Isolation and characterization of a Chinese hamster ovary cell mutant ...

Cytometry 17:33-38 (1994)

0 1.994 Wiley-Liss, Inc.

Isolation and Characterization of a Chinese Hamster Ovary Cell Mutant With Improved Staining for Indo-1' Eric D. Wieder2 and Michael H. Fox3 Department of Radiological Health Sciences and Graduate Program in Cell and Molecular Biology, Colorado State University, Fort Collins, Colorado Received for publication October 12, 1993; accepted April 14, 1994

-

Chinese hamster ovary (CHO) 10B2 cells do not stain well with indo-1 and thus cannot be used for experiments to measure intracellular calcium using this dye. We have isolated a mutant CHO cell line (CHO IS1) that stains quite well with indo-1 and that has virtually identical growth characteristics and heat sensitivity as the parent line. The mutant was isolated by sorting individual mutagenized cells with high indo-1 fluorescence and cloning them. Since it has been reported that cells with multiple drug resistance (MDR+) can pump out various fluorescent dyes, the mutant and parent lines were characterized for Hoechst 33342 staining, Adriamycin toxicity, and P-glycoprotein expression, which are markers of the MDR phenotype. P-Glycoprotein was measured with the C219 antibody using flow cytometry. Multidrug-resistant

Chinese hamster ovary (CHO) cells have been widely characterized not only in terms of their survival response to ionizing radiation and to hyperthermia, but genetically as well (6,20). The potential role of changes in intracellular calcium ( C A f + , )a s a n important factor in determining heat sensitivity (15,18,19,23)led us to attempt to measure C a t + , in CHO cells by flow cytometry. Indo-1 is the dye of choice for measuring C a + + i by flow cytometry (3,7,9,21). After much effort, we, along with others (24) (M. Borrelli, personal communication), discovered that CHO 10B2 cells simply do not stain well with indo-1. Staining with indo-1 is accomplished by incubating cells in the presence of the methoxyester form of indo-1 (jndo-liAM). This form of indo-1 is lipophilic; i t freely passes through cell membranes. Once in the cell, the ester linkages are hydrolyzed by cellular esterases, leaving the charged form of indo-1 trapped in the cell. Differences in esterase activity can result in differences in stainability for dyes of this type (9). Since we have extensive experience successfully staining CHO

cells (CHRC5)were used as positive controls. The IS1 cells stained as well with Hoechst 33342 as fixed 10B2 cells, and much better than unfixed 10B2 cells. The Is1 cells were 10- to 30-fold more sensitive to Adriamycin than the 10B2 cells, and both cell lines were much more sensitive than the CHRC5cells. The amount of P-glycoprotein was similar in both lOB2 and IS1 cell lines, but was about fivefold lower than the CHRC5 cells. Thus, the poor staining for indo-1 in the 10B2 cells may not be caused by the P-glycoprotein MDR pump, but by a different efflux pathway. Alternatively, the P-glycoprotein may be altered and less efficient in the CHO IS1 cells. 0 1994 Wiley-Liss, Inc.

Key terms: Mutant isolation, flow cytometry, multidrug resistance

cells with other dyes that require esterase activity, we did not believe that the poor staining for indo-1 could be attributed to low esterase activity. An alternative hypothesis is that the indo-1 is somehow pumped out of CHO cells. Krishan showed that drug efflux blockers such a s verapamil and trifluoperazine, which increase cellular sensitivity to certain drugs, improve the vital staining of DNA with Hoechst 33342 (H0342) in drug-resistant cells (12). The Hoechst dye is actively pumped out of drug-resistant cells, which leads to the poor DNA staining. When we attempted to stain CHO cells with indo-1 in the presence of verapamil, the cells stained

'This work was supported by PHS grants CA25636 and CA09236 awarded by the National Cancer Institute, DHHS. 'Current address: University of Colorado Health Sciences Center, Denver, CO 80262. 3Address reprint requests to Dr. Michael H. Fox, Department of Radiological Health Sciences, Colorado State University, Fort Collins, CO 80523.

