Transformation frequencies were measured in CHO mutant EM9 after transfection with intact or modified plasmid pSV2-gpt. The mutant and wild-type strain behaved similarly under all conditions except when homologous recombination was required to produce an intact plasmid. Therefore, the defect of the mutant which renders it slow in DNA strand break rejoining and high in sister chromatid exchange induction reduces its ability to recombine foreign DNA molecules.Recombination of genomic sequences in mammalian cells is important in development and differentiation (25,26), the generation of sister chromatid exchange (14), and chromosomal translocations in cancer (41), among other important cellular processes. However, the mechanisms involved in recombination in mammalian cells have not been elucidated to the extent that they have been in procaryotes and simple eucaryotes (22,30). Recently, numerous studies have described systems for the study of recombination and ligation in mammalian cells. These systems generally involve the addition of foreign DNAs, such as animal viruses (38-40) or recombinant plasmids with selectable markers (5,6,17,21,27), and have shown that mammalian cells are able to efficiently reconstitute intact genes from gene fragments via homologous or nonhomologous recombination.In the present study, transformation to mycophenolic acid resistance of CHO mutant EM9 was measured after transfection with intact or modified pSV2-gpt, which codes for bacterial guanosine phosphoribosyltransferase (18,19), to assess integration, ligation, and recombination frequencies of the mutant compared with the wild-type strain, AA8 (36). Mutant EM9 has been characterized as having enhanced sensitivity to some, but not all, alkylating agents (9, 33, 37), slight sensitivity to X-irradiation (37), high base-line and induced sister chromatid exchange frequencies, and hampered ability to rejoin DNA strand breaks (33), making it an interesting candidate for the study of processing of exogenous DNA.A diagram of pSV2-gpt and relevant restriction sites is shown in Fig. 1A. For transfection experiments, DNA was calcium phosphate precipitated (35) and added to monolayers (2 x 106 cells per dish) of AA8 and EM9 cells obtained from suspension cultures grown in alpha-MEM (KC Biologicals) supplemented with 10% fetal bovine serum (Flow Laboratories, Inc.) (alpha-MEMFBS) as described previously (34). After DNA exposure, the cell monolayers were rinsed twice with phosphate-buffered saline and incubated in alpha-MEMFBS for 24 h. They were then transferred to roller tubes for suspension growth. After the appropriate time for expression of the Gpt phenotype, cells were counted and plated for plating efficiency measurements in alpha-MEMFBS (300 cells per dish) or for selection (105 to 106 cells per dish) in MAXTA medium (alpha-MEMFBS