Glucocorticoid Effect on Oncogene/Growth Gene Expression in Human T Lymphoblastic Leukemic Cell Line CCRF-CEM SPECIFIC c-myc mRNA SUPPRESSION BY DEXAMETHASONE*

Glucocorticoids induce growth inhibition and even-tually cause cell lysis in certain sensitive leukemic cells. To investigate how glucocorticoids interact with cell growth pathways, we studied the expression of 14 growth-related genes in dexamethasone-treated CEM-C7A cells, a steroid-sensitive clone of the CCRF-CEM cell line, and in several closely related clones. The 14 genes studied were chosen to represent four different levels of mitogenic signal transduction. Detectable mRNA levels were found for 8 of the 14 genes, but among these only c-myc expression was obviously suppressed by dexamethasone. The c-myc mRNA levels declined abruptly during the first 12 h after addition of 1 PM dexamethasone, and maximal suppression oc- curred by 18 h. This change was not seen in the C7A controls, in the glucocorticoid-resistant, receptor-de-ficient clone ICR-27, or in the glucocorticoid-resistant, receptor-positive clone C1. H.lO, a hybrid clone between C1 and ICR-27, showed restoration of the sen- sitive phenotype, and in H.10 cells the c-myc mRNA was also suppressed by dexamethasone. Our results suggest that: 1) functional glucocorticoid receptor is required for inducing c-myc suppression. 2) In dexa- methasone-resistant cells with functional receptors c-myc is not suppressed.

It was recently reported that glucocorticoids seem to interact with pathways of importance to the control of cell growth by suppressing the expression of c-myc, c-myb, and c-Ki-ras in the mouse lymphoma cell line S49 (Eastman-Reks and Vedeckis, 1986) and of c-myc in P1798 cells (Forsthoefel and Thompson, 1987). In S49 cells, the oncogene suppression occurred much earlier than did DNA fragmentation, which was not seen in these cells until 24-48 h following addition of dexamethasone (Vedeckis and Bradshaw, 1983). To see whether these potentially important findings are general for glucocorticoid-sensitive lymphoid cells, and if so, whether the suppression of such oncogenes requires functional glucocorticoid receptor, we have employed clones from the human lymphoblastic leukemic cell line CCRF-CEM. Several cell clones with different well-studied phenotypes of steroid sensitivity have been isolated.
C7 is a glucocorticoid-sensitive clone with normal, functional glucocorticoid receptors (Norman and Thompson, 1977). Clones ICR-27 and C1 are both resistant to the lytic effect of glucocorticoids. ICR-27 is a receptor-deficient clone , while CI contains normal, functional receptors (Zawydiwski et al., 1983). In a previous study, we fused C1 and ICR-27 cells and found complementation, with restoration of the glucocorticoid-sensitive phenotype in the C1xICR-27 hybrids (Yuh and Thompson, 1987).
Various oncogenes/growth genes, chosen to include several recognized categories (Bishop, 1985;Smith et al., 1986) were selected for study. Interleukin-2, transferrin receptor, c-erb-B, c-src, and c-ab1 were chosen from the categories of growth factors, growth factor receptors, and receptor-related membrane tyrosine kinases (Doolittle e t ul., 1983;Downward e t al., 1984;Waterfield et al., 1983). From the rm family we picked c-Ki-ras and c-H-ras (Parada et al., 1982;Santos et al., 1982;Chang et al., 1982), and from the serine protein kinase family we picked c-mos and c-ruf (Kloetzer et ut., 1984;Moelling et al., 1984;Papkoff et al., 1983). As representatives of the nuclear protein family we examined c-erb -A, c-fos, c-myc, and c-myb (Mellon et al., 1978;Gonda and Bishop, 1983;Van Beverer et al., 1983;Vennstrom and Bishop, 1982). Using these genes as landmarks, we tried to investigate at which level(s) glucocorticoids generate their effects and whether there is a cascade of sequential changes between different categories of genes. The specificity of the positive finding(s) in steroid-sensitive C7A was investigated by comparing with ICR-27 (the steroid-resistant, receptor-deficient clone), C1 (steroid-resistant, receptor-positive clone), and ICR-27xCl:H.lO (steroid-sensitive, receptor-positive hybrids between . In this report we describe our major finding, the suppression of c-myc mRNA by glucocorticoid (Thompson et al., 1989), and briefly describe the results with the other 13 genes. The DNA probes used for hybridization were either purchased from Oncor