34

WIEUER AND FOX

well. Unfortunately, verapamil is a C a t channel blocker, which makes its use in staining CHO cells with indo-1 to measure Ca' + i changes of little value. A common pathway for drug resistance is the multidrug resistance (MDR) phenotype, a spontaneous mutation that confers resistance t o various anticancer drugs (4,lO). The mechanism for MDR is the elevated expression of P-glycoprotein, a membrane-spanning protein with a structure like a membrane transport molecule, including a n ATP binding site (2). It is not known whether the drugs bind directly to P-glycoprotein, or if they bind to a carrier protein that then is recognized by P-glycoprotein and pumped out of cells (10). A Chinese hamster drug-resistant cell line was isolated by Ling and Thompson (17) and characterized for sensitivity to various drugs. Ling's group also developed a monoclonal antibody ((3219) against a 170 kD glycoprotein known as P-glycoprotein, which is responsible for multidrug resistance (11).It binds to a cytoplasmic epitope of P-glycoprotein, and it recognizes both MDRl and MDR2 isoforms of P-glycoprotein (4). Thus, this cell line makes a n appropriate positive control to test the Adriamycin sensitivity and P-glycoprotein levels in Chinese hamster ovary cells. The level of P-glycoprotein has also been measured using (2219 with flow cytometry (14). The goal of this work was to isolate a n indo-1 staining CHO mutant and to determine whether or not the mechanism for poor staining in CHO cells was related to the MDR phenotype. +

Indo-1 Staining Cells were plated in T-75 flasks so t h a t 2 x lo6 cells were present a t the time of the experiment. To load the cells with indo-1, cells were trypsinized, centrifuged, and resuspended in 1 ml HEPES-buffered F12 medium (10 mM HEPES, 2 mM NaHC03, 10% FBS). IndoliAM was added (6 pM from a 1 mM stock solution in DMSO with 5% w/v pluronic F-127, a non-ionic detergent), and the cells were incubated for 45-60 min a t 37°C. After staining, the cells were centrifuged and resuspended in medium for 15 min at 37"C, and then run on the flow cytometer. Staining with indo-1 did not affect the intracellular pH of the cells, as measured with flow cytometry using SNARF-1 (25) (data not shown). Flow Cytometry All flow cytometric measurements and cell sorting were done on a n EPICS V (Coulter, Hialeah, FL). Indo1-stained cells were excited with the UV doublet (351364 nm) of a n argon laser (Coherent, Palo Alto, CA). To screen for bright-staining mutants, fluorescence above 408 nm was measured in each cell. An autoclone unit (Coulter) was used to sort single cells into each well of 96-well tissue flasks.

Replica Plating and Mutant Isolation The original replica plating technique to isolate mutants was first developed by Lederberg and Lederberg (16). The first mammalian cell replica plate technique was described by Goldsby and Zipser (5). A modification of the replica plating technique was perfected by MATERIALS AND METHODS Stackhouse and Bedford (22). After a single cell was Cell Culture sorted into each well of several 96-well plates containCHO 10B2 cells were originally obtained from Dr. ing 300 pl F12 mediumiwell, the cells were allowed to W.C. Dewey (University of California, San Francisco, grow for 5-7 days for colony formation. A monolayer of CA). Cells were routinely grown in Ham's F12 (Gibco, Cytodex-1 microcarrier beads (Pharmacia, Uppsala, Grand Island, NY) medium supplemented with 10% Sweden) was added to each well, and the cells were fetal bovine serum (FBS, Irvine Scientific, Santa Ana, allowed to grow (and attach to the beads) for 3 more CA). The medium was maintained a t pH 7.3 by adding days. The beads were removed from each well by rap14mM bicarbonate in a 5% CO, atmosphere. All cul- idly pipeting the medium up and down, and put into a tures were grown in a 37°C humidified incubator. Chi- replica 96-well plate. Then the original plate was nese hamster AUXBl and CHRC5 cells were kindly stained with 6 pM indo-llAM for 30 min a t 37°C with provided by Dr. V. Ling (Ontario Cancer Institute, On- the cells still attached. The colonies were screened for tario, Canada). EJ30 human fibroblasts were kindly indo-1 staining under a fluorescence microscope and provided by R.T. Johnson (Cambridge University, UK), the brightest colonies were noted. The replica was aland the BALB 3T3 cells were obtained from the Amer- lowed to grow for a few days, and the colonies of interican Tissue Culture Collection. These cell lines were est were trypsinized, grown up, and further tested for grown in the same medium and conditions a s the CHO indo-1 staining. Those cells confirmed as indo-staining cells. mutants were then tested for hyperthermia survival characteristics. Cell Mutagenesis Hyperthermia Survival Procedures CHO 10B2 cells were mutagenized according to the Cells were trypsinized (0.3% trypsin-versene) from procedure described by Jeggo and Kemp (8). ICR-191mutagenized cells (24 h in 2 pgiml ICR-191) and EMS- stock flasks 14-20 h r prior to heating. After determin(N-N-ethylmethane sulfonate) mutagenized cells (24 h ing the concentration of cells using a n electronic cell in 300 Fgiml EMS) were kindly provided by M. Stack- counter (Particle Data, Inc., Elmhurst, IL), appropriate numbers of cells were plated into T-25 tissue culture house (Colorado State University).

