PRC2-independent chromatin compaction and transcriptional repression in cancer.

The silencing of large chromosomal regions by epigenetic mechanisms has been reported to occur frequently in cancer. Epigenetic marks, such as histone methylation and acetylation, are altered at these loci. However, the mechanisms of formation of such aberrant gene clusters remain largely unknown. Here, we show that, in cancer cells, the epigenetic remodeling of chromatin into hypoacetylated domains covered with histone H3K27 trimethylation is paralleled by changes in higher-order chromatin structures. Using fluorescence in situ hybridization, we demonstrate that regional epigenetic silencing corresponds to the establishment of compact chromatin domains. We show that gene repression is tightly correlated to the state of chromatin compaction and not to the levels of H3K27me3-its removal through the knockdown of EZH2 does not induce significant gene expression nor chromatin decompaction. Moreover, transcription can occur with intact high-H3K27me3 levels; treatment with histone deacetylase inhibitors can relieve chromatin compaction and gene repression, without altering H3K27me3 levels. Our findings imply that compaction and subsequent repression of large chromatin domains are not direct consequences of PRC2 deregulation in cancer cells. By challenging the role of EZH2 in aberrant gene silencing in cancer, these findings have therapeutical implications, notably for the choice of epigenetic drugs for tumors with multiple regional epigenetic alterations.

Large-scale identification of mammalian proteins localized to nuclear sub-compartments.

Many nuclear components participating in related pathways appear concentrated in specific areas of the mammalian nucleus. The importance of this organization is attested to by the dysfunction that correlates with mis-localization of nuclear proteins in human disease and cancer. Determining the sub-nuclear localization of proteins is therefore important for understanding genome regulation and function, and it also provides clues to function for novel proteins. However, the complexity of proteins in the mammalian nucleus is too large to tackle this on a protein by protein basis. Large-scale approaches to determining protein function and sub-cellular localization are required. We have used a visual gene trap screen to identify more than 100 proteins, many of which are normal, located within compartments of the mouse nucleus. The most common discrete localizations detected are at the nucleolus and the splicing speckles and on chromosomes. Proteins at the nuclear periphery, or in other nuclear foci, have also been identified. Several of the proteins have been implicated in human disease or cancer, e.g. ATRX, HMGI-C, NBS1 and EWS, and the gene-trapped proteins provide a route into further understanding their function. We find that sequence motifs are often shared amongst proteins co-localized within the same sub-nuclear compartment. Conversely, some generally abundant motifs are lacking from the proteins concentrated in specific areas of the nucleus. This suggests that we may be able to predict sub-nuclear localization for proteins in databases based on their sequence.

Unusual chromosome architecture and behaviour at an HSR.

Amplification of sequences within mammalian chromosomes is often accompanied by the formation of homogeneously staining regions (HSRs). The arrangement of DNA sequences within such amplicons has been investigated, but little is known about the chromosome structure or behaviour of these unusual regions. We have analysed the metaphase chromosome structure of the dihydrofolate reductase (DHFR) amplicon of CHOC400 cells. The chromatin in this region contains hyperacetylated nucleosomes yet, at the same time, appears to be densely packed like heterochromatin. The region does not bind heterochromatin proteins. We show that the dense packing of the region is restricted to DNA located close to the chromosome core/scaffold. In contrast, levels of the chromosome scaffold protein topoisomerase II at HSRs are the same as those found at other euchromatic locations. Metaphase chromosome condensation of the HSR is shown to be sensitive to topoisomerase II inhibitors, and sister chromatids often appear to remain attached within the HSRs at metaphase. We suggest that these features underlie anaphase bridging and the aberrant interphase structure of the HSR. The DHFR amplicon is widely used as a model system to study mammalian DNA replication. We conclude that the higher-order chromosome structure of this amplicon is unusual and suggest that caution needs to be exercised in extrapolating data from HSRs to normal chromosomal loci.

HuCHRAC, a human ISWI chromatin remodelling complex contains hACF1 and two novel histone-fold proteins.

Chromatin remodelling complexes containing the nucleosome-dependent ATPase ISWI were first isolated from Drosophila embryos (NURF, CHRAC and ACF). ISWI was the only common component reported in these complexes. Our purification of human CHRAC (HuCHRAC) shows that ISWI chromatin remodelling complexes can have a conserved subunit composition in completely different cell types, suggesting a conserved function of ISWI. We show that the human homologues of two novel putative histone-fold proteins in Drosophila CHRAC are present in HuCHRAC. The two human histone-fold proteins form a stable complex that binds naked DNA but not nucleosomes. HuCHRAC also contains human ACF1 (hACF1), the homologue of Acf1, a subunit of Drosophila ACF. The N-terminus of mouse ACF1 was reported as a heterochromatin-targeting domain. hACF1 is a member of a family of proteins with a related domain structure that all may target heterochromatin. We discuss a possible function for HuCHRAC in heterochromatin dynamics. HuCHRAC does not contain topoisomerase II, which was reported earlier as a subunit of Drosophila CHRAC.

The reticulocalbin gene maps to the WAGR region in human and to the Small eye Harwell deletion in mouse.

