Cytological, flow cytometric, and molecular analysis of the rapid evolution of mammalian chromosomes containing highly amplified DNA sequences

Transfection of a mouse dihydrofolate reductase (DHFR) cDNA contained in a plasmid "expression vector" into DHFR deficient Chinese hamster cells, followed by progressive selection of cells in increasing concentrations of methotrexate (MTX), leads to marked amplification of the exogenous DHFR sequences in the recipient hamster cells. This gene amplification is evident at the cytological level, in the form of homogeneously staining chromosomal regions (HSRs), at a gene expression level, in the form of fluorescein-methotrexate binding, and at the DNA level. Flow sorting, based on variable fluorescein-MTX binding, or direct cellular cloning, followed by chromosome analysis, revealed intercellular heterogeneity of HSRs in size and distribution. This suggested that there was a rapid evolution of HSRs in MTX-resistant transfectants. Chromosomal analysis of HSR evolution in situ, by examining individual colonies presumably derived from one or a few cells, underscored this impression of chromosome structural fluidity. Rates of HSR change in excess of 0.01 per cell division, increased by low doses of the recombinogen, mitomycin C, were detected. The Chinese hamster DHFR transfectants described should be amenable to detailed, coordinate cytological and molecular characterization. Such an analysis should contribute to an understanding of processes such as homologous recombination in mediating HSR evolution in mammalian chromosomes.     

Cytological,flow cytometric,and molecular analysis of the rapid evolution of mammalian chromosomes containing highly amplified DNA sequences

Summary Transfection of a mouse dihydrofolate reductase (DHFR) cDNA contained in a plasmid expression vector into DHFR deficient Chinese hamster cells, followed by progressive selection of cells in increasing concentrations of methotrexate (MTX), leads to marked amplification of the exogenous DHFR sequences in the recipient hamster cells. The gene amplification is evident at the cytological level, in the form of homogeneously staining chromosomal regions (HRSs), at a gene expression level, in the form of fluorescein-methotrexate binding, and at the DNA level. Flow sorting, based on variable fluorescein-MTX binding, or direct cellular cloning, followed by chromosome analysis, revealed intercellular heterogeneity of HSRs in size and distribution.This suggested that there was a rapid evolution of HSRs an MTX-resistant transfectants. Chromosomal analysis of HSR evolution in situ, by examining individual colonies presumably derived from one or a few cells, underscored this impression of chromosome structural fluidity. Rates of HSR change in excess of 0.01 per cell division, increased by low doses of the recombinogen, mitomycin C were detected. The Chinese hamster DHFR transfectants described should be amenable to detailed, coordinate cytological and molecular characterization. Such an analysis should contribute to an understanding of processes such as homologous recombination in mediating HSR evolution in mammalian chromosomes.     

Polymorphic HSRs in chromosome 1 of the two semispecies Mus musculus musculus and M. m. domesticus have a common origin in an ancestral population

HSRs (homogeneously staining regions) are the cytological correlates of DNA amplification. In the house mouse, Mus musculus, many populations are polymorphic for the presence or absence of HSRs on chromosome 1. In the semispecies M. m. domesticus the amplified DNA is present within one HSR, whereas in M. m. musculus chromosomes 1 with two HSRs are found. Hybridization of HSR-specific probes to Southern blots of HSR-carrying genomic DNAs from different localities and semispecies revealed similar complex band patterns. the remaining variation is restricted to sequences with a low degree of amplification. Variation is higher between semispecies than within one semispecies. It is assumed that HSRs are derived from one original amplification event and that unequal recombination is the mechanism underlying the length variation of HSRs present today in both semispecies. Evidence from G-banding and in situ hybridization shows that the two HSRs of M. m. musculus originated from a single HSR by means of a paracentric inversion, where one break-point was located within the single HSR and the second outside the HSR. As a consequence of the paracentric inversion the two HSRs of M. m. musculus are permanently linked together. Since exchange of genes between the two semispecies is restricted to a narrow hybrid zone the amplification that gave rise to the HSR most probably occurred prior to the divergence into the semispecies M. m. domesticus and M. m. musculus about 1 million years ago.by D. Schweizer     

