Vol. 3, No. 11

MOLECULAR AND CELLULAR BIOLOGY, Nov. 1983, p. 1967-1974

0270-7306/83/111%7-08$02.00/0 Copyright © 1983, American Society for Microbiology

Genetic Effects of Chromosomal Rearrangements in Chinese Hamster Ovary Cells: Expression and Chromosomal Assignment of TK, GALK, ACPI, ADA, and ITPA Loci RAYMOND L. STALLINGS,' GERALD M. ADAIR,' JEANETTE SICILIANO,2 JEFFREY GREENSPAN,2 AND MICHAEL J. SICILIANO2* The University of Texas System Cancer Center, Research Division, Smithville, Texas 78957,' and Department of Genetics, University of Texas, Houston, Texas 770302 Received 3 June 1983/Accepted 26 August 1983

Polyethylene glycol-mediated fusion of Chinese hamster ovary (CHO) cells with mouse C11D cells produced interspecific somatic cell hybrids which slowly segregated CHO chromosomes. Cytogenetic and isozyme analysis of HAT- and bromodeoxyuridine-selected hybrid subclones and of members of a hybrid clone panel retaining different combinations of CHO chromosomes enabled provisional assignments of the following enzyme loci to CHO chromosomes: TK, GALK, and ACPJ to chromosome 7; TK and GALK to chromosome Z13; ACPI, ADA, and ITPA to chromosome Z8; and ADA and ITPA to chromosome Z9. These genetic markers reflect the origin of each of these Z group chromosomes and indicate the functional activity of alleles located on rearranged chromosomes. Identification of diploid electrophoretic shift mutations for ADA and ITPA was consistent with those observations. Assignment of the functional TK locus in TK+'- CHO-AT3-2 cells indicated that gene deletion may be responsible for TK hemizygosity in this subline.

During the past few years, over 35 gene loci have been assigned to the chromosomes of euploid Chinese hamster cells (3, 4, 8, 12, 13, 18, 24-27). The location and functional behavior of alleles of these loci in CHO cells, the most widely used Chinese hamster line for somatic cell genetic and mutation research, are not as well defined since CHO cells have undergone stable cytogenetic alterations resulting from deletions, translocations, and other types of chromosomal rearrangements. Most of the original Chinese hamster chromosomal material can be accounted for in CHO cells either as intact normal chromosomes or as identifiable portions of rearranged Z group chromosomes (6, 28). However, on the basis of G-banding patterns, the origins of some of the Z group chromosomes and the fates of several Chinese hamster chromosomes or chromosomal regions are not at all clear (6, 28). These events may have led to position effects, rendering many loci functionally haploid and perhaps accounting for the high frequency of autosomal recessive mutations in CHO cells (20). The mapping of gene loci in CHO cells should answer that question as well as verify genetically the origin of Z group chromosomes. Thus far, the assignment of gene loci has been limited to four chromosomes of CHO cells: the Z2, which differs from the euploid chromosome

2 by having a large deletion from ql to q24; chromosome 8, which is present in two normal copies; chromosome 9, which is present in only a single copy with no identifiable remnants of its homolog; and the X chromosome, present in only a single copy (6, 28). The functional ploidy of loci associated with these chromosomes in CHO cells has been shown, both by concordant segregation and mutational analysis, to be constitutive with respect to their chromosomal assignments (4, 18, 26, 27). It remains to be determined whether loci present on CHO chromosomes resulting from rearrangements are constitutively expressed or perhaps modified in their expression owing to position effects. In this paper, we provisionally map alleles for five Chinese hamster loci (thymidine kinase, TK, EC 2.7.1.21; galactokinase, GALK, EC 2.7.1.6; acid phosphatase 1, ACP1, EC 3.1.3.2; adenosine deaminase, ADA, EC 3.5.4.4; and inosine triosphosphatase, ITPA, EC 3.6.1.19) onto CHO chromosomes by biochemical and cytogenetic analysis of CHO x mouse LTKCllD somatic cell hybrids. In euploid Chinese hamster cells, three of these loci (TK, GALK, and ACPI) are syntenic on chromosome 7 (24) and the other two (ADA and ITPA) have been assigned to chromosome 6 (25). Mapping of these markers in CHO cells allowed us to geneti-

1%7

1968

STALLINGS ET AL.

MOL. CELL. BIOL.

cally trace the origins of several extensively protocols and the procedures involved in the subserearranged CHO Z group chromosomes. These quent cloning of mutagenized cell populations and data, plus information obtained from electropho- isolation of electrophoretically detectable mutations been described previously (19). retic shift mutations induced in CHO cells by have Clonal from mutagenized populations were ethyl methanesulfonate, allowed us to ascertain screened isolates for electrophoretically detectable variations the effects of the rearrangements on gene in the products of the 48 enzyme loci. For various expression. In addition, cytogenetic analysis of reasons, not every clone was screened for the prodthe TK+- heterozygote CHO-AT3-2 and map- ucts of all 48 loci. However, no clone was screened for ping of the functional TK allele in this cell line less than 30 loci, and >95% of the clones were allowed further localization of the recent region- screened for 40 ± 2 loci. A clone seen to have a variant al mapping assignment of TK on the long arm of was regrown and analyzed for confirmation. Subclones of confirmed clones were analyzed for determichromosome 7 (13). nation of the heritability of the mutation.