Co. (v-sis, v-erb-B, v-erb-A, v-H-ras, v-mos, v-src, and cmyc third exon), Oncogene Science, Inc. (human c-ab1 pr-l), American Type Culture Collection , or were gifts from other researchers (interleukin-2, transferrin receptor, v-abl-K2, v-myc, v-H-ras, a-tubulin). The interleukin-2 probe was a 0.85-kb EcoRI-BglII fragment of MMT/IL-2/3.2 from Dr. Gerald R. Crabtree, Stanford University School of Medicine, Palo Alto, CA (Holbrook et al., 1984). The v-abl probe was a 1.45-kb EcoRI fragment of v-ab1 from Dr. Steve Tronick, National Cancer Institute, Bethesda, MD (Reddy et al., 1983). The v-myc probe was a 1.0-kb SalI-BamHI fragment from Dr. J. Michael Bishop, University of California, Hooper Research Foundation, San Francisco, CA (Alitalo et al., 1983). The a-tubulin cDNA in pILaT3 provided by Dr. Phillip A. Sharp, Massachusetts Institute of Technology, Cambridge, MA (Lemischka et al., 1981). Transferrin receptor genes were probed with mouse transferrin cDNA provided by Dr. Frank H. Ruddle, Yale University School of Medicine, New Haven, CT (McClelland et al., 1984). Bacteria containing plasmids were grown in Luria-Bertain medium, and plasmids were harvested as described (Maniatis et al., 1982). Plasmids of each probe were then digested with appropriate restriction enzyme(s), fractionated by electrophoresis on 1% agarose gel. Specific DNA fragments containing the appropriate probe were eluted from gel and purified through a NACS Prepace cartridge. Each probe was labeled with ["'PIdCTP by the nick-translation method to a specific activity of 20-350 X 10' cpmlpg of input DNA. The labeled probe was separated from unincorporated ['"PIdCTP by passing through a G-50 column.

Materials
Cell and Cell Culture-The clonal cell lines used in this study were all derived from the human lymphoblastic cell line CCRF-CEM (Foley et al., 1965). The derivation of the glucocorticoid-resistant clonal lines ICR-27 and CEM-C1, and the glucocorticoid-sensitive parental CEM-C7 and ICR-27xCl:H.lO hybrids have been described previously (Norman and Thompson, 1977;Harmon and Thompson, 1981;Zawydiwski et al., 1983;Yuh and Thompson, 1987). To ensure maximal homogeneity of each parental clone, we recloned each in semisolid agarose medium and selected clones that had their classic parental growth characteristics, determined by growth and viability curves f M dexamethasone, for this study; i.e. C7A from C7, C1-1 from C1, and ICR-27 TK.3.2 from ICR-27. All CEM cells were grown in RPMI 1640 medium containing heat-inactivated 10% fetal bovine serum, 50 mM Tricine, and 0.5 g/liter NaHCO:, (with the final pH value adjusted to 7.4). Cell growth was measured by counting cells on a Coulter counter, as described previously (Yuh and Thompson, 1987). Cell morphology was observed every day by light microscopy. After 4 days culture, an aliquot of cells was stained with trypan blue for 1-2 min, and the unstained cells were taken to be viable. Cells were maintained in suspension in 2-liter Corning roller bottles in a 37 "C incubator. Culture medium was changed every 48 h for 6 days in each experiment to keep cells in the log phase of proliferation. The last change of medium was 16 h before the addition of dexamethasone. The cell concentration at the time of adding dexamethasone was 2-4 X 105/ml. Dexamethasone in ethanol or ethanol only was added and cells were harvested at various times thereafter. The final concentration of ethanol never exceeded 0.01%.
For cell cycle distribution analysis, cells in suspension were pelleted by centrifugation at 250 X g for 10 min. The cell pellets were treated with 0.5% pepsin and 3-5 drops 0.1% ribonuclease A for 5 min, fixed by adding ethanol to a final concentration of 70%, and stored at 4 "C. For flow cytometry, ethanol-fixed cells were stained with ethidium bromide for 20 min, and analyzed on a Counter TPS-1 (Coulter Electronics, Hialeah, FL).
RNA Extraction and Purification-Total cellular RNA was purified

TABLE I
Growth control and protooncogenes whose expression is unaffected by dexamethusone in CEM-C7 cells The RNAs were detected on Northern blots, loading 10-20 pg of total CEM-C7 cell RNA/lane and normalizing the data to the atubulin signal, as in Figs. 1 and 2. Cells were exposed to lo-' M dexamethasome or vehicle only before RNA extraction, as described under "Experimental Procedures." No RNA was detected for interleukin-2, c-sis, c-abl, c-mos, c-fos, or c-erb-A genes, under various stringencies. Drobes.

