Somatic Cell Hybrid Analysis Research Articles

Specific Detection of CD56 (NCAM) Isoforms for the Identification of Aggressive Malignant Neoplasms with Progressive Development

Alternative splicing of transcripts from many cancer-associated genes is believed to play a major role in carcinogenesis as well as in tumor progression. Alternative splicing of one such gene, the neural cell adhesion molecule CD56 (NCAM), impacts the progression, inadequate therapeutic response, and reduced total survival of patients who suffer from numerous malignant neoplasms. Although previous investigations have determined that CD56 exists in three major isoforms (CD56(120kD), CD56(140kD), and CD56(180kD)) with individual structural and functional properties, neither the expression profiles nor the functional relevance of these isoforms in malignant tumors have been consistently investigated. Using new quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) strategies and novel CD56 isoform-specific antibodies, CD56(140kD) was shown to be exclusively expressed in a number of highly malignant CD56(+) neoplasms and was associated with the progression of CD56(+) precursor lesions of unclear malignant potential. Moreover, only CD56(140kD) induced antiapoptotic/proliferative pathways and specifically phosphorylated calcium-dependent kinases that are relevant for tumorigenesis. We conclude, therefore, that the specific detection of CD56 isoforms will help to elucidate their individual functions in the pathogenesis and progression of malignant neoplasms and may have a positive impact on the development of CD56-based immunotherapeutic strategies.

RNA interference-based strategies directed against Bcl-xL and MCL1 for the treatment of malignant pleural mesothelioma

The MCL1 gene, recently identified in a myeloid leukemia cell line, has sequence similarity to BCL2, the gene at the t(14;18) translocation in follicular lymphoma. The chromosomal location of MCL1 has now been determined. The human locus (MCL1) was mapped to the long arm of human chromosome 1q21, using the methods of in situ hybridization and somatic cell hybrid analysis. In the mouse, MCL1-related sequences were mapped to positions on two mouse chromosomes (chromosomes 3 and 5), using haplotype analysis of an interspecific cross. The location of the locus on mouse chromosome 3 (Mcl1) was homologous to that of MCL1 on human chromosome 1; the second locus (Mcl-rs on mouse chromosome 5) may represent a pseudogene. The proximal long arm of human chromosome 1, where MCL1 is located, is duplicated and/or rearranged in a variety of preneoplastic and neoplastic diseases, including hematologic diseases and solid tumors. MCL1 is thus a candidate gene for involvement in cancer.

Two novel translocation breakpoints upstream of SOX9 define borders of the proximal and distal breakpoint cluster region in campomelic dysplasia

The semilethal skeletal malformation syndrome campomelic dysplasia (CD) with or without XY sex reversal is caused by mutations within the SOX9 gene on 17q24.3 or by chromosomal aberrations (translocations, inversions or deletions) with breakpoints outside the SOX9 coding region. The previously published CD translocation breakpoints upstream of SOX9 fall into two clusters: a proximal cluster with breakpoints between 50-300 kb and a distal cluster with breakpoints between 899-932 kb. Here, we present clinical, cytogenetic and molecular data from two novel CD translocation cases. Case 1 with karyotype 46,XY,t(1;17)(q42.1;q24.3) has characteristic symptoms of CD, including mild tibial bowing, cryptorchidism and hypospadias. By standard fluorescence in situ hybridization (FISH) and by high-resolution fiber FISH, the 17q breakpoint was mapped 375 kb from SOX9, defining the centromeric border of the proximal breakpoint cluster region. Case 2 with karyotype 46,X,t(Y;17)(q11.2;q24.3) has the acampomelic form of CD and complete XY sex reversal. By FISH and somatic cell hybrid analysis, the 17q breakpoint was mapped 789 kb from SOX9, defining the telomeric border of the distal breakpoint cluster region. We discuss the structure of the 1 Mb cis-control region upstream of SOX9 and the correlation between the position of the 14 mapped translocation breakpoints with respect to disease severity and XY sex reversal.

An Entirely Novel Form of α Thalassemia in Patients from the South Pacific Linked to Chromosome 16.

