Isolation and Characterization of Dexamethasone-Resistant Mutants

Vol. 1, No. 6

MOLECULAR AND CELLULAR BIOLOGY, June 1981, p. 512-521

0270-7306/81/060512-10$02.00/0

Isolation and Characterization of Dexamethasone-Resistant Mutants from Human Lymphoid Cell Line CEM-C7 JEFFREY M. HARMON* AND E. BRAD THOMPSON Laboratory of Biochemistry, Division of Cancer Biology and Diagnosis, National Cancer Institute, Bethesda, Maryland 20205

Received 4 February 1981/Accepted 30 March 1981

Fifty-four independent dexamethasone-resistant clones were isolated from the clonal, glucocorticoid-sensitive human leukemic T-cell line CEM-C7. Resistance to 1 uAM dexamethasone was acquired spontaneously at a rate of 2.6 x 10-5 per cell per generation as determined by fluctuation analysis. After mutagenesis with N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), the phenotypic expression time for dexamethasone resistance was determined to be 3 days. Spontaneous acquisition of resistance to 0.1 mM 6-thioguanine appeared to occur at a much slower rate, 1.6 x 10-6 per cell per generation. However, the expression time after MNNG mutagenesis for this resistant phenotype was greater than 11 days, suggesting that the different rates of acquisition for the two phenotypes measured by fluctuation analysis were the results of the disparate expression times. The mutagens ICR 191 and MNNG were effective in increasing the dexamethasoneresistant fraction of cells in mutagenized cultures; ICR 191 produced a 35.6-fold increase, and MNNG produced an 8.5-fold increase. All the spontaneous dexamethasone-resistant clones contained glucocorticoid receptors, usually less than half of the amount found in the parental clone. They are therefore strikingly different from dexamethasone-resistant clones derived from the mouse cell lines S49 and W7. Dexamethasone-resistant clones isolated after mutagenesis of CEMC7 contained, on the average, lower concentrations of receptor than did those isolated spontaneously, and one clone contained no detectable receptor. These results are consistent with a mutational origin for dexamethasone resistance in these human cells at a haploid or functionally hemizygous locus. They also suggest that this is a useful system for mutation assay. The potent lympholytic effect of glucocorticoid steroids on certain lymphoid cell lines (15, 22) has provided a simple system with which to select steroid-nonresponsive variants (2, 3, 23, 29) and thus to attempt to identify genetically distinct components in the pathway of steroid hormone action. Steroid-resistant variants isolated from the mouse cell line S49 arose spontaneously at a rate of 3.5 x 10-6 per cell per generation (20) and were predominantly glucocorticoid receptor negative (r-) (30). Nearly all the others were characterized as having receptors with a decreased capacity for nuclear transfer and a low affinity for deoxyribonucleic acid (nt-) (9, 30) or an unusually high degree of nuclear transfer with correspondingly higher than normal affinity for deoxyribonucleic acid (nt1) (30). In a second murine cell line, W7, Bourgeois and Newby (3) demonstrated that complete steroid resistance could only be acquired by selection against increasing concentrations of dexamethasone in two steps. Each step 512

involved the loss of approximately half the dexamethasone receptor sites of the cells. On the basis of these data it was postulated that the difference between S49 and W7 was an effect of receptor gene dosage. Though among the second-step steroid-resistant variants of W7 a few nt- clones were also observed (4, 23), defects at steps in the pathway of steroid action other than those directly attributable to receptor function have not been conclusively demonstrated. In the cases of both the nt- and the nte phenotypes, demonstration of altered receptor properties provided strong evidence for the mutational origin of resistance (23, 30, 37). In the case of the r- phenotypes, a mutational origin was inferred from the stability of phenotype, the increased frequency of steroid resistance after mutagenesis, and the rate of spontaneous acquisition of steroid resistance (3, 4, 29). The apparent uniformity in the phenotype of steroid resistance in S49 and W7 raised the question of whether such uniformity would ex-

