Involvement of a cGMP-dependent pathway in the natriuretic peptide-mediated hormone-sensitive lipase phosphorylation in human adipocytes.

Our previous studies have demonstrated that natriuretic peptides (NPs), peptide hormones with natriuretic, diuretic, and vasodilating properties, exert a potent control on the lipolysis in human adipocytes via the activation of the type A guanylyl cyclase receptor (1, 2). In the current study we investigated the intracellular mechanisms involved in the NP-stimulated lipolytic effect in human preadipocytes and adipocytes. We demonstrate that the atrial NP (ANP)-induced lipolysis in human adipocytes was associated with an enhanced serine phosphorylation of the hormone-sensitive lipase (HSL). Both ANP-mediated lipolysis and HSL phosphorylation were inhibited in the presence of increasing concentrations of the guanylyl cyclase inhibitor LY-83583. ANP did not modulate the activity of the cAMP-dependent protein kinase (PKA). Moreover, H-89, a PKA inhibitor, did not affect the ANP-induced lipolysis. On primary cultures of human preadipocytes, the ANP-mediated lipolytic effect was dependent on the differentiation process. On differentiated human preadipocytes, ANP-mediated lipolysis, associated with an increased phosphorylation of HSL and of perilipin A, was strongly decreased by treatment with the inhibitor of the cGMP-dependent protein kinase I (cGKI), Rp-8-pCPT-cGMPS. Thus, ANP-induced lipolysis in human adipocytes is a cGMP-dependent pathway that induces the phosphorylation of HSL and perilipin A via the activation of cGKI. The present study shows that lipolysis in human adipocytes can be controlled by an independent cGKI-mediated signaling as well as by the classical cAMP/PKA pathway.

surgery. Their mean age was 48.4 Ϯ 3.8 years, and their mean body mass index was 26.3 Ϯ 0.9 kg/m 2 . The study was approved by the ethical committees of Toulouse and Frankfurt am Main University Hospitals.
Human Mature Adipocyte Preparation-Isolated adipocytes were obtained according to the method of Rodbell (18) by collagenase digestion of adipose fragments in Krebs-Ringer bicarbonate buffer containing albumin (3.5 g/100 ml) and glucose (6 mmol/liter) at pH 7.4 and under gentle shaking at around 60 cycles/min at 37°C. Then the fat cells were filtered through a silk screen and washed three times with Krebs-Ringer bicarbonate buffer to eliminate collagenase.
Human Preadipocyte Preparation and Culture-The isolation of human fat cell precursors was performed as described by Hauner et al. (19). All of the visible fibrous material and blood vessels were discarded from adipose tissue samples. The remaining adipose tissue was cut into small pieces and digested in phosphate-buffered saline containing 0.15% collagenase and 2% bovine serum albumin under gentle shaking at 120 cycles/min at 37°C. After centrifugation, the stromal cell fraction was resuspended and incubated in an erythrocyte-lysing buffer (155 mmol/liter NH 4 Cl, 5.7 mmol/liter K 2 HPO 4 , 0.1 mmol/liter EDTA) and then filtered through a nylon mesh (150 m). After additional centrifugation, washing, and filtration steps, stromal cells were suspended in Dulbecco's modified Eagle's medium/Ham's F-12 medium supplemented with 10% fetal calf serum (60000 cells/cm 2 ). After 24 h of incubation, the medium was changed into medium consisting on Dulbecco's modified Eagle's medium/Ham's F-12 medium supplemented with 33 mol/liter biotin, 17 mol/liter panthotenate, 10 g/ml human transferrin, 50 g/ml gentamycin (basal medium) in the presence of 66 nmol/liter insulin, 1 nmol/liter triiodothyronine, 100 nmol/liter cortisol, and, for the first 3 days, 1 g/ml ciglitazone. The medium was changed every 2 days. Before the experiments, the cells were placed in basal medium for 24 h, washed with phosphate-buffered saline, and either lysed for Western blot analysis or treated for lipolysis experiments.