We describe the localization of the gene encoding reticulocalbin, a Ca2+-binding protein of the endoplasmic reticulum, on human chromosome 11p13 midway between the WT1 and the PAX6 genes and show that it is hemizygously deleted in WAGR individuals. The mouse reticulocalbin gene is also shown to map to the region of conserved synteny on mouse chromosome 2 and to be deleted in the Small eye Harwell (SeyH) mutation. Loss of the reticulocalbin gene could contribute to the early lethality of SeyH and SeyDey homozygotes.

Factors affecting the timing and imprinting of replication on a mammalian chromosome.

Fluorescence in situ hybridisation has been used to follow replication of the short arm of human chromosome 11 using chromosome anomalies to distinguish the maternally-and paternally-derived homologues. The temporal difference in replication timing within and between chromosomes has been estimated by combining S phase detection with dual colour fluorescence in situ hybridisation. Proximal regions of 11p, including the WT1 gene, tend to replicate earlier on the maternally-derived chromosome than on the paternally-derived homologue. More distal parts of 11p (including the IGF2 gene) have the opposite imprint. The average difference in replication timing between homologous loci in the population of cells is small compared to the differences between loci along a single chromosome. The imprint is not strictly adhered to since many nuclei have hybridisation patterns opposite to the trend within the population. The nature of the imprinting signal has been investigated. Absolute replication time, but not the imprint, was affected by azacytidine, an inhibitor of DNA methylation. The replication imprint was modified by treatments that inhibit histone deacetylation. We suggest that replication imprinting reflects differences in chromatin structure between homologues.

Identification of MIC 11 antigen as an epitope of the CD59 molecule.

The MIC 11 antigen is expressed on human cells and is characterized by reaction with a monoclonal antibody (mAb), 16.3A5. The gene controlling MIC 11 was recently mapped to the p13 region of chromosome 11 within 500 kb of the gene encoding CD59, a complement regulatory protein. The present report investigates the antigenic relationship between these cell-membrane determinants and sets out evidence that MIC 11 and CD59 are encoded by the same gene. Western blotting of human erythrocyte membrane proteins and purified membrane CD59 showed that 16.3A5 anti-MIC 11 antibody bound to a 19-24,000 MW band with the characteristic appearance of CD59 protein, and gave staining patterns identical to those obtained with the CD59 antibody, BRIC 229. The binding of 16.3A5 monoclonal IgG to purified urine-derived CD59 in enzyme-linked immunosorbent assay (ELISA) was inhibited by YTH 53.1 rat CD59 antibody, indicating that the MIC 11 epitope is the same as, or close to, that recognized by CD59 antibodies such as YTH 53.1, BRIC 229 and 2/24. Prior exposure of erythrocytes to 16.3A5 anti-MIC 11 also reduced the ability of the CD59 antibodies, BRIC 229 and YTH 53.1, to block the complement-inhibiting function of membrane CD59. Anti-MIC 11 antibody alone, however, had no inhibitory effect on CD59 function. This may be due to its relatively low binding affinity or to some slight difference in epitope specificity. Further studies using immunofluorescence showed that the MIC 11 epitope, like CD59, is absent from EBV-B cells lacking GPI-anchored proteins and from a B-cell line specifically deficient in CD59 protein. Overall, the results provide strong evidence that MIC 11 is a determinant on the CD59 molecule.

The expression of the Wilms' tumour gene, WT1, in the developing mammalian embryo.

In the developing mouse, the Wilms' tumour gene, WT1, is first expressed in the intermediate mesenchyme lateral to the coelomic cavity (13 somite, early 9 dpc embryo). A few hours later, it is present around all the cavity and in the urogenital ridge (the earliest mesonephric tubules) and the differentiating heart mesothelium. By 11 dpc, expression is in the uninduced metanephric mesenchyme and in the presumptive motor neurons of the spinal cord. By 12.5 dpc, WT1 expression has increased in the induced mesenchyme of the kidney and a day later is particularly marked in the nephrogenic condensations. At 13.5 dpc, WT1 is briefly expressed in some differentiating body-wall musculature, while two days later, there is a small domain of expression in the roof of the fourth ventricle of the brain. By day 20, however, expression has become restricted to the kidney glomeruli. RNA-PCR analysis on 12.5 dpc embryos and on adult tissues shows that WT1 is weakly expressed in both eye and tongue. The expression pattern in human embryos (28-70 days) is very similar to that in the equivalent mouse stages (10-15 dpc). The results indicate that WT1 is mainly present in mesodermally derived tissues, although exceptions are ectodermally derived spinal cord and brain. The data indicate that WT1 plays a role in mediating some cases of the mesenchyme-to-epithelial transition, but its expression elsewhere argues that it has other tissue-specific roles in development.

Submicroscopic deletions at the WAGR locus, revealed by nonradioactive in situ hybridization.