VOL. 1, 1981

DEXAMETHASONE RESISTANCE IN CEM-C7 CELLS

tend to resistant variants derived from other steroid-sensitive cells. That such uniformity might not extend beyond these two cell lines was suggested by results with mouse L cells and rat HTC cells. L cells selected for lack of growth inhibition by glucocorticoids continue to elaborate other steroid-inducible functions (35). Likewise, HTC cells identified by lack of plasminogen activator response to steroids retain other steroid-inducible functions (6), as do HTC cells identified by lack of induction of tyrosine aminotransferase (34). Presumably, expression of these responses requires the presence of functional glucocorticoid receptors. In addition, various steroid-resistant human leukemic celLs obtained directly from patients have been found to possess glucocorticoid receptors which in some cases have been shown capable of undergoing nuclear translocation (8). To explore systematically the cellular mechanisms for the acquisition of steroid resistance in human leukemic cells, free from the complications of in vivo influences, we have chosen to examine the acquisition of steroid resistance in the glucocorticoid-sensitive human leukemic cell line CEM-C7 (22). Fiftyfour independent, spontaneously resistant clones were isolated and characterized. None of these clones was devoid of glucocorticoid receptor. Sixty-one clones obtained after mutagenesis with N-methyl-N'-nitro-N-nitrosoguanidine or ICR 191 contained, on average, less receptor than did the spontaneous isolates, but only one contained no receptor. Several properties of these clones suggest an essential difference from S49 and W7 cells. MATERIALS AND METHODS Chemicals. Dexamethasone, 6-thioguanine (6TG),

agarose (EEO type II) and MNNG were obtained from Sigma Chemical Co., St. Louis, Mo. ICR 191 was obtained from Polyscience Inc., Warrington, Pa. [12(n)-3H]dexamethasone (38 Ci/mmol) was purchased from Amersham Corp., Arlington Heights, Ill. MNNG dissolved in 5 mM sodium acetate, pH 4.5, at a concentration of 20 ug/ml was prepared immediately before use, as was ICR 191, which was dissolved in water at a concentration of 100 pg/ml. Cell growth and culture. The isolation and characterization of the glucocorticoid-sensitive human leukemic cell line CEM-C7 have been previously described (14, 22). Cultures of CEM-C7 were routinely grown in RPMI 1640 (National Institutes of Health Media Unit, Bethesda, Md.) containing 10% fetal calf serum (North American Biologicals, Inc., Miami, Fla.) at 37°C in a humidified atmosphere of 95% air-5% CO2 as stationary suspensions. There was little or no clumping of cells. Cell concentrations were determined with a Coulter Counter (model ZB; Coulter Electronics, Inc., Hialeah, Fla.), and cultures were routinely maintained at between 105 and 106 cells per ml. Isolation of clones. Cultures of CEM-C7 were

513

diluted to a density of 1 cell per ml and distributed in 0.2-ml portions into the wells of 96-hole microtiter tissue culture plates (Falcon Plastics, Oxnard, Calif.); this ensured that no well received more than one cell. Plates were incubated for 10 to 14 days and examined microscopically for the presence of colonies. Each well containing a well-defined colony of 500 to 1,000 cells was gently agitated with a Pasteur pipette, and the contents were delivered into a 25-cm2 tissue culture flask containing 5 ml of medium. The plating efficiency of cells plated by this method was approximately 50%. Alternatively, when a larger number of cells were to be plated, plating was performed in 0.225% agarose over a feeder layer of human diploid fibroblasts (ArMor) in 60-mm petri dishes as previously described (14). In 14 to 20 days, cells plated in this manner formed spherical colonies 0.5 to 1 mm in diameter which could easily be isolated with a Pasteur pipette or stained and counted. The plating efficiency for cells plated by this method was 40 to 60%. Fluctuation analysis. For each fluctuation experiment, a fresh subclone (500 to 1,000 cells) of CEM-C7 isolated from a microtiter well was dispersed into twenty 25-cm2 tissue culture flasks containing 5 ml of medium. One week later these cultures were transferred to 75-cm2 tissue culture flasks with the addition of 5 ml of fresh medium. Each of the cultures was allowed to grow to a density of 5 x 105 to 10 x 105 cells per ml and then plated in semisolid medium as described above (five plates per culture) in the presence of the selective agent at a density of 105 cells per 60-mm plate. Samples of each culture were also plated in the absence of the selective agent (250 cells per plate) to determine the absolute plating efficiency of each culture. Colonies appearing after 2 weeks were stained and counted. Isolation of spontaneous dexamethasone-resistant clones. One colony from each set of replicate plates of each fluctuation experiment was isolated with a Pasteur pipette and returned to liquid culture, thus ensuring the independence of the resistant clones. The frequency of resistant clones was independent of the number of cells plated and thus not subject to crossfeeding phenomena. Isolation of mutagen-induced dexamethasoneresistant clones. Cultures of CEM-C7 cells were treated with 1 pg of ICR 191 per ml for 18 h or 1 pg of MNNG per ml for 2 h. Cells were washed twice in complete medium, and a portion of them was plated to determine the extent of mutagen toxicity. The remaining ceels were grown for 5 days in complete medium in the absence of dexamethasone and then plated in agarose containing 1 pM dexamethasone at 105 cells per plate. Cells were also plated in the absence of dexamethasone to determine the absolute plating efficiency. The resistant fraction was calculated as the fraction of cells plated in the presence of 1 AM dexamethasone which formed colonies, divided by the absolute plating efficiency of the cells. Colonies appearing on plates containing dexamethasone were isolated for further analysis. Measurement of glucocorticoid receptor content. The quantity of glucocorticoid receptor was measured by a modification of the whole-cell binding assay of Sibley and Tomkins (30). Cells were harvested by