Lipolysis Measurement in Human Preadipocytes or Mature Adipocytes-Isolated mature adipocytes were brought to a suitable dilution (2000 -3000 cells/assay) in Krebs-Ringer bicarbonate buffer and incubated with pharmacological agents at the indicated concentrations for 90 min at 37°C. For lipolysis measurements, 10-day differentiated human adipocytes were incubated for 6 h in basal medium at 37°C in both the presence and absence of the pharmacological agents. At the end of the incubation, 20 -50-l aliquots of the infranatant were taken for glycerol determination (used as the lipolytic index) (20). Total lipid content was determined gravimetrically after solvent extraction.
Determination of cGMP Concentrations-Isolated mature adipocytes were preincubated in 1 ml of Krebs-Ringer bicarbonate buffer for 15 min at 37°C in the presence of 0.1 mmol/liter isobutylmethylxanthine (nonspecific PDE inhibitor). The cells were further incubated for 10 min both in the presence of and absence of 10 nmol/liter ANP. The reaction ceased with the addition of a solution of chloroform, methanol, 1 N HCl (2 volumes/1 volume/0.1 volume). To measure cGMP content according to the kit manufacturer instructions (Cayman Chemical Company, Ann Arbor, MI), the aqueous phase of each sample was freeze-dried and redissolved in the appropriate buffer after centrifugation (5000 rpm, 5 min).
RNA Extraction and RT-PCR Analysis-Total RNAs were extracted from human mature adipocytes or preadipocytes using the RNeasy kit (Qiagen). RNA concentrations were determined using a fluorometric assay (Ribogreen). For the RT, 1-2 g of total RNA were incubated with 200 units of reverse transcriptase (SuperScript II; Invitrogen), dNTP (0.5 mmol/liter), hexamer (25 ng/liter), dithiothreitol (0.01 mol/liter), and reaction buffer in a final volume of 20 l at 42°C for 50 min and then at 70°C for 15 min. In some reaction mixtures, total RNA was omitted to ensure the lack of amplification of contaminating genomic DNA, which was always the case in the present study. After a final denaturation at 94°C for 4 min, 10 l of cDNA was subjected to PCR consisting of a denaturation at 94°C for 1 min followed by 1 min of hybridization with specific primers at a variable temperature that depends on the target gene, and 90 s of elongation at 72°C for n cycles. The last cycle ended with 7 min of elongation at 72°C. The annealing temperature (T) and the number (n) of PCR cycles (optimized for each couple of primer): cGKI, 54°C for 40 cycles; cGKII, 52°C for 40 cycles; and GAPDH, 56°C for 35 cycles (Table I).
The PCR contained 0.4 mol/liter of each primer, dNTP (0.2 mmol/ liter), MgCl 2 (1.5mmol/liter) reaction buffer, and 5 units of Taq polymerase (Promega) in a final volume of 50 l. The amplified cDNAs were size-fractioned by 1.5% agarose gel electrophoresis and visualized under UV with a 1% ethidium bromide staining. The gels were scanned and analyzed using the National Institutes of Health Image program. Individual data were calculated as means of four independent experiments and expressed as the ratios of cGKI or cGKII over GAPDH expression.
Real Time Quantitative PCR Assay-RNA (2 g) was reverse transcribed using the ThermoScript RT system (Invitrogen) according to the manufacturer's instructions (Random Hexamers and dNTPs were also supplied by Invitrogen). Reverse transcription was also performed without Thermoscript enzyme on RNA samples to provide a control for contamination of samples with genomic DNA. PCR primers were designed using Primer Express software according to the recommendations of Applied Biosystems. Optimum primer concentrations were determined by performing PCR with a range of primer concentrations and comparing the rate of product accumulation (Table II).
Amplification reaction was performed in duplicate on 20 ng of the cDNA sample (5 l) in a final volume of 26 l in 96-well Optical reaction plates (Applied Biosystems) in a GeneAmp 5700 sequence detection system. For GC-A, GC-B, and NPr-C, the PCR mixture contained 900 nmol/liter forward and reverse primer mix (8 l) and 13 l of SYBR Green PCR Master Mix (Applied Biosystems) containing a fluorescence dye, SYBR Green, which upon binding to double-stranded DNA exhibits fluorescence enhancement. The enhancement of fluorescence is proportional to the initial concentration of the cDNA. For ribosomal RNA control (18 S rRNA), PCR mixture contained 8 l of primers and fluorogenic probe mix (Applied Biosystems) and 13 l of TaqMan Universal PCR Master Mix (Applied Biosystems). All of the reactions were performed under the same conditions: 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. The results were analyzed with the GeneAmp 5700 software, and all of the values were normalized to levels of the ribosomal RNA control (18 S rRNA).