Fluorescence in situ hybridization (FISH) with biotin-labeled probes mapping to 11p13 has been used for the molecular analysis of deletions of the WAGR (Wilms tumor, aniridia, genitourinary abnormalities, and mental retardation) locus. We have detected a submicroscopic 11p13 deletion in a child with inherited aniridia who subsequently presented with Wilms tumor in a horseshoe kidney, only revealed at surgery. The mother, who has aniridia, was also found to carry a deletion including both the aniridia candidate gene (AN2) and the Wilms tumor predisposition gene (WT1). This is therefore a rare case of an inherited WAGR deletion. Wilms tumor has so far only been associated with sporadic de novo aniridia cases. We have shown that a cosmid probe for a candidate aniridia gene, homologous to the mouse Pax-6 gene, is deleted in cell lines from aniridia patients with previously characterized deletions at 11p13, while another cosmid marker mapping between two aniridia-associated translocation breakpoints (and hence a second candidate marker) is present on both chromosomes. These results support the Pax-6 homologue as a strong candidate for the AN2 gene. FISH with cosmid probes has proved to be a fast and reliable technique for the molecular analysis of deletions. It can be used with limited amounts of material and has strong potential for clinical applications.

Modulation of DNA binding specificity by alternative splicing of the Wilms tumor wt1 gene transcript.

The technique of whole-genome polymerase chain reaction was used to study the DNA binding properties of the product of the wt1 gene. The zinc finger region of this gene is alternatively spliced such that the major transcript encodes a protein with three extra amino acids between the third and fourth fingers. The minor form of the protein binds specifically to DNA. It is now shown that the major form of wt1 messenger RNA encodes a protein that binds to DNA with a specificity that differs from that of the minor form. Therefore, alternative splicing within the DNA binding domain of a transcription factor can generate proteins with distinct DNA binding specificities and probably different physiological targets.

Role for the Wilms tumor gene in genital development?

Detailed molecular definition of the WAGR region at chromosome 11p13 has been achieved by chromosome breakpoint analysis and long-range restriction mapping. Here we describe the molecular detection of a cytogenetically invisible 1-megabase deletion in an individual with aniridia, cryptorchidism, and hypospadias but no Wilms tumor (WT). The region of overlap between this deletion and one associated with WT and similar genital anomalies but no aniridia covers a region of 350-400 kilobases, which is coincident with the extent of homozygous deletion detected in tumor tissue from a sporadic WT. A candidate WT gene located within this region has recently been isolated, suggesting nonpenetrance for tumor expression in the first individual. The inclusion within the overlap region of a gene for WT predisposition and a gene for the best-documented WT-associated genitourinary malformations leads us to suggest that both of these anomalies result from a loss-of-function mutation at the same locus. This in turn implies that the WT gene exerts pleiotropic effect on both kidney and genitourinary development, a possibility supported by the observed expression pattern of the WT candidate gene in developing kidney and gonads.

Aniridia, Wilms' tumor and human chromosome 11.

Aniridia-a developmental abnormality of the eye in which the iris is apparently absent-has been shown to be genetically associated with Wilms' tumor (an embryonic nephroblastoma) in the WAGR syndrome. Genetic and cytogenetic evidence points to band p13 of human chromosome 11 as the localization of the genes responsible for these defects. Deleted chromosomes 11 from WAGR patients and clinically associated translocations involving 11p13 have been used to map and order genes and anonymous DNA markers around the WAGR locus refining the localization of the aniridia and Wilms' tumor genes to within about 1 million base pairs of DNA.

CpG islands surround a DNA segment located between translocation breakpoints associated with genitourinary dysplasia and aniridia.

We have isolated a DNA segment absent from all the constitutionally deleted chromosomes 11 of our patients with Wilms tumor. This marker separates two balanced translocations that break in band 11p13: the distal one associated with aniridia (AN2), and the proximal one with genitourinary dysplasia (GUD). The GUD breakpoint maps within the smallest region of overlap (SRO) for the Wilms tumor (WT) gene locus, thus strengthening the previous suggestion of an association between Wilms tumor and other abnormalities of the genitourinary system. The 11p13 translocation breakpoint associated with T-cell acute lymphatic leukemia (T-ALL) is centromeric to the SRO and separated from the WT locus by at least one known gene. This region of the human genome (11p13) is rich in CpG islands that potentially identify genes, some of which may be involved in the various phenotypes associated with the WAGR syndrome. This is consistent with the proposition that the majority of human genes are in G-negative bands.

Long-range structure of H-ras 1-selected transgenomes.

We have used chromosome-mediated gene transfer (CMGT) and whole cell fusion to derive human-mouse hybrid cells carrying reduced human chromosomes 11, by selecting for expression of the transforming H-ras 1 oncogene. To realize the full potential of these somatic cell genetic techniques as resources for enriched DNA probe isolation and the fine structure mapping of chromosomes, the nature of any molecular rearrangements that may accompany the process of DNA transfer must be understood. We have analyzed the long-range structure of our transgenomes by pulsed field gel electrophoresis (PFGE) and show here that, whereas during cell fusion several megabase pairs (Mb) of DNA can be transferred intact, multiple rearrangements of DNA accompany CMGT even in transgenomes where other methods of analysis gave no indication of such molecular scrambling.