrids, containing Xq27.3-qter as the sole human DNA, will aid the search for DNA associated with the fragile X site and will augment the high resolution genomic analysis of Xq28, including the identification of candidate genes for genetic-disease loci mapping to this region.     

Localization of loci for hypoxanthine phosphoribosyltransferase and glucose-6-phosphate dehydrogenase and biochemical evidence of nonrandom X chromosome expression from studies of a human X-autosome translocation.

We report a unique and complex karyotypic rearrangement involving chromosomes X, 3, 7, and 21. Blood cells and fibroblasts from the proband do not express the maternal allele for glucose-6-phosphate dehydrogenase (G6PD), providing biochemical evidence for nonrandom expression of X-linked genes in balanced X-autosome translocations. The break point on the X chromosome, at the junction of Xq27-Xq28, separates the loci for hypoxanthine phosphoribosyltransferase (HPRT) and G6PD. Studies of mouse-human hybrids derived from the proband's cells indicate that G6PD, at q28, is clearly distal to all other X loci now assigned. From these and previous studies, we can localize HPRT to that segment between Xq26 and Xq27. The studies also provide further evidence for the stability of the inactive X phenotype in hybrid cells.     

Localized Derepression on the Human Inactive X Chromosone in Mouse-Human Cell Hybrids.

Evidence for derepression of the gene for hypoxanthine phosphoribosyltransferase (HPRT; IMP: pyrophosphate phosphoribosyltransferase, EC 2.4.2.8) on the human inactive X chromosome was obtained in hybrids of mouse and human cells. The mouse cells lacked HPRT and were also deficient in adenine phosphoribosyltransferase (APRT; AMP: pyrophosphate phosphoribosyltransferase; EC2.4.2.7). The human female fibroblasts were HPRT-deficient as a consequence of a mutation on the active X but contained a normal HPRT gene on the inactive X. The two human X chromosomes were further distinguished by differences in morphology: the inactive X was morphologically normal while the active X included most of the long arm of autosome no. 1 translocated to the distal end of the X long arm. Forty-one hybrid clones were first isolated by selection for the presence of APRT; when these clones were selected for HPRT, six of them yielded derivatives having human HPRT with incidences of about 1 in 10-6 APRT-selected hybrid cells. The HPRT-positive derivatives contained a normal-appearing X chromosome indistinguishable from the inactive X of the parental human fibroblasts. The active X with the translocation was not found in any of the HPRT-positive hybrid cells. Human phosphoglycerokinase (ATP:3-phospho-D-glycerate 1-phosphotransferase. EC 2.7.2.3) and glucose-6-phosphate dehydrogenase (D-glucose 6-phosphate: NADP 1-oxidoreductase, EC 1.1.1.49), which are specified by X-chromosomal loci, were not detected in the hybrids expressing HPRT even though they contained an apparently intact X chromosome. The observations are most simply explained by the infrequent, stable derepression of inactive X chromosome segments that include the HPRT locus but not the phosphoglycerokinase and glucose-6-phosphate dehydrogenase loci.     

Genetics of cell surface receptors for bioactive polypeptides: Binding of epidermal growth factor is associated with the presence of human chromosome 7 in human-mouse cell hybrids

Mouse A9 cells, L-cell-derived mutants deficient in hypoxanthine phosphoribosyltransferase (HPRT; IMP:pyrophosphate phosphoribosyltransferase, EC 2.4.2.8) were found to be incapable of binding ¹²⁵I-labeled epidermal growth factor (EGF) to the cell surface. The A9 cells were fused with human diploid fibroblasts (WI-38) possessing EGF-binding ability, and human-mouse cell hybrids (TA series) were isolated after hypoxanthine/aminopterin/thymidine/ouabain selection. Analyses of isozyme markers and chromosomes of four representative clones of TA hybrids indicated that the expression of EGF-binding ability is correlated with the presence of human chromosome 7 or 19. Four subclones were isolated from an EGF-binding-positive line, TA-4, and segregation of EGF-binding was found to be concordant with the expression of human mitochondrial malate dehydrogenase (MDHM; L-malate:NAD⁺ oxidoreductase, EC 1.1.1.37), a marker for chromosome 7, but not with glucosephosphate isomerase (GPI; D-glucose-6-phosphate ketol-isomerase, EC 5.3.1.9), a marker for chromosome 19. Furthermore, evidence from 27 clones of AUG hybrids that were produced between A9 and another human fibroblast line, GM1696, carrying an X/7 chromosome translocation indicated that EGF-binding ability segregates together with human MDHM and two X-linked markers, HPRT and glucose-6-phosphate dehydrogenase (G6PD; D-glucose-6-phosphate:NADP⁺ 1-oxidoreductase, EC 1.1.1.49), that are located on the translocation chromosome 7p⁺. These results permit assignment of the gene, designated EGFS, which is associated with the expression of EGF-binding ability, to human chromosome 7 and its localization to the p22-qter region. Because the EGF receptor is reported to be a glycoprotein the EGFS could be either a structural gene(s) for receptor protein or a gene(s) for modifying the receptor protein through glycosylation.     

