Validation of a flow cytometric in vitro DNA repair (UDS) assay in rat hepatocytes.

An in vitro flow cytometric (FCM) DNA repair assay has been developed and validated by comparison to conventional autoradiography (ARG). Both assays measure unscheduled DNA synthesis (UDS). Cultures of hepatocytes from young male Sprague-Dawley rats were exposed to a battery of 26 chemicals plus bromodeoxyuridine (BrdUrd) or 3H-thymidine (3H-dT) for 18-20 h before harvest. Selection of test chemicals was based upon both their genotoxicity classifications and carcinogenicity bioassay results in male rats. DNA repair in chemically treated cultures was detected flow cytometrically by measuring the uptake of BrdUrd in non-replicating (G1, G2, mitotic and 4C) cells. Intracellular levels of incorporated BrdUrd were visualized by immunochemical labeling with fluorescein isothiocyanate (FITC), and total cellular DNA content was simultaneously estimated by counterstaining samples with the nucleic acid intercalator, propidium iodide (PI). Information was obtained from 10(4) cells/sample. Since repairing cells incorporate significantly less BrdUrd per unit of time than replicating cells, low intensity BrdUrd-FITC fluorescent signals from repairing cells are readily discriminated from high intensity signals from replicating cells when displayed on linear univariate histograms. Further distinction between repairing and replicating cells was achieved by displaying the DNA contents of all cells on linear bivariate histograms. Thus, repairing cells were resolved without subjecting these cultures to agents which suppress replicative synthesis (e.g., hydroxyurea). Results from these concurrent FCM and ARG investigations include the following: (1) conclusions (DNA repair positive or negative) were in agreement, with one exception, cinnamyl anthranilate, for which cytotoxic doses produced a positive FCM response, but lack of intact hepatocytes in parallel ARG preparations prevented analysis; (2) similar sensitivities for most of the positive chemicals were reported; (3) a high correlation (85%) exists between the reported genotoxicity classification and these DNA repair results in the absence of overt cytotoxicity; (4) a poor correlation exists between these DNA repair results and hepatocarcinogenesis (only 4/11 liver carcinogens tested positive) or overall carcinogenesis in the male rat (only 9/21 carcinogens tested positive). This FCM assay provides a rapid, sensitive, safe and reliable means of identifying agents which induce DNA repair in mammalian cells.

Detection of DNA adducts by 32P-postlabeling and multifraction contact-transfer thin-layer chromatography.

32P-Postlabeling is a sensitive method for detecting DNA adducts. Large bulky adducts, particularly from polycyclic aromatic compounds, are readily detected using this technique. Detection of small modifications, such as methylations, has often required specific additional enrichment procedures prior to 32P-postlabeling. We report the use of a single analytical procedure that can detect DNA adducts of a wide range of sizes and hydrophobicities (exemplified by adducts produced with methyl methanesulfonate, diepoxybutane, styrene oxide, or benzo[a]-pyrene). This 32P-postlabeling/thin-layer chromatography procedure is particularly useful when examining the potential of novel compounds or their metabolites to form DNA adducts.

Phosphatase activity in commercial spleen exonuclease decreases the recovery of benzo[a]pyrene and N-hydroxy-2-naphthylamine DNA adducts by 32P-postlabeling.

Spleen exonuclease, which degrades nucleic acids into single 3'-nucleotides, is used in the detection of DNA adducts by 32P-postlabeling. Contamination of the exonuclease with phosphatase activity can reduce the recovery of benzo[a]pyrene and N-hydroxy-2-naphthylamine DNA adducts by 32P-postlabeling. Four preparations of spleen exonuclease containing varying levels of phosphatase activity (< 1-62% of the unmodified 3'-nucleotides being dephosphorylated) were used to hydrolyze the DNA. The exonuclease with the lowest phosphatase activity produced a recovery of up to 9.60 mumol of benzo[a]pyrene adducts per mole of DNA. Recovery of benzo[a]pyrene adducts was reduced to 0.56 mumol of adduct per mole of DNA using the exonuclease with the highest phosphatase activity. Phosphatase in the exonucleases also dephosphorylated N-hydroxy-2-naphthylamine DNA adducts. Surprisingly, recovery of these DNA adducts was nearly 10 times greater using nuclease P1 than when using 1-butanol extraction for adduct enrichment, since arylamine DNA adducts have previously been reported to be poorly detected by 32P-postlabeling after nuclease P1 treatment. Our data indicate that the hydrolysis of DNA by spleen exonuclease may be an important source of variability in both qualitative and quantitative analysis of adducts by 32P-postlabeling.

