Differentiation-specific increase of cAMP-dependent protein kinase in the 3T3-L1 cells.

The specific activity of CAMP-dependent protein kinase was studied during the differentiation of 3T3-Ll fibroblasts into adipocytes. 8-Azido-adenosine 3’:5’monophosphate, a photoaffinity-labeling analogue of CAMP, was used to quantitate the regulatory subunits of the type I and type 11 CAMP-dependent protein kinase in cell extracts. The catalytic subunit of CAMP-dependent protein kinase was studied by assaying for CAMPdependent histone kinase activity. A significant increase in CAMP-dependent protein kinase activity was observed during the adipose development of 3T3-Ll cells, promoted by 3-isobutyl-1-methylxanthine (IBMX), dexamethasone, and insulin. This increase was attributed primarily to a 3to 6-fold increase in the type I CAMP-dependent protein kinase, as demonstrated either by the incorporation of 8-azidoadenosine 3’:5’-monophosphate into the 47,000-dalton regulatory subunit of the type I kinase or by the separation and analysis of the type I and type I1 kinases through DEAE-Sephacel column chromatography. The time course of increase in the type I CAMP-dependent protein kinase preceded increases both in malic enzyme and glutamine synthetase activities and was maximal at 5 days after initiation of the hormone treatment. The induction of CAMP-dependent protein kinase does not appear to involve an increase in intracellular CAMP. The addition of either 1 m~ dibutyryl CAMP or 0.5 m~ IBMX had little or no effect on the expression of CAMP-dependent protein kinase activity in the 3T3-Ll cells, although a marked increase in CAMP phosphodiesterase activity was observed in cells treated with dibutyryl CAMP. The increased expression of the type I CAMP-dependent protein kinase in 3T3-Ll cells appears to be determined by the operation of the intrinsic differentiation program in these cells; it was observed in both the hormone-induced as well as spontaneously differentiated 3T3-Ll cells but was not observed in a control clone of 3T3, the 3T3-C2 cells, treated with IBMX-dexamethasone and insulin. These results suggest that the 3T3-Ll cells may provide an ideal system for studying the mechanism of regulation of CAMP-dependent protein kinase and its functional significance during development and differentiation.

A significant increase in CAMP-dependent protein kinase activity was observed during the adipose development of 3T3-Ll cells, promoted by 3-isobutyl-1-methylxanthine (IBMX), dexamethasone, and insulin. This increase was attributed primarily to a 3-to 6-fold increase in the type I CAMP-dependent protein kinase, as demonstrated either by the incorporation of 8-azidoadenosine 3':5'-monophosphate into the 47,000-dalton regulatory subunit of the type I kinase or by the separation and analysis of the type I and type I1 kinases through DEAE-Sephacel column chromatography.
The time course of increase in the type I CAMP-dependent protein kinase preceded increases both in malic enzyme and glutamine synthetase activities and was maximal at 5 days after initiation of the hormone treatment. The induction of CAMP-dependent protein kinase does not appear to involve an increase in intracellular CAMP. The addition of either 1 m~ dibutyryl CAMP or 0.5 m~ IBMX had little or no effect on the expression of CAMP-dependent protein kinase activity in the 3T3-Ll cells, although a marked increase in CAMP phosphodiesterase activity was observed in cells treated with dibutyryl CAMP.
The increased expression of the type I CAMP-dependent protein kinase in 3T3-Ll cells appears to be determined by the operation of the intrinsic differentiation program in these cells; it was observed in both the hormone-induced as well as spontaneously differentiated 3T3-Ll cells but was not observed in a control clone of 3T3, the 3T3-C2 cells, treated with IBMX-dexamethasone and insulin. These results suggest that the 3T3-Ll cells may provide an ideal system for studying the mechanism of regulation of CAMP-dependent protein kinase and its functional significance during development and differentiation.
The conversion of 3T3-Ll cells from fibroblasts to adipocytes represents a spontaneous and heritable phenotype; this conversion is favored by cessation of cell growth or treatment of the cells with various hormones, drugs, or nutrients (1-5).