Expression of an X-linked gene from an inactive human X chromosome in mouse-human hybrid cells: further evidence for the noninactivation of the steroid sulfatase locus in man.

Somatic cell hybrid clones were derived from the fusion of hypoxanthine phosphoribosyltransferase (HPRT; EC 2.4.2.8)-deficient mouse cells and two different human fibroblast strains, each carrying an X chromosome-autosome translocation. One of these had an X/11 translocation [46,X,t(X;11)(p21;q13)] and the other had an X/19 translocation [46,X,t(X;19)(q22;q13)]. The structurally normal human X chromosome is the late-replicating (genetically inactive) chromosome in these two cell strains; the rearranged X chromosome is early replicating (genetically active). One primary hybrid clone carrying both the translocated X chromosome and the structurally normal X chromosome was isolated in hypoxanthine/aminopterin/thymidine medium from each of these two cell fusion experiments. These clones were then selected in medium containing 8-azaguanine to achieve the loss of the active human HPRT locus. Five subclones from the cell hybrid with the X/11 translocation failed to express two known human X-chromosome markers [glucose-6-phosphate dehydrogenase (G6PD; EC 1.1.1.49) and phosphoglycerate kinase (PGK; EC 2.7.2.3)] but did express human microsomal steroid sulfatase (STS; sterol-sulfate sulfohydrolase, EC 3.1.6.2). Three of these were cytogenetically analyzed and found to contain a structurally normal human X chromosome but not the X/11 translocation. Two subclones were isolated in 8-azaguanine from the hybrid with the X/19 translocation. Cytogenetic analysis of these two clones showed the presence of a structurally normal human X chromosome; the X/19 translocation was not present. They did not express human G6PD, PGK, or HPRT but did express human STS. These results indicate that human STS is expressed from a locus on the inactive human X chromosome and support our earlier finding that the STS locus escapes X-inactivation in man.     

Cytogenetic and molecular studies on a recombinant human X chromosome: implications for the spreading of X chromosome inactivation.

A pericentric inversion of a human X chromosome and a recombinant X chromosome [rec(X)] derived from crossing-over within the inversion was identified in a family. The rec(X) had a duplication of the segment Xq26.3----Xqter and a deletion of Xp22.3----Xpter and was interpreted to be Xqter----Xq26.3::Xp22.3----Xqter. To characterize the rec(X) chromosome, dosage blots were done on genomic DNA from carriers of this rearranged X chromosome using a number of X chromosome probes. Results showed that anonymous sequences from the distal end of the long arm to which probes 4D8, Hx120A, DX13, and St14 bind as well as the locus for glucose-6-phosphate dehydrogenase (G6PD) were duplicated on the rec(X). Mouse-human cell hybrids were constructed that retained the rec(X) in the active or inactive state. Analyses of these hybrid clones for markers from the distal short arm of the X chromosome showed that the rec(X) retained the loci for steroid sulfatase (STS) and the cell surface antigen 12E7 (MIC2); but not the pseudoautosomal sequence 113D. These molecular studies confirm that the rec(X) is a duplication-deficiency chromosome as expected. In the inactive state in cell hybrids, STS and MIC2 (which usually escape X chromosome inactivation) were expressed from the rec(X), whereas G6PD was not. Therefore, in the rec(X) X chromosome inactivation has spread through STS and MIC2 leaving these loci unaffected and has inactivated G6PD in the absence of an inactivation center in the q26.3----qter region of the human X chromosome. The mechanism of spreading of inactivation appears to operate in a sequence-specific fashion. Alternatively, STS and MIC2 may have undergone inactivation initially but could not be maintained in an inactive state.     

Reexpression of the rat hypoxanthine phosphoribosyltransferase gene in rat-human hybrids

Fusion of hypoxanthine phosphoribosyltransferase (HPRT)^- rat hepatoma cells with HPRT⁺ human fibroblasts yielded hybrid clones that grew in HAT selective medium and contained all the rat chromosomes and one to nine human chromosomes. Among the retained chromosomes was the human X chromosome. In all clones backselected in medium containing 8-azaguanine, human X chromosome was absent. Electrophoretic analysis revealed that, without exception, hybrid clones growing in HAT medium had an active HPRT enzyme, either human or rat, or both. When these clones were backselected in 8-azaguanine, they did not show HPRT enzyme activity. Hybrids that contained the human X chromosome also had human glucose-6-phosphate dehydrogenase. The observed reexpression of rat HPRT in hybrid cells derived from HPRT^- rat cells suggests that a genetic factor from the human cell determined the expression of the rat structural gene for HPRT.     

Human X-Linked genes regionally mapped utilizing X-autosome translocations and somatic cell hybrids.

Human genes coding for hypoxanthine phosphoribosyltransferase (HPRT, EC 2.4.2.8; IMP:pyrophosphate phosphoribosyltransferase), glucose-6-phosphate dehydrogenase (G6PD, EC 1.1.1.49; D-glucose-6-phosphate:NADP+ 1-oxidoreductase), and phosphoglycerate kinase (PGK, EC 2.7.2.3; ATP:3-phospho-D-glycerate 1-phosphotransferase) have been assigned to specific regions on the long arm of the X chromosome by somatic cell gentic techniques. Gene assignment and linear order were determined by employing human somatic cells possessing an X/9 translocation or an X/22 translocation in man-mouse cell hybridization studies. The X/9 translocation involved the majority of the X long arm translocated to chromosome 9 and the X/22 translocation involved the distal half of the X long arm translocated to 22. In each case these rearrangements appeared to be reciprocal. Concordant segregation of X-linked enzymes and segments of the X chromosome generated by the translocations indicated assignment of the PGK gene to a proximal long arm region (q12-q22) and the HPRT and G6PD genes to the distal half (q22-qter) of the X long arm. Further evidence suggests a gene order on the X long arm of centromere-PGK-HPRT-G6PD.     

X chromosome reactivation in oocytes of Mus caroli.

Mature mammalian oocytes have both of their X chromosomes active, while somatic cells from the same individual have one of their X chromosomes in an inactive state. We asked whether the X chromosomes of the germ cells never undergo inactivation in their ontogeny or whether inactivation of an X chromosome does occur but is followed by a subsequent reactivation event. Our approach has used an electrophoretic polymorphism for the X-linked enzyme glucose-6-phosphate dehydrogenase (G6PD) in the mouse species Mus caroli. G6PD is dimeric, and a heterodimer is produced in cells from heterozygous females if and only if both X chromosomes are active. Ovaries from heterozygous fetuses at different gestational ages were dissected and either studied cytologically or pressed between microscopy slides to obtain germ cell-rich and germ cell-poor preparations. No heterodimer band was detected on the 10th day of development in germ cell-rich preparations. On subsequent days, an increasingly intense heterodimer band was detected, which, by the 13th day, was approximately twice as intense as the corresponding homodimer bands. Consideration of (i) the G6PD activity per germ cell and per somatic cell and (ii) the percentage of germ cells in the germ cell-rich preparations indicated that a heterodimer band should have been visible on the 10th day had both X chromosomes been active. Cytological examinations showed that the earliest germ cells enter meiotic prophase on the eleventh day. These results demonstrate that oogonia have a single active X chromosome and that the inactive X chromosome is reactivated at or, more likely, shortly before entry into meiotic prophase.     

Transfer of the human gene for hypoxanthine-guanine phosphoribosyltransferase via isolated human metaphase chromosomes into mouse L-cells.

We have transferred the human gene for hypoxanthine-guanine phosphoribosyltransferase (HPRT, EC 2.4.2.8; IMP:pyrophosphate phosphoribosyltransferease) via isolated metaphase chromosomes from human HeLa S3 cells into murine A9 cells which lack functional murine HPRT activity, using the technique of McBride and Ozer (Proc, Nat. Acad. Sci. USA 70, 1258-1262, 1973). Three transformed clones were isolated which contained human HPRT activity as determined by electrophoretic and immunochemical assays. Twenty human isozymes other than HPRT whose genes have been assigned to 14 human chromosomes were found to be absent in our transformed clones. Moreover, the human isozymes of hlucose-6-phosphate dehydrogenase (EC 1.1.1.49; D-glucose 6-phosphate:NADP 1-oxidoreductase) and phosphoglycerate kinase (EC 2.7.2.3;ATP:3-phospho-D-glycerate 1-phosphotransferase), whose genes have been linked with the HPRT gene to the long are of the human X chromosome, were also absent. On the basis of the known linkage relationships of the three markers, we thereby suggest that the transferred piece of human genetic material is smaller than 20% of the human X chromosome or less than 1% of the human genome. This estimate assumes a normal syntenic relationship for the long arm of the X chromosome in HeLa S3 cells. In agreement with this conclusion, no human chromosomes could be detected in our transformed clones. When grown under nonselective conditions about 3% of the gene transfer cells lost the human HPRT marker per cell generation. Transformants that had lost human HPRT activity were subjected to hypoxanthine-aminopterin-thymidine selection. The frequency of revertants to the HPRT(+) phenotype was less than 1 x 10(-6), and two revertants that were obtained possessed the mouse electrophoretic phenotype. These results argue against a stable integration of the human donor genetic material into the mouse recipient genome.     

Clonal nature of Philadelphia chromosome-positive and -negative chronic myelogenous leukemia by DNA hybridization analyses

Evidence for the clonal nature of chronic myelogenous leukemia (CML) has been obtained primarily from studies of black females expressing polymorphic glucose-6-phosphate dehydrogenase (G6PD) isoenzymes where, instead of the heterozygous pattern normally found as a result of random X chromosome inactivation, exclusive expression of only one G6PD allele has been demonstrated in leukemic cell populations. We report here the use of two other molecular approaches to examine clonality of peripheral blood cells in patients with CML. The first of these is based on the analysis of consistent differential methylation patterns associated with active and inactive X chromosomes within the region spanned by a BamHI restriction fragment length polymorphism (RFLP) at the hypoxanthine phosphoribosyltransferase (HPRT) locus. By this method, three heterozygous females gave results consistent with monoclonal origin of the disease, including one patient lacking the Philadelphia chromosome (Ph1) normally associated with CML. In the other two patients, both of whom had Ph1-positive CML, clonality was confirmed by the demonstration of simple gene rearrangements by Southern hybridization with a breakpoint cluster region (bcr) probe from chromosome 22.     

Studies of the locus for androgen receptor: localization on the human X chromosome and evidence for homology with the Tfm locus in the mouse.

We have established a cell line from mouse kidney cells expressing the tfm mutation and showed that these cells lack androgen binding activity. A subclone of these simian virus 40 (SV40)-transformed cells (6TGR-SV-tfm) selected in 6-thioguanine and lacking hypoxanthine phosphoribosyltransferase was used to produce a series of mouse--human hybrids containing the normal human X chromosome or various X autosome-translocation chromosomes (expressing only segments of the human X chromosome). When the androgen receptor locus (AR) was present in the hybrid, the number of receptor sites and kinetics of binding were similar to that in the human parental cells. Analysis of hybrids with partial human X chromosomes by using assays for X chromosome-linked enzymes and for the androgen receptor protein indicate that the AR locus on the human X chromosome is near the centromere between Xq13 and Xp11 and is proximal to the locus for phosphoglycerate kinase. Hybrids derived from 6TGR-SV-tfm mouse cells and human labial fibroblasts from an XY individual with the ar- form of androgen insensitivity have no binding activity. The lack of complementation indicates that the X chromosome-linked mutations in mouse and man affect homologous loci and supports the evolutionary conservation of X chromosomal loci in mammals; however, the position of the locus on the human X chromosome indicates that intrachromosomal rearrangement has occurred.     

Mitotic Separation of Two Human X-Linked Genes in Man—Mouse Somatic Cell Hybrids

SIX INTERSPECIFIC SOMATIC HYBRID CELL LINES WERE DERIVED FROM A MOUSE LINE DEFICIENT IN HYPOXANTHINE: guanine phosphoribosyltransferase (HGPRT) and human diploid cells with normal enzyme activity. Human HGPRT was present in all six hybrids and the clones derived from them. However, in two of the six, and in some clones from another two, human glucose-6-phosphate dehydrogenase (G6PD) was absent. Since the structural loci for both these enzymes are X-linked in man, these findings suggest that these two loci have separated quite frequently through chromosome breakage and that they must be rather far apart on the X chromosome.     

Comparative gene mapping: order of loci on the X chromosome is different in mice and humans.

For comparative studies we have used the somatic cell hybridization approach to regionally map genes on the mouse X chromosome. Fibroblasts from a mouse with the balanced reciprocal translocation T(XD;16B5)16H were fused with a Chinese hamster cell line (V79/380-6) deficient in activity of the enzyme hypoxanthine phosphoribosyltransferase (HPRT). Interpecific cell hybrids were initially selected for retention of the mouse translocation chromosome carrying the Hprt gene. Subsequently, hybrid clones were counterselected to force segregation of this chromosome. Selected and counterselected hybrid clones were analyzed for their chromosome content by trypsin/Giemsa banding and for expression of the mouse forms of the X-linked enzymes HPRT and alpha-galactosidase (GALA) by isoelectric focusing. The results indicate that the breakpoint on the mouse X chromosome (in band XD) has separated the genes for HPRT (Hprt) and for GALA (Ags). Hprt is proximal to the breakpoint in region Xcen-XD and Ags is distal in region XD-Xter. The gene order in the mouse (centromere-Hprt-Ags) is therefore inverted when compared to the order of the homologous loci on the long arm of the human X (centromere-GALA-HPRT).     

Assignment of the gene for cytoplasmic superoxide dismutase (Sod-1) to a region of chromosome 16 and of Hprt to a region of the X chromosome in the mouse

In the search for homologous chromosome regions in man and mouse, the locus for cytoplasmic superoxide dismutase (SOD-1; superoxide:superoxide oxidoreductase, EC 1.15.1.1) is of particular interest. In man, the SOD-1 gene occupies the same subregion of chromosome 21 that causes Down syndrome when present in triplicate. Although not obviously implicated in the pathogenesis, SOD-1 is considered to be a biochemical marker for this aneuploid condition. Using a set of 29 mouse-Chinese hamster somatic cell hybrids, we assign Sod-1 to mouse chromosome 16. Isoelectric focusing permits distinction between mouse and Chinese hamster isozymes, and trypsin/Giemsa banding distinguishes mouse from Chinese hamster chromosomes. The mouse fibroblasts used were derived from a male mouse carrying Searle's T(X;16)16H reciprocal translocation in which chromosomes X and 16 have exchanged parts. Analysis of informative hybrids leads to regional assignment of Sod-1 to the distal half of mouse chromosome 16 (16B4 --> ter). Because the Chinese hamster cell line (380) used for cell hybridization is deficient in hypoxanthine phosphoribosyltransferase (HPRT; IMP: pyrophosphate phosphoribosyltransferase, EC 2.4.2.8), that part of the mouse X chromosome carrying the complementing Hprt gene can be identified by selection in hypoxanthine/aminopterin/thymidine medium and counterselection in 8-azaguanine. Mouse Hprt is on the X(T) translocation product containing the proximal region X cen --> XD.     

Differential methylation of hypoxanthine phosphoribosyltransferase genes on active and inactive human X chromosomes.

Previous theoretical considerations and some experimental data have suggested a role for DNA methylation in the maintenance of mammalian X chromosome inactivation. The isolation of specific X-encoded sequences makes it possible to investigate this hypothesis directly. We have used cloned fragments of the human hypoxanthine phosphoribosyltransferase (HPRT) gene and methylation-sensitive restriction enzymes to study methylation patterns in genomic DNA of individuals with different numbers of X chromosomes and in somatic cell hybrid lines containing human X chromosomes that are either active or inactive or have been reactivated by treatment with 5-azacytidine. The results of these analyses show that there is hypomethylation of active X chromosomes relative to inactive X chromosomes in the 5' region of this gene. In the middle region of the gene, however, a site that is consistently undermethylated on inactive X chromosomes was identified. Taken together, the data suggest that the overall pattern of methylation, rather than methylation of specific sites, plays a role in the maintenance of X chromosome inactivation.     

Characterization of a cloned DNA sequence that is present at centromeres of all human autosomes and the X chromosome and shows polymorphic variation.

We have identified a human DNA recombinant (p308) with a 3.0-kilobase (kb) BamHI insert that hybridizes in situ exclusively to the centromeric region of all human autosomes and the X chromosome. This highly repetitive sequence is significantly enriched on several chromosomes, most prominently on chromosome 6. In all individuals, the majority of genomic repeats are organized as tandem 3.0-kb BamHI repeats, each containing one Taq I site; the others are organized into BamHI and Taq I repeats of variable size that have some chromosome specificity. Using mouse-human hybrids, we have defined the specific organization of this sequence on chromosomes 6, 3, and X. In some individuals, there are differences in the number and nature of the tandem repeats. These polymorphisms segregate in families as if chromosome specific. Although variable from one chromosome to another, 308 contains sequences homologous to DNA present in centric heterochromatin of essentially all human chromosomes and is evolutionarily conserved. Therefore, a significant component of pericentric DNA is similar for all human chromosomes.