Immunochemical quantitation of bromodeoxyuridine: application to cell-cycle kinetics.

We have described several laboratory procedures for the immunochemical staining of the halopyrimidines, BrdUrd and IdUrd, in cell suspensions for flow cytometry and a method for staining histological sections on slides. Halogenated pyrimidine quantitation allows cell-cycle parameters, including total cell-cycle time, phase durations, and growth fraction to be determined. We have presented some flow cytometric data to demonstrate the use of these methods in determining bivariate BrdUrd/DNA histograms with CHO cells and in kinetic studies with the brown Norway rat myeloid leukemia model.

Statistical confirmation that immunofluorescent detection of DNA repair in human fibroblasts by measurement of bromodeoxyuridine incorporation is stoichiometric and sensitive.

Diploid human fibroblasts (IMR-90 cells), grown to confluency and growth-arrested by serum starvation, were irradiated with a variety of doses of UV light (0.025-40 J/m2) or incubated with broad dose ranges of four direct-acting mutagens [ethyl methanesulfonate (EMS), ICR-170, methyl methanesulfonate (MMS), and 4-nitroquinoline oxide (4-NQO)] and pulsed with a thymidine analog, 5-bromodeoxyuridine (BrdUrd) to detect evidence of DNA repair. These cells and appropriate controls were immunochemically stained and subjected to flow cytometric analysis to quantify BrdUrd incorporation into DNA and simultaneously measure cellular DNA content. Since the maximal quantity of BrdUrd incorporated with repairing cells is profoundly less than the amount incorporated during replicative synthesis and flow cytometric analysis collects information on a cell-to-cell basis, data collection using linear histograms succeeded both in revealing repairing cellular populations and eliminating replicative cells from the analysis. Technical modifications necessary to achieve stoichiometry with UV-irradiated IMR-90 fibroblasts included a 48h repair (and pulse) period, followed by denaturing cellular DNA for 15 min at 90 degrees C. The limit of detection was equal to or below the lowest dose investigated (0.025 J/m2). DNA repair was also detected with cultures incubated with low doses of all chemicals and pulsed with BrdUrd for a 24 h period. The limits of detection were near or below 500 microM EMS, 5 microM MMS, 0.25 microM 4-NQO, and 0.1 microM ICR-170.

Chromosomal abnormalities and gene amplification in renal cancers induced in rats by nickel subsulfide.

Nickel subsulfide (alpha Ni3S2) was administered to male Fischer-344 rats by unilateral intrarenal (i.r.) injection (20 mg per rat) to establish the time-course of alpha Ni3S2-induced erythrocytosis and to identify chromosomal abnormalities and molecular genetic aberrations in ensuing renal cancers. Blood hematocrit values were increased in alpha Ni3S2-treated rats during two to 36 weeks post-injection, attained a maximum of 77 percent (SD +/- 5) at 16 weeks (vs 51 +/- 3 percent in vehicle controls), and returned to baseline at 40 weeks. Within 21 months, malignant neoplasms (five sarcomas, one carcinoma) occurred in the injected kidneys of 6/28 alpha Ni3S2-treated rats (vs 0/13 controls). Cytogenetic analyses of direct preparations or primary cell cultures showed prominent chromosomal aberrations in three neoplasms, with rearranged marker chromosomes, polyploidy, and in one case an homogeneously staining region (HSR). Assays for gene amplification were performed with probes for murine erythropoietin (EPO) gene and H-ras, c-fos, c-myc, and N-myc oncogenes, using deoxyribonucleic acid (DNA) samples isolated from injected kidneys and renal neoplasms, as well as from the contralateral, non-injected kidneys. No consistent pattern was found; in one sarcoma, N-myc was amplified six-fold and c-fos was amplified two-fold; in another sarcoma, H-ras, c-fos, and EPO were amplified two-fold. This study shows that (a) karyotypes of 3/6 renal neoplasms of alpha Ni3S2-treated rats contained prominent marker chromosomes, (b) oncogene amplification was noted in 2/6 renal neoplasms, and (c) the EPO gene was not consistently amplified in DNA from the injected kidneys of alpha Ni3S2-treated rats during the initiation of erythrocytosis, or in subsequent renal neoplasms.

A human c-erbA oncogene homologue is closely proximal to the chromosome 17 breakpoint in acute promyelocytic leukemia.

