Somatomedins (insulin-like growth factors), but not growth hormone, are mitogenic for chicken heart mesenchymal cells and act synergistically with epidermal growth factor and brain fibroblast growth factor

Chicken, ovine or human growth hormones have no mitogenic effect on chicken heart mesenchymal cells, which are proliferatively quiescent at low culture densities in medium containing heparinized, heat-defibrinogenated rooster plasma at 10%. Sm-C/IGF-I (15 ng/ml; 2 nM), MSA/rIGF-II (50 ng/ml; 7 nM), insulin (10,000 ng/ml; 1750 nM) or proinsulin (16,000 ng/ml; 1750 nM), however, cause these cells to increase threefold in number during four days of incubation. While EGF alone at 100 ng/ml causes threefold multiplication at four days and brain FGF causes a sixfold increase, EGF acts synergistically with Sm-C/IGF-I, MSA/rIGF-II, insulin or proinsulin to cause 18-fold multiplication, and brain FGF acts synergistically with IGFs to cause 20-fold multiplication. EGF and brain FGF, however, show no mitogenic synergy. Addition to the plasma-containing culture medium of a monoclonal antibody to Sm-C/IGF-I nearly abolishes the mitogenic effect of added EGF or brain FGF but does not affect the autonomous (mitogenic hormone-independent) proliferation of RSV-infected chicken heart mesenchymal cells. These findings support the somatomedin hypothesis for growth hormone action and suggest that potentiation of the activity of other mitogenic hormones, like EGF and FGF, makes a significant contribution to control of cell proliferation by the GH/IGF axis.     

Activation of signal transduction pathways protects quiescent Balb/c-3T3 fibroblasts against death due to serum deprivation.

Platelet-derived growth factor (PDGF), epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), and insulin protect density-inhibited murine Balb/c-3T3 fibroblasts against death by distinctive mechanisms. Determination of the cell survival-enhancing activity of growth factors by cell enumeration and neutral red uptake measurement gives equivalent results. PDGF displays a steep dose-response relationship in the 1-5 ng/ml range. The other factors display shallow log-linear relationships in the following ranges: EGF: 0.2-5 ng/ml; IGF-1: 2-80 ng/ml; and insulin: 57-4,500 ng/ml. Agonists that lead to the activation of protein kinase A, including forskolin, 8-bromoadenosine 3':5'-cyclic monophosphate (Br-cAMP) and N6,2'-O-dibutyryladenosine 3':5'-cyclic monophosphate (db-cAMP), markedly increase both short-term (5-h) and long-term (20-h) survival of cells. 2-Isobutyl-1-methylxanthine (IBMX) markedly enhances short-term survival, but its effect decays with time. The protein kinase C agonist 12-O-tetradecanoyl phorbol-13-acetate (TPA) has a moderate protective effect at concentrations of 16-32 nM, and 64 nM TPA is highly effective. The synthetic diaclglycerols 1,2-dioctanoylglycerol (DiC8) and 1-oleoyl-2-acetylglycerol (OAG) and the calcium ionophore ionomycin show low activity. Supplementation of EGF with a protein kinase A or C agonist results in a varying additive increase in short-term (5-h) cell survival and supplementation of EGF + insulin or PDGF + EGF + insulin increases further the already high level of protection given by the growth factor combinations. Combining a protein kinase A and a protein kinase C agonist in the absence of growth factors gives an approximately additive increase in cell survival. Results obtained with kinase, RNA, and protein synthesis inhibitors suggest that: 1) activated protein kinase C catalyzes one or more phosphorylation events in quiescent Balb/c-3T3 cells that lead to gene expression with the protein product(s) mediating protection of quiescent cells against death, and 2) phosphorylation events catalyzed by protein kinase A largely serve to protect cells by a mechanism not requiring de novo RNA and protein biosynthesis.     