* This work was supported by a grant from the American Cancer Society, Massachusetts Chapter. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Concomitant with the morphological appearance of large fat droplets in the differentiating 3T3-Ll cells, there are increases in the activities of enzymes involved in the synthesis and accumulation of lipid (6-ll), as well as that of enzymes involved in mobilization of the stored fat (11)(12). In addition, the hormone sensitivity of the differentiated cell differs remarkably from that of its undifferentiated counterpart (for review, see Ref. 13). The appearance of these intracellular enzymes and cell surface receptors in the differentiated cells is presumably coupled to changes in the pattern of cellular metabolism, uiz. from one directed mainly to growth to one designed to maximize triglyceride synthesis and utilization.
The primary interest of this laboratory has been the role of cAMP and CAMP-dependent protein kinase in cell function. Since (a) cAMP and CAMP-dependent protein kinase appear to have important functions in controlling the activity of triglyceride lipase (12,(14)(15)(16) as well as the activities of key glycolytic enzymes (for review, see Ref. 17) and ( b ) there are significant increases in triglyceride lipase and various lipogenic and glycolytic enzymes upon differentiation of the 3T3-Ll cells (for review, see Ref. I), it appears likely that modulation of the type and amount of intracellular CAMP-dependent protein kinase may have a pivotal function in the differentiation of 3T3-Ll cells. In view of these considerations, I commenced a study on the CAMP-dependent protein kinase of the undifferentiated and differentiated 3T3-Ll cells. The purpose of this report is to describe the induction of a type I CAMPdependent protein kinase upon adipose conversion of the 3T3-L1 cells.
Cell Growth and Differentiation-The 3T3-Ll and 3T3-C2 fihroblasts from Swiss mouse embryo were provided by Dr. Howard Green, Department of Physiology, Harvard Medical School. Cells were grown in Dulbecco's modified Eagle's medium (with 6500 mg/liter of glucose and 5 mM glutamine) containing 10% newborn calf serum and supplemented with 50 units/ml of penicillin and 50 p g / d of streptomycin.
Cells were plated at a density of approximately 1-10 X lo3 cells/cm2 and were refed at 3-4-day intervals.
To promote differentiation of the 3T3-Ll fibroblast to adipocyte, confluent monolayers (designated as day 0) were treated for 2 days Protein Kinase and 3T3 Differentiation 299 with 0.5 mu IBMX, 0.25 p~ dexamethasone, followed by refeeding the cells with fresh growth medium supplemented with 10 pg/ml of insulin. Appearance of morphologically differentiated cells was apparent on day 5 (i.e. 3 days after feeding the cells with insulincontaining medium) and was 70-80% complete on day 7. In some cases, cells were allowed to differentiate spontaneously by maintaining the confluent culture at 37 "C and refeeding at 3-4-day intervals until >50% of the cells developed the characteristic accumulation of multilocular fat droplets of differentiated cells (approximately 3-4 weeks postconfluency). Preparation of Cytosol Fraction-Unless otherwise stated, undifferentiated 3T3-Ll cells represent untreated fibroblastic cells at the stationary phase of growth with a cell density of approximately 6-8 X 10" cells/cm' (generally 5-7 days after plating). Differentiated cells generally represent IBMX-dexamethasone and insulin-treated cells at day 7 after initiation of the hormone treatment.
Cells were rinsed twice with 5 ml of phosphate-buffered saline (0.15 M NaCl, 10 mM sodium phosphate, pH 7.4) and scraped off the dishes with a rubber policeman. Cells were sedimented by centrifuging at 1,000 X g for 2 min and washed once with PBS. The cell pellet, obtained from five to ten 100-mm plates, was resuspended in 1.5 to 2 ml of 10 mu Tris-HC1 (pH 7.4) containing 1 mM EDTA, 50 pg/ml of PMSF. Cells were broken by forcing the cell suspension through a 27 gauge %-inch needle. The 150,000 X g supernatant obtained from cell homogenate was defined as the cytosol fraction. (It should be noted that, under the conditions used for preparing the cytosol fraction, actin is primarily associated with the particulate fraction.) For routine analysis, cytosol preparations were dialyzed against 10 Tris-HC1 (pH 7.4) containing 1 mu EDTA, 50 pg/ml of PMSF, 1 mM dithiothreitol. Protein concentration was determined by the method of Lowry et al. (19) using bovine .serum albumin as the standard.