Identifying a single-copy DNA sequence associated with the expression of a heterochromatic gene, the light locus of Drosophila melanogaster

Although little is known about the molecular organization of most genes within heterochromatin, the unusual properties of these chromosomal regions suggest that genes therein may be organized and expressed very differently from those in euchromatin. We report here the cloning, by P transposon tagging, of sequences associated with the expression of the light locus, an essential gene found in the heterochromatin of chromosome 2 of Drosophila melanogaster. We conclude that this DNA is either a segment of the light locus, or a closely linked, heterochromatic sequence affecting its expression. While other functional DNA sequences previously described in heterochromatin have been repetitive, light gene function may be associated, at least in part, with single-copy DNA. This conclusion is based upon analysis of DNA from mutations and reversions induced by P transposable elements. The cloned region is unusual in that this single-copy DNA is embedded within middle-repetitive sequences. The in situ hybridization experiments also show that, unlike most other sequences in heterochromatin, this light-associated DNA evidently replicates in polytene chromosomes, but its diffuse hybridization signal may suggest an unusual chromosomal organization.     

Metaphase chromosome structure. Involvement of topoisomerase II

SCI is a prominent, 170,000 Mr, non-histone protein of HeLa metaphase chromosomes. This protein binds DNA and was previously identified as one of the major structural components of the residual scaffold structure obtained by differential protein extraction from isolated chromosomes. The metaphase scaffold maintains chromosomal DNA in an organized, looped conformation. We have prepared a polyclonal antibody against the SC1 protein. Immunolocalization studies by both fluorescence and electron microscopy allowed identification of the scaffold structure in gently expanded chromosomes. The micrographs show an immunopositive reaction going through the kinetochore along a central, axial region that extends the length of each chromatid. Some micrographs of histone-depleted chromosomes provide evidence of the substructural organization of the scaffold; the scaffold appears to consist of an assembly of foci, which in places form a zig-zag or coiled arrangement. We present several lines of evidence that establish the identity of SC1 as topoisomerase II. Considering the enzymic nature of this protein, it is remarkable that it represents 1% to 2% of the total mitotic chromosomal protein. About 60% to 80% of topoisomerase II partitions into the scaffold structure as prepared from isolated chromosomes, and we find approximately three copies per average 70,000-base loop. This supports the proposed structural role of the scaffold in the organization of the mitotic chromosome. The dual enzymic and apparent structural function of topoisomerase II (SC1) and its location at or near the base of chromatin loops allows speculation as to its involvement in the long-range control of chromatin structure.     

Preferential, cooperative binding of DNA topoisomerase II to scaffold-associated regions.

DNA elements termed scaffold-associated regions (SARs) are AT-rich stretches of several hundred base pairs which are known to bind specifically to nuclear or metaphase scaffolds and are proposed to specify the base of chromatin loops. SARs contain sequences homologous to the DNA topoisomerase II cleavage consensus and this enzyme is known to be the major structural component of the mitotic chromosome scaffold. We find that purified topoisomerase II preferentially binds and aggregates SAR-containing DNA. This interaction is highly cooperative and, with increasing concentrations of topoisomerase II, the protein titrates quantitatively first SAR-containing DNA and then non-SAR DNA. About one topoisomerase II dimer is bound per 200 bp of DNA. SARs exhibit a Circe effect; they promote in cis topoisomerase II-mediated double-strand cleavage in SAR-containing DNA fragments. The AT-rich SARs contain several oligo(dA).oligo(dT) tracts which determine their protein-binding specificity. Distamycin, which is known to interact highly selectively with runs of A.T base pairs, abolishes the specific interaction of SARs with topoisomerase II, and the homopolymer oligo(dA).oligo(dT) is, above a critical length of 240 bp, a highly specific artificial SAR. These results support the notion of an involvement of SARs and topoisomerase II in chromosome structure.     