Although a great deal is known about the molecular mechanism underlying α thalassemia (thal), which is characterized by reduced or absent α-globin expression, the molecular basis of α thalassemia in Pacific Islanders is not fully understood. Previously single α gene deletion alleles (−α3.7 and −α4.2) were the only types of α thalassemia identified in this population, however these genotypes do not account for several cases of Hb H disease in the Melanesian Islanders of Papua New Guinea (PNG) and Vanuatu. To date, no mutations have been found in the α globin gene cluster of these individuals. To determine the cause of α thalassemia in this population, we studied 35 individuals from three unrelated families of Pacific Islanders. 16 of them had classical Hb H disease characterized by mild to moderate anemia (Hb; 10.3 ± 2.3 g/dL), microcytosis (MCV; 66 ± 3.5 fL), hypochromia (MCH; 19.8 ± 1.6 pg), positive Hb H by electrophoresis (5–18%) and numerous Hb H inclusion bodies in peripheral blood. Gene mapping revealed that six had normal genotypes (αα/αα) while the rest were heterozygotes for α+ thalassemia (−α/αα). Preliminary sequence analyses demonstrated that the α globin genes and their regulatory elements were entirely normal casting some doubt on whether this novel form of α thal is genuinely linked to chromosome 16. Using a combination of a long range haplotype analysis by MALDI-TOF to simultaneously genotype 80 SNPs covering 350 kb of the telomeric end of chromosome 16 (16p; 13.3) and several tandem repeat markers (VNTRs and HVRs), we demonstrated that the α thal phenotype is strongly associated with a single haplotype (IIIa;PNG) found in these cases. Consistent with the Hb H patients, individuals from these families and another 15 unrelated individuals (4.7%) identified in a population survey (n = 315) who were hypochromic (MCH < 25 pg), despite having normal α genes (αα/αα) and iron status, were also found to carry the same type of chromosome 16 linked to the affected haplotype. These findings indicate that this form of α thalassemia is unequivocally linked to chromosome 16.Subsequent linkage disequilibrium analysis (LD) in these cases limited the chromosomal interval containing the mutation to within 169 kb from the 16p-telomere. Extensive genomic mapping using more than 10 restriction enzymes and 37 plasmid and PCR probes together with Pulse Field Gel Electrophoresis (PFGE), covering 400 kb of 16p, excluded small deletions, insertions, rearrangements and chromosomal truncations in the affected segment of chromosome 16. Direct genomic sequencing of collectively 40 kb of the α globin genes, all known multispecies conserved sequences (MCS-R, 1–4) demonstrated no apparent mutation. Interestingly, interspecific somatic cell hybrid analysis of the affected chromosome 16 in a mouse erythroid leukemia (MEL) background did not recapitulate the phenotype of α thalassemia. These results suggest a novel mechanism, either at the genetic or epigenetic level, underlying the reduced expression of the α globin genes in the Pacific Islanders and understanding such mechanism might provide more insights on the complexity of stage- and tissue-specific gene regulation in human erythroid environment.

Molecular analysis of three gain-of-function CHO mutants that add the bisecting GlcNAc to N-glycans.

LEC10 Chinese hamster ovary (CHO) cells are gain-of-function mutants that express N-acetylglucosaminyltransferase III (GlcNAc-TIII), the glycosyltransferase that adds the bisecting GlcNAc to complex N-glycans. LEC10 cells are useful for glycosylation engineering of recombinant glycoproteins, including antibody therapeutics, for defining lectin recognition specificities and for determining biological functions of the bisecting GlcNAc. We show that three CHO mutants, LEC10, LEC10A, and LEC10B, arose due to transcriptional activation of the quiescent CHO Mgat3 gene. They each express Mgat3 gene transcripts of approximately 4.7 kb at different levels (LEC10B > LEC10 > LEC10A). Southern analyses gave a single band in LEC10, LEC10A, and parent CHO DNA with four restriction enzymes but an additional band with three of them in LEC10B genomic DNA, indicative of a duplication event in LEC10B. The deduced amino acid sequence of the Mgat3 gene expressed in each CHO mutant and in parent CHO genomic DNA is identical. However, 5' UTR sequences differ with LEC10 and LEC10B containing a 5' UTR segment of the Atf4 gene downstream of the Mgat3 gene in human and mouse. Somatic cell hybrid analysis indicated that the LEC10B Mgat3 gene was induced by a cis mechanism. LEC10B glycoproteins bound more erythroagglutinin lectin (E-PHA) than LEC10 glycoproteins and MALDI-TOF MS revealed a broad spectrum of complex, bisected N-glycans expressed by the LEC10B mutant. LEC10B is therefore the cell line of choice for producing recombinant glycoproteins carrying bisected N-glycans and for investigating biological functions of the bisecting GlcNAc.