Covalent Incorporation of 8-N3-["2PJcAMP-Covalent binding of 8-N&*P]cAMP was performed as described previously (20, 21). Briefly, the standard reaction mixture (final volume, 0.1 ml) contained 50 mM 2-(N-morpholino)ethanesulfonic acid (pH 6.2), 10 MgC12, 0.5 m~ IBMX, 0.1 nM to 1 PM 8-No-["*P]cAMP (specific activity, 4-10 Ci/mmol), and various amounts of cytosol protein up to 200 pg. For studying specificity of the incorporation of 8-Nn-["P]cAMP, 50 p~ cAMP was added to the assay mixture. Samples were incubated in the dark at 4 "C for 60 min to achieve equilibrium binding and were then photolyzed for 10 min with a Mineralite UVS-11 hand lamp at a distance of 4 cm. Samples were then prepared for sodium dodecyl sulfate-polyacrylamide gel electrophoresis to determine the incorporation of radioactivity into protein bands according to methods previously described (20).
Histone Kinase Assay-Histone kinase activity was measured by the method of Witt and Roskoski (22) as previously described (20). One unit of histone kinase activity was defined as the amount of enzyme which catalyzed the incorporation of 1 pmol of ,'2P into histone in 1 min at 30 "C.
DEAE-Sephacel Column Chromatography-DEAE-Sephacel was pre-equilibrated with 10 mu Tris-HC1 (pH 7.4). 1 mM dithiothreitol, 50 g / m l of PMSF and was packed into columns (1.5 X 6 cm). For the separation of type I and type I1 CAMP-dependent protein kinases in cell extract, a 150,000 X g supernatant prepared from 3T3-Ll cells was loaded onto a DEAE-Sephacel column. The column was washed with 10 ml of the equilibration buffer followed by elution of the enzyme with a linear gradient (0-0.3 M) of NaCl in the same buffer (total volume of gradient was 30 ml). Fractions of approximately 0.75 ml were collected and were assayed for CAMP-dependent histone kinase and ['HICAMP-binding activities.
Malic Enzyme Assay-Malic enzyme activity present in cell ex-

40.000
FIG. 1. Patterns of (A) the covalent incorporation of 8-Ns-['*PICAMP and (B) the Coomassie blue-stained SDS-polyacrylamide slab gel of cytosol proteins from 3T3-Ll fibroblasts and adipocytes. To promote adipose conversion, confluent 3T3-Ll monolayers (day 0) were treated with IBMX (0.5 m~) and dexamethasone (0.25 p ) for 2 days, followed by refeeding the cells with an insulincontaining growth medium. Cells were harvested on day 7 of the treatment schedule (adipocytes). Parallel untreated cell culture was used as the control (fibroblasts). The cytosol preparations obtained were used for studying the incorporation of ~-NS-[~~P]CAMP, which was done under standard conditions using 1 p~ 8-Nn-[32P]cAMP, in the absence and presence of 50 CM CAMP. Upon completion of the incorporation reaction, samples were subjected to SDS-polyacrylamide (5-15% gradient) slab gel electrophoresis, protein staining, and The assay is based on the generation of NADPH as measured by spectrophotometric absorbance at 340 nm. Results, determined by the malate-dependent rate of NADPH formation, were expressed in nanomoles of NADPH formed at 37 "C/min/mg of protein.
Glutamine Synthetase Assay-Glutamine synthetase activity present in cell extracts was determined by a modified version of the yglutamyl transferase assay previously described (24). Briefly, the reaction mixture (total volume, 1 m l ) contained 50 m~ irnidazolehydrochloride (pH 6.8), 50 mM hydroxylamine, 100 m~ L-glutamine, 25 mM potassium arsenate, 0.2 mM ADP, 0.5 mM MnCL, 100-300 pg of cell extract protein. The reaction was carried out at 37 "C for 45-60 min and was terminated by the addition of 1 ml of a solution containing 0.37 M ferric chloride, 0.3 M trichloroacetic acid, 0.6 M hydrochloric acid. After removal of denatured protein by centrifugation, the absorbance at 505 nm was recorded. y-Glutamylhydroxymate was used as the standard. One unit of enzyme activity was defined as that needed to produce 1 pmol of y-glutamylhydroxymate in 1 min at 37 OC.
Other Meth~ds-[~H]cAMP-binding activity was determined according to the methods of Gilman (25); all results were corrected for nonspecific binding, determined from the amount of radioactivity retained in the presence of 50 p~ CAMP. Phosphodiesterase activity was measured according to the method of Thompson et al. (26), activity was expressed in picomoles of [3H]cAMP hydrolyzed/min/ mg of protein. Measurement of cellular cAMP concentration was done according to methods of Brown et al. (27) as previously described (21); the assay was based on the competition of binding of [3H]cAMP and cAMP to a preparation of CAMP-binding protein, in this case, the regulatory subunit of a partially purified type I1 CAMP-dependent protein kinase from bovine heart.