514

HARMON AND THOMPSON

MOL. CELL. BIOL.

centrifugation and suspended in RPMI 1640 contain- tion of resistance is a random process occurring ing 50 mM Tricine, 0.5 g of sodium bicarbonate per in the absence of steroid. Calculated rates of liter, and 10% fetal calf serum, pH 7.2, at a density of acquisition of the resistance phenotype in these 107 cells per ml. Ten microliters of 5 uM [3H]dexaexperiments were 1.98 x 10-5 and 3.32 x methasone in 95% ethanol and 10 pl of 95% ethanol two were added to duplicate 0.98-ml samples of this sus- 10-5 per cell per generation (Table 2). In the same series of experiments the rate of pension. Two additional samples received 10 p1 of 5 uM [3H]dexamethasone and 10 pA of 1 mM unlabeled acquisition of Dexr was compared with that for dexamethasone in 95% ethanol. The presence of the acquisition of resistance to 6TG (6TGr). 6TGr 200-fold excess of unlabeled steroid effectively com- results from mutation in the X-chromosomal peted for all of the binding of [3H]dexamethasone to locus encoding the enzyme hypoxanthine phosspecific glucocorticoid receptors. Cells were incubated phoribosyltransferase and thus represents mufor 1 h at 37°C and then harvested by centrifugation tation in a functionally hemizygous gene (7). at 1,200 x g for 1 min. They were washed three times in 3.0 ml of Hanks balanced salt solution (Na2PO4, 48 Cells from each of the 20 subcultures of experimg/liter, NaHCO3, 350 mg/liter; NaCl, 8 g/liter, ment 2 of Table 2 were also plated in medium MgSO4, 97.7 mg/liter; KH2PO4, 60 mg/liter, KCl, 400 containing 0.1 mM 6TG. The results (Table 2 mg/liter, CaCl2, 140 mg/liter; glucose, 1 g/liter) and and Fig. 2c) showed that the rate of acquisition finally suspended in 1.6 ml of Hanks balanced salt of Dexr appeared to be approximately 20 times solution. A 0.2-ml amount of this suspension was used greater than that for 6TGr (3.32 x 10-5 versus for the determination of cell concentration, and 1.0 ml 1.6 x 10-6 per cell per generation). This differwas assayed for radioactivity by liquid scintillation ence in rate could have reflected increased mumethods in 10 ml of Aquasol, using a Beckman LS tational susceptibility of the gene(s) encoding 9000 liquid scintillation counter programmed to cal- the lympholytic response to steroids; a large culate disintegrations per minute. The difference in disintegrations per minute per cell between those sam- number of such genes, each being required for ples incubated with 50 nM [3H]dexamethasone alone sensitivity; or a nonmutational origin for steroid and those containing the 200-fold excess of unlabeled resistance. Alternatively, the difference could steroid represented the binding of [3H]dexamethasone have resulted from the nature of the fluctuation to specific glucocorticoid receptors. Using the specific analysis itself. The experiments performed to activity of the [3H]dexamethasone, the number of measure the rate of acquisition of resistance receptors per cell was calculated, assuming that each detected only those cells phenotypically able to receptor binds one steroid molecule. survive selection, as opposed to those possibly