Frequent translocations occur between low copy repeats on chromosome 22q11.2 (LCR22s) and telomeric bands of partner chromosomes.

The chromosome 22q11.2 region is susceptible to rearrangements, mediated by low copy repeats (LCR22s). Deletions and duplications are mediated by homologous recombination events between LCR22s. The recurrent balanced constitutional translocation t(11;22)(q23;q11) breakpoint occurs in an LCR22 and is mediated by double strand breaks in AT-rich palindromes on both chromosomes 11 and 22. Recently, two cases of a t(17;22)(q11;q11) were reported, mediated by a similar mechanism (21). Except for these constitutional translocations, the molecular basis for non-recurrent, reciprocal 22q11.2 translocations is not known. To determine whether there are specific mechanisms that could mediate translocations, we analyzed cell lines derived from 14 different individuals by genotyping and FISH mapping. Somatic cell hybrid analysis was carried out for four cell lines. In five cell lines, the translocation breakpoints occurred in the same LCR22 as for the t(11;22) translocation, suggesting that similar molecular mechanisms are responsible. An additional three occurred in other LCR22s, and six were in non-LCR22 regions, mostly in the proximal half of the 22q11.2 region. The translocation breakpoints on the partner chromosomes were all located in the telomeric bands, proximal to the most telomeric unique sequence probe, in eight cell lines and distal to those loci in six. Therefore, several of the breakpoints were found to occur in the vicinity of highly dynamic regions of the genome, 22q11.2 and telomeric bands. We hypothesize that these regions are more susceptible to breakage and repair, resulting in translocations.

Application of spectral karyotyping to the analysis of the human chromosome complement of interspecies somatic cell hybrids

Mouse–human somatic cell hybrids have been extensively used in the molecular genetic dissection of human disease-related chromosome rearrangements because of their ability to selectively and randomly eliminate human chromosomes. This technology allows the isolation of structural chromosome abnormalities, which then allows determination of the precise molecular address of chromosome breakpoints associated with deletions and translocations, down to the nucleotides involved. The main confounding problem with the analysis of somatic cell hybrids is determining the exact chromosome complement unequivocally and quickly. Spectral karyotyping can identify each of the individual human chromosomes in a normal metaphase spread, as well as structural chromosome rearrangements—although, because of potential cross-hybridization between the human probe and mouse DNA sequences during the hybridization reaction, it has not been determined whether the same analysis will selectively identify human chromosomes on a mouse background. We show (to our knowledge, for the first time) that, under modified conditions of chromosomal in situ suppression hybridization, the standard spectral karyotyping probe does not cross-react with mouse chromosomes and can be used to identify subtle structurally rearranged chromosomes in hybrid cells. This analysis allows for the rapid and unequivocal identification of the human chromosome complement in these hybrids, as well as structural chromosome rearrangements that occur between mouse and human chromosomes that might otherwise confound the analysis.

Tratamiento de úlceras vasculares crónicas con equivalentes cutáneos obtenidos mediante ingeniería tisular

Split hand/split foot (SHSF; also known as ectrodactyly) is a human developmental disorder characterized by missing central digits and other distal limb malformations. An association between SHSF and cytogenetically visible rearrangements of chromosome 7 at bands q21-q22 provides compelling evidence for the location of a causative gene at this location, and the locus has been designated SHFD1. In the present study, marker loci were localized to the SHFD1 critical region through the analysis of somatic cell hybrids derived from individuals with SHSF and cytogenetic abnormalities involving the 7q21-q22 region. Combined genetic and physical data suggest that the order of markers in the SHFD1 critical region is cen-D7S492-D7S527-(D7S479-D7S491)-SHFD1-++ +D7S554-D7S518-qter. Dinucleotide repeat polymorphisms at three of these loci were used to test for linkage of SHSF to this region in a large pedigree that demonstrates autosomal dominant SHSF. Evidence against linkage of the SHSF gene to 7q21-q22 was obtained in this pedigree. Therefore, combined molecular and genetic data provide evidence for locus heterogeneity in autosomal dominant SHSF. We propose the name SHSF2 for this second locus.