35

INDO-1 STAINING CHO MUTANT

flasks (Costar, Cambridge, MA). The flasks were sealed and submerged in a 45.0" 5 0.05"C waterbath for various times and then returned to the 37°C incubator. The cells were allowed to grow for 7-12 days to allow for colony formation. The cells were then fixed in a1 3:l methano1:acetic acid fixative and stained with crystal violet for colony counting. Colonies with more than 50 cells were counted. A multiplicity correction was made for each experiment, and the surviving fraction was calculated,

Growth Curves Cells (5 x lo4) were plated into several T-25 flasks with 5 ml of F12 medium and allowed to grow at 37°C. At various times after seeding, flasks were trypsinized and the concentration of cells in each flask was calculated based on the counts from the electronic cell counter. Hoechst 33342 DNA Staining Cells were grown in a monolayer, and 10-20 p M H0342 (Calbiochem, La Jolla, CA) was added from a stock solution in distilled H20. The cells were incubated for 1.5h r a t 37"C, then trypsinized, resuspended in phosphate-buffered saline (PBS) on ice, and run on the flow cytometer (same setup as for indo-1 fluorescence). For drug-efflux blocker experiments, 15 pM trifluoperazine or 100 pM verapamil (Sigma, St. Louis, MO) was added during the Hoechst incubation, as described by Krishan (12). Alternatively, cells were fixed in 70% ethanol, rinsed, and incubated in 10-20 pM H0342 for 15-20 min a t room temperature prior to flow cytometric analysis. Adriamycin Toxicity Assays Appropriate numbers of cell were plated into T-25 flasks, and the cells were allowed to grow for 14-20 hr. The medium was removed, and new medium containing various concentrations of Adriamycin (Sigma) was added. The cells were incubated for 1 h r at 37"C, and the flasks were washed twice with fresh medium. The cells were grown, fixed, and counted with the same procedure a s hyperthermia survival measurements. P-Glycoprotein Measurements Exponentially-growing cells (1 x lo6) were trypsinized, centrifuged, and resuspended in 2 ml cold PBS, and then ice-cold methanol was added dropwise while vortexing gently to a final concentration of 70% methanol. Cells were fixed in methanol for 15 min a t -20"C, and then centrifuged and resuspended in 1 ml TPBS (PBS with 0.05% Triton X-100). C219 antibody labeled with FITC (Signet Laboratories, Dedham, MA) was added to give a final concentration of 3.3 pgiml, and the cells were incubated for 30 min at 4"C, then rinsed in cold TPBS, and resuspended in 0.5 ml TPBS. Flow cytometry measurements were made immediately, using 500 mW a t 488 nm and a 515 nm blocking filter. A Cicero data collection and analysis system (Cy-

1600

_-I_

m

-1

w

1400

4

1200

I

1 EJ30

I1

; 0

1000

E

800

2

600

A00 250 0

FLUORESCENCE INTENSITY FIG.1. Comparison of indo-l staining intensity in four different cell lines. Cells were labeled with indo-1 for 45 min, rinsed, and then measured. The fluorescence intensity from indo-l was measured with a 3-decade log amplifier.

tomation, Ft. Collins, CO) was interfaced to the EPICS V for data collection and analysis.

