Acquisition of Increased Hormone Sensitivity during in Vitro Adipocyte Development*

Murine 3T3-L1 fibroblasts enter a differentiation program subsequent to prolonged maintenance in the confluent state and develop into adipocytes. The hormone sensitivity of adenylate cyclase and the physiological responsiveness to insulin were compared in 3T3-L1 preadipocytes and adipocytes. The following observations, comprising several distinct categories of hormone responsiveness, were made. (a) (2.5 micronM) isoproterenol stimulated adenylate cyclase 15-fold in adipocyte homogenates, but only 2.5-fold in preadipocyte preparations, suggesting a considerable magnification in beta-adrenergic responsiveness during development. (b) A totally new control element, adrenocorticotropic hormone responsiveness, was incorporated into the adenylate cyclase system of the adipocytes. (c) Sensitivity to prostaglandin E1 was observed in both preadipocytes and adipocytes, but no change in responsiveness could be detected in the differentiated cells. (d) Glucagon-sensitive adenylate cyclase could not be detected in either preadipocytes or adipocytes. (e) Both preadipocytes and adipocytes possess considerable insulin binding activity, but near physiological levels of insulin stimulate the conversion of glucose to CO2 and lipid only in the differentiated cells.


SUMMARY
Murine 3T3-Ll fibroblasts enter a differentiation program subsequent to prolonged maintenance in the confluent state and develop into adipocytes. The hormone sensitivity of adenylate cyclase and the physiological responsiveness to insulin were compared in 3T3-Ll preadipocytes and adipocytes. The following observations, comprising several distinct categories of hormone responsiveness, were made. (a) (2.5 PM) isoproterenol stimulated adenylate cyclase 15-fold in adipocyte homogenates, but only 2.5-fold in preadipocyte preparations, suggesting a considerable magnification in padrenergic responsiveness during development. (b) A totally new control element, adrenocorticotropic hormone responsiveness, was incorporated into the adenylate cyclase system of the adipocytes. (c) Sensitivity to prostaglandin E, was observed in both preadipocytes and adipocytes, but no change in responsiveness could be detected in the differentiated cells. (d) Glucagon-sensitive adenylate cyclase could not be detected in either preadipocytes or adipocytes. (e) Both preadipocytes and adipocytes possess considerable insulin binding activity, but near physiological levels of insulin stimulate the conversion of glucose to CO, and lipid only in the differentiated cells.
Green and Kehinde (l-3) established and cloned several sublines of mouse 3T3 fibroblasts (4) that are capable of differentiating into adipocytes in vitro. During the course of exponential growth and the approach to confluence, preadipose 3T3-Ll cells are indistinguishable from previously established 3T3 lines with respect to growth rate, sensitivity to densitydependent inhibition, cell morphology (l-4), and the specialized biosynthesis of collagen (5). When these cells achieve confluence and are subsequently maintained in the resting state by replenishing the medium every 2 days, a significant proportion spontaneously enters a differentiation program that parallels the development of mammalian adipose tissue (1,3,5 25, pp. 3554.3557, 1911 Prrnted L,, U.S.A. the preadipose cells retract their processes, pass through several stages of morphological differentiation that mirror mammalian adipocyte development, and finally emerge as enlarged and rounded adipocytes that remain attached to the substrate (3).
Several biochemical correlates of elevated triglyceride formation have been established. 3T3-Ll adipocytes incorporate palmitate, glucose (2), and acetate (2, 6) into triglycerides at rates that are lo-to loo-fold greater than those observed in preadipose 3T3-Ll cells and other 3T3 fibroblasts. Moreover, the 50-fold increment in de ~OL'O lipogenesis indicated by [14Clacetate incorporation into triglycerides appears to be accomplished by a coordinate 40-to 50-fold increase in the activities of ATP-citrate lyase, acetyl-CoA carboxylase, and fatty acid synthetase (6).
The spontaneous adipose conversion process occurs over a period of 2 to 4 weeks after the cells reach confluence. Treatment of confluent 3T3-Ll cells with either 20 to 30% serum (51, 0.18 PM insulin (2), 0.5 mM 1-methyl-3-isobutylxanthine (71, 0.3 PM prostaglandin F,, (71, 33 PM biotin (61, or combinations of insulin and any of the other agents (6, 7) significantly accelerates the expression of the fat cell phenotype. Conversely, adipocyte differentiation and development are blocked when the cells are grown in the presence of 5 PM bromodeoxyuridine (2, 5). Since (a) mature mammalian fat cells contain adenylate cyclases that are activated by a number of lipolytic polypeptide hormones and p-adrenergic agents (see Ref. 8 for review) as well as insulin receptors that control lipogenesis (9) and (b) triglyceride synthesis in 3T3-Ll adipocytes is enhanced by insulin and inhibited by epinephrine (2), we commenced investigations on the development, regulation, and physiological coupling of hormone receptors and effector systems in 3T3-Ll cells. In this report, the hormone sensitivity of adenylate cyclase in intact and broken cell preparations and the physiological responsiveness to insulin are compared in confluent, preadipose cells, and differentiated 3T3-Ll adipocytes.  (7). Confluent 3T3-Ll cells were exposed to 0.5 rnM 1-methyl-3-isobutylxanthine and 0.2 PM insulin in standard medium for 48 h. Subsequently, the medium was removed, the cells were washed with isotonic phosphate-buffered saline, pH 7.4, and were then refed with medium containing 0.2 PM insulin. Cells were refed with medium containing insulin every 2 days. Differentiation was monitored by phase contrast microscopy. Cells containing multiple fat droplets were scored as differentiated.
Accumulation of lipid in the droplets was verified by staining with oil red 0 (1). At least 80% of the cells in differentiating cultures could be identified as adipocytes 7 to 8 days after the termination of treatment with lmethyl-3-isobutylxanthine.
Unless otherwise indicated, experiments on adipocytes were carried out on cultures that were 85 to 95% differentiated.
Adipocytes were compared to confluent preadipose 3T3-Ll cells (preadipocytes) that had not been treated with l-methyl-3-isobutylxanthine or insulin, but were fed every 2 days with standard medium.