MATERIALS AND METHODS Cell lines and culture conditions. Cells were maintained in either Dulbecco modified Eagle medium supplemented with 5% fetal calf serum and 5% newborn calf serum or alpha-modified minimal essential medium supplemented with 10% fetal calf serum. The two CHO sublines used to construct most of the CHO x mouse cell hybrids in this study were: TGA 102a/9 HPRT- (26) and EDGAL1-1, a GALK'-derivative of CHO-AT3-2 APRT+'- TK'- (2); the mouse parental strain was LTK- C1lD (7). Our hybrid clone panel also included several members resulting from the hybridization of LTK- CllD cells with herpes simplex virus TK-transformed TK- derivatives of Toronto strain CHO cells (1) obtained from Michael Gottesman, National Institutes of Health, Bethesda, Md. Hybrids were formed by polyethylene glycol-induced cell fusion (7) and selected in HAT medium (100 ,uM hypoxanthine, 10 pLM aminopterin, 10 ,uM thymidine) or HAT medium plus 1 mM ouabain, as previously described (18). Hybrid subclones were isolated either by cloning in microwell test plates or by plating hybrid cell populations into HAT or bromodeoxyuridine (BUdR) (50 pLg of 5-BUdR per ml) selection medium and picking resistant colonies (24). Chromosome analysis. G-band chromosome analyses were performed on conventionally prepared, airdried chromosome preparations, as previously described (14). Normal Chinese hamster chromosomes were classified and numbered according tQ standard nomenclature (11), and rearranged CHO marker chromosomes were designated by the nomenclature of Deaven and Petersen (6). Isozyme analysis. Extracts from parental cells, hybrid subclones, and drug-resistant hybrid segregants were screened for the presence of Chinese hamster and mouse isozyme gene products by vertical starch gel electrophoresis and histochemical staining, as previously described (23, 24). Interspecific electrophoretic mobility differences were sufficient to resolve 32 isozyme gene products of Chinese hamster and mouse loci (24); four which were of particular interest in this study were ACP1, ADA, GALK, and ITPA. Enzyme nomenclature and abbreviations follow the recommendations for an international system for human gene nomenclature (16) since homologies between Chinese hamster and human isozymes are easily recognized (23). Mutation analysis. CHO cell populations were mutagenized with ethyl methanesulfonate at doses yielding 1 to 50% survival. The details of our mutagenesis

RESULTS Cytogenetic characterization of CHO sublines used for hybridization. G-band chromosome analyses were performed on the three CHO sublines used for hybridization. TGA 102a/9 and CHO-AT3-2 cells both had modal karyotypes similar to those described by Deaven and Petersen (6), with several minor differences. We have previously described the differences in TGA 102a/9 (26). In CHO-AT3-2, differences include additions to the Xq, 5p, and Z6 chromosomes, a missing Z12 chromosome, and interstitial deletions from the long arms of chromosomes 7 and Z4. Derivations of the Toronto strain of CHO cells retained modal karyotypes identical to that published by Worton et al. (28). Segregation of chromosome 7-linked markers in BUdR- or HAT-selected hybrid subclones. The genes for TK, GALK, and ACPI have been assigned to Chinese hamster chromosome 7. Only one copy of chromosome 7 can be identified in CHO, but material from its homolog has been suggested to be present on the Z10 chromosome (6). The isolation of TK+'- CHO cell heterozygotes (2, 20) and comparisons of TK segregation frequencies in presumptive TK'- x TK-1- and TK+1+ x TK-1- CHO cell hybrids (2, 21) indicate that the TK locus and perhaps other chromosome 7-linked markers are present in two copies in CHO cells. To determine which CHO chromosomes carry genes for TK, GALK, and ACPI, nine independent TGA 102a/9 x LTK- CllD hybrid clones (designated with the prefix cc) were removed from HAT selection and grown for 2 weeks to allow segregation of the chromosomes carrying the TK genes. Each hybrid clone was then plated into either BUdR or HAT. A total of 33 BUdRand 34 HAT-resistant colonies were picked and grown to mass culture for isozyme analysis. Zymograms demonstrating the segregation of Chinese hamster from mouse gene products as conducted in our laboratory have been published previously (24, 25). Hamster isozyme GALK was segregated from all BUdR-selected clones and retained in all HAT-resistant clones.