RESULTS We have previously shown that glucocorticoids exert a potent lympholytic effect on the clonal human T-cell leukemic cell line CEM-C7, killing virtually all of the treated cells (14, 22). However, when cells are plated in semisolid medium in the presence of 1 uM dexamethasone, a small number form colonies. When these colonies were isolated and tested for steroid sensitivity, they were found to be completely resistant to the cytolytic effects of dexamethasone. The lack of dexamethasone inhibition of growth of several such clones is shown in Fig. 1. This resistance phenotype (Dexr) was extremely stable. Cultures of resistant clones grown in the absence of steroid for 18 months showed no significant reduction in plating efficiency in the presence of steroid (Table 1). To ascertain if the steroid resistance phenotype was acquired in the absence of selective pressure and to determine the rate of acquisition of the resistance phenotype, Luria-Delbruck fluctuation analyses (19) were performed as described in Materials and Methods. The results from two such experiments are shown in Table 2 and Fig. 2a and b. In both experiments x2 analysis of the data demonstrated that acquisi-

containing mutations in the gene(s) encoding products responsible for sensitivity, but whose products had yet to be degraded or diluted. Accordingly, the phenotypic expression times of both the Dexr and 6TGr phenotypes were measured after mutagenesis with MNNG (Fig. 3). Dexr was expressed rapidly after mutagenesis, with a stable Dexr fraction of -1.6 x 10-3 achieved 2 to 4 days after mutagenesis. In contrast, 6TG resistance was acquired much more slowly. Four days after mutagenesis, at a time when a maximal Dexr fraction was already observed, there was essentially no increase in the 6TGr fraction. An increase in 6TGr cells was not observed until 7 days after mutagenesis, and even as long as 11 days after mutagenesis a stable 6TGr fraction had not been reached. The results of this experiment therefore indicated a significantly longer expression time for 6TGr than for Dexr and suggested that the apparent difference in rate of acquisition of the two resistant phenotypes measured by fluctuation analysis may have simply reflected the large difference in expression time for the two phenotypes. If this was the case and the actual mutation rates were similar, after a large number of generations, the spontaneous Dexr fraction would have been nearly equal to the spontaneous 6TGr

VOL. 1, 1981

DEXAMETHASONE RESISTANCE IN CEM-C7 CELI5

515

0 -

100

3R43

C7

80-

50

30

-

20 -P

10

1

2

8

x

E 300 -J

31S741

ys

200-

00

50~~~~~~~~~~~~ 30

-

20

10

TIME OF EXPOSURE

(days)

FIG. 1. Growth of glucocorticoid-sensitive and glucocorticoid-resistant clones in dexamethasone. The glucocorticoid-sensitive clone CEM-C7and three steroid-resistant subclones, 3R7, 3R43, and 4R4, weregrown dexamethasone. Cell densities uere measured with a in the absence (0) or presence of (A) 0.02 or (0) 1 Coulter Counter. TABLE 1. Plating efficiencies of steroid-sensitive and steroid-resistant clones in the presence of dexamethasonea

Plating efficiency of clone: Dexamethasone CEM-C7

3R7

3R43

4R4

None 0.65 ± 0.06 0.44 ± 0.05 0.24 ± 0.01 0.58 ± 0.09 1 ,uM 0.74 ± 0.07 x 10 3b 0.39 ± 0.13 0.18 ± 0.01 0.48 ± 0.05 'Two hundred fifty cells were plated in quintuplicate in 60-mm dishes as described in the text. Colonies present after 2 weeks were stained and counted. Results are expresaed as the mean ± standard deviation of the absolute plating efficiency in each experiment. b 105 celLs were plated in each dish.

fraction. Indeed, when a subclone of CEM-C7 (CEM-C7-6S) was grown for 100 generations and assayed for Dexr and 6TG' fractions, it was found that the Dex' fraction (1.1 x 10-3) was only 2.4fold greater than the 6TGrfraction (4.6 x 10-')

(Table 3) Using this method, the rates of acquisition of the Dexr and 6TG' phenotypes were calculated to be 1.5 x 10-5 per cell per generation and 6.3 x 10' per cell per generation, respectively (Table 3). Although rates calculated in

MOL. CELL. BIOL.