c-myc mRNA in C7-
. a-t u b.

FIG. 1. Dexamethasone suppression of c-myc RNA levels in
C7A cells. Ten pg of poly(A)+-enriched RNA from C7A cells, which had been grown in the presence (+) or absence (-) of 1 p~ dexamethasone for 12 h, were hybridized with v-myc and a-tubulin (atub.) probes as described under "Experimental Procedures." The fluorescent photograph of the RNA gel, stained with ethidium bromide, is shown in panel a. The autoradiographic result is shown in panel b. mRNA Suppression by Dexamethasone using the guanidine hydrochloride/CsCl method as described by Kantor et al. (1980). For each cell sample, 2-3 X 10' cells were harvested from the medium by centrifugation. Cell pellets were then washed three times with Dulbecco's phosphate-buffered saline (Sigma) in 0.1% diethylpyrocarbonate-treated water. Five pI of vanadyl ribonucleoside complex (Berger and Birkenmeier, 1979) was added to each final pellet. Cells were disrupted by freezing a t -70 "C (Revco freezer) for at least 2 h and then thawing on ice for 0.5 h. Following this, 0.9 ml of 6 M guanidine HC1 with 50 mM EDTA was added, and the whole solution was agitated on a Vortex mixer in a 15-ml Corning centrifuge tube for 1 min; then 100 pl of 2 M potassium acetate (pH 5) was added and the mixture was again agitated for another 10 min. The suspension was mixed with 2 ml of 1% N-laurosylsarcosine in 0.1 mM EDTA (pH 8), 1.08 g of CsCI, and agitated until the CsCl dissolved. This mixture was carefully layered over a 2-ml cushion of 5.6 M CsCl in a 5-ml centrifuge tube and centrifuged for 18 h a t 45,000, 22 "C in an SW-50 rotor. The RNA pellet at the bottom was then resuspended in 200 pg of Tris-EDTA (10 mM Tris, 1 mM EDTA) buffer and precipitated by ethanol. Partial purification of poly(A)+enriched RNA was carried out using oligo(dT)-cellulose columns as described (Maniatis et al., 1982) except that lithium instead of sodium salts were used.
Detection of RNA Transcripts-Samples of total RNA (20 pgllane) or poly(A)+-enriched RNA (10 pg/lane) were dissolved in 8% formaldehyde and 67.4% formamide and denatured by incubating a t 65 "C for 15 min, then fractionated on a 1% formaldehyde-agarose gel in MOPS (pH 7.4) and transferred to nitrocellulose filters as described (Thomas, 1980). The filters were prehybridized a t 65 'C for 4-8 h in a solution of 4 X standard saline citrate (SSC, 1 X SSC = 0.15 M NaCl + 0.015 M sodium citrate), 50 mM sodium phosphate (pH 6.5), 200 pg/ml denatured fish sperm DNA, 5 X Denhardt's solution (1 X Denhardt's = 1% Ficoll 400, 1% polyvinylpyrrolidone, 1% bovine serum albumin), and 0.01% sodium dodecyl sulfate. The nitrocellulose filters were then hybridized in fresh hybridizing buffer containing nick-translated :'2P-labeled DNA probes (1 X 10" cpm/ml, with specific activity 210' cpm/pg for the myc probe) for 36 h a t 65 "C. For myc, the probe chosen was to v-myc (1.2 kb between SalI and BamHI sites) or the third exon of the myc gene (the 1.8-kb fragment released by ChI and EcoRI). The hybridization buffer contained 4 X SSC, 50 mM sodium phosphate (pH 6.5), 100 pg/ml denatured fish sperm DNA, and 1 X Denhardt's solution. Filters were then washed twice for 5 min with 2 X SSC, a t 22 "C for 30 min, for 30 min with 2 X SSC and 0.1% sodium dodecyl sulfate a t 60-65 "C, and finally with 0.1 X SSC at 22 "C. The dried, hybridized filters were exposed to Kodak XAR-5 films with intensifying screens a t -70 "C for autoradiographic detection of DNA-RNA hybrids. The size of transcripts was determined relative to 18 S and 28 S ribosomal RNA bands which had  dicated (myc, a-tub). Panels c show a briefer film exposure of the a-tubulin signal.
been measured with RNA sizing markers as 1.8 and 4.5 kb, respectively. Integrated densitometric tracings were carried out with the Compuset program of the Beckman DU-8B spectrophotometer scanning a t 500 nm wavelength. Results were normalized to the amounts of a-tubulin mRNA on the same filters which were probed with labeled a-tubulin cDNA after removal of the c-myc probe and the subsequent radioautographic signal measured independently by densitometry on the same lane of each time point. The final results, after normalization, were then plotted as percentage change relative to the hybridization signal a t 0 h for each probe.
