Retinoic Acid Increases Cyclic AMP-dependent Protein Kinase Activity in Murine Melanoma Cells*

The influence of all trans-retinoic acid on cyclic AMP metabolism was examined in Bl6-FI mouse melanoma cells. Treatment of these cells with retinoic acid re- sulted in a dose-dependent inhibition of cell growth which was accompanied by a concentration-dependent increase in both basal and cyclic AMP-stimulated protein kinase activity. Intracellular levels of cyclic AMP, however, were not altered by retinoid treatment. A protein kinase-deficient variant of Bl6-FI (MR-4) did not exhibit decreased growth or increased protein kinase activity in response to retinoic acid treatment. At least 24 h of incubation was required before increased protein kinase activity could be detected in treated B16-F1 cells. Retinoic acid treatment increased the V,, of protein kinase, but the K,, for cyclic AMP activation was not altered. These findings suggest that in B16 mouse melanoma cells, cyclic AMP-dependent protein kinase may be a target for the growth inhibitory effects of the retinoid.

was not altered. These findings suggest that in B16 mouse melanoma cells, cyclic AMP-dependent protein kinase may be a target for the growth inhibitory effects of the retinoid.
Vitamin A has been demonstrated to be required for maintenance of normal epithelial cell differentiation (1). Analogs of vitamin A (retinoids) were found to affect glycosylation reactions (2), increase iodinatable fibronectin ( 3 ) , increase cell-to-substratum adhesiveness (4), inhibit growth (5, 6), increase tyrosinase and melanin biosynthesis in mouse melanoma cells ( 7 ) , and reverse the tumor promoter induction of ornithine decarboxylase activity in mouse epidermal cells (8). Although cytoplasmic binding proteins for retinol and retinoic acid have been found in many cell types, it is not understood how the interaction of retinoids with specific receptor proteins ultimately results in a cellular response. The cyclic AMP system has also been implicated in r e plating the differentiated functions of many cell types (9-12). In particular, cyclic AMP has been shown to inhibit growth and stimulate melanin production in murine melanoma cells (11,13). Thus, its action in these cells mimics that of retinoids ( 7 ) . Therefore, we undertook a study to determine whether there was any interrelationship between the actions of retinoids and the cyclic AMP system. * This work was supported by Grant BC-317E from the American Cancer Society. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