HARMON AND THOMPSON

516

TABLE 2. Rate of acquisition of Dexr and 6TGrphenotypes Phenotype

tPheno-

Est Expt

nlo.

Mean no. of colonies/105 platedacells

X2b

Variance

1.98 x 10-5 31.9 (18 df,c P = 0.02) 5.8 9.7 1 3.32 X 10-5 17.72 296.3 334.4 (19 df, P < 0.005) 2 1.6 X 10-6 25.4 343.3 (19 df, P < 0.005) 2 1.48 6TGr a Corrected for plating efficiency in the absence of steroid of cells from each of the cultures plated. b Calculated by the maximal-likelihood method of Lea and Coulson (16). 'One culture was lost to contamination. DeXr

7

a. 6 5 0

4

1.2

z 0.0 o F

3

L.i

2

0.8

z

0.6 n

n

_

0 4 8 12 16M2( 24 28323640 M

4

56

646872

u) 4

,

bb. 3

DAYS AFTER MUTAGENESIS UJ

2 I

z 0

I

0 4 8 12 1620242832364044485256606468 72

C.

15 4

FIG. 3. Phenotypic expression times for Dexr and 6TG' after mutagenesis with MNNG. Cells of a fresh subclone of CEM-C7 were treated with 1 pg of MNNG per ml for 2 h. Cells were washed free of mutagen and grown for various lengths of time before plating in 1 pM dexamethasone (a) or 0.1 mM 6TG (A). At each time point cells were also plated in the absence of selective agent to determine absolute plating efficiencies. Results are expressed as relative plating efficiencies in the presence of the selective agent with the resistant background on day zero subtracted.

3

0 4 8 12 16 20 24 28 32 36 40

44

48 52 56 60 64 68 72

RESISTANT COLONIES/105 CELLS

FIG. 2. Fluctuation analysis of dexamethasone resistance and 6TG resistance. Luria-Delbruck fluctuation analyses were performed on fresh subclones of CEM-C7 as described in the text. Colonies appearing on plates containing dexamethasone (experiments 1 [a] and 2 [bl of Table 2) or 6TG (c) were stained and counted. Cells from each subculture were also assayed for absolute plating efficiency, and the results were calculated as resistant colonies per 105 cells based on relative plating efficiency. Results are expressed as the number of cultures containing a given number of resistant clones per 105 cells after correcting for plating efficiency.

this manner are subject to larger errors, the results suggested that the rates of acquisition of the Dexr and 6TGr phenotypes were more similar than those derived from the fluctuation analysis and entirely consistent with Dexr being acquired through mutation at one or a few haploid or functionally hemizygous loci. Quantitation of steroid receptors in Dexr subelones. Each of 54 DexT clones isolated as described in Materials and Methods was assayed for its glucocorticoid receptor content (Fig. 4). In contrast to the results reported for the murine lymphoid cell lines S49 and W7, where the principal phenotype was complete loss of receptor (4, 23, 30) each of the Dexr clones contained measurable quantities of steroid receptors. Compared with the receptor content of the steroidsensitive parent CEM-C7, the relative receptor quantities in the resistant subclones varied be-

VOL. 1, 1981


TABLE 3. Acquisition of resistance in long-term culturesa fraction (acquisition rate) Resistant No. of generations 6Mr DeX' 4.1 x 10-4 (1.6 x 10-5) 1.1 x 10-3 (1.5 x 10-5) 4.6 x 10-4 (6.3 x 106) aCells were plated in 1 LM dexamethasone or 0.1 mM 6TG, and the number of colonies present after 2 weeks was counted. Results were corrected for the plating efficiency of unselected controls. The numbers in parentheses are rates of acquisition of resistance calculated from the equation