Amplification of a plasmid bearing a mammalian replication initiation region in chromosomal and extrachromosomal contexts

Amplified genes in cancer cells reside on extrachromosomal double minutes (DMs) or chromosomal homogeneously staining regions (HSRs). We used a plasmid bearing a mammalian replication initiation region to model gene amplification. Recombination junctions in the amplified region were comprehensively identified and sequenced. The junctions consisted of truncated direct repeats (type 1) or inverted repeats (type 2) with or without spacing. All of these junctions were frequently detected in HSRs, whereas there were few type 1 or a unique type 2 flanked by a short inverted repeat in DMs. The junction sequences suggested a model in which the inverted repeats were generated by sister chromatid fusion. We were consistently able to detect anaphase chromatin bridges connected by the plasmid repeat, which were severed in the middle during mitosis. De novo HSR generation was observed in live cells, and each HSR was lengthened more rapidly than expected from the classical breakage/fusion/bridge model. Importantly, we found massive DNA synthesis at the broken anaphase bridge during the G1 to S phase, which could explain the rapid lengthening of the HSR. This mechanism may not operate in acentric DMs, where most of the junctions are eliminated and only those junctions produced through stable intermediates remain.     

Destabilization of the adenosine deaminase gene sequences in rat-rat somatic cell hybrids

Rat hepatoma cells amplified for adenosine deaminase (ADA) gene sequences show the amplified DNA on large, homogeneously staining regions (HSRs). The amplified cells are stable in the absence of selection for 12 mo without loss of ADA activity or gene sequences. However, in hybrids formed between an amplified cell line with a prominent HSR and a nonamplified cell line, rapid loss of ADA activity, as well as gene sequences, occurs. Karyotype analyses of the hybrids indicate that the HSR structures are no longer visible in a large percentage of the hybrid metaphase spreads and appear to have been replaced by DNA structures that resemble double minutes. Our data provide evidence that the extent of the breakdown of the HSR in the hybrids may be affected by the presence of an active adenosine kinase or the level of ATP in the cells and additional unidentified factors are present in the hybrids that affect the integrity of the HSR structure. There is no evidence for a specific trans-acting factor in nonamplified cells that regulates gene amplification.     

Chromosome architecture can dictate site-specific initiation of DNA replication in Xenopus egg extracts

Xenopus egg extracts initiate DNA replication specifically at the dihydrofolate reductase (DHFR) origin locus with intact nuclei from late G1-phase CHO cells as a substrate, but at nonspecific sites when purified DNA is assembled by the extract into an embryonic nuclear structure. Here we show that late G1-phase CHO nuclei can be cycled through an in vitro Xenopus egg mitosis, resulting in the assembly of an embryonic nuclear envelope around G1-phase chromatin. Surprisingly, replication within these chimeric nuclei initiated at a novel specific site in the 5' region of the DHFR structural gene that does not function as an origin in cultured CHO cells. Preferential initiation at this unusual site required topoisomerase II-mediated chromosome condensation during mitosis. Nuclear envelope breakdown and reassembly in the absence of chromosome condensation resulted in nonspecific initiation. Introduction of condensed chromosomes from metaphase- arrested CHO cells directly into Xenopus egg extracts was sufficient to elicit assembly of chimeric nuclei and preferential initiation at this same site. These results demonstrate clearly that chromosome architecture can determine the sites of initiation of replication in Xenopus egg extracts, supporting the hypothesis that patterns of initiation in vertebrate cells are established by higher order features of chromosome structure.     

Novel DNA sequences at chromosome 10q26 are amplified in human gastric carcinoma cell lines: molecular cloning by competitive DNA reassociation.