Immunoprecipitation-Human mature fat cells or preadipocytes were lysed at 4°C in lysis buffer. The lysates were centrifuged (10 min, 10,000 ϫ g, 4°C), and a preclearing step was performed on supernatants by incubation for 1 h with protein G-Sepharose at 4°C under constant shaking. After a brief centrifugation, the precleared extracts were incubated for 2 h at 4°C under constant shaking with a polyclonal rabbit anti-phosphoserine antibody or with the chicken polyclonal antibody against human HSL, and the immune complexes were recovered by 1 h of incubation with protein G-Sepharose at 4°C under shaking and analyzed by Western blot.
Western Blot Analysis-40 g of cell lysate protein were separated by SDS-PAGE under denaturing conditions and transferred to nitrocellulose membranes, and Ponceau staining was performed to verify equal loading of the lanes. The membranes were blocked overnight at 4°C with a blocking buffer (50 mmol/liter Tris-HCl (pH 7.5), 200 mmol/liter NaCl, 0.05% polyethylene-sorbitan monolaurate (Tween), 3% bovine serum albumin, and 10% horse serum), incubated with a primary antibody for 90 min, rinsed, blocked for 40 min, and incubated with secondary antibody for 60 min. The immunocomplexes were visualized using the chemiluminescence reagent kit. The autoradiographs were scanned by an imaging densitometer and quantified using the National Institutes of Health Image program. Drugs and Chemicals-(Ϫ)-Isoproterenol hydrochloride (nonselective ␤-adrenergic receptor agonist), isobutylmethylxanthine (nonselective phosphodiesterease inhibitor), insulin, and bovine serum albumin (fraction V) were from Sigma-Aldrich. Crude collagenase, enzymes for glycerol assays, and tablets of protease inhibitors came from Roche Applied Science. Human-ANP (1-28) was from Neosystem Laboratories (Strasbourg, France). The cell culture reagents came from either Invitrogen or Roche Applied Science. The ERK, phospho-ERK, p38, and phospho-P38 antibodies were from Cell Signaling Technology (New England Biolabs). The PKA activity reagents came from Calbiochem (VWR, Fontenay sous Bois, France). The polyclonal rabbit antibody against phosphoserine came from Zymed Laboratories (Clinisciences, Montrouge, France). The polyclonal chicken antibody against the HSL was a gift from Dr. Cecilia Holm (Department of Cell and Molecular Biology, Lund University, Lund, Sweden). The polyclonal guinea pig antibody against perilipin came from ProgenBiotechnik (Heidelberg, Germany). The peroxidase-conjugated antibodies were purchased from Chemicon (Euromedex, Souffelweyersheim, France), and the ECL kit was from Amersham Biosciences. The prestained protein markers and the protein determination kit were obtained from Bio-Rad. RNA concentration was determined using Ribogreen (Molecular Probes, Leiden, The Netherlands). RT-PCR reagents and protease inhibitor tablets (Complete Mini) were from Roche Applied Science.
Data Analysis-The values are given as the means Ϯ S.E. of n separate experiments. Differences were considered significant when p Ͻ 0.05. The concentration response curves were fitted using nonlinear regression, and the EC 50 values (half-maximal effective drug concentration) were calculated using the program Prism (GraphPad Software, San Diego, CA).

HSL Phosphorylation after ANP Exposure on Human
Adipocyte-Because the serine phosphorylation of HSL is the limiting step of HSL activation and lipolysis, we first studied the effects of treatments for increasing times (5, 10, and 30 min, respectively) with 10 nmol/liter ANP or 100 nmol/liter isoproterenol (␤-adrenergic receptor agonist) on the serine phosphorylation rate of HSL in human mature adipocytes. The selected concentrations are known to promote stimulation of lipolysis to a similar extent. As shown in Fig. 1A, Western blot analysis using an antibody directed against human HSL, performed on immunoprecipitates obtained with an antibody directed against phosphoserine residues, evidenced one band, the molecular mass of which corresponded to HSL (88 kDa). ANP, as well as isoproterenol, induced a time-dependent phosphorylation of HSL on serine residues (Fig. 1A). The kinetics of HSL phosphorylation observed for the both lipolytic agents exhibited statistically significant differences. Although exposure to isoproterenol was associated with a transient phosphorylated state of HSL, ANP-induced HSL phosphorylation, which was maximal after 10 min of incubation (3-fold increase compared with control), remained elevated for the following 20 min (p Ͻ 0.05, n ϭ 6) (Fig. 1B). Moreover, the same experiments performed on rat adipocytes, known as a nonresponsive species regard-ing ANP-mediated lipolysis (2), showed that ANP was unable to promote HSL phosphorylation (data not shown).