35 100

21n2

IM2 Ml _M \N2N,

mutation rate =

g where M = number of mutants, N = number of cells, and g = number of generations. 10 z 8l

0 D

0

0

20 40 60 80 100>100 % WILD TYPE RECEPTOR

FIG. 4. Glucocorticoid receptor contents of spontaneously derived Dexr clones. Steroid-resistant clones were isolated in four different experiments. A small number of CEM-C7 cells (1,200 sites per cell) glucocorticoid binding capacity. Pfahl et al. (23) have reported that 84% of spontaneously derived Dexr clones of W7 and S49 contained no detectable receptor. Sibley and Tomkins (29) reported that 79% of S49 Dexr clones lacked receptors. The remaining Dexr clones generally had less receptor than the sensitive parents and were characterized as having decreased nuclear translocation of steroid-receptor complexes with decreased affinity for deoxyribonucleic acid (nt-) (9, 23, 30) or higher than nornal levels of nuclear translocation with increased affinity for deoxyribonucleic acid (nt') (30, 37). Preliminary characterization of our Dexr clones suggests that these clones contain receptors defective in forming the "activated" complexes prerequisite for nuclear translocation (26; J. M. Harmon, T. Schmidt, and E. B. Thompson, manuscript in preparation) rather than the postactivation defects in nuclear interaction described previously. Thus, although it seems probable that, as in the murine cell lines S49 and W7, steroid resistance in CEM-C7 is a result of defective steroid receptors or defective factors modulating receptor activity, it appears that the specific targets for resistance differ. Glucocorticoid-nonresponsive variants isolated from mouse L cells (13, 18) contained markedly reduced levels of glucocorticoid binding activity. Likewise, HTC cells selected for loss of steroid-inducible mouse mammary tumor virus have lost steroid binding capacity (11). Stevens and Stevens (32) have shown that a steroidresistant line of the mouse transplantable lymphoma P1798 contains glucocorticoid receptors significantly smaller than those of the glucocorticoid-sensitive parent line. These receptors have higher affinity for nuclei and deoxyribonucleic acid and could be similar to the nt' receptors of S49 (31). Venetianer et al. (35) have described L-cell variants which show reduced

MOL. CELL. BIOL.

growth sensitivity to dexamethasone and yet retain normal receptor concentration and continue to express other steroid-specific responses. However, the extent of these responses is reduced, and thus it is possible that they contain partially defective receptors. HTC variants isolated for lack of induction of tyrosine aminotransferase retained inducibility of glutamine synthetase and of mouse mammary tumor virus and suppressibility of a-aminoisobutyric acid transport and plasminogen activator (34). In these cases the sequential selection of low inducers makes it unlikely that lack of response to steroid is the result of a single genetic lesion. HTC cells with nonsuppressible plasminogen activator but inducible tyrosine aminotransferase and suppressible a-aminoisobutyric acid transport have also been isolated (6). However, since these variants arose at a spontaneous rate of 3.5 X 1o-3 to 4.0 x 1O-3 per cell per generation (28) and since this resistance was acquired through loss of an inducible plasminogen activator inhibitor (27), it is unlikely that receptor mutants occurring at a rate similar to that which we have determined for CEM-C7 or that measured for S49 or W7 would have been detected. Thus, it appears that in those cases where a mutational basis for glucocorticoid resistance can be inferred, the steroid receptor is the target. Androgen receptors in patients with testicular feminization have been reported to be absent or defective, thus establishing a correlation between defective steroid response and altered receptor function for this genetic disorder (10). A tamoxifen-resistant clone of MCF-7 also contains defective estrogen receptors (M. Lippman, personal communication). These results suggest that loss of receptor function may be the predominant mechanism of loss of steroid response for other classes of steroid hormones as well. A large number of Dexr clones have been isolated from the human leukemic cell line CEM-C7. Resistance is acquired in a manner consistent with somatic mutation, although direct proof must await detailed biochemical analyses. This system appears to have several advantages for use in human cell mutation assays: Dexr is a phenotype with short expression time, and there is no apparent requirement for residual receptor function. Thus, the entire locus can serve as a target for mutation. The abilities of ICR 191 and MNNG to induce mutation demonstrate the susceptibility of this locus to mutagenesis. Patients with acute lymphoblastic leukemias treated with corticosteroids alone almost uniformly become refractory to therapy (36). It has been reported that in childhood acute lymphoblastic leukemias, those patients with low receptor levels are least likely to respond to


