A model for random genetic damage directing selection of diploid or aneuploid tumours.

OBJECTIVES: To test whether genetic instability may determine whether tumours become aneuploid or diploid. MATERIALS AND METHODS: We have identified genes needed for cell survival or replication by combining Affymetrix gene expression array data from 12 experimental cell lines with in silico GEO+GNF and expO databases. Specific loss of heterozygosis (LOHs), chromosomal abnormalities (called derivative chromosomes) and numbers of normal homologues were identified by SNP and SKY analyses. Random gene losses were calculated under the assumption that bi-allelic MMR gene inactivation causes a 20-fold increase in rate of gene loss. RESULTS: There were ∼1.23 × 10(4) genes widely dispersed throughout the genome and possibly expressed by all cells for survival or proliferation, many of these genes performed housekeeping functions. Conservation of the genes may explain the complete haploid genomes found for 15 different cell types and derivative chromosomes selectively retained in aneuploid cancer cell lines after LOH formations, and normal homologue losses. Loss of cell survival/replication genes was calculated to be higher in colon stem cells of carriers of MMR gene mutations than carriers of APC gene mutations. CONCLUSION: Random loss of cell survival/replication genes was calculated to be low enough for colon stem cells with APC gene mutations to 'select' LOH and derivative chromosome combinations favouring tumour cell proliferation. However, cell survival/replication gene loss was calculated to be too high for colonic stem cells lacking MMR genes to survive chromosomal instability, explaining why MMR mutations only produce tumours with diploid chromosome cells.

A model for genetic complementation controlling the chromosomal abnormalities and loss of heterozygosity formation in cancer.

The relationship between the apparently random chromosomal changes found in aneuploidy and the genetic instability driving the progression of cancer is not clear. We report a test of the hypothesis that aneuploid chromosomal abnormalities might be selected to preserve cell-survival genes during loss of heterozygosity (LOH) formations which eliminate tumor suppressor genes. The LOHs and structurally abnormal chromosomes present in the aneuploid LoVo (colon), A549 (lung), SUIT-2 (pancreas), and LN-18 (glioma) cancer cell lines were identified by single nucleotide polymorphisms (SNPs) and Spectral Karyotyping (SKY). The Mann-Whitney U and chi square tests were used to evaluate possible differences in chromosome numbers and abnormalities between the cell lines, with two-tailed P values of <0.01 being considered significant. The cell lines differed significantly in chromosome numbers and frequency of structurally abnormal chromosomes. The SNP analysis revealed that each cell line contained at least a haploid set of somatic chromosomes, consistent with our hypothesis that cell-survival genes are widely scattered throughout the genome. Further, over 90% of the chromosomal abnormalities seemed to be selected, often after LOH formation, for gene-dosage compensation or to provide heterozygosity for specific chromosomal regions. These results suggest that the chromosomal changes of aneuploidy are not random, but may be selected to provide gene-dosage compensation and/or retain functional alleles of cell-survival genes during LOH formation.

DNA damage and p53 induction do not cause ZD1694-induced cell cycle arrest in human colon carcinoma cells.

Using four complementary approaches, ie., cell synchronization, bromodeoxyuridine labeling, and DNA and Western blot analyses, we investigated the underlying mechanism of cell cycle perturbation in response to ZD1694, a quinazoline-based antifolate thymidylate synthase inhibitor. With a single exposure at a concentration of 1 microM for 2 h, ZD1694 completely inhibits thymidylate synthase over 72 h and causes a sustained growth for at least 120 h, DNA damage, and p53 induction in human carcinoma cells. Although these cells displayed an S-phase block with the precise terminal arrest point depending on the timing of drug treatment in the cell cycle, their DNA-replicating machinery associated with polymerase alpha was preserved intact. When supplemented with exogenous dThd, these cells resumed an apparently normal S-phase progression for at least 4 h. Kinetic analyses based on synchronized cells indicate that S-phase arrest occurs first, preceding the induction of DNA double strand breaks and p53/p21. SW480 cells, in which p53mu failed to transduce p21, also exhibited the mode of S-phase arrest, essentially indistinguishable from that displayed by HCT-8 cells expressing the functional p53 (p53wt). That the DNA replication process is prerequisite for DNA double strand breaks was indicated by the following: (a) DNA damage occurred only when cells treated with ZD1694 progressed through S phase; and (b) the inhibition of DNA polymerase alpha by aphidicolin-blocked DNA damage. Based on the above, we conclude that S-phase arrest by ZD1694, with a subsequent damage of DNA double strands, is caused by the block of DNA synthesis in the middle of replication due to dTTP depletion and not by p53-mediated G1-G2 checkpoint mechanisms or p21-induced inactivation of the DNA-replicating machinery.