As an alternative, "dot blot" analyses were carried out, using serial dilutions of total cell RNA probed with the c-myc third exon cDNA, as above. This probe was then removed, and the blots were reprobed with nick-translated cDNA prepared from total poly(A)+ RNA from CEM cells. Autoradiographs of each probed filter were obtained and scanned densitometrically; all data were found to be in the linear response range. The densitometric data from 3 serial dilutions of cmyc and poly(A)+ probed filters were summed, and the c-myc data normalized to the poly(A)+ data, to eliminate variations due to differences in RNA on the filter. (In fact, qualitatively the results were the same without the normalization step.) Data were then plotted as amount of normalized myc RNA a t each time point.
DNA Extraction and DNA Integrity Test-During each total cellular RNA extraction, the DNA layer between the 2-ml 5.6 M CsCl "cushion" and the sample solution was removed by sterile Pasteur pipette. Each sample was dialyzed overnight in 0.1 X Tris-EDTA buffer and extracted with phenol/chloroform twice. Ten pg of DNA from each time point was fractionated by electrophoresis in an 0.8% agarose gel, stained with ethidium bromide, and examined with UV illumination.

RESULTS
Dexamethasone Suppresses c-myc mRNA Expression in C7A Cells-Among the 14 genes examined, we found that the RNA of eight could be detected in our CEM cells, and of these only c-myc mRNA was strongly suppressed by 1 p~ dexamethasone in C7A. Table I lists the data for the constitutively expressed genes and names the remainder. A Northern blot of poly(A)+-enriched RNA from cells incubated 12 h with dexamethasone showed that the 2.4-kb c-myc mRNA was suppressed to less than 20% of the level in control C7A cells (Fig. 1). In order to detect how fast this message was regulated, detailed time course studies were performed, using total cellular RNA. Since fresh serum has been reported to simulate c-myc expression in some cells, we minimized this possible perturbation by changing medium 16 h before the addition of dexamethasone and confirmed the fact that the cell cycle patterns were identical in control and treated cells during the 12 h following dexamethasone addition, monitoring the subsequent cell cycle changes by flow cytometry (data not shown). The c-myc mRNA level in treated C7A cells dropped abruptly during the first 12 h, compared with control cells (Fig. 2). The visual data show clearly that the c-myc signal in the dexamethasone-treated cells falls in a few hours after the addition of hormone. Examination of the a-tubulin signal in all our experiments showed it to vary in a random fashion, as would be expected from small variations in the RNA present on the filter, with one exception. The only non-random variation in a-tubulin was a diminution at later (36-48 h) times, when the culture conditions employed caused even the control cells to be reaching late stationary growth phase. After normalization with a-tubulin, the relative amount of c-myc mRNA was found to be significantly diminished by 1 h after dexamethasone administration and to have reached a minimum by 12- 18 h (Fig. 3). When RNA from dexamethasone-treated uersus control cells was compared by "slot blot" technique and probed with a c-myc fragment specific for the third axon of myc, similar results were obtained (data not shown).
Dose-response studies (Fig. 4) showed that the down-regulation of c-myc mRNA occurred across a concentration range of dexamethasone consistent with occupancy of the glucocorticoid receptors known to exist in these cells (Zawydiwski et d., 1983). Some effect was seen after as little as io-' M dexamethasone, with a maximum being reached as the hormone concentration approached

M.
Glucocorticoid Receptors and "Lysis Function" Required for c-myc Suppression-We used two steroid-resistant clones of CEM cells to investigate further whether functional glucocorticoid receptor is necessary and sufficient for inducing the cmyc suppression. The first mutant tested was the glucocorticoid-resistant, receptor-deficient cell clone ICR-27 TK.3.2. In these cells, c-myc expression was not affected by dexamethasone treatment. After normalization to a-tubulin or to poly(A)+ RNA, the results show that c-myc expression showed only random variations from the controls (Fig. 5). This result suggests that functional glucocorticoid receptors are required for inducing c-myc suppression. To see whether receptor alone was enough, the glucocorticoid-resistant, receptor-positive clone C1-1 was tested. This clone does not show growth inhibition or lysis by dexamethasone, and in it the c-myc expression also did not decrease in response to dexamethasone, only showing a lessening at the last time point, as the cells reached late stationary phase (Fig. 6). Thus c-myc expression in either type of resistant cells is not suppressed quickly by dexamethasone, as it is in the wild-type clone. The results suggest that the presence of glucocorticoid receptor is necessary but not sufficient to cause c-myc expression. Also necessary is an intact response system that leads to growth inhibition and eventual cell lysis, a system not fully defined biochemically but for convenience termed the lysis function.