VOL. 1, 1981

therapy (17). Whether the acquired steroid resistance of leukemic cells in culture represents an in vitro model for the study of acquired resistance to chemotherapy is a subject of continued investigation. ACKNOWLEDGMENTS A portion of this work was performed while J.M.H. was a postdoctoral fellow in cancer research of the Damon RunyonWalter Winchell Cancer Fund (grant DRG-116F). We thank Thomas Schmidt for assistance in performing some of the receptor assays and J. Regan for editorial assistance in the preparation of this manuscript. LITERATURE CITED 1. Albertini, R. J., and R. DeMars. 1973. Detection and quantification of x-ray-induced mutation in cultured, diploid human fibroblasts. Mutat. Res. 18:199-224. 2. Baxter, J. D., A. W. Harris, G. M. Tomkins, and M. Cohn. 1971. Glucocorticoid receptors in lymphoma cells in culture: relationship to cell killing activity. Science 171:189-191. 3. Bourgeois, S., and R. F. Newby. 1977. Diploid and haploid states of the glucocorticoid receptor of mouse lymphoid cell lines. Cell 11:423-430. 4. Bourgeois, S., R. F. Newby, and M. Huet. 1978. Glucocorticoid resistance in murine lymphoma and thymoma lines. Cancer Res. 38:4279-4284. 5. Buchwald, M. 1977. Mutagenesis at the ouabain-resistance locus in human diploid fibroblasts. Mutat. Res.

44:401-412. 6. Carlson, S. A., and T. D. Gelehrter. 1977. Hormonal regulation of membrane phenotype. J. Supramol. Struct. 6:325-331. 7. Caskey, C. T., and G. D. Kruh. 1979. The HPRT locus. Cell 16:1-9. 8. Crabtree, G., K. Smith, and A. Munck. 1978. Glucocorticoid receptors and sensitivity of isolated human leukemia and lymphoma cells. Cancer Res. 38:4268-4272. 9. Gehring, U., and G. M. Tomkins. 1974. A new mechanism for steroid unresponsiveness: loss of nuclear binding activity of a steroid hormone receptor. Cell 3:301306. 10. Griffin, J. E., and J. D. Wilson. 1980. The syndrome of androgen resistance. N. Engl. J. Med. 302:190-209. 11. Grove, J. R., B. S. Dieckmann, T. A. Schroer, and G. M. Ringold. 1980. Isolation of glucocorticoid-unresponsive rat hepatoma cells by fluorescence-activated cell sorting. Cell 21:47-56. 12. Gupta, R. S., and L Siminovitch. 1978. Isolation and characterization of mutants of human diploid fibroblasts resistant to diphtheria toxin. Proc. Natl. Acad. Sci. U.S.A. 75:3337-3340. 13. Hackney, J. F., S. R. Gross, L Aronow, and W. B. Pratt. 1970. Specific glucocorticoid-binding macromolecules from mouse fibroblasts growing in vitro. A possible steroid receptor for growth inhibition. Mol. Pharmacol. 6:500-512. 14. Harmon, J. M., M. R. Norman, B. J. Fowlkes, and E. B. Thompson. 1979. Dexamethasone induces irreversible G, arrest and death of a human lymphoid cell line. J. Cell. Physiol. 98:267-278. 15. Horibata, K., and A. W. Harris. 1970. Mouse myelomas and lymphomas in culture. Exp. Cell Res. 60:61-77. 16. Lea, D. E., and C. A. Coulson. 1949. The distribution of the number of mutants in bacterial populations. J. Genet. 49:264-284. 17. Lippman, M., G. Konior-Yarbro, and B. Leventhal. 1978. Practical implications of glucocorticoid receptors in human leukemia. Cancer Res. 38:4251-4256. 18. Lippman, M. E., and E. B. Thompson. 1974. Steroid