Glucocorticoids regulate the secretion of a 21kDa-IGF-binding protein by sheep adipose tissue explants

Using a solution phase assay we have demonstrated that sheep adipose tissue explants secrete insulin-like growth factor binding proteins (IGFBPs) when cultured in serum-free medium over a 24 h period. Further, we demonstrate that secretion of IGFBP(s) is inhibited (up to 50%) by incubation of the cultures in the presence of 10^–8M dexamethasone. This inhibitory effects is overcome when insulin (10 ng/ml) and ovine growth hormone (100 ng/ml) are incubated together (but not separately) with glucocorticoid. Further characterisation of this IGF binding activity by high performance size exclusion chromatography and Western ligand blot analysis indicated that under our culture conditions sheep adipose tissue explants secrete one predominant 21 kDa IGFBP and it is this BP which is hormonally regulated as described above. We discuss our results in the context of endocrine/paracrine/autocrine control of adipose tissue metabolism and differentiation.Abbreviations IGF insulin-like growth factor - IGF-BP insulin-like growth factor binding proteins - DX dexamethasone - GH ovine growth hormone - CM conditioned medium     

Aromatase activity in granulosa cells: Regulation by growth factors

This study describes the effects of insulin, insulin-like growth factor 1 (IGF₁), and epidermal growth factor (EGF) on the aromatase activity of granulosa cells isolated from immature rat ovaries. None of the growth factors alone influenced the basal level of aromatase activity, but did modulate follicle-stimulating hormone (FSH)-induced aromatase activity. Insulin and IGF₁ augmented the action of a sub-optimal concentration of FSH (5 ng/mL) on aromatase activity in a dose-dependent manner. In contrast, EGF (1–10 ng/mL) was effective in inhibiting aromatase activity maximally stimulated by FSH. Since insulin and IGF₁ had opposing actions to those of EGF on FSH-induced aromatase activity, we examined the interactions between the growth factors. EGF inhibited the actions of both FSH and insulin on aromatase activity. Both IGF₁ and EGF increased the [³H]thymidine incorporation into the DNA of bovine granulosa cells , IGF₁ being a more potent mitogen. Whereas EGF inhibited the actions of IGF₁ on aromatase activity, it did not inhibit the effects of IGF₁ on the growth of granulosa cells. In summary, growth factors influence both the differentiation and growth of granulosa cells, and may be important regulators of follicular development.     

Growth of human foreskin fibroblasts in a serum-free,defined medium without platelet-derived growth factor

The hormones which support growth, in vitro, of normal, neonatal human foreskin fibroblasts were determined. Wheresas thrombin and hydrocortisone were major growth stimulants, platelet-derived growth factor was not. Human foreskin fibroblasts grew in a serum-free, biochemically defined medium consisting of epidermal growth factor (100 ng/ml), insulin (100 ng/ml), trasferrin (10 μg/ml), thrombin (1 μg/ml), ascorbic acid (10 μg/ml), and hydrocortisone (5 × 10^?5M) in a 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F-12, supplemented with ovalbumin (1 mg/ml) and trace elements. The growth achieved was comparable to that achieved with 5% fetal bovine serum. Neither platelet-derived growth factor, fibroblst growth factor, nor somatomedin activity increased proliferation. This serum-free medium designated Defined Medium F, provides a biochemically defined system for growth and limited subcultivation of human foreskin fibroblasts in vitro.     

Insulin alters the effects of follicle stimulating hormone on aromatase in bovine granulosa cells in vitro

It is known that follicle-stimulating hormone (FSH) and insulin stimulate estradiol secretion from cultured non-luteinizing granulosa cells. The interaction between these hormones is less well understood. Granulosa cells from small (2-4 mm) bovine follicles were cultured in serum-free medium to determine if cytochrome P450 aromatase activity is regulated by FSH in the presence of different concentrations of insulin. Insulin significantly stimulated aromatase activity in the absence of FSH. There was a significant interaction between insulin and FSH on aromatase activity, such that FSH stimulated activity at low (0.5, 1 and 10 ng/ml) doses of insulin, whereas at higher (100 ng/ml) doses of insulin FSH failed to stimulate aromatase activity. To determine if the lack of a response to FSH with higher doses of insulin is related to gene expression, the effect of FSH on P450 aromatase mRNA levels was measured. An 'uncoupling' of mRNA and enzyme activity was observed for cells cultured with 100 ng/ml insulin, as FSH significantly increased P450 aromatase mRNA abundance without affecting estradiol secretion or aromatase activity. We conclude that in the presence of high doses of insulin, FSH decreases aromatase activity, and an uncoupling of P450 aromatase mRNA and aromatase activity occurs. This may have implications for infertility treatments when there is a risk of hyperinsulinemia.     