RESULTS Initial attempts to find a n indo-staining CHO mutant proved quite difficult, as the rate of false positives detected by the assay was quite high. Therefore, we enriched the population for indo-staining mutants by doing several rounds of cell sorting prior to screening. The brightest cells were sorted and grown up for further enrichment. By the third round of enrichment, the number of brightly staining cells increased from 2% to more than 90% of the population. Several mutants that stained brightly with indo-1 were isolated. These were further screened to test their response to hyperthermia, and all of them had a sensitivity close to t h a t of wild-type cells. Finally, one of the most brightly staining mutants was selected that had a growth rate identical to that of the wild-type cells. Both the mutant and the wild-type cell lines had a doubling time of 12.5 h r and a plating efficiency of about 85%.This mutant was designated CHO IS1 (indo staining). A comparison of indo-1 staining intensity in several cell lines is shown in Figure 1. The CHO IS1 cells are about 16-fold brighter than the CHO 10B2 cells and similar to 3T3 cells. The EJ30 cells are about 3 times as bright as the CHO IS1 cells, however. The mutant phenotype is maintained in the IS1 cells for a t least 20 passages in culture. The survival of CHO IS1 cells is virtually identical to that of CHO 10B2 cells after heating a t 42°C (Fig. 2) or 45°C (Fig. 3). The IS1 cells develop thermotolerance during 42°C heating similar to the 10B2 cells, and also develop thermotolerance after a n initial heat treatment a t 45°C (data not shown). The heat sensitivity was a n important consideration in isolating the cell

36

WIEDEK AND FOX

>

\

1

@

3

ln

1 o-2

I o - ~ 6 0 1

0

2

4

0

10

T I M E AT 42” C (Hr) FIG.2. Surviving fraction of CHO 10B2 ( 0 ) and CHO IS1 (0)cells after heating at 42.0”C. The data are pooled from three experiments, and the error bars (ignored when they are smaller than the symbols) are the standard error of the mean.

line, since CHO cells are widely used in hyperthermia experiments. The effect of drug efflux blockers on vital DNA staining with H0342 was tested in IS1 cells and compared with wild-type CHO cells (Fig. 4).While the wild-type CHO cells had poor staining without drug efflux blockers, the IS1 cells needed no blockers to stain well with Hoechst 33342. The sensitivity to the chemotherapy agent Adriamycin was tested in both cell lines (Fig. 5). This drug is often used as a test for the MDR phenotype (13). As a positive control, MDRt cells (CHRC5)were used. The parental cell type of this mutant (AUXB1) was also evaluated. The CHO IS1 cells were at least tenfold more sensitive to 3 ygiml Adriamycin than the CHO 10B2 cells. The CHO 10B2 cells were also very sensitive compared with the CHRC5 MDR cells, and were the same a s the AUXBl cells. P-Glycoprotein was also measured for both normal CHO and the IS1 mutant cell lines, a s well as the AUXBl and CHRC5cells (Fig. 6). The CHRC5cells had about 5 times the level of P-glycoprotein as the CHO 10B2 cells. There was little difference among the 10B2, IS1, and AUXBl cells, however. +

DISCUSSION We have reported the isolation and partial characterization of a Chinese hamster ovary mutant that stains well with indo-1 (16-fold greater than CHO 10B2 cells) and can therefore be used to measure Ca’ using indo-1. This is important, because the CHO 10B2 cells do not stain adequately with indo-1 and therefore cannot be used. We have also tested CHO K1 cells and hamster V79 cells and they also do not stain well with

10

20

30

T I M E AT 45” C (Min) FIG.3. Surviving fraction of CHO 1082 ( 0 ) and CHO IS1 io) cells after heating at 45°C. The data are pooled from three experiments, and the error bars (ignored when smaller than the symbols) are the standard error of the mean.

indo-1. Whether this is a general phenomenon in hamster cells is not known. The isolation of this mutant was based upon sorting individual mutagenized cells by intensity of indo-1 fluorescence. This is a novel method for mutant isolation and is very effective in isolating mutants with a high level of a fluorescent compound. Of particular interest in isolating this mutant was the necessity of similar responses to hyperthermia as the parental CHO cells, since we were interested in studying the effects of hyperthermia on Ca’ ,. The CHO IS1 mutant has virtually identical sensitivity to hyperthermia a t both 42°C and 45°C a s the parental CHO cells, and it has the same growth rate. We have used this Chinese hamster mutant in analyzing the effects of hyperthermia on C a + + , ,and we report our results in a separate publication (E.D. Wieder and M.H. Fox: “The role of intracellular calcium in the cellular response to hyperthermia,” submitted to Int. J. Hyperthermia). Since Krishan has shown that drug efflux via the MDR pathway prevents adequate cell staining of DNA by H0342 (121, the hypothesis that CHO 10B2 cells had poor staining with indo-1 due to a n efflux pump seemed plausible. Recently, Krishan discovered that drug-resistant cells (MDR’) also stain poorly with indo-1 (13). The Hoechst data support the hypothesis that the mechanism for increased indo-1 staining is the inactivation of a drug-efflux pump, since the IS1 cells had much better DNA staining with H0342 than the 10B2 cells. The fact that 10B2 cells treated with verapamil or trifluoperazine stained as well with H0342 a s IS1 cells also supported the hypothesis that the differ-