Changes of an androgen-dependent nuclear protein during functional differentiation and by dedifferentiation of the dorsolateral prostate of rats.

Nuclei of the dorsolateral prostate of rats contain a large amount of androgen-dependent non-histone protein (20K-NHP) (mol. wt. not equal to 20,000; pI not equal to 11.5) (Matuo et al. (1]. Its content in the nuclei increased most markedly during 4-8 weeks of age, when functional differentiation of the prostate was most active on the basis of the changes of major cytosol proteins and zinc. Nuclei of the Dunning tumors originating in the dorsolateral prostate were found to lack 20K-NHP regardless of androgen dependency, indicating the disappearance of the 20K-NHP from the nuclei by dedifferentiation. These suggest that the 20K-NHP is an important nuclear protein for differentiation of the dorsolateral prostate cells.

Hybridization of nuclear matrix attached deoxyribonucleic acid fragments.

Annealing studies were performed on DNA fragments associated with rat and mouse liver interphase nuclear matrix and the metaphase scaffold of Chinese hamster DON cells. Matrix and scaffold bound DNA fragments, reassociated with an excess of total genomic DNA, displayed kinetics virtually identical with total nuclear DNA probes. Moreover, both the extent and kinetics of these hybridizations were independent of the matrix DNA fragment size (less than 350--5000 base pairs) and the method of nuclease digestion used in their preparation (DNase I, micrococcal nuclease or endogenous digestion). The repetitive DNA component of the matrix DNA was examined by reacting discrete sizes of matrix DNA fragments (less than 350--5000 base pairs) from mouse liver with a library of cloned repetitive sequence DNA fragments which included mouse major satellite sequences. Our results demonstrate that short DNA fragments anchored to the nuclear matrix contain these cloned sequences is similar proportion of total nuclear DNA and, when viewed in light of the annealing results, indicate that matrix DNA is not enriched in either repetitive or unique sequences. Furthermore, the matrix DNA fragments appear to contain the entire sequence complexity of the genome. Finally, we hybridized both matrix and total nuclear DNA fragments with cDNA to total nuclear polyadenylated RNA. The kinetics and extent of hybridization indicate that most, if not all, of the actively transcribed DNA sequences are present in similar concentrations. We conclude that in the overall organization of eukaryotic DNA within the nucleus, the repeating domains or loops which have been demonstrated by a number of investigators are not anchored at specific attachment sequences in interphase cells or during mitosis. These findings are discussed with regard to current concepts of eukaryotic DNA loop organization.

Disappearance of a structural chromatin protein A24 in mitosis: implications for molecular basis of chromatin condensation.

A chromatin protein, A24, a conjugate of histone H2A and evolutionally conserved ubiquitin, was virtually the only structural polypeptide that was present in interphase but missing in mitosis of a Chinese hamster cell line (DON). Because a 10% increase in the H2A/DNA ratio observed in interphase-mitosis transition explained the stoichiometric conversion of A24 to H2A, it appears that ubiquitin bound to H2A of nucleosomal surfaces in interphase is released at mitosis whereas the total H2A remains as a structural component of nucleosomes. Regardless of protein synthesis, ubiquitin was again bound to H2A when cells entered the G1 phase. Based on the electrostatic nature of the COOH-terminal region of H2A, where ubiquitin binds, and the mitosis-specific rise of covalently linked phosphates in histones H1 and H3, we propose that an ionic interaction between the positively charged H2A COOH-terminal regions on fibers and negatively charged phosphates linked to serine or threonine of H1 and H3 molecules on adjacent fibers could generate an assembly of chromatin fibers in mitosis.

