Recombination in Mammalian Cells - NCBI

5 downloads 616 Views 2MB Size Report
unrepaired UV damage and that increased DNA repair in highly transcribed alleles removes ...... Bohr, V. A., C. A. Smith, D. S. Okumoto, and P. C. Hanawalt.
Vol. 14, No. 1

MOLECULAR AND CELLULAR BIOLOGY, Jan. 1994, p. 391-399

0270-7306/94/$04.00+0 Copyright X 1994, American Society for Microbiology

Preferential Repair of UV Damage in Highly Transcribed DNA Diminishes UV-Induced Intrachromosomal Recombination in Mammalian Cells WIN PING DENG AND JAC A. NICKOLOFF* Department of Cancer Biology, Harvard University School of Public Health, Boston, Massachusetts 02115 Received 26 January 1993/Returned for modification 27 June 1993/Accepted 27 September 1993

The relationships among transcription, recombination, DNA damage, and repair in mammalian cells were investigated. We monitored the effects of transcription on UV-induced intrachromosomal recombination between neomycin repeats including a promoterless allele and an inducible heteroallele regulated by the mouse mammary tumor virus promoter. Although transcription and UV light separately stimulated recombination, increasing transcription levels reduced UV-induced recombination. Preferential repair of UV damage in transcribed strands was shown in highly transcribed DNA, suggesting that recombination is stimulated by unrepaired UV damage and that increased DNA repair in highly transcribed alleles removes recombinogenic lesions. This study indicates that the genetic consequences of DNA damage depend on transcriptional states and provides a basis for understanding tissue- and gene-specific responses to DNA-damaging agents.

Recombination systems regulate gene expression of specific loci by transferring alleles from silent to active regions, as in yeast mating-type switching (21), and by assembling functional genes from physically separated components, as in mammalian immunoglobulin loci (65). Also, inappropriate recombination can alter gene expression, as exemplified by the activation of oncogenes following chromosomal rearrangements produced by aberrant immunoglobulin V(D)J site-specific recombination (11). This study was designed to clarify further the relationships among UV damage and repair, transcription, and recombination in mammalian cells (diagrammed in Fig. 1). In particular, we were interested in how transcription-enhanced repair might affect UV-induced recombination in Chinese hamster ovary (CHO) cells. We have investigated these effects by using cell lines carrying two neomycin genes (neo), one of which (MMTVneo) was regulated by the dexamethasone (DEX)-responsive mouse mammary tumor virus (MMTV) promoter. Since both high-level transcription (38) and UV-induced DNA damage (3) stimulate recombination, it was thought that UV damage and transcription might have additive or synergistic effects on recombination. Instead, we found that UV-induced recombination levels with highly transcribed alleles are lower than with identical alleles transcribed at low levels. We have correlated the reduction in UV-induced recombination with enhanced repair of transcribed strands in highly transcribed DNA. These results are explained by a model in which transcription-enhanced repair reduces recombination by removing recombinogenic lesions in DNA.

There are complex relationships among DNA damage and repair, recombination, and transcription. It is well established that recombination is stimulated by DNA strand breaks (27, 31, 32, 40-42, 44, 46, 59; reviewed in reference 63) and by UV-induced or chemically induced DNA damage (2, 3, 37, 68, 71). Recombination may be stimulated by excision repair of UV or chemical damage since excision repair exposes single-stranded regions that could pair with homologous DNA. Recombination is also stimulated by transcription. Intra- and interchromosomal recombination is enhanced by transcription in the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe (17, 24, 60, 64, 69, 70), and transcription stimulates both extra- and intrachromosomal recombination in mammalian cells (38, 41). Transcription also enhances site-specific recombination at immunoglobulin gene loci (1, 4, 30) and at yeast mating-type loci (26). Besides influencing recombination rates, transcriptional activity also influences DNA repair. Several studies have demonstrated increased repair of UV damage in transcribed loci relative to the genome overall or to silent loci (5, 29, 43; reviewed in references 19 and 20). Increased repair in transcriptionally active loci is due principally to preferential repair of transcribed strands in Escherichia coli (34) and mammalian cells (28, 35). Recent studies have suggested that preferential repair in procaryotes is apparently mediated by the Mfd protein, which displaces stalled RNA polymerase and recruits UvrA to repair lesions (56), whereas in eucaryotes, ERCC-3, which is active in excision repair, has been shown to be an intrinsic part of the RNA polymerase complex (55). DNA damage and transcription are linked in two other ways. UV damage blocks transcription (reviewed in reference 53), and UV, ionizing radiation, and chemical damage stimulates expression of specific damage-inducible genes (7, 8, 14, 15, 22, 50), the products of which may be directly or indirectly involved in DNA repair.