MATERIALS AND METHODS
B16-FI cells were obtained through the courtesy of Dr. I. J. Fidler, Frederick Cancer Research Center, Frederick, Md. The protein kinase-deficient variant of FI (MR-4) was obtained in our laboratory through selective cloning procedures as previously described (14). Both cell lines were routinely maintained in minimal essential medium plus Earle's salts supplemented with nonessential amino acids, L-glutamine (2 mM), sodium pyruvate (1 m), antibiotics (50 pg/ml of streptomycin sulfate, 50 units/& of penicillin G), and 10% fetal bovine serum (heat-inactivated, GIBCO), and incubated in a 37"C, 95% air, 5% CO2 humidified atmosphere. Growth inhibition in both cell lines was determined by seeding 2.0 X lo5 cells onto 90-mm tissue culture dishes (Falcon, Oxnard, Calif.). Twenty-four hours after seeding, triplicate plates were used to determine cell number through the use of a hemacytometer. Also, at this time, one-half of the dishes were refed with medium supplemented with either M d trans-retinoic acid (Sigma) or various concentrations of retinoic acid (lo-' to M). Control plates received medium plus solubilization vehicle (0.1% ethanol). At 24,48,72, and 96 h of incubation with or without retinoic acid, triplicate plates were analyzed for cell number. Also, after 48 h of incubation, control and treated plates were refed with the appropriate medium. All manipulations involving cells and retinoic acid were performed under indirect light.
Protein kinase activity in control and retinoic acid-treated cells was assayed by the method of Corbin et al. (15). using a 10-min incubation time. Briefly, this involved washing the cells three times with 4 ml of 50 mM phosphate buffer (pH 6.8), scraping the cells from the dishes with a rubber policeman and sonicating the resultant suspension (1.5 ml) for 30 s at setting No. 3 in a model WI85 sonicator (Heat Systems, Plainview, N. Y.). The reaction mixture consisted of 50 mM phosphate buffer (pH 6.8), 0.2 m~ [y-'"P]ATP (180 cpml pmol), 0.5 mg of histone (type 11-A) & various concentrations of cyclic AMP, and 20 p1 of cell homogenate. The reaction was terminated by pipetting 50 pl of the reaction mixture onto Whatman 3" filter paper discs (2.3 mm diameter; Whatman, Inc., Clifton, N. J.). The discs were dropped into ice-cold 10% trichloroacetic acid and washed in sequence with 10% trichloroacetic acid, 95% ethanol, and ether. After drying, the filters were placed in scintillation vials containing Omnifluor (New England Nuclear, Boston, Mass.) and counted. Activity was assessed in the absence or presence of to 5 X lo-'' M cyclic AMP. The samples were corrected for endogenous protein phosphorylation in the absence of histone and cyclic AMP. Under our assay conditions, activity was linear with respect to protein concentration.
AU chemical reagents were obtained from Sigma Chemical Co., St.
Louis, Mo. [y-,"'P]ATP specific activity, 10 Ci/mmol, was obtained from New England Nuclear, Boston, Mass. Fig. 1 illustrates the effect of IO-' M retinoic acid on B16-F, cell growth. Within 24 h after the addition of retinoic acid to the culture medium, growth inhibition was evident. The relative growth inhibition increased with time and, at 4 days of retinoic acid treatment, reached 71%. In contrast, there was little, if any, growth inhibition in MR-4 cells treated in the same manner at any time point examined (Fig. 2).

RESULTS
The growth inhibition of B16-FI cells induced by retinoic acid was concentration-dependent ( Fig. 3 ) with significant inhibition occurring at 10" M. Also, it should be noted that the degree of inhibition obtained by M retinoic acid did vary somewhat between separate experiments. For example, after 48 h of treatment (Fig. l ) , 54% inhibition was achieved, while in a separate experiment (Fig. 3 ) , 70 to 75% inhibition was achieved. It appeared that the condition of the ceUs (subconfluent versus confluent) used to set up the experiment influenced the degree of retinoic acid-induced growth inhibi- tion. Fig. 3 also demonstrates that the MR-4 cells were not inhibited at any concentration of retinoic acid tested. We next determined whether retinoic acid might have caused growth inhibition by increasing intracellular cyclic AMP levels since we have shown (16) that the latter inhibit B16-F1 cell growth. However, in agreement with the results reported by Lotan for S91 murine melanoma (7), we were unable to observe an increase in B16-F1 cyclic A M P levels at any concentration of retinoic acid tested (data not shown). To examine the possible involvement of the cyclic A M P system further, we determined the activity of cyclic AMPdependent protein kinase, the only known mediator of cyclic AMP action, in control and retinoic acid-treated cells.  (Fig. 4B). In other experiments (data not shown), we found that treating a B16-F1 cell sonicate directly with M retinoic acid for as long as 6 h at 37°C did not alter protein kinase activity.
Retinoic acid stimulation of B16-F, protein kinase activity does not occur immediately but, rather, requires more than 6 h of incubation (Fig. 5). Significant stimulation of protein kinase is obtained by 24 h of treatment and this level is more or less maintained through 48 h of treatment. In some experiments where cells were refed after 48 h of incubation with fresh medium containing retinoic acid, a further stimulation of protein kinase was observed at 72 h. or 10 pM retinoic acid. Cells were then harvested and processed for protein kinase activity as described in the text. The experiment was repeated two additional times with similar results.
The ability of various concentrations of cyclic AMP to activate protein kinase from control and 48-h retinoic acidtreated cells was examined (Fig. 6). The K , for cyclic AMP activation of the protein kinases from both control and treated cells is quite similar (-2.0 X 10" M ) , suggesting at least superficially that there has been no alteration in the regulatory subunit of the protein kinase from retinoic acid-treated cells.