Quantitative conservation of chromatin-bound RNA polymerases I and II in mitosis. Implications for chromosome structure.

RNA synthesis almost ceases in mitosis. It is ambiguous whether this temporal, negative control of RNA synthesis is solely because of the nature of chromosomes per se, (i.e., their condensed state), or to a physical loss of RNA polymerases along with other nuclear proteins which have been shown to pass into the cytoplasm in mitosis, or to their combined feature. Aside from such regulatory considerations, a question has also been raised as to whether RNA polymerases are constituents of metaphase chromosomes. To clarify these aspects of RNA polymerase-chromatin interaction in mitosis, the enzymes in chromosomes were quantitated and their levels compared to those in interphase nuclei and cells at various phases of the cell cycle. The results show that the amounts of form I, form II, and probably form III enzymes bound to a genome-equivalent of chromatin stay constant during the cell cycle. Thus, the mechanism for the negative control of RNA synthesis in mitosis appears to exist in the chromosomes per se, but not to be directly related to the RNA polymerase levels. This quantitative conservation of chromatin-bound RNA polymerases implies that they may persist as structural components of the chromosomes in mitosis.

Isolation and characterization of nonhistone chromosomal protein C-14 which stimulates RNA synthesis.

The nonhistone chromatin protein, C-14, was extracted from chromatin of Novikoff hepatoma ascites cells and isolated in high purity as shown by its migration as a single dense spot on two-dimensional polyacrylamide gels. Its mobility on sodium dodecyl sulfate gels is consistent with a molecular weight of approximately 70 000. The amino acid composition shows that protein C-14 has an acidic:basic amino acid ratio of 1.8. Its amino terminal amino acid is lysine. Protein C-14 stimulated the incorporation of [3H]UMP into RNA by approximately 30% when added to naked DNA and homologous RNA polymerase I. A 30% stimulation of [3H]UMP incorporation into RNA was also found when protein C-14 was added to an E. coli RNA polymerase system containing either E. coli or Novikoff hepatoma DNA.

Fidelity of ribosomal ribonucleic acid synthesis by nucleoli and nucleolar chromatin.

Isolated nucleoli, nucleolar chromatin, and nucleolar DNA were used as templates for DNA synthesis in appropriately supplemented systems in which RNA polymerases other than RNA polymerase I were blocked by alpha-amanitin. With the aid of nucleotide analysis, DNA-RNA hybridization, and homochromatography fingerprinting, it was found that isolated nucleoli and nucleolar chromatin serve primarily as templates for synthesis of rRNA. However, the products formed with purified nucleolar DNA as a template do not contain the specific rRNA oligonucleotides nor are they appreciably hybridized to the rDNA region on cesium chloride gradients. These results indicate that whole nucleoli and nucleolar chromatin contain control mechanisms that restrict readouts by RNA polymerase I of nucleolar DNA to rDNA.

Induction of prophase in interphase nuclei by fusion with metaphase cells.

Fusion of an interphase cell with a metaphase cell results in profound changes in the interphase chromatin that have been called "chromosome pulverization" or "premature chromosome condensation" In addition to the usual light microscopy, the nature of the changes has been investigated in the present study with electron microscopy and biochemical techniques Metaphase and interphase cells were mixed and fused at 37 degrees C by means of ultraviolet-inactivated Sendai virus. After cell fusion, morphological changes in interphase nuclei occurred only in binucleate cells which contained one intact set of metaphase chromosomes Irrespective of the nuclear stage at the time of cell fusion, the morphologic changes that occurred 5-20 min later simulated very closely a sequence of events that characterizes the normal G(2)-prophase transition. Radioautography revealed that, late in the process, substantial amounts of RNA and probably protein were transferred from the interphase nucleus into the cytoplasm of fused cells. Thus, the findings indicate the existence in metaphase cells of factor(s) which are capable of initiating biochemical and morphological events in interphase nuclei intrinsic to the normal mitotic process.

Enzyme inactivation with ultraviolet laser energy (2650 Angstroms).

Inactivation of rat heart lactate dehydrogenase was accomplished by irradiation of the enzyme in solution with a frequency quadrupled neodymium glass laser.