VOL. 3, 1983

GENE MAPPING IN CHO CELLS

1969

Hamster isozyme ACP1 was missing in all presence or absence of the Zl or Z8 chromoBUdR-selected clones but retained in only 22 of some. There were 26 such 7 or 7q- segregants, the 34 HAT-selected clones. These results indi- for which the data are analyzed in Table 3. The cate that TK and GALK have remained linked in Z8 chromosome shows nearly perfect concorCHO cells and that one ACPI allele is no longer dant segregation with ACPI expression (25 of 26, 96%), whereas the Zl chromosome is discorlinked to these loci. A total of 17 HAT- and 9 BUdR-resistant dant in 7 of the 26 clones. The deletion of the clones were subjected to G-band cytogenetic 7q- chromosome (Fig. 1) does not appear to analysis. These data, along with segregation affect ACP1 since it is expressed in clones information for GALK and ACPI, are summa- R3/B3, R4/B2, and R3/H2 (Table 2), which segrized in Table 1. As expected from the TK regated chromosome Z8 and retained chromoassignment in euploid Chinese hamster cells some 7q-. We therefore assigned the alleles for (24), all BUdR-resistant colonies segregated ACPI to chromosomes 7 and Z8 of CHO cells. Based on G-banding patterns, the long arms of hamster chromosome 7 along with GALK and ACPJ. The Zl, Z8, and Z13 chromosomes were chromosomes Z8 and Z9 originated from Chialso missing in all BUdR' colonies, making them nese hamster chromosome 6 (6, 28) (Fig. 2). candidates for the second TK, GALK, and ACPI ACPJ is most likely located on the short arm of alleles in CHO cells. The chromosome carrying chromosome Z8, which contains material of the second allele for TK would need to be unknown origin (6, 28) (Fig. 2). We also deterpresent in all six HAT-resistant colonies which mined whether the origins of the Z8 and Z9 lost chromosome 7 (Table 1). The only candi- chromosomes could be genetically verified (see date chromosome so present is Z13. Since ham- below). Genetic evidence for the chromosome 6 origin ster isozyme GALK was also present in all six of these colonies, we assigned the alleles for TK of the Z8 and Z9 chromosomes in CHO cells. and GALK to chromosomes 7 and Z13 of CHO ADA and ITPA have been previously assigned to cells. Chinese hamster isozyme ACP1 was miss- Chinese hamster chromosome 6 (25). In CHO ing in three (cc7/H8, cc12/H5, and ccl3/H8) of cells, since no normal number 6 chromosomes the six HATr colonies which segregated chro- are present (6, 28), one would expect these loci mosome 7 and retained chromosome Z13, sug- to be located on Z group chromosomes. The gesting that the second allele for ACPJ is sepa- results from the mutation experiments indicate rated from TK and GALK in CHO cells and that alleles for both of these loci should be might be on the Zl or Z8 chromosome. Since the located on more than one Z group chromosome. three colonies without hamster isozyme ACP1 Among the 527 clones picked and analyzed for were also missing both chromosomes Zl and Z8, electrophoretic shift variations at over 40 enadditional clones, with segregation data for zyme loci after ethyl methanesulfonate exposure, two (EMS2-36 and AEM-37) were variant ACPJ, are described below. A second series of hybrids, which had been for ADA, and one (2HEME-12) was variant for generated to map the functional TK allele in ITPA. Mutations detected for other loci have TK`' heterozygous CHO-AT3-2, provided data either been previously described (17-19, 26) or confirming the synteny of TK with the Z13 will be presented elsewhere after subcloning for chromosome and allowed mapping of ACPI in verification. The ADA and ITPA variants were CHO cells. These hybrids were the products of verified as mutations by recovery of the variant fusions of LTK- CllD cells with EDGAL1-1, a patterns in all subclones (e.g., Fig. 3). The derivative of CHO-AT3-2. Segregation of ham- variant patterns were consistent with induced ster chromosomes and ACPI in seven BUdR- heterozygosity for mutations of both loci. Since and seven HAT-selected hybrids (designated ADA is monomeric (22), two-banded patterns with prefix R) is shown in Table 2. Chromosome are expected for heterozygotes, and since ITPA Z13 was missing in all BUdR-resistant colonies is dimeric (10), heterozygotes should be three and retained in all HAT-resistant colonies, banded (15). The appearance of the heterozywhereas the 7q- chromosome was present or gous patterns (Fig. 3) and the subclone data absent independent of the selection. These re- allow us to conclude that both loci are functionsults confirm our assignment of TK to chromo- ally diploid in CHO cells (17) and that their some Z13 and demonstrate that the functional alleles have not been inactivated by position TK allele in CHO-AT3-2 cells is on this chromo- effects despite their locations on Z group chrosome. mosomes. The presence of an ACPI allele on either To map ADA and ITPA in CHO cells, we Zl or Z8 is best resolved by examining the established a clone panel consisting of 77 subcolonies shown in Tables 1 and 2 which segre- clones derived from 10 independent TGA 102a/9 gated chromosomes 7 and 7q- for the expres- x LTK- C11D, 6 Toronto strain CHO x LTKsion of the hamster ACP1 isozyme relative to the CllD, and 18 CHO-AT3-2 x LTK- CliD inde-

1970

MOL. CELL. BIOL.

STALLINGS ET AL.

++++++++I+++I I+++

.) c) .)

I

E