37

INDO-1 STAINING CHO MUTANT

CHANNEL NUMBER

FIG.4. Hoechst 33342 fluorescence histograms of DNA content in CHO 10B2 cells (top row) and CHO IS1 cells (bottom row). A: Hoechst 33342 staining in unfixed cells with no drugs; B: with 100 pM verapamil; C: with 15 pM trifluoperazine. D: Cells fixed in 70% ethanol prior to staining.

2500

0

2 1

2000

w

u 1500

0 [r

;

1000

2

3 500

0 10

101

102

10'

FLUORESCENCE INTENSITY lo-* 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

FIG.6. Flow cytometry measurement of P-glycoprotein with C219 antibody in CHO 10B2, CHO IS1, AUXB1, and CH'C5 cell lines. All cells were fixed in 7 0 9 cold methanol prior to labeling with antibody.

ADRIAMYCIN (pg/rnl) FIG.5. Toxicity of Adriamycin to CHO 10B2 (o),CHO IS1 (el, AUXB1 (01and , CH"C5 (V)cells. The cells were treated with various concentrations of Adriamycin for 1 hr a t 37"C, and then rinsed and incubated for colony formation. The data are pooled from two experiments. The error bars represent the standard error of the mean.

ence between the 10B2 and IS1 cells was in the MDR pathway. The Adriamycin toxicity studies show that normal CHO 10B2 cells are quite sensitive to Adriamycin compared with the MDR CHRC5cells, and virtually identic.al to AUXBl cells, implying that they are MDR-. The IS1 cells are a t least 30-fold more sensitive to Adriamycin than the 10B2 cells. Thus, there may be a range of MDR activity. The P-glycoprotein measurements, however, show very little difference in the amount of P-glycoprotein found in AUXB1, CHO 10B2, and CHO IS1 cells. The amount of P-glycoprotein in the IS1 cells +

thus cannot explain the difference in Adriamycin sensitivity between IS1 and 10B2 cells. This may mean that another drug efflux pathway is involved in pumping out indo-1 in CHO 10B2 cells. A novel ATP-dependent drug efflux pathway for extruding bis-carboxyethyl-carboxyfluorescein (BCECF) in epithelial cells was recently reported (1).It is unlikely that this particular pump could be operative in CHO cells, since they retain BCECF quite well. It could also be that the amount of P-glycoprotein is the same in the 10B2 and IS1 cells, but that the P-glycoprotein is altered in the IS1 cells so that it does not function effectively. Since the C2 19 antibody recognizes the P-glycoprotein from both MDRl and MRD2 gene isoforms, but only the MDRl gene product is the functioning P-glycoprotein (4), this may also be the explanation. Further experiments are necessary to distinguish among these alternatives.

38

WIEDER AND FOX

Our results indicate that the CHO 10B2 cells may have an efflux pump independent of the MDRi pathway that is capable of extruding indo-1, H0342, and, to some extent, Adriamycin. This pump is inactivated in the CHO IS1 cell line. Alternatively, the P-glycoprotein pump may be altered to function less efficiently. Since CHO 10B2 cells are also poorly stained with fura-2 (another calcium indicator) and SBFI (a sodium indicator) and the IS1 cells stain well with these dyes (unpublished observation), the CHO IS1 cell line should prove useful in future studies in which fluorescence measurements of Ca ' +,or N a + are desired.

ACKNOWLEDGMENTS The authors thank George Amorino and Kim Fagen for their help with the Adriamycin toxicity and P-glycoprotein experiments. We also thank Dr. Victor Ling for kindly providing the CHRC5 and AUXBl cells and Dr. R.T. Johnson for kindly providing the human EJ30 cells.