RESULTS
Cytosolic CAMP-binding proteins of 3T3-Ll fibroblasts and adipocytes were studied by the photoactivated incorporation of 8-N3-["PICAMP and were characterized by 1) the pattern of incorporation (Fig. l), 2) the apparent affinity of this incorporation (Fig. 2), and 3) a quantitative comparison of the amount of 8-N3-["*P]cAMP covalently incorporated to the amount of reversibly bound [3H]cAMP.
The pattern of labeling with 8-N3-[32P]cAMP of cytosol proteins from 3T3-Ll fibroblasts and adipocytes and the corresponding Coomassie blue stained SDS-polyacrylamide slab gel are shown in Fig. 1, A and B, respectively. Three protein bands with apparent M, = 47,000, 52,000, and 54,000 present in cytosols of the undifferentiated and differentiated 3T3-Ll cells incorporated 8-N3-[32P]cAMP. The incorporation of 8-N3-[32P]~AMP into the 47,000-dalton protein of the differentiated cells was significantly greater than that of the undifferentiated cells, occasionally, as in the autoradiogram illustrated in Fig. IA, a slight increase in the incorporation of 8-N3-[3'P] cAMP into the 52,000-and 54,000-dalton dimer was also observed. The 47,000-, 52,000-, and 54,000-dalton 8-N3-[32P] CAMP-labeled proteins represent specific CAMP-binding proteins; the addition of a 50X excess cAMP to the binding assay mixture inhibited the incorporation of 8-N3-['*P]cAMP into these three proteins. The term RI will be used to refer to the 47,000-dalton protein, while the term Rn will be used to refer to the 52,000-54,000-dalton dimer.
To determine whether the increased incorporation of 8-N3-[32P]cAMP into cytosolic proteins of the differentiated adipocytes when compared to that of the undifferentiated fibroblasts was attributable to an increase in the amount of the protein or an increase in the affinity of the protein for 8-N3-[32P]cAMP, the concentration-dependent incorporation of 8-N3-[32P]~AMP into cytosol proteins of the 3T3-Ll fibroblasts and adipocytes was studied. Results are shown in the form of autoradiograms and semilog dose-response plots in Fig. 2. The apparent dissociation constants ( K d ) for the incorporation of 8-N3-["*P]cAMP into RI and R11 of the 3T3-Ll fibroblasts were similar to that of the 3T3-Ll adipocytes; they were, for RI, 1.7 X lo-' M (fibroblast) and 2.3 X M (adipocyte) and for RII, 1.1 X 10" M (fibroblast) and 1.1 X lo-' M (adipocyte). pocytes were, for RI, 3.1 pmol/mg of protein and for RII, 1.35 pmol/mg of protein. In both the fibroblasts and adipocytes, the total amounts of 8-N3-[32P]cAMP incorporated into RI and R I~ CAMP-binding proteins were in agreement with the amounts of specifically bound [3H]cAMP ( Table I).