Hormone-sensitive particulate cAMP phosphodiesterase activity in 3T3-L1 adipocytes. Regulation of responsiveness by dexamethasone

Confluent 3T3-L1 fibroblasts incubated for 72 h with methylisobutylxanthine, dexamethasone, and insulin differentiate and acquire phenotypic characteristics of mature adipocytes, including hormone-sensitive cAMP phosphodiesterase activity located in a particulate fraction of homogenates. About 10 days after initiating differentiation, a maximally effective concentration of insulin (100 pM) increased particulate cAMP phosphodiesterase activity 40 to 60% in 8 min; activation persisted for at least 30 min in the presence of insulin. Incubation of adipocytes for 6-8 min with agents that increased cAMP, e.g. 1 microM epinephrine, 0.1 microM isoproterenol, corticotropin (2 mu units/ml), or thyroid-stimulating hormone (15 ng/ml), also increased particulate phosphodiesterase activity 40-60%. Changes in phosphodiesterase activity produced by epinephrine tended to lag behind changes in cAMP. Insulin, epinephrine, and corticotropin increased Vmax, not Km (0.5 microM), for cAMP. Particulate phosphodiesterase activity, solubilized with detergent, eluted in a single peak from DEAE-Bio-Gel. Insulin and epinephrine increased the activity eluted in this peak. Neither insulin nor lipolytic hormones increased activity in soluble fractions from differentiated cells or particulate or soluble fractions from undifferentiated cells. Incubation of adipocytes for 48 h with 1 microM dexamethasone prevented insulin-induced activation of the particulate phosphodiesterase and did not alter basal activity. After incubation for 72 h with 0.1 microM dexamethasone, insulin and epinephrine activation were abolished. These effects of dexamethasone on hormonal regulation of particulate phosphodiesterase activity could account for some of the so-called permissive effects of glucocorticoids on cAMP-mediated processes as well as the "anti-insulin" effects of glucocorticoids.     

Mitogen-activated S6 kinase is stimulated via protein kinase C-dependent and independent pathways in Swiss 3T3 cells

Soluble extracts prepared from quiescent Swiss mouse 3T3 cells that had been briefly exposed to various mitogens exhibited a 2- to 3-fold elevation in phosphorylating activities toward ribosomal protein S6 and a synthetic peptide, Arg-Arg-Leu-Ser-Ser-Leu-Arg-Ala (RRLSSLRA), patterned after a phosphorylation site sequence from S6. Optimal activation of the phosphorylating activity occurred within 15-20 min of exposure of the cells to platelet-derived growth factor (10 ng/ml), epidermal growth factor (100 nM), and insulin (100 nM), and 2-5 min after 12-O-tetradecanoylphorbol-13-acetate (TPA) (100 nM) treatment. Fractionation of the cytosolic extracts from mitogen- or TPA-treated cells on Sephacryl S-300, TSK-400, and DEAE-Sephacel columns gave results suggesting that a single stimulated kinase accounted for the enhanced S6 and RRLSSLRA phosphorylating activities. The mitogen-activated kinase had an apparent Mr of about 85,000 as determined with Sephacryl S-300, but eluted with an apparent Mr of 26,000 from a TSK-400 high pressure liquid chromatography column. The S6 kinase was also stimulated in cytosols from insulin-like growth factor 1- (100 nM), vasopressin- (250 nM), prostaglandin F2 alpha- (250 nM), and 10% fetal calf serum-treated cells but not from quiescent cells exposed to beta-transforming growth factor (2 ng/ml). TPA, vasopressin and prostaglandin F2 alpha appeared to stimulate this kinase via a protein kinase C-dependent mechanism, since the responses to these hormones, but not to platelet-derived growth factor, epidermal growth factor, and insulin, were lost in protein kinase C-depleted cells.     

The effect of transforming growth factor beta on the intensity of cellular DNA synthesis in relation to the type and conditions of cultivation

A study was made of the influence of transforming growth factor beta (0.05-50.00 ng/ml) on the intensity of 3H-thymidine incorporation in the DNAs of pseudonormal mouse fibroblasts of NIH 3T3 line, of cells of NRK-49F line from normal rat kidney, and of cells of A-549 line from human lung adenocarcinoma. The experiments were carried out in the absence and in the presence of epidermal growth factor (5 ng/ml), and(or) insulin (1 micrograms/ml), as well as in the presence of different concentrations of fetal calf serum, and while using different time of incubation of cells with the above mentioned growth factors. It was shown that depending on the culture conditions the transforming growth factor beta exerted stimulatory, inhibitory or no action on the intensity of DNA synthesis in the cells of the same type. An attempt was made to analyse the reasons, which may significantly determine the direction of regulatory influence of the transforming growth factor beta on DNA replication in the cells.     