Assignment of mitochondrial glycerol-3-phosphate acyltransferase (GPAM) gene to bovine chromosome 26 (26q22) by in situ hybridization and confirmation by somatic cell hybrid mapping

Glycerol-3-phosphate acyltransferase catalyzes the first step of glycerolipid biosynthesis. It plays a key role in the regulation of cellular triacylglycerol and phospholipid levels. There are two isoforms of glycerol-3-phosphate acyltransferase in mammals, a mitochondrial and a cytosolic form. The mitochondrial form (GPAM) prefers saturated fatty acyl-CoA as a substrate, whereas the cytosolic enzyme uses both saturated and unsaturated fatty acyl-Co. The GPAM gene has been cloned and sequenced in rat (NM017274) and mouse (NM008149). Human GPAM has been located on chromosome 10 (HSA10q24→q26) and mouse Gpam on MMU19 (Welch et al., 1998) whereas the bovine GPAM gene remains unassigned. We used the mouse Gpam gene sequence to design specific primers in order to partially amplify the bovine gene. We report the localization of the bovine GPAM gene on bovine (BTA) chromosome 26q22 by fluorescence in situ hybridization and somatic cell hybrid analysis. Materials and methods

Isolation of expressed sequence tags of skeletal muscle of neonatal healthy and splay leg piglets and mapping by somatic cell hybrid analysis

We have isolated 14 differentially displayed and 10 further expressed sequence tags (ESTs) from Musculus biceps femoris of newborn healthy and splay leg piglets. By comparison with EMBL/GenBank data we could identify nine porcine homologues to human genes (TATA box binding protein associated factor B TAF1B; B-cell CLL/lymphoma 7B BCL7B; pyruvate dehydrogenase kinase, isoenzyme 4 PDK4; ribosomal protein S10 RPS10; SPARC-like 1 SPARCL1; epithelial protein lost in neoplasm beta EPLIN; N-myc downstream-regulated gene 2 NDRG2; pleiomorphic adenoma gene like 2 PLAGL and, BCL-2 associated transcription factor short form BTFS). Eight fragments correspond to uncharacterized ESTs and 7 ESTs had no significant match with database sequences. These data provide the first expression profiles in skeletal muscle of neonatal piglets and are a basis for candidate gene investigations for congenital splay leg in piglets. Eleven ESTs were physically mapped to porcine chromosomes 1, 3, 4, 5, 7, 8, 9 and 10 and contribute to the comparative map of humans and pigs.

Congenital alpha(2)-plasmin inhibitor deficiencies: a review.

Congenital alpha(2)-plasmin inhibitor deficiencies: a review.

Assignment of six genes to bovine chromosome 13

Six previously unassigned genes have been assigned to the BTA13 by means of somatic cell hybrid analysis. Polymerase chain reaction primer design for the amplification of AHCY, PLTP, PPGB and PCK1 was based on nucleic acids sequence information of homologous genes on HSA20. Primers for PDYN and ASIP were designed from porcine and bovine sequence information, respectively. Homology was established by sequence analysis. These assignments support the previous finding of a conserved syntenic relationship between HSA20 and BTA13 and contribute to the understanding of the chromosomal evolution of BTA13. This is significant since the prion protein gene, thought to play a key rule in the development and course of bovine spongiforme encephalopathy is also located on BTA13.

Development and assignment of bovine-specific PCR systems for the Texas nomenclature marker genes and isolation of homologous BAC probes.

In 1996, Popescu et al. published the Texas standard nomenclature of the bovine karyotype in which 31 marker genes, already mapped in man, were chosen to permit unambiguous identification and numbering of each bovine chromosome. However, specific PCR systems were not available for each marker gene thus preventing the assignment of part of these markers by somatic cell hybrid analysis. In addition, some difficulties remained with the nomenclature of BTA25, BTA27 and BTA29. In this work, specific PCR systems were developed for each of the marker genes except VIL1 (see results), from either existing bovine or human sequences, and a bovine BAC library was screened to obtain the corresponding BAC clones. These PCR systems were used successfully to confirm the assignment of each marker gene (except for LDHA, see results) by analysis on the INRA hamster-bovine somatic cell hybrid panel. The difficulties observed for LDHA and VIL1 are probably due to the fact that these genes belong to large gene families and therefore suggest that they may not be the most appropriate markers for a standardisation effort. This panel of BACs is available to the scientific community and has served as a basis for the establishment of a revised standard nomenclature of bovine chromosomes.