The catalytic activities of CAMP-dependent protein kinase present in extracts of the 3T3-Ll fibroblasts and adipocytes were studied and compared. Fig. 3 shows the effect of varying TABLE I Regulatory a n d catalytic subunit activities of CAMP-dependent protein kinase from cytosols of 3T3-Ll fibroblasts a n d adipocytes Regulatory subunit activity was measured both by the covalent incorporation of ~-N S -[~~P ] C A M P into the 47,000-dalton protein (RI) and the 52,000-54,000-dalton dimer (Rn), and the binding of [3H] CAMP. The amounts of cytosol protein used for 8-N#P]cAMP incorporation and [3H]cAMP binding were 200 and 100 pg, respectively. Both assays were carried out under standard conditions according to methods described. The catalytic subunit activity was determined by the CAMP-dependent histone kinase assay using histone f2b as the substrate protein. Results represent the means of six indeDendent exneriments f standard deviation. cytosol protein concentrations on histone kinase activity in the absence and presence of 5 PM CAMP. In the range of protein concentration tested (10-50 pg/lOO pl), the specific CAMP-dependent protein kinase activity ( i e . that which is activated by CAMP end inhibited by the protein kinase inhibitor (28) Results of the incorporation of ~-NS-[~*P]CAMP and histone kinase activity indicated the presence of an increased amount of CAMP-dependent protein kinase in the differentiated 3T3-L1 adipocytes when compared to the undifferentiated cells. This was further examined by chromatographing cytosol preparations over DEAE-Sephacel columns to resolve the type I and type I1 CAMP-dependent protein kinases. Elution profiles of histone kinase activity, assayed in the absence and presence of 5 PM CAMP, and [3H]cAMP-binding activity of the 3T3-Ll fibroblasts and adipocytes are shown in Fig. 4. A minor and a major peak of CAMP-dependent protein kinase activity, referred to as the type I and type I1 CAMP-dependent protein kinases according to their order of elution from the DEAE-Sephacel column with increasing salt concentration, was obtained using cytosol of the undifferentiated cells. The elution profile of [3H]cAMP-binding activity coincided with that of the histone kinase activity, with a minor peak eluting at 0.11 M NaCl and a major peak eluting at 0.25 M NaCl. Differentiation of the 3T3-Ll fibroblasts to adipocytes was associated with a marked increase (approximately 3-fold) of the type I CAMP-dependent protein kinase peak eluted from the DEAE-Sephacel column; a minor increase (approximately 10-20%) in the type I1 CAMP-dependent protein kinase was also observed. Studies on the incorporation of 8-N3-["*P]cAMP into proteins present in the peak fractions of the type I and type I1 kinases eluted from DEAE-Sephacel columns demonstrated that the 47,000-dalton and the dimer of 52,000-54,000-dalton 8-N3-[3'P] CAMP-labeled proteins represent, respectively, regulatory subunits of the type I and type I1 CAMP-dependent protein kinases (Fig. 5). In addition, the relative amounts of 8-N3-[3'P] cAMP incorporated into these proteins mirrored their respective protein kinase activities. Since neither free regulatory nor free catalytic subunit activity of CAMP-dependent protein kinase was detected in the DEAE-Sephacel column eluate, it would seem reasonable that the incorporations of 8-N3-[32P] cAMP into the 47,000-dalton protein and the 52,000-54,000- Autoradiogram illustrating the pattern of incorporation of 8-Ns-[S*P]cA" using cytosol and the partially purified type I and type I1 CAMP-dependent protein kinases from the 3T3-Ll fibroblasts and adipocytes. The sources for the type I and type I1 protein kinases of the fibroblasts (Fibro) were fractions 15 and 3 4 , respectively, of the DEAE-Sephacel column eluate illustrated in Fig. 4 and fractions 16 and 30 for the 3T3-LI adipocytes (Adip). Aliquots of the cytosol preparations containing 100 pg of protein, and 150-pl aliquots of the column fractions were assayed for the incorporation of 8-N,#*P]cAMP according to methods described. The locations of the 47,000-dalton and the 52,000-and 54,000-dalton regulatory subunits, designated as RI and RII, respectively, are indicated by the arrows. dalton dimer may serve as correct indices for the relative concentrations of type I and type I1 CAMP-dependent protein kinases present in cell extracts.

Time Course of Increase of CAMP-dependent Protein Kinase during Differentiation of the 3T3-LI Cells-The time course of increase in CAMP-dependent protein kinase, as monitored by the binding either of [3H]cAMP or 8-N3-[32P]
cAMP and histone kinase activity, was studied and compared to the time course of changes in cellular protein, malic enzyme, and glutamine synthetase (Figs. 6 and 7). The addition of insulin to the IBMX-dexamethasone-primed 3T3-Ll cells resulted in a rapid increase in total cellular protein; maximal increase (approximately a doubling) was observed 2 to 3 days after the addition of insulin (day 4-5) and declined slightly thereafter. Previous studies have demonstrated increases in malic enzyme (7, 8) and glutamine synthetase (24, 29) activities during adipose conversion of the 3T3-Ll cells. Malic

3T3-Ll cells, confluent 3T3-Ll fibroblasts (lane A )
were either treated with IBMX-dexamethasone and insulin for 7 days according to proenzyme may participate in lipogenesis by its capacity to generate reducing equivalents as NADPH, while glutamine synthetase is essential for supplying amide nitrogen for the synthesis of cellular proteins and nucleic acids. In the results illustrated in Fig. 6, the increase in malic enzyme appeared to parallel the increase in glutamine synthetase; detectable increases of both enzyme activities were observed on day 7 and remained on the upswing on day 11, the longest time point studied. This is to be contrasted with the increase in [3H] CAMP-binding and histone kinase activities, which rose in consonance, reaching a maximum at approximately day 5, and CAMP-binding and histone kinase activities (compare Figs. 6  and 7 ) . No signifcant change either in the CAMP-dependent protein kinase or glutamine synthetase activity was observed in the control samples over the entire culture period, while slight increases in total cellular protein and malic enzyme activity were observed.