Adenylate
Cyclase Assay The standard assay conditions are essentially those described by Salomon et al. (10). Assays were performed for 10 min at 37" in 50 ~1 as 10 microunits of insulin/ml (70 PM), the comparisons to be presented here were made at the near saturating concentration of 360 microunitsiml (2.5 nM1. All studies on insulin sensitivity were performed on cells which had either never been exposed to added insulin (preadipocytes) or were withdrawn from insulin for 24 h prior to assay (adipocytes).

Determination of Cyclic AMP Content
Trichloroacetic acid-free cell extracts were acetylated (111 and the amount of cyclic AMP present was determined by radioimmunoassay (12). ""I-P'-0-succinyl cyclic AMP tyrosine methyl ester, anti-cyclic AMP antibody, and goat anti-rabbit y-globulin were obtained from Collaborative Research.
The iodinated product was purified by chromatography on cellulose (14) and had a specific activity of 206 ~CiI~g.

Insulin
Binding Assay Assessment of insulin binding by intact cells was carried out according to Thomopoulos et al. (15) using 5 x lo" cells/assay and lziIinsulin at a concentration of 0.33 nM.
-The absolute values obtained using the preceding assays varied as much as 100% from one preparation of cells to another. However, the degree of adenylate cyclase stimulation and alteration in insulin binding during differentiation were highly consistent.  (Table  I). In parallel with the preadipocytes, adipocyte cyclase was not affected by glucagon (Table I) were directly added to 60-mm culture dishes containing standard growth medium and 0.1 rnM 1-methyl-3-isobutylxanthine. After 5 min, the medium was aspirated, and the cells were washed once with 5 ml of phosphate-buffered saline (~10 s) and then extracted with 3 ml of 5% trichloroacetic acid. Acid-insoluble macromolecules were sedimented at 2,000 x g for 5 min at 4". Pellets were dissolved in 0.  (Tables I and II). Responsiveness to these two agents defines two categories of developmental change in hormone sensitivity.
In the case of the P-adrenergic agonist, heightened responsiveness represents a considerable magnification of a pre-existing, P-agonist-sensitive system such that enzyme activity is stimulated 15-fold (Table I) (21,22). Occupation of these receptors by insulin is coupled, by unknown mechanisms, to the stimulation of glucose uptake and metabolism, the enhancement of lipogenesis and the accumulation of triglycerides and a variety of other metabolic effects (23,24).
To determine whether 3T3-Ll cells also provide and single cell suspensions were prepared. Suspended cells were sedimented at 200 x g and washed once with 3 ml of the insulin binding medium (100 rnM Hepes, 120 rnM NaCl, 1.2 rnM MgCl,, 5 rnM KCl, 10 rnM glucose, 1 mM EDTA, 16 mM sodium acetate, and 1% bovine serum albumin, pH 7.9). Cells were then resuspended in insulin binding medium to a density of 1 x 106/ml and binding assays were carried out for 90 min at 18" (15, 25). Nonspecific binding was equal to 10 to 12% of the specific insulin l&&g and was subtracted from the total amount bound. Da%. cqe presented gs the percentage of total lz51insulin bound per lo6 cells.
,_ -._ Stage of cell differentiation 'Z51-insulin bound/lo6 cells % Confluent preadipocytes 4.8 25% differentiated 9.0 60% differentiated 6.7 tally differentiated cells. This finding agrees with the data of Thomopoulos et al. (15) who have recently reported the appearance of a high level of insulin binding activity when 3T3 flbroblasts reach confluence. Although COs production and lipid synthesis in cultures containing 60 to 75% differentiated cells were comparable or higher than the values given in Table   III, insulin binding activity declined 28% in the same cultures.
In several additional experiments, insulin binding activity consistently peaked when the cultures were 25 to 40% differentiated and slowly diminished thereafter; no obvious direct correlation between physiological responsiveness and binding activity was found. However, more detailed analyses of insulin binding may reveal differences not apparent from our measurements. The 3T3-Ll cells should provide an excellent system for studying insulin receptor synthesis and the coupling of these receptors to physiological effecters. The initial data presented here suggest that the relationship between binding and insulin responsiveness in these cells may be quite complex.