521

receptors and the mechanism of specificity of glucocorticoid responsiveness of somatic cell hybrids between hepatoma tissue culture cells and mouse fibroblasts. J. Biol. Chem. 249:2483-2488. 19. Luria, S. E., and M. Delbruck. 1943. Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491-511. 20. Mankovitz, R., M. Buchwald, and R. M. Baker. 1974. Isolation of ouabain-resistant human diploid fibroblasts. Cell 3:221-226. 21. Moehring, T. S., and J. M. Moehring. 1977. Selection and characterization of cells resistant to diphtheria toxin and pseudomonas exotoxin A: presumptive translational mutants. Cell 11:447-454. 22. Norman, M. R., and E. B. Thompson. 1977. Characterization of a glucocorticoid-sensitive human lymphoid cell line. Cancer Res. 37:3785-3791. 23. Pfahl, M., R. J. Kelleher, Jr., and S. Bourgeois. 1978. General features of steroid resistance in lymphoid cell lines. Mol. Cell. Endocrinol. 10:193-207. 24. Pious, D., and D. Soderland. 1977. HLA variants of cultured human lyinphoid cells: evidence for mutational origin and estimation of mutation rate. Science 197: 769-771. 25. Scatchard, G. 1949. The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 51:660-672. 26. Schmidt, T. J., J. M. Harmon, and E. B. Thompson. 1980. 'Activation-labile' glucocorticoid-receptor complexes of a steroid resistant variant of CEM-C7 human lymphoid cells. Nature (London) 286:507-510. 27. Seiffert, S. C., and T. D. Gelehrter. 1978. Mechanism of dexamethasone inhibition of plasminogen activator in rat hepatoma cells. Proc. Natl. Acad. Sci. U.S.A. 75: 6130-6133. 28. Seiffert, S. C., and T. D. Gelehrter. 1979. Isolation of rat hepatoma variants selectively resistant to dexamethasone inhibition of plasminogen activator. J. Cell. Physiol. 99:333-342. 29. Sibley, C. H., and G. M. Tomkins. 1974. Isolation of lymphoma cell variants resistant to killing by glucocorticoids. Cell 2:213-220. 30. Sibley, C. H., and G. M. Tomkins. 1974. Mechanisms of steroid resistance. Cell 2:221-227. 31. Sibley, C. H., and K. R. Yamamoto. 1979. Mouse lymphoma cells: mechanisms of resistance to glucocorticoids, p. 357-376. In J. D. Baxter and G. G. Rousseau (ed.), Glucocorticoid hormone action. Springer-Verlag New York, Inc., New York. 32. Stevens, J., and Y.-W. Stevens. 1979. Physiochemical differences between glucocorticoid-binding components from the corticoid-sensitive and -resistant strains of mouse lymphoma P1798. Cancer Res. 39:40114021. 33. Thilly, W. G., J. G. DeLuca, H. Hoppe IV, and B. W. Penman. 1976. Mutation of human lymphoblasts by methylnitrosourea. Chem. Biol. Interact. 15:33-50. 34. Thompson, E. B., D. K. Granner, T. Gelehrter, J. Erickson, and G. L Hager. 1979. Unlinked control of multiple glucocorticoid-induced processes in HTC cells. Mol. Cell. Endocrinol. 15:135-150. 35. Venetianer, A., K. Bajnoczky, A. Gal, and E. B. Thompson. 1978. Isolation and characterization of L cell variants with altered sensitivity to glucocorticoids. Som. Cell Genet. 4:513-530. 36. Wolff, J., C. Brubaker, M. Murphy, M. Pierce, and N. Severo. 1967. Prednisone therapy of acute childhood leukemia: prognosis and duration of response in 330 treated patients. J. Pediatr. 70:662-631. 37. Yamamoto, K. R., M. R. Stampfer, and G. M. Tomkins. 1974. Receptors from glucocorticoid-sensitive lymphoma cells and two classes of insensitive clones: physical and DNA-binding properties. Proc. Natl. Acad. Sci. U.S.A. 71:3901-3906.