Specificity of the Increase in CAMP-dependent Protein Kinase in 3T3-Ll Adipocytes-
The conversion of 3T3-Ll fibroblasts into adipocytes represents a spontaneous and heritable event, which can be accelerated by treating the cells with IBMX-dexamethasone and insulin. The results obtained thus far demonstrated an increase in CAMP-dependent protein kinase activity in cells treated with IBMX-dexamethasone and insulin, it remains to be demonstrated that such an increase is indeed related to adipose conversion per se rather than to some effects of IBMX, dexamethasone, or insulin, unrelated to adipose conversion. In view of this consideration, confluent 3T3-Ll cells were allowed to age and spontaneously differentiate in the normal serum-containing growth medium for a period of 4 weeks; at this point, cells, which appeared to be 50-60% differentiated under a phase contrast microscope, were harvested and analyzed for the incorporation of 8-Ns-["2P]cAMP. Fig. 8 illustrates the pattern of incorporation of 8-N3-[32P]~AMP into cytosolic proteins of confluent-undifferentiated, IBMX-dexamethasone and insulin-induced differentiated, and spontaneously differentiated 3T3-Ll cells. An increase in the incorporation of 8-N&'P]cAMP into the 47,000-dalton protein was observed in the hormone-induced as well as the spontaneously differentiated 3T3-Ll cells over that of the control. The results suggest that the increased incorporation of 8-N3-[32P]cAMP into Rl, and hence the increase in type I CAMP-dependent protein kinase, is directly related to cell differentiation rather than to the means used to promote cell differentiation.
Several observations suggest that the differentiation of 3T3-L1 fibroblasts to adipocytes may involve modulation of intracellular cAMP concentration. For example, adipocyte conversion of the 3T3-Ll fibroblast can be facilitated by treating the confluent cells with either prostaglandin El (30) or phosphodiesterase inhibitors (13,31). In addition, previous studies (32, 33) have demonstrated increased level of cAMP in the 3T3 fibroblasts upon cessation of growth at confluency. These results, in conjunction with the ability of cAMP to induce differentiation of the mouse neuroblastoma cells (for review, see Ref. 34) and the enhancement of CAMP-binding activity upon neuroblastoma differentiation (21,35,36), prompted me to investigate the possible role of cAMP in the induction of type I CAMP-dependent protein kinase during 3T3-Ll adipo-Protein Kinase and 3T3 Differentiation cyte conversion. In Table 11, five different cell extracts obtained from five groups of cells subjected to various treatments were examined for (a) malic enzyme activity (as an index of lipogenic activity), ( b ) cellular cAMP level, (c) CAMP-dependent protein kinase activity, as measured by the binding of [3H]cAMP and histone kinase assay, and (d) cAMP phosphodiesterase activity. The five groups of cells being examined were: 1) untreated 3T3-U cells harvested at confluency (day O), 2) IBMX-dexamethasone and insulin-induced differentiated 3T3-U adipocytes harvested on day 7 of the treatment, 3) confluent 3T3-Ll cells treated for 2 days with 0.5 m~ IBMX followed by refeeding the cells with normal growth medium and harvesting on day 7, 4) confluent 3T3-Ll cells treated for 2 days with 1 mM dibutyryl CAMP followed by refeeding the cells with normal growth medium and harvesting on day 7, and 5) confluent 3T3-Ll cells treated for 7 days with 1 m~ dibutyryl CAMP.