Hormonal control of the replication of human fetal fibroblasts: role of somatomedin C/insulin-like growth factor I

Sparse cultures of fetal and postnatal human fibroblasts were equivalent in their responsiveness to the mitogenic action of somatomedin C/insulin-like growth factor I (SM-C/IGF-I). At both developmental stages, the addition of SM-C/IGF-I (100 ng/ml) increased cell number at day 3 1.4-fold in serum-free medium and 2-fold in the presence of 0.25% human hypopituitary serum. Furthermore, dose-response curves indicated that there was no difference in the sensitivity of fetal and postnatal fibroblasts to the growth-promoting effects of SM-C/IGF-I, with a half-maximal response occurring at 6 ng/ml SM-C/IGF-I. This biological action of SM-C/IGF-I correlated with SM-C/IGF-I binding to fetal and postnatal fibroblast monolayers. Epidermal growth factor (EGF) and platelet-derived growth factor (PDGF) also stimulated replication of fetal and postnatal fibroblasts. The mitogenic effects of SM-C/IGF-I, EGF, and PDGF were additive. Dexamethasone, which alone had no effect, was synergistic with SM-C/IGF-I in stimulating replication of postnatal fibroblasts. The combination of SM-C/IGF-I (100 ng/ml), dexamethasone (10(-7) M), EGF (10 ng/ml), and PDGF (5 ng/ml) had the same mitogenic effectiveness as 10% calf serum (CS) in postnatal cells. In marked contrast, there was no mitogenic interaction between SM-C/IGF-I and dexamethasone in fetal fibroblasts. In fetal cells, SM-C/IGF-I + EGF + PDGF +/- dexamethasone could only account for 50% of the activity of 10% CS. Moreover, fetal cells were 50-100% more responsive than postnatal cells to the proliferative effect of serum.     

Lack of association of epidermal growth factor-, insulin-, and serum-induced mitogenesis with stimulation of phosphoinositide degradation in BALB/c 3T3 fibroblasts

The hypothesis that inositol phospholipid degradation is a step in the mechanism by which epidermal growth factor (EGF) stimulates mitogenesis in confluent monolayers of quiescent BALB/c 3T3 fibroblasts was tested. The maximum mitogenic response (a nearly 30-fold increase in incorporation of [3H]thymidine) occurred at 1 ng/ml EGF (0.16 nM). This degree of stimulation corresponded to 60% of that elicited by 10% serum. To determine whether EGF stimulated formation of inositol phosphates via degradation of polyphosphoinositides, the intracellular levels of [3H] inositol phosphates and [3H]phosphoinositides were determined after EGF addition to BALB/c 3T3 fibroblasts prelabeled with [3H]inositol. These experiments were performed under conditions designed to mimic exactly those conditions used to study mitogenesis. The results demonstrated that 10% serum or 10 ng/ml of platelet-derived growth factor, but not as much as 50 ng/ml EGF or 10 micrograms/ml insulin, increased the levels of inositol phosphates via degradation of phosphoinositides in the presence of 10 mM Li+. The serum-induced effects occurred in 30 s, the earliest time investigated. Phorbol dibutyrate (100 nM), alone or in conjunction with EGF (10 ng/ml), failed to stimulate inositol phospholipid degradation. However, phorbol dibutyrate inhibited the serum-induced stimulation. Finally, fetal bovine serum dialyzed so as to retain peptide mitogens lost almost 70% of the capacity to stimulate degradation of inositol phospholipids while remaining as mitogenic as the control serum. Thus, stimulation of inositol phospholipid degradation is an unlikely component in the mechanism by which EGF and probably insulin and serum stimulate mitogenesis in BALB/c 3T3 fibroblasts.     