A human cDNA library was screened for sequences homologous to the erbA gene of avian erythroblastosis virus (AEV). One such clone, cHerbA-1, was used to map the chromosomal location of highly homologous human sequences that were found to be present on chromosome 17 as judged by Southern blot screening of a panel of mouse-human hybrid cell lines segregating human chromosomes. cHerbA-1 was hybridized in situ to metaphase chromosomes from a normal male subject and from a female patient with an acute promyelocytic leukemia (APL) having the typical t(15;17) translocation. The results localized the cellular c-erbA sequences on chromosome 17 to the q21-q24 region of normal chromosomes and indicated that the c-erbA sequences remained on the 17q- chromosome in the APL cells, suggesting that they could be assigned to the 17(q21-q22) region. For additional data, we hybridized human neoplastic cells derived from a poorly differentiated acute leukemia carrying a t(17;21) translocation with thymidine kinase (TK)-deficient LMTK- mouse cells. A resulting hybrid, containing only the 21q+ chromosome, did not have human c-erbA sequences. Since the breakpoint on 17q in this translocation was similar to that in the APL t(15;17) translocation, this supported the assignment of c-erbA to the q21-q22 region of chromosome 17. The apparent close proximity of the c-erbA sequences to the chromosomal breakpoints in these two leukemias suggests a possible role for this oncogene homologue in the development of these neoplasms.

The 2p breakpoint of a 2;8 translocation in Burkitt lymphoma interrupts the V kappa locus.

The majority of chromosomal rearrangements observed in Burkitt lymphomas involve a translocation between 8q and 14q, while the remaining minority carry variant translocations between chromosome 8 and either 2 or 22. We have studied the JI Burkitt lymphoma cell line carrying the variant 2;8 chromosome translocation using a combination of high-resolution and molecular cytogenetic techniques. We have determined that the chromosome 2 breakpoint of the 2;8 translocation in these cells is in the distal portion of 2p11.2. In situ hybridization of a DNA probe for kappa light chain variable (V kappa) region demonstrated that this 2p11.2 breakpoint is within the V kappa region. There was significant hybridization of the probe to both the 2p- and 8q+ chromosomes, with 23% of all grains considered to be specific for V kappa located over the middle one-third of the long arm of the 8q+ chromosome. Thus, there is translocation of the entire kappa constant (C kappa) region and a portion of the region carrying V kappa genes from 2p to a region 3' of the c-myc oncogene on the involved chromosome 8, resulting in transcriptional activation of the c-myc that is quite distant from the 5' end of the C kappa gene. These results provide direct evidence for translocation-related rearrangement of the kappa immunoglobulin gene cluster in this Burkitt lymphoma and for the assignment of the V kappa locus to 2p11.2.

In situ hybridization and translocation breakpoint mapping. I. Nonidentical 22q11 breakpoints for the t(9;22) of CML and the t(8;22) of Burkitt lymphoma.

In situ chromosomal hybridization of a probe for part of the lambda light chain constant region (C lambda) has demonstrated that the 22q11 breakpoints of chronic myelogenous leukemia (CML) t(9;22) and Burkitt lymphoma t(8;22) are not identical. For CML, the breakpoint is distal to the IGLC genes, whereas for Burkitt lymphoma it is proximal. The study provides direct evidence for regional assignment of the IGLC gene cluster to 22q11.

Inherited XX sex reversal in the cocker spaniel dog.

Nine XX true hermaphrodites and two XX males were discovered in a family of American cocker spaniels. The true hermaphrodites were partially-masculinized females with ovotestes; the XX males had malformed male external genitalia and cryptorchid aspermatogenic testes. Wolffian and Mullerian duct derivatives were present in both true hermaphrodites and XX males. All four sires of sex-reversed dogs were normal XY males; five of the dams were anatomically normal females and one was an XX true hermaphrodite. A second true hermaphrodite reproduced as a female, producing anatomically normal offspring. All matings that produced sex-reversed offspring were consanguineous. Matings of the parents of sex-reversed cocker spaniels to normal beagles with no family history of intersexuality produced only normal offspring. Examination of G-banded karyotypes of the affected animals, their parents, and siblings, revealed no structural anomalies of the chromosomes that were consistently associated with sex-reversal. In assays for serologically-detectable H-Y antigen, the group of XX true hermaphrodites and the group of XX males had mean levels of the antigen not significantly different from that in normal male controls. Female parents of sex-reversed dogs and some of their female siblings were typed H-Y antigen positive, but the mean level of the antigen in this group was less than that of normal male controls. It is proposed that XX sex reversal in cocker spaniels is due to a mutant gene which when homozygous in females, results in a level of H-Y antigen similar to that found in normal males and the gonads develop as ovotestes or testes. When the gene is heterozygous in females, the level of serologically-detectable H-Y antigen is lower than that found in normal males and the gonads develop as normal ovaries. The persistence of Mullerian structures in the presence of testicular tissue suggests that Mullerian inhibiting substance is deficient or ineffective in its action in this condition.