Differentiation, as monitored either by the morphological appearance of fat droplets or the marked increases in malic enzyme and CAMP-dependent protein kinase activity, was observed only in those cells that underwent the routine IBMX-dexamethasone and insulin treatment. The intracellular cAMP level and cAMP phosphodiesterase activity of the differentiated cells did not differ significantly from that of the control. Treatment of confluent 3T3-Ll cells with either 0.5 m~ IBMX or 1 mM dibutyryl cAMP for 2 days had little or no effect on the activities of malic enzyme and CAMP-dependent protein kinase 5 days later, In fact, both malic enzyme and CAMP-dependent protein kinase appeared to decrease significantly in cells treated with 1 mM dibutyryl CAMP for the entire 7-day period. These results suggested that, under the experimental conditions and treatment regimens used, raising cAMP concentration (either by the inhibition of phosphodiesterase or by the addition of dibutyryl CAMP) had either no or negative effect on the expression of malic enzyme and CAMP-dependent protein kinase in the 3T3-L1 cells. A previous study has noted a similar effect of cAMP on mdic enzyme activity (37). These actions of cAMP were to be contrasted with the effect of cAMP on phosphodiesterase activity. A significant increase in cAMP phosphodiesterase activity was observed in cells treated with 1 mM dibutyryl cAMP either for 2 or 7 days. The result is reminiscent of that obtained with the S-49 mouse lymphoma cells (38) and the mouse neuroblastoma cells (39); it further substantiates the notion that cAMP may be self-regulated through the induction of phosphodiesterase activity.
A potentially useful approach in delineating the significance of the induction of type I CAMP-dependent protein kinase in the differentiation of 3T3-Ll c e k is to examine this parameter in two related cell lines, one endowed with the ability to undergo adipose conversion under the appropriate experimental conditions and the other not. To this end, the 3T3-Ll and the 3T3-C2 cells were used (4). Results of the incorporation of 8-N3-f3'P]cAMP into cytosols of control and IBMX-dexamethasone and insulin-treated 3T3-Ll and 3T3-C2 cells are presented in Fig. 9. Unlike the 3T3-Ll cells, treatment of the 3T3-C2 cells, a cell line with a much lower intrinsic ability to undergo adipose conversion, did not result in increased incorporation of 8-N3-f3'P)cAMP into the 47,000-dalton protein.

DISCUSSION
Homogenates, particulate fractions, and cytosol fractions of the 3T3-Ll fibroblasts and adipocytes were examined for CAMP-dependent protein kinase activity and for the incorporation of 8-N&2P]cAMP into specific CAMP-receptor proteins. Greater than 80% of these activities found in the homogenate was recovered in the cytosol fraction (data not shown). Results presented in this study are limited to the cytosolic CAMP-dependent protein kinase, although qualitatively similar results were obtained using the particulate fractions.
Analysis o E cytosolic CAMP-dependent protein kinase of 3T3-Ll fibroblasts and adipocytes indicates a significant increase during adipose development. This increase in CAMPdependent protein kinase activity was rapid and appeared to be determined by the operation of the intrinsic adipocyte differentiation program. The results presented in this study are in contrast to the results of a previous study by Kawamura et al. (12), which reported that the level of CAMP-dependent protein kinase activity in 3T3-Ll cells did not increase during differentiation. The reason for this discrepancy is not known.
The enhanced CAMP-dependent protein kinase activity in the differentiated 3T3-Ll adipocyte is due primarily to a 3-to 6-fold increase in the type I CAMP-dependent protein kinase. This conclusion is based on 1) the separation and analysis of the type I and type I1 enzymes by DEAE-Sephacel column chromatography, and 2) the incorporation of 8-Ns-[32P]cAMP into the 47,000-dalton protein and the 52,000-54,000-dalton dimer, representing regulatory subunits of the type I and type I1 kinases, respectively. Previous studies on the CAMP-binding proteins of bovine cardiac muscle have also demonstrated the occurrence of both a 52,000-and a 54,000-dalton 8N3-[32P]

TABLE 11