LITERATURE CITED 1. Allen CN, Harpur ES, Gray TJB, Simmons NL, Hirst BH: Efflux of bis-carboxyethyl-carboxyfluorescein(BCECF) by a novel ATPdependent transport mechanism in epithelial cells. Biochem Biophys Res Commun 172:262-267, 1990. 2. Bradley G. Guranka PF, Ling V: Mechanisms of multidrug resistance. Biochem Biophys Acta 948:87-128, 1988. 3. Chused TM, Wilson HA, Greenblatt D, Ishida Y, Edison LJ, Tsien RY, Finkelman FD: Flow cytometric analysis of murine splenic B lymphocyte cytosolic free calcium response to anti-IgM and antiIgD. Cytometry 8:396-404, 1987. 4. Dalton WS, Miller TP: Multidrug resistance. Prin Prac Oncol 5:l-13, 1991. 5. Goldsby RA, Zipser E: The isolation and replica plating of mammalian cell clones. Exp Cell Res 54:271-275, 1969. 6. Gottesman MM: Chinese hamster mutants. In: Molecular Cell Genetics, Vol. 6, Gottesman MM (ed). Academic Press, San Francisco, 1973, 264. 7. Grynkiewicz G, Poenie M, Tsien RY: A new generation of Ca2 + indicators with greatly improved fluorescence properties. J Biol Chem 260:3440-3450, 1985. 8. Jeggo PA, Kemp LM: X-ray sensitive mutants of Chinese hamster ovary cell line. Isolation and cross-sensitivity to other DNAdamaging agents. Mutat Res 112:313-327, 1983. 9. June CH, Rabinovitch PS: Flow cytometric measurement of cellular ionized calcium concentration. Pathol Immunopathol Res 6:409-432, 1988.

10. Kartner N, Ling V: Multidrug resistance in cancer. Sci Am 260: 44-51, 1989. 11. Kartner N, Evernden-Porelle D, Bradley G, Ling V: Detection of P-glycoprotein in multidrug resistant cell lines by monoclonal antibodies. Nature 316320-823, 1985. 12. Krishan A: Effect of drug efflux blockers on vital staining of cellular DNA with Hoechst 33342. Cytometry 8:642-645, 1987. 13. Krishan A: Rapid determination of cellular resistance-related drug efflux in tumor cells. In: Methods in Cell Biology, Vol. 33, Flow Cytometry, Darzynkiewicz 2, Crissman HA (eds).Academic Press, San Diego, 1990, pp 491-500. 14. Krishan A, Sauerteig A, Stein JH: Comparison of three commercially available antibodies for flow cytometric monitoring of P-glycoprotein expression in tumor cells. Cytometry 12731-742, 1991. 15. Landry J , Crete P, Lamarche S, Chretien P: Activation of Ca' dependent processes during heat shock: Role in cell thermoresistance. Radiat Res 113:426-436, 1988. 16. Lederberg J , Lederberg EM: Replica plating and indirect selection of bacterial mutants. J Bacteriol 63:399-406, 1952. 17. Ling V, Thompson LH: Reduced permeability in CHO cells as a mechanism of resistance to colchicine. J Cell Physiol83:103-116, 1974. 18. Malhotra A, Heynen MLP, Lepock J R Role of extracellular calcium in the hyperthermic killing of CHL V79 cells. Radiat Res 112:478-489, 1987. 19. Mikkelsen RB, Reinlib L, Donowitz R, Zahniser D: Hyperthermia effects on cytosolic ICa"1: Analysis at the single cell level by digitized imaging microscopy and cell survival. Cancer Res 51: 359-364, 1991. 20. Puck TT: The Mammalian Cell as a Microorganism: Genetic and Biochemical Studies in Vitro. Holden-Day Inc., San Francisco, 1972. 21. Rabinovitch PS, June CH, Grossman A, Ledbetter JA: Heterogeneity among T cells in intracellular free calcium responses after mitogen stimulation with PHA or anti-CD3. Simultaneous use of indo-1 and immunofluorescence with flow cytometry. J Immunol 137:952-961, 1986. 22. Stackhouse MA, Bedford JS: IRS-20: An ionizing radiation sensitive mutant of CHO cells. I. Isolation and initial characterization. Radiat Res 136:241-249, 1993. 23. Stevenson MA, Calderwood SK, Hahn GM: Rapid increases in inositol triphosphate and intracellular Ca after heat shock. Biochem Biophys Res Commun 137:826-833, 1986. 24. Vidair CA, Wang Z, Dewey WC: Noninvolvement of the heatinduced increase in the concentration of intracellular free Ca' in killing by heat and induction ofthermotolerance. Radiat Res 124: 156-164, 1990. 25. Wieder ED, Hang H, Fox MH: Measurement of intracellular pH using flow cytometry with carboxy-SNARF-1. Cytometry 14: 916-921, 1993. +

+

+