To test the temporal relation between c-myc suppression and DNA fragmentation, a process found to be induced by glucocorticoids in rodent lymphoid cells, DNA was extracted from cells at each time point, fractionated by electrophoresis on 0.8% agarose gel, and visualized by staining with ethidium. Northern blot technique ( a ) or dot blot ( b ) . In a, Northern blots such as those of Figs. 1 and 2 were quantified and the data plotted as described in the legend to Fig. 3. In b, serial dilution dot blots of treated and control cells were tested. The myc densitometric data for each were normalized to the poly(A)+ RNA content of each dot. The normalized myc data is expressed as percent of control at each time point.
(data not shown). This phenomenon was not seen in C7A control DNAs, which were extracted from cells incubated with the ethanol vehicle only.
To further examine the relation between myc suppression, glucocorticoid receptors, and the lysis function, we employed a somatic cell hybrid between ICR-27 and C1 cells (clone H.lO). In this hybrid we had found that the lysis function could be restored by complementation between ICR-27 and C1 (Yuh and Thompson, 1987) with receptors supplied (presumably) by C1 and lysis function by ICR-27. In H.10 cells, c-myc expression was promptly suppressed, in a pattern similar to that of dexamethasone-treated C7A cells (Fig. 7). The c-myc RNA expression dropped abruptly in the first 12 h and reached a level below 10% of zero time control after 24 h. Although the suppression of c-myc by glucocorticoid is not found in either of the resistant ICR-27 and C1 parental cells, it is restored in their dexamethasone-sensitive hybrid. Therefore it seems that supplying glucocorticoid receptors from C1 cells not only allows activation of the lysis function of ICR-27 cells (Yuh and Thompson, 1987), but also permits early suppression of c-myc expression, emphasizing the close correlation between the two.

DISCUSSION
Among the eight detectable oncogene/growth-related genes tested, only c-myc is suppressed by dexamethasone in C7A cells. This suppression occurs only in those cells with both functional glucocorticoid receptors and the intact lysis function. The down-regulation of c-myc mRNA levels occurs rapidly, within a very few hours of addition of dexamethasone.
DNA fragmentation, cell accumulation in G1, and cell death all occur much later. This could mean that the c-myc effect is just a paraphenomenon, with no causal meaning for eventual cell death, but it could also mean that the c-myc suppression is an important, leading component in the cell growth effects of dexamethasone. Theories of the reasons why steroids kill cells are of two types, that the hormones induce a lethal process, or alternatively that they suppress a process required for sustaining cell integrity. The lag between suppression of c-myc RNA and the other effects might easily be due to the time required for the product of c-myc RNA to reach a critical minimum. The known short half-lives of c-myc and p65, the nuclear peptide product of the third exon (Hann et al., 1983;Dani et al., 1984;Watt et al., 1985), would fit with this possibility. It seems relevant that in prior experiments we found that washing C7 cells free of dexamethasone at any time before 18-24 h allowed the cells to continue to grow or clone unabated. After that time, increasing cell death occurred if the steroid remained present. This onset of cell death is strikingly similar to the time course over which c-myc mRNA reaches its nadir. As to the mechanism of myc regulation by glucocorticoids, we do not yet know in these cells whether the level is regulated transcriptionally or post-transcriptionally, directly or indirectly. Further experiments are required to address these issues.
One problem in these experiments was to find a constantly expressed internal control mRNA against which to quantitate the amount of oncogene RNA. We therefore took two approaches. We chose a-tubulin because we found that there were no striking consistent early changes in a-tubulin mRNA in either dexamethasone-treated or control C7A cells. As described by Eastman-Reks and Vedeckis (1986), however, there is often a decrease in a-tubulin mRNA at later times, especially as the cell density exceeded 1 X 106/ml, when these cells are approaching stationary growth phase. Also, in the instances where a-tubulin mRNA goes down with time, visual inspection shows that both bands of ribosomal RNA are also diminishing. Since this occurred both in dexamethasonetreated and control groups, we believe it is a general, normal physiological change. Therefore, it seems that the marker chosen is a valid measure of total RNA in a given lane. In any case, whether or not the data is normalized to a-tubulin, the early, specific c-myc effect is obvious (Fig. 2). Repeat experiments done in dot blot style, and utilizing total poly(A)+ RNA instead of a-tubulin to quantify the RNA/blot, also clearly show the rapid, specific, dose-dependent inhibition of myc (Figs. 4, 5, and 7).