The effect of tumor necrosis factor/cachectin on follicle-stimulating hormone-induced aromatase activity in cultured rat granulosa cells

We investigated the effects of tumor necrosis factor (TNF)/cachectin on follicle-stimulating hormone (FSH)-induced aromatase activity in cultured rat granulosa cells using the stereospecific transfer of 3H from [1 beta-3H] androstenedione into 3H2O. TNF (10 pg/ml-10 ng/ml) inhibited FSH (250 ng/ml)-induced aromatase activity in a concentration-dependent manner, and 10 ng/ml of TNF completely abolished the FSH-induced aromatase activity. A time course analysis of the effects of TNF showed that TNF had no effect on induced aromatase activity, but inhibited the further induction of the enzyme by FSH. TNF (10 ng/ml) also inhibited the ability of TGF beta (1 ng/ml) to enhance aromatase activity and increase progesterone synthesis. Thus, TNF is a component of the complex array of proteins that modulate ovarian function and, as such, may play a physiological role in the regulation of the granulosa cell. In view of its association with cachexia, it may also play a pathophysiological role in the suppression of reproductive function during chronic illness.     

The primary culture of mouse adipocyte precursor cells in defined medium.

1. A defined medium supporting the proliferation and differentiation of adipocyte precursors isolated from inguinal fat pads of 8-12-day-old mice was developed. 2. It consists of a 1:1 mixture of DME and WAJC404A media supplemented with insulin (10 micrograms/ml), transferrin (10 micrograms/ml), fibroblast growth factor (10 ng/ml) and high density lipoproteins (HDL) (90 micrograms protein/ml). 3. DME-F12 medium (1:1 mixture) used as a nutrient mixture in the defined medium of rat and human adipocyte precursors was inadequate for cultivating mouse adipocyte precursors. 4. HDL had a definite beneficial effect on both preadipocyte growth and differentiation. 5. Differentiation was enhanced by addition of dexamethasone (10(-9) M) but could be almost completely inhibited by transforming growth factor beta 1 (TGF-beta 1). 6. TGF-beta 1 was shown to be effective only when present in the early stages of differentiation.     

Characterization of the growth of murine fibroblasts that express human insulin receptors. II. Interaction of insulin with other growth factors

The effects of insulin-like growth factor-1 (IGF-1), epidermal growth factor (EGF), platelet-derived growth factor (PDGF), and insulin on DNA synthesis were studied in murine fibroblasts transfected with an expression vector containing human insulin receptor cDNA (NIH 3T3/HIR) and the parental NIH 3T3 cells. In NIH 3T3/HIR cells, individual growth factors in serum-free medium stimulated DNA synthesis with the following relative efficacies: insulin greater than or equal to 10% fetal calf serum greater than PDGF greater than IGF-1 much greater than EGF. In comparison, the relative efficacies of these factors in stimulating DNA synthesis by NIH 3T3 cells were 10% fetal calf serum greater than PDGF greater than EGF much greater than IGF-1 = insulin. In NIH 3T3/HIR cells, EGF was synergistic with 1-10 ng/ml insulin but not with 100 ng/ml insulin or more. Synergy of PDGF or IGF-1 with insulin was not detected. In the parental NIH 3T3 cells, insulin and IGF-1 were found to be synergistic with EGF (1 ng/ml), PDGF (100 ng/ml), and PDGF plus EGF. In NIH 3T3/HIR cells, the lack of interaction of insulin with other growth factors was also observed when the percentage of cells synthesizing DNA was examined. Despite insulin's inducing only 60% of NIH 3T3/HIR cells to incorporate thymidine, addition of PDGF, EGF, or PDGF plus EGF had no further effect. In contrast, combinations of growth factors resulted in 95% of the parental NIH 3T3 cells synthesizing DNA. The independence of insulin-stimulated DNA synthesis from other mitogens in the NIH 3T3/HIR cells is atypical for progression factor-stimulated DNA synthesis and is thought to be partly the result of insulin receptor expression in an inappropriate context or quantity.     

Negative regulation of hepatocyte growth factor gene expression in human lung fibroblasts and leukemic cells by transforming growth factor-beta 1 and glucocorticoids.