Molecular cloning of genomic sequences altered in cancer cells is believed to lead to the identification of new genes involved in the initiation and progression of the malignant phenotype. DNA amplification is a frequent molecular alteration in tumor cells, and is a mode of proto-oncogene activation. The cytologic manifestation of this phenomenon is the appearance of chromosomal homogeneously staining regions (HSRs) or double minute bodies (DMs). The gastric carcinoma cell line KATO III is characterized by a large HSR on chromosome 11. In-gel renaturation analysis confirmed the amplification of DNA sequences in this cell line, yet none of 42 proto-oncogenes that we tested is amplified in KATO III DNA. We employed the phenol-enhanced reassociation technique (PERT) to isolate 21 random DNA fragments from the amplified domain, and used 6 of them to further clone some 150 kb from that genomic region. While in situ hybridization performed with some of these sequences indicated that in KATO III they are indeed amplified within the HSR on chromosome 11, somatic cell hybrid analysis and in situ hybridization to normal lymphocyte chromosomes showed that they are derived from chromosome 10, band q26. The same sequences were found to be amplified in another gastric carcinoma cell line, SNU-16, which contains DMs, but were not amplified in other 70 cell lines representing a wide variety of human neoplasms. One of these sequences was highly expressed in both KATO III and SNU-16. Thus, the cloned sequences supply a starting point for identification of novel genes which might be involved in the pathogenesis of gastric cancers, and are located in a relatively unexplored domain of the human genome.     

Cloning of DNA from double minutes of Y1 mouse adrenocortical tumor cells: Evidence for gene amplification

We have isolated a metaphase chromosome fraction highly enriched in double minutes (dm) from a mouse adrenocortical tumor cell line (Y1-DM). We have cloned DNA from this dm-enriched fraction in the λ vector Charon 4A, and have characterized two randomly chosen recombinant bacteriophage clones from this dm DNA library. When ³²P-labeled DNA from each recombinant was hybridized to Southern blots of restriction endonuclease-digested DNA from different mouse cell lines, large differences were seen in the intensity of the resulting autoradiographic images, depending on the source of the genomic DNA. A very strong signal was obtained with DNA from the Y1-DM cells and with DNA from a related Y1 subline that lacks dm but contains a marker chromosome bearing a large homogeneously staining region (HSR). Hybridization to DNA from parental inbred mice and from two unrelated mouse cell lines produced a significantly weaker signal than that obtained with DNA from the Y1 cells, but the DNA fragments from these sources were of similar size. Based on results from filter hybridization analysis, we estimate that sequences homologous to the cloned fragments are approximately 100- to 200-fold more abundant in the genome of the Y1-DM cells than in the parental mouse cells. The data are consistent with the hypothesis that dm and HSRs in these cells contain amplified genes.     

Satellite DNA sequences flank amplified DHFR domains in marker chromosomes of mouse fibrosarcoma cells

This study centers on marker chromosomes carrying expanded chromosomal regions which were observed in two independent derivatives of the AA12 murine fibrosarcoma line, the 10^–3 M MTX-res H2 and the 5×10^–7 M MTX-res E. Previous characterization of the marker chromosomes of MTX-res variants showed their common derivation from a marker chromosome (m) of the parental line, endowed with two interstitial C-bands. Cytogenetic evidence pointed to one C-band ofm as the site involved in the chromosomal rearrangements leading to the HSR/ASR chromosomes. ISH of a³H-labeled satellite DNA probe allowed satellite sequences flanking the HSR/ASR in the marker chromosomes, where the C-band was no longer visible, to be detected. FISH experiments using biotinylated DHFR and satellite DNA probes showed that the respective target sequences are contiguous in new marker chromosomes. They also allowed inter- and intrachromosomal rearrangements to be seen at DHFR amplicons and satellite sequences. Double-color FISH using digoxygenated satellite DNA and biotinylated pDHFR7 showed that in a marker chromosome from the H2 cell line the two target sequences are not only adjacent, but closer than 3 Mb, as indicated by overlapping of the different fluorescence signals given by the two probes. Another marker chromosome in the E variant was shown to display a mixed ladder structure consisting of a head-to-head tandem of irregularly-sized satellite DNA blocks, with two symmetrical interspersed DHFR clusters.Abbreviations DHFR dihydrofolate reductase - MTX Methotrexate - HSR Homogeneously Staining Region - ASR Abnormally Staining Region - DM Double Minute - ISH In Situ Hybridization - FISH FluorescenceIn Situ Hybridization     