It has been suggested that c-myc expression may be tied to the cell cycle, and therefore that any agent which affects the cycle must perforce affect c-myc mRNA levels. Recent studies suggested, however, that c-myc levels may be constant throughout the cell cycle Hann et al., 1985). Furthermore, dexamethasone does not alter the cell cycle in CEM-C7 cells for at least 12 h. We monitored the state of cells in the cycle in the present work by cytofluorometric analysis, and the proportion of cells in different phases of the cell cycle was assessed. Although in the first 12 h after adding vehicle or vehicle plus hormone there were slight increases of cells in S phase, they occurred in both the dexamethasone-treated and control groups.
Expression of c-myc is characteristic of normal proliferating cells and is selectively switched off when cells enter the quiescent nonproliferative state (G,/G, phase) that accompanies terminal differentiation (Kronke et aL, 1985;Einat et al., 1985). In several hematopoietic cell lines, when differentiation was induced by certain reagents, the c-myc mRNA level was suppressed (Kronke et al., 1987;Norwell et al., 1983;Lachman and Skoultchi, 1984;Campisi et al., 1984;Reistma et al., 1983). In mouse erythroleukemia cells, dimethyl sulfoxide induces a cessation of proliferation and appearance of differentiated phenotypes (Friend et al., 1971;Marks and Rifkind, 1978). Before these differentiated phenotypes ap-peared, a rapid decrease of c-myc mRNA level was noted (Lachman and Skoultchi, 1984). On the other hand, when the cells had been transfected with constitutively expressed cmyc, dimethyl sulfoxide could not induce differentiation in mouse erythroleukemia cells (Coppela and Cole, 1985;Dmitrovsky et al., 1986;Prochownick et al., 1986). The ability of cells to proliferate and to differentiate seems to be two coupled and closely related functions. In our experiments, c-myc is specifically suppressed in CEM cells with an intact lysis function. Although suppression of c-myc does not necessarily have a causal relationship with lysis of human leukemic lymphoblasts, our results certainly raise this possibility. It is known that malignant cells have difficulty in doing what normal differentiating cells do well, i.e. shutting down the processes necessary for continued growth in a well organized way. The kinetics of myc RNA suppression that we observe are consistent with the myc protein falling, after a time, below a critical level necessary to sustain growth or prevent some deleterious cell process from occurring. CEM cells, when stopped from growth as a threshold level of c-myc (and perhaps other growth-sustaining gene products) is passed, simply may be unable to avoid setting into motion the processes of "programmed cell death" (Ucker, 1987). Our preliminary finding that DNA hydrolysis only becomes manifest late in the course of events would be consistent with this view. It is also possible that loss of c-myc is directly deleterious, as for instance, by altering Caz+ channels (Caffrey et al., 1987). We saw no dexamethasone effects on c-ras or c-myb, and therefore we can rule them out as essential components of cell growth/lysis effects in CEM cells, although they were down-regulated by dexamethasone in S49 cells (Eastman-Reks and Vedeckis, 1986).
Aside from its relation to the growth effects of glucocorticoids on CEM cells, the down-regulation of c-myc mRNA levels we observe provides an interesting system in which to study the negative regulation of a gene by these hormones. Gene down-regulation by steroids is less well understood than gene induction, but has lately received increasing interest and is of great importance in fully appreciating their overall actions. The regulation of c-myc expression may be complex and offers multiple levels for potential control (Greenberg and Ziff, 1984;Blanchard et al., 1985;Bentley and Groudine, 1986;Nepveu and Marcu, 1986). We assayed in several critical experiments the myc RNA product that follows the point of known attenuation in the transcription of this gene by using the cDNA corresponding to the third exon of the gene. Thus it seems unlikely that the regulation we observed was due to a step-down in transcripts during elongation. In recent reports, steroid down-regulation of myc has been suggested to occur at transcriptional (Forsthoefel and Thompson, 1987) or transcriptional plus post-transcriptional levels (Quarmby et al., 1987). The magnitude and kinetics of the response in CEM cells offer encouragement for further study of the mechanism.