Hepatocyte growth factor (HGF), a mesenchymal-derived factor which regulates growth, motility, and morphogenesis of epithelial and endothelial cells, functions as a hepatotrophic and renotrophic factor for regeneration of the liver and kidney. We have now obtained evidence that transforming growth factor-beta 1 (TGF-beta 1) and glucocorticoids are negative regulators for HGF gene expression. When TGF-beta 1 or dexamethasone was added to cultures of MRC-5 human embryonic lung fibroblasts and HL-60 human promyelocytic leukemic cells, the amount of HGF secreted into the culture medium was inhibited to 30-40% of that of control cultures by 10 ng/ml TGF-beta 1 and to 40-50% by 10(-6) M dexamethasone. The inhibitory effect of TGF-beta 1 and dexamethasone on HGF synthesis in MRC-5 cells was additive, thereby suggesting that TGF-beta 1 and dexamethasone exert effects through distinct mechanisms. Hydrocortisone also inhibited HGF synthesis with the same potency as dexamethasone; however, testosterone, estriol, and beta-estradiol had no effect. The rate of HGF synthesis in MRC-5 cells, as measured by pulse labeling with [35S]methionine and subsequent immunoprecipitation, was suppressed to 30-40% of the control with 10 ng/ml TGF-beta 1, and to 30-45% by 10(-6) M dexamethasone. HGF mRNA levels in MRC-5 cells and HL-60 cells were dose-dependently suppressed by TGF-beta 1 and dexamethasone; 10 ng/ml TGF-beta 1 suppressed HGF mRNA levels to 32% and 35% of control culture, respectively, in MRC-5 cells and HL-60 cells, and 10(-6) M dexamethasone suppressed to 43% and 38%, respectively. Thus, TGF-beta 1 and glucocorticoids seem to inhibit HGF synthesis by suppressing the expression of the HGF gene. We propose that a negative regulation of HGF gene expression by TGF-beta 1 or glucocorticoids may be involved in physiological or pathological processes during tissue regeneration.     

Purification of ovine transferrin and study of the hormonal control of its secretion in enriched cultures of ovine Sertoli cells.

Ovine transferrin (o-transferrin) was purified from sheep serum by fractionated precipitation with ammonium sulphate, ion-exchange chromatography on DEAE trisacryl and finally by affinity chromatography on Affigel blue to remove albumin. Ovine transferrin was identified by its apparent molecular weight in sodium dodecyl sulphate polyacrylamide gel electrophoresis and by its N-terminal amino-acid sequence. The procedure presented in this report permits the preparation of highly purified o-transferrin with a good recovery (52% of initial total immunoactivity). An antiserum against o-transferrin was then raised in rabbits, using this highly purified preparation. A specific radioimmunoassay was set up using 125I-labelled o-transferrin. Its detection threshold (4 ng/ml) was low enough to measure o-transferrin in spent culture media of ovine Sertoli cells, which ranged between 15 and 600 ng/ml. Sheep seminiferous tubule cells, containing approximately 80% Sertoli cells, were cultured at a high density (1.5 x 10(6) cells/cm2) on a thin layer of reconstituted basement membrane. Kinetic studies showed that basal daily secretion of o-transferrin was reduced by half (-49%) between Day 1 and Day 2 of culture, and progressively decreased thereafter. Under FIRT (500 ng ovine follicle-stimulating hormone (FSH)/ml + 10 micrograms insulin/ml + 500 ng retinol/ml + 5 x 10(-7) mol/l testosterone) stimulation, the ratio of stimulated to basal secretions increased 11-fold between Day 1 (1.1) and Day 6 (12). When 10% fetal calf serum was added, mean o-transferrin secretion was a third of that in serum-free medium, suggesting that fetal calf serum contains factors that inhibit secretion of ovine Sertoli cell transferrin. In the presence of serum, the ratio of FIRT-stimulated to basal secretions doubled between Day 1 (1.0) and Day 4-6 (2.0). Between Days 2 and 4 of culture, insulin had a slight stimulatory effect on o-transferrin secretion (128% of control at 10 micrograms insulin/ml), as well as epidermal growth factor (124% of control at 50 ng/ml). Testosterone at up to 5 x 10(-7) mol/l had no effect; 500 ng retinol/ml doubled o-transferrin secretion (218% of control) as did 500 ng FSH/ml (220% of control). A combination of retinol and FSH increased the secretion 4-fold, indicating that maximal stimulation of o-transferrin secretion by ovine Sertoli cells requires the combined actions of mechanisms dependent and independent of cAMP.