Studies of He-T DNA sequences in the pericentric regions of Drosophila chromosomes

He-T DNA is a complex set of repeated DNA sequences with sharply defined locations in the polytene chromosomes of Drosophila melanogaster. He-T sequences are found only in the chromocenter and in the terminal (telomere) band on each chromosome arm. Both of these regions appear to be heterochromatic and He-T sequences are never detected in the euchromatic arms of the chromosomes (Young et al. 1983). In the study reported here, in situ hybridization to metaphase chromosomes was used to study the association of He-T DNA with heterochromatic regions that are under-replicated in polytene chromosomes. Although the metaphase Y chromosome appears to be uniformly heterochromatic, He-T DNA hybridization is concentrated in the pericentric region of both normal and deleted Y chromosomes. He-T DNA hybridization is also concentrated in the pericentric regions of the autosomes. Much lower levels of He-T sequences were found in pericentric regions of normal X chromosomes; however compound X chromosomes, constructed by exchanges involving Y chromosomes, had large amounts of He-T DNA, presumably residual Y sequences. The apparent co-localization of He-T sequences with satellite DNAs in pericentric heterochromatin of metaphase chromosomes contrasts with the segregation of satellite DNA to alpha heterochromatin while He-T sequences hybridize to beta heterochromatin in polytene nuclei. This comparison suggests that satellite sequences do not exist as a single block within each chromosome but have interspersed regions of other sequences, including He-T DNA. If this is so, we assume that the satellite DNA blocks must associate during polytenization, leaving the interspersed sequences looped out to form beta heterochromatin. DNA from D. melanogaster has many restriction fragments with homology to He-T sequences. Some of these fragments are found only on the Y. Two of the repeated He-T family restriction fragments are found entirely on the short arm of the Y, predominantly in the pericentric region. Under conditions of moderate stringency, a subset of He-T DNA sequences cross-hybridizes with DNA from D. simulans and D. miranda. In each species, a large fraction of the cross-hybridizing sequences is on the Y chromosome.     

Chromosome anomalies associated with amplification of the adenosine deaminase gene (ADA) in rat hepatoma cells

Two independently selected series of rat hepatoma cell lines resistant to the drug deoxycoformycin (dCF) were analyzed karyotypically. Several forms of homogeneously staining regions (HSRs) were present on metaphase chromosomes of these cells. In some instances HSRs comprised nearly an entire chromosome, which are among the largest chromosomes in the karyotype. Stable resistance to dCF is acquired in rat cells by overproduction of the enzyme adenosine deaminase (ADA) as a result of amplification of ADA gene sequences. We have localized the amplified ADA gene sequences to HSRs on metaphase chromosomes from both series of dCF-resistant cell lines by in situ hybridization. Based upon the number of ADA gene sequences present and the lengths of the HSRs, we have estimated the size of the amplified unit to range from 450 to 1,000 kb.     