Association of amplified oncogene c-myc with an abnormally banded chromosome 8 in a human leukaemia cell line.

Several unusual chromosome structures have been described in drug-resistant cell lines and in certain tumours. These structures include elongated homogeneously staining regions (HSRs), small extrachromosomal paired chromatin bodies (double minutes, DMs) and abnormally banded regions (ABRs) with strong but anomalous band patterns. There is evidence that these are alternative forms of gene amplification, with HSRs breaking down to form DMs, and DMs integrating into the chromosome to generate HSRs and ABRs. Recently, it was demonstrated that, compared with several normal and leukaemia human cells, DNA sequences representing the human homologue of the onc gene of the avian myelocytomatosis virus (MC29), the so-called c-myc gene, were amplified in HL-60 cells. This is a human promyelocytic leukaemia cell line established in the laboratory of one of us (R.C.G.) at the National Cancer Institute (Bethesda, Maryland) in 1977, and widely used for studies on myeloid and monocytic differentiation. Amplification of the gene was present in primary leukaemic cells of the patient, and DMs were noted in some of these cells as well as in early passages of the HL-60 line. No structure resembling HSRs or ABRs were noted in karyotypic studies at this early stage and there were no alterations involving the long arm of chromosome 8 (8q), to which the c-myc gene has recently been mapped. We have now re-examined the karyotype of the HL-60 line, using cells frozen at various times during its continuous passage at the Wistar Institute (Philadelphia, Pennsylvania) to look for chromosomal abnormalities that might be associated with the amplification of c-myc. We find that, beginning in 1979, HL-60 cells at the Wistar Institute no longer had DMs, but did show an abnormal 8q+ chromosome, replacing a normal chromosome 8, and representing an ABR reflecting the site of myc gene amplification.

Amplified C lambda and c-abl genes are on the same marker chromosome in K562 leukemia cells.

The human leukemia cell line K562, derived from a patient with Philadelphia chromosome-positive chronic myelogenous leukemia, contains amplified c-abl oncogenes and unrearranged C lambda genes. Using in situ hybridization techniques, we have determined that the amplified c-abl and C lambda DNA sequences of K562 cells are both located on the same abnormal acrocentric marker chromosome, which may represent an altered Philadelphia chromosome.

H-Y antigen and sex determination in animals.

In all instances described, H-Y antigen was found when testes were present. Extragonadal tissues in androgen-insensitive individuals with TFS were typed H-Y+, indicating that this cell surface component is hormone-independent. Thus, association of H-Y antigen and primary sex determination is reinforced. But variation in the degree of masculinization observed in certain mammalian intersexes (e.g., polled goats) is not explained by simple H-Y antigen dosage effects. Additional genetic and/or intrauterine modifiers may influence both gonadal and somatic tissue target cell sensitivity to hormones.

Breeding studies reveal segregation of a canine Robertsonian translocation along Mendelian proportions.

Six matings of golden retriever-type mongrel dogs with a 13/23 Robertsonian translocation were studied. Three matings between normal dogs and dogs heterozygous for the translocation produced 12 normals and 11 heterozygotes. In three matings between heterozygotes, the distribution of normal, heterozygous, and homozygous proteny was Mendelian (8:12:5). No gross phenotypic anomalies were observed with the exception of congenital inguinal hernia in two female homozygotes. Nondisjunction in the form of trisomic or monosomic individuals for either autosome 13 or 23 was not observed.

Genetic basis of XX male syndrome and XX true hermaphroditism: evidence in the dog.

Serological analysis of white blood cells from the members of a family of American cocker spaniels indicates that a form of abnormal sexual development, in which individuals with a female karyotype have testes or ovotestes, is caused by anomalous transmission of male-determining H-Y genes.

The Giemsa banding pattern of the canine karyotype.

The canine metaphase karyotype consists of 78 chromosomes. All autosomes exhibit telocentric or acrocentric configurations gradually diminishing in size. These features make identification of homologous pairs by conventional analysis difficult. Chromosome preparations were derived from short-term cultures of peripheral blood lymphocytes obtained from clinically normal dogs representing at least four breeds. Most components of the canine karyotype can be distinguished readily. No significant G-banding pattern variations were detected in the individuals screened. An idiogrammatic interpretation of the banding pattern is presented. Apart from bands, other characteristic morphologic features were found which aid in identification. The G-banding pattern of the canine metacentric X is quite similar to that of the banded human X. The canine Y is a minute metacentric having two positive bands.