Development and assignment of bovine-specific PCR systems for the Texas nomenclature marker genes and isolation of homologous BAC probes

In 1996, Popescu et al. published the Texas standard nomenclature of the bovine karyotype in which 31 marker genes, already mapped in man, were chosen to permit unambiguous identification and numbering of each bovine chromosome. However, specific PCR systems were not available for each marker gene thus preventing the assignment of part of these markers by somatic cell hybrid analysis. In addition, some difficulties remained with the nomenclature of BTA25, BTA27 and BTA29. In this work, specific PCR systems were developed for each of the marker genes except VIL1 (see results), from either existing bovine or human sequences, and a bovine BAC library was screened to obtain the corresponding BAC clones. These PCR systems were used successfully to confirm the assignment of each marker gene (except for LDHA, see results) by analysis on the INRA hamster-bovine somatic cell hybrid panel. The difficulties observed for LDHA and VIL1 are probably due to the fact that these genes belong to large gene families and therefore suggest that they may not be the most appropriate markers for a standardisation effort. This panel of BACs is available to the scientific community and has served as a basis for the establishment of a revised standard nomenclature of bovine chromosomes.

Assignment of T cell receptor (TCR) alpha-chain gene (A), beta-chain gene (B), gamma-chain gene (G), and delta-chain gene (D) loci on swine chromosomes by in situ hybridization and radiation hybrid mapping

Using the cDNA sequence of porcine T-cell receptor (TCR) α-, β-, γ-, and δ-chain genes, we screened a porcine BAC library to isolate clones containing these genes. The isolated BAC clones were confirmed to carry these TCR genes by partial nucleotide sequencing. Among the clones obtained in the present screening, two clones carried both the TCR α-chain gene (TRA) and the TCR δ-chain gene (TRD) while one clone each carried only the sequence of either TRA or TRD. This observation demonstrated that TRA and TRD are localized in close proximity on a swine chromosome. Also two clones contained the sequence of the TCR β-chain gene (TRB), and two clones contained the sequence of TCR γ-chain gene (TRG). Fluorescence in situ hybridization using the above BAC clones as probes revealed that TRA and TRD, TRB, and TRG loci reside on swine chromosomes 7q15.3→q21, 18q11.3→q12, and 9q21→q22, respectively. The chromosome positions of TRA and TRB are consistent with those determined by somatic cell hybrid analysis (Rettenberger et al., 1996). In addition, RH mapping of these genes was performed using the INRA-University of Minnesota RH panel DNA. The result confirmed the position of TRA and TRB reported earlier (http://imprh.toulouse.inra.fr/), and further demonstrated that TRG was located 11 cR away from genetic marker SW989 toward the marker S0019.

Assignment1 of TACSTD1 (alias TROP1, M4S1) to human chromosome 2p21 and refinement of mapping of TACSTD2 (alias TROP2, M1S1) to human chromosome 1p32 by in situ hybridization

TACSTD1 (alias TROP1, M4S1; OMIM 600718) and TACSTD2 (alias TROP2, M1S1; OMIM 137290) encode cellsurface glycoproteins expressed in epithelial cells. Trops are calcium signal transducers (Ripani et al., 1998), are overexpressed by human carcinomas and play a role in the control of tumor cell growth (manuscript in preparation). TACSTD2 (originally called M1S1; Fornaro et al., 1995) is responsible for the gelatinous drop-like corneal dystrophy (OMIM 204870) (Tsujikava et al., 1999). Although placed within a YAC/BAC/ cosmid contig on 1p (Tsujikava et al., 1999), its chromosome location still is undecided at 1p32→p31 or 1p13→q12 (Linnenbach et al., 1993; UniGene Hs.23582). By fluorescence in situ hybridization (FISH) we assign TACSTD2 to 1p32. The growth-suppressing properties of TACSTD2 and the frequent deletion of 1p32 in human tumors (White et al., 1999) suggest that this gene is a candidate to be a novel oncosuppressor. TACSTD1 (formerly called M4S1) shares a 49%-similar cDNA sequence with TACSTD2. TACSTD1 was localized to 4q by human/rodent somatic cell hybrid analyses (Linnenbach et al., 1993) (UniGene Hs.692). We reassign TACSTD1 to human chromosome 2p21, a region which contains the MSH2 and MSH6 DNA mismatch repair genes (Papadopoulos et al., 1995). In agreement with the HUGO Gene Nomenclature Committee, the M4S1 and M1S1 genes have been renamed TACSTD (tumor-associated calcium signal transducer) 1 and 2, respectively.