MATERIALS AND METHODS CHO cell lines. The construction and characterization of CHO line Klc derivatives containing recombination substrates was described previously (38). Both lines used in this study have single copies of recombination substrates diagrammed in Fig. 2, and MMTV neo transcription levels are rapidly increased by the addition of 1 ,uM DEX (38). Recombination assays. Culture conditions were described previously (41). Harvested cells were counted, and 2 x 106

* Corresponding author. Mailing address: Department of Cancer Biology, Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02115. Phone: (617) 432-1184. Fax: (617) 432-3950 or (617) 739-8348.

391

392

DENG AND NICKOLOFF

MOL. CELL. BIOL.

UV Damage

Recombination -

8 10 12 14 16 18

UV Dose

0

4-6

(Jm2)

18

UV Dose (Jm2)

FIG. 3. Effects of DEX-enhanced transcription on UV-induced intrachromosomal recombination. (A and B) Recombination frequencies at various UV doses were determined with four independent populations each of MSGneo2 lines 3-6 (A) and 4-6 (B) in the presence or absence of DEX. Two determinations were made with each population; shown are average recombination frequencies + standard deviations. (C) Recombination frequencies were determined at various UV doses with a single population of SVneo2 as described for panels A and B. Values represent averages standard deviations of three determinations. (D) Reductions in UV-induced recombination with DEX-treatment for MSGneo2 lines 3-6 and 4-6 and SVneo2 were quantified as follows. For each dose, UV inductions of recombination were calculated as the ratio of the recombination frequency at that dose to the frequency without UV, with separate UV inductions calculated for DEX-treated and untreated cultures. Ratios of UV inductions without DEX to UV inductions with DEX at each dose are plotted versus dose for SVneo2 (closed symbols) and the two MSGneo2 lines (open symbols). Included in panel D are data from additional experiments with MSGneo2 lines (noted in the text) that are not shown in panels A and B.

somal recombination. To determine the effects of low and high transcription levels on UV-induced recombination, we measured recombination frequencies in four independently derived populations of MSGneo2 lines 3-6 and 4-6 irradiated with 0 to 18 J of UV per m2 and either treated or not treated with DEX. Treatment with DEX alone enhanced recombination by 5- to 10-fold. UV irradiation also enhanced recombination (Fig. 3), consistent with the report by Bhattacharyya et al. (3), with UV doses of 18 J/m2 enhancing recombination by 1,100- and 4,300-fold for lines 3-6 and 4-6, respectively. Although treatment with either DEX or UV alone stimulated recombination, UV-induced recombination levels at doses above 6 J/m2 were sharply reduced by treatment with DEX, with 18 J/m2 stimulating recombination only 43- to 53-fold. These results suggest that increasing transcription diminishes UV-induced recombination. A detailed view of transcriptional effects on UV-induced recombination at lower doses was obtained by repeating these experiments with single populations of MSGneo2 lines 3-6 and 4-6. Results similar to those shown in Figs. 3A and B were obtained (data not shown), indicating that equivalent

394

DENG AND NICKOLOFF

recombination frequencies result with UV doses of 5 to 6 J/m2 in the presence or absence of DEX. The reconstruction experiments described above rule out the possibility that the different recombinational responses to UV in the presence and absence of DEX are due to differential UV sensitivities among various recombinant products. Also, although recombination frequencies in DEX-treated cultures are reduced slightly by the small increase in UV survival, this effect is insufficient to account for the large DEX-dependent reductions in UV-induced recombination shown in Fig. 3A and B. For example, DEX has no effect on survival at UV doses