DISCUSSION
The data presented in this communication would suggest that cyclic AMP-dependent protein kinase is involved in mediating at least some of the actions of retinoic acid in the B16-F, murine melanoma cell line. The evidence for this is as follows. (a) Retinoic acid inhibits growth and increases tyrosinase activity.' Cyclic AMP duplicates both of these effects ration.
' K. W. Ludwig,B. Lowev,and H . M. Niles,cells (16). (b) Retinoic acid significantly increases the activity of cyclic AMP-dependent protein kinase which is the major if not the only receptor responsible for the effects of this cyclic nucleotide. (c) Retinoic acid does not inhibit growth or enhance protein kinase activity in a variant of B16-FI cells (MR-4) which is resistant to the growth-inhibiting effects of melanocyte-stimulating hormone and has deficient protein kinase activity.
The mechanism by which retinoic acid increases protein kinase activity in B16-Fl cells is not clear. There are specific cytoplasmic receptor proteins for retinoic acid (17). It is thought that the retinoic acid.receptor protein complex is transported to the nucleus where it modifies gene transcription analogous to the action of steroid hormones (18,19). Our data would fit this model in that there is a long lag period subsequent to retinoic acid treatment before increased protein kinase activity is observed. However, we do not yet have evidence suggesting that increased protein kinase activity after retinoic acid treatment is due to de novo enzyme synthesis. In regard to the model proposed for retinoid action and its similarity to steroid hormone action, it is interesting that Fuller et al. (20) have found a specific regulation of the amount of type I cyclic AMP-dependent protein kinase by testosterone especially in the prostate.
Lotan et al. (5,6) have reported that retinoids inhibit the growth of many, but not all, cells in tissue culture. In particular, they demonstrated that B16 mouse melanoma cells were inhibited by retinoic acid in a concentration-dependent fashion. However, at a retinoic acid concentration of M, they could not observe growth inhibition until 3 days of treatment, whereas we found significant growth inhibition at 24 h of treatment. This discrepancy may be due to the fact that we used a different growth medium and different treatment protocol than Lotan et al. (6). Whether the ability of retinoids to inhibit the growth of cells in culture has any relationship to their ability to inhibit in vivo chemical carcinogenesis (21-23) remains to be determined.
A variety of reports have provided evidence that cyclic AMP-dependent protein kinase is a mediator of growth inhibition. These include: the increase in cyclic AMP binding protein and protein kinase activity in mammary tumors regressing due to ovariectomy (241, the inability of cyclic AMP to inhibit cell growth in a protein kinase-deficient variant of S49 lymphoma cells (25), the inability of adrenocorticotropic hormone (ACTH) and melanocyte-stimulating hormone (MSH) to inhibit growth in protein kinase-defective variants of Y1 adrenal cells (26) and B16 mouse melanoma cells (14), respectively, and the correlation between the ability of various cyclic AMP analogs to inhibit growth in human carcinoma cell cultures and their ability to activate cyclic AMP-dependent protein kinase from the same cell line (27). There are, however, some reports which have correlated increased cyclic AMP-dependent protein kinase activity with stimulation of growth. These include the activation of protein kinase by growth hormone in the liver and the adrenal (28), an increase in protein kinase during mitosis in Chinese hamster ovary (CHO) cells (29), and an early activation of protein kinase in mitogen-stimulated human lymphocytes (30). However, for the most part, these increases in protein kinase activity were restricted to type I protein kinase and it was shown that treatment of human lymphocytes with analogs of cyclic AMP which blocked the mitogenic response resulted in activation of both types I and I1 protein kinase (30). Whether the increase in protein kinase activity in retinoic acid-treated B16-F, cells is restricted to type I or type I1 protein kinase remains to be determined. Also, future studies should be directed to other cell lines where retinoids have been shown to affect