Role of cGMP in ANP-induced Lipolysis on Human Fat Cells-To further characterize the intracellular pathway involved in the ANP-mediated HSL phosphorylation, human fat cells were treated with increasing concentrations (1-100 mol/liter) of the GC inhibitor, LY-83583. As shown in Fig A, human mature adipocytes were exposed to 10 nmol/liter ANP or 100 nmol/liter isoproterenol (ISO) for the time indicated, and the HSL phosphorylation profile was analyzed by immunoprecipitation experiments using an antibody directed against phosphoserine followed by a Western blot analysis using an antibody directed against human HSL. A representative blot of six independent experiments is shown. IP, immunoprecipitation; IB, immunoblot. B, densitometric analysis of the blots were performed. The results are expressed as fold increases of basal HSL phosphorylation and are the means Ϯ S.E. from six independent experiments. *, p Ͻ 0.05 ANP (q) versus isoproterenol (E).

Role of PKA in the ANP-induced Lipolysis on Human
Fat Cells-To rule out possible cross-talk between the cGMP-and the cAMP-dependent pathways, we measured the effect of ANP on PKA activity in human mature adipocytes. As depicted in Fig. 3A, ANP, even at high concentration (1 mol/liter), did not affect the activity of PKA in human mature fat cells. However, 0.1 mmol/liter forskolin, a direct activator of adenylyl cyclase, led to a 7.5-fold increase of the PKA activity compared with untreated cells (p Ͻ 0.001, n ϭ 6). In addition, pretreatments of the human mature adipocytes for 30 min with increasing concentrations (0.5-10 mol/liter) of the PKA inhibitor H89 did not affect the lipolysis stimulated by 10 nmol/liter ANP, whereas it inhibited in a concentration-dependent manner the lipolysis induced by isoproterenol (from 0.95 Ϯ 0.02 mol glycerol/100 mg lipid (isoproterenol alone) to 0.74 Ϯ 0.03 mol glycerol/100 mg lipid (isoproterenol plus H-89), p Ͻ 0.001, n ϭ 6).
Subtype of cGK Expressed in Human Adipocytes-The most likely intracellular target of cGMP is the cGK. Because the presence of cGK has never been reported in adipocytes, we first characterized the expression of cGK in mature human fat cells. As shown in Fig. 4, the PCR product of 469 bp corresponding to cGKI was found in human mature adipocytes as well as in mouse heart and kidney (positive control); however, the PCR product of 726 bp corresponding to cGKII was detected in Inset, human mature isolated fat cells were exposed to 10 nmol/liter human ANP in the presence or in the absence of 10 mol/liter LY-83583, and cGMP concentrations were determined. The data represent the means Ϯ S.E. from six independent experiments, ***, p Ͻ 0.001 ANP versus basal; ##, p Ͻ 0.001 ANP versus ANP plus LY-83583. B, human mature isolated fat cells were exposed to 10 nmol/liter human ANP in the presence or absence of 10 mol/liter LY-83583. HSL phosphorylation profile was analyzed by immunoprecipitation using an antibody directed against phosphoserine followed by a Western blot analysis using an antibody directed against human HSL. The results are expressed as fold increases of basal HSL phosphorylation and are the means Ϯ S.E. from six independent experiments. *, p Ͻ 0.05 ANP versus basal HSL phosphorylation; ##, p Ͻ 0.01 ANP versus ANP plus LY-83583. IP, immunoprecipitation; IB, immunoblot.
cGMP/cGKI Involvement in NP-mediated HSL Phosphorylation mouse heart and kidney but was not detected in human fat cells.
Ontogenesis of the ANP-induced Lipolysis System in Primary Cultures of