Effects of various treatments on the expression of (a) malic enzyme, (b) cellular cAMP-level, (e) CAMP-dependent protein kinase, and (dl phosphodiesterase activities in 3T3-Ll cells 3T3-Ll ceUs were plated in 100-mm dishes. Upon reaching con-refeeding the cells with normal growth medium and harvesting on fluency, cells were divided into five groups (each group consisted of day 7, and 5) confluent 3T3-Ll cells treated for 7 days with 1 mM six 100-mm plates), and were treated accordingly: 11 untreated 3T3-dibutyryl CAMP. Of each of these five groups of cells, one 100-mm L1 cells harvested at confluency (day 0 ) , 2) IBMX-dexamethasone plate was processed for cAMP determination, while the remaining (DXM) and insulin-induced differentiated 3T3-LI adipocytes har-five plates were used-to obtain a cytosol fraction. The malic enzyme, vested on day 7 of the treatment, 3) confluent 3T3-LI cells treated CAMP-dependent protein kinase, and phosphodiesterase activities for 2 days with 0.5 mM IBMX followed by refeeding the cells with present in these cytosol preparations were determined according to normal growth medium and harvesting on day 7,4) confluent 3T3-L1 methods described. cells treated for 2 days with 1 m~ dibutyryl CAMP followed by cAMP binding protein, identified as regulatory subunits of the type I1 kinase of bovine heart (40). Unlike the enzyme from bovine cardiac muscle, the 52,000-and 54,000-dalton proteins of the 3T3-Ll cells were apparently not interconvertable by phosphorylation and dephosphorylation of the proteins. The apparent affiity for the incorporation of 8-N,-["P] cAMP into the 52,000-and 54,000-dalton proteins appeared similar. Furthermore, these two proteins co-eluted, upon DEAE-Sephacel column chromatography, with the type I1 CAMP-dependent protein kinase. Based on these observations, and the difficulty involved in resolving and separately quantitating the 52,000-and 54,000-dalton ~-Ns-[~~P]cAMP binding proteins, the amounts of radioactivity incorporated into these two proteins were summed, representing the total CAMP-binding activity of the type I1 kinase.
The significance of the presence of a small but persistent amount of type I CAMP-dependent protein kinase in the undifferentiated 3T3-Ll cells and the mechanisms of its increased expression upon adipose conversion are not clear. The 3T3-Ll cell is a line of 3T3 cells selected for its propensity to undergo adipocyte conversion. While >90% of the cells appeared fibroblastic and >60-80% of the cells possessed fat droplets in those cultures labeled, respectively, as "undifferentiated" and "differentiated," the two morphologically distinct cell forms can and do co-exist in any given culture. It is possible that the type I CAMP-dependent protein kinase of the undifferentiated culture may exist exclusively in the minority of cells already committed to differentiation. In this regard, immunofluorescent localization of the regulatory subunits of type I and type I1 CAMP-dependent protein kinase at a single cell level would be useful in resolving this problem.
The catalytic subunits of the type I and type I1 CAMPdependent protein kinase appear similar (41), and it has been difficult, if not impossible, to assign specific functions to each of the two enzymes. In this regard, the de novo synthesis of enzymes that are subjected to regulation by cAMP and CAMPdependent protein kinase and the increased expression of type I CAMP-dependent protein kinase in the differentiated 3T3-L1 cells in a rapid (days rather than months) and predictable manner provides a unique opportunity to examine the functions of a specific CAMP-dependent protein kinase in the regulation of specific biochemical events. Differentiation-dependent regulation of CAMP-dependent protein kinase is not unique to the 3T3-Ll cells. Changes in CAMP-binding and CAMP-dependent protein kinase activities during growth and differentiation of the water mold, Blastocladialla ernersonii, have previously been demonstrated (42). Studies on the differentiating sperm cells from mouse testis showed a remarkable decrease in the ratio of type I/type I1 CAMP-dependent protein kinase upon conversion of pachytene spermatocytes to elongating spermatids (43). The expression of the type I and type I1 CAMP-dependent protein kinases also appears to be tightly geared to cell cycle. Studies by Costa et al. (44) have demonstrated a cell cycle-dependent oscillation of the type I and type I1 kinase activity in synchronized Chinese hamster ovary cells. In the systems discussed above, differentiation appears to bring about a concerted alteration in the expression of the regulatory and catalytic subunits of CAMP-dependent protein kinase. This is to be contrasted with results obtained in the mouse neuroblastoma cells. Studies from various laboratories, including our own, have demonstrated enhanced CAMP-binding activity unassociated with CAMP-dependent protein kinase in the differentiated neuroblastoma cells (21,35,36,39,45).
CAMP is an important regulator in many biological functions. It is perhaps not surprising that cell differentiation, a process which can involve global changes in cellular function and metabolism, is associated with specific changes in intracellular CAMP-dependent protein kinase or CAMP-binding activities. The 3T3-Ll cells may provide an ideal system for studying the mechanisms and functional significance of the changes in CAMP-dependent protein kinase activity during development and differentiation.