Occurrence and evolution of homogeneously staining regions may be due to breakage-fusion-bridge cycles following telomere loss

Chromosomes with homogeneously staining regions (HSR) were analysed in a subclone of the H4 rat hepatoma cell line, where they represent amplification of the ribosomal RNA (rRNA) genes. Detailed G-band analysis of the subclone revealed that an HSR on the short arm of chromosome 3 became unstable and changed its position within the chromosome. The evolution of this marker chromosome was associated with the terminal deletion of the normal long arm of the HSR-bearing chromosome 3 and may have involved ring formation as a result of fusion between the HSR on the short arm and the broken end of the long arm. Evidence was obtained for breakage at different sites within the ring, producing chromosomes with HSRs located terminally on either the long arms or both arms. The terminally located HSR underwent elongation in some cells presumably as a result of a breakage-fusion-bridge cycle characteristic of instability due to telomeric loss. It is suggested that terminally located HSRs may generally occur this way.     

Meiotic events at the centromeric heterochromatin: histone H3 phosphorylation, topoisomerase IIα localization and chromosome condensation

Mechanisms of chromosome condensation and segregation during the first meiotic division are not well understood. Resolution of recombination events to form chiasmata is important, for it is chiasmata that hold homologous chromosomes together for their oppositional orientation on the meiotic metaphase spindle, thus ensuring their accurate segregation during anaphase I. Events at the centromere are also important in bringing about proper attachment to the spindle apparatus. This study was designed to correlate the presence and activity of two proteins at the centromeric heterochromatin, topoisomerase II alpha (TOP2A) and histone H3, with the processes of chromosome condensation and individualization of chiasmate bivalents in murine spermatocytes. We tested the hypothesis that phosphorylation of histone H3 is a key event instigating localization of TOP2A to the centromeric heterochromatin and condensation of chromosomes as spermatocytes exit prophase and progress to metaphase. Activity of topoisomerase II is required for condensation of chromatin at the end of meiotic prophase. Histone H3 becomes phosphorylated at the end of prophase, beginning with its phosphorylation at the centromeric heterochromatin in the diplotene stage. However, it cannot be involved in localization of TOP2A, since TOP2A is localized to the centromeric heterochromatin throughout most of meiotic prophase. This observation suggests a meiotic function for TOP2A in addition to its role in chromatin condensation. The use of kinase inhibitors demonstrates that phosphorylation of histone H3 can be uncoupled from meiotic chromosome condensation; therefore other proteins, such as those constituting metaphase-promoting factor, must be involved. These results define the timing of important meiotic events at the centromeric heterochromatin and provide insight into mechanisms of chromosome condensation for meiotic metaphase.     

Molecular cytogenetic characterization of the 11q13 amplicon in head and neck squamous cell carcinoma

Amplification of 11q13 DNA sequences and overexpression of CCND1 are common findings in head and neck squamous cell carcinoma (HNSCC), identified in about 30% of the cases. However, little is known about initiation of the amplification and the organization of the amplicon. In order to study the structure of the amplicon in more detail and to learn more about the mechanisms involved in its initiation, prometaphase, metaphase, and anaphase fluorescence in situ hybridization (FISH) with 40 BAC clones spanning a 16-Mb region in chromosome bands 11q12.2 to 11q13.5 was performed in nine HNSCC cell lines with homogeneously staining regions. FISH analysis showed that the size of the amplicon varied among the nine cell lines, the smallest being 2.12 Mb and the largest 8.97 Mb. The smallest overlapping region of amplification was approximately 1.61 Mb, covering the region from BAC 729E14 to BAC 102B19. This region contained several genes previously shown to be amplified and overexpressed in HNSCC, including CCDN1, CTTN, SHANK2, and ORAOV1. The cell lines were also used to study the internal structure of the amplicon. Various patterns of amplified DNA sequences within the amplicon were found among the nine cell lines. Even within the same cell line, different amplicon structures could be found in different cell populations, indicating that the mechanisms involved in the development of the amplicons in HNSCC were more complex than previously assumed. The frequent finding of inverted repeats within the amplicons, however, suggests that breakage-fusion-bridge cycles are important in the initiation, but the fact that such repeats constituted only small parts of the amplicons indicate that they are further rearranged during tumor progression.