Mapping of 13 horse genes by fluorescence in-situ hybridization (FISH) and somatic cell hybrid analysis.

We report fluorescence in-situ hybridization (FISH) and somatic cell hybrid mapping data for 13 different horse genes (ANP, CD2, CLU, CRISP3, CYP17, FGG, IL1RN, IL10, MMP13, PRM1, PTGS2, TNFA and TP53). Primers for PCR amplification of intronic or untranslated regions were designed from horse-specific DNA or mRNA sequences in GenBank. Two different horse bacterial artificial chromosome (BAC) libraries were screened with PCR for clones containing these 13 Type I loci, nine of which were found in the libraries. BAC clones were used as probes in dual colour FISH to confirm their precise chromosomal origin. The remaining four genes were mapped in a somatic cell hybrid panel. All chromosomal assignments except one were in agreement with human-horse ZOO-FISH data and revealed new and more detailed information on the equine comparative map. CLU was mapped by synteny to ECA2 while human-horse ZOO-FISH data predicted that CLU would be located on ECA9. The assignment of IL1RN permitted analysis of gene order conservation between HSA2 and ECA15, which identified that an event of inversion had occurred during the evolution of these two homologous chromosomes.

Mapping of 14 expressed sequence tags (ESTs) from porcine skeletal muscle by somatic cell hybrid analysis.

Chromosomal assignments are reported for fourteen porcine expressed sequence tags (ESTs)--CALM1, CRYAB, MYH7, MYL1, PDK4, PGAM2, PYGM, REV3L, RFC1, SLN, SPTBN1, SRM160, TPM1 and YWHAG. The ESTs were derived from our porcine skeletal muscle cDNA library. The ESTs sequences selected for mapping included the presence of the 3'-untranslated region. The assignments were performed using two independent somatic cell hybrid panels providing the possibility of confirmation of the results obtained. The observed localizations are compared with the locations predicted from heterologous (human-pig, pig-human) chromosome painting data and knowledge of the map locations of the human homologues. These results add new information to the porcine genome transcript map.

Localization of the human SP3 gene to chromosome 7p14–p15.2. The lack of expression in multiple sclerosis does not reflect abnormal gene organization

Sp3 belongs to a large family of transcription factors that bind GC/T box elements. We have previously demonstrated the deficient expression of Sp3 in peripheral blood mononuclear cells (PBMC) from most patients with multiple sclerosis (MS). In the current study, the Sp3 gene was assigned to chromosome 7 by using somatic cell hybrid mapping and analysis of a chromosome 7-specific cosmid library. The position of Sp3 was refined to 7p14–p15.2 by fluorescence in situ hybridization (FISH). Southern blot and polymerase chain reaction analysis of genomic DNA failed to demonstrate a detectable difference between MS and control PBMC.

Determination of X-Chromosome Inactivation Status Using X-Linked Expressed Polymorphisms Identified by Database Searching

The large number of redundant sequences available in nucleotide databases provides a resource for the identification of polymorphisms. Expressed polymorphisms in X-linked genes can be used to determine the inactivation status of the genes, and polymorphisms in genes that are subject to inactivation can then be used as tools to examine X-chromosome inactivation status in heterozygous females. In this study, we have identified six new X-linked single-nucleotide polymorphisms and determined the inactivation status of these genes by examination of expression patterns in female cells previously demonstrated to have skewed inactivation, as well as by analysis of somatic cell hybrids retaining the inactive human X chromosome. Expression was seen from both alleles in females heterozygous for the RPS4X gene, confirming the previously reported expression from the inactive X chromosome. Expression of only a single allele was seen in females heterozygous for polymorphisms in the BGN, TM4SF2, ATP6S1, VBP1, and PDHA1 genes, suggesting that these genes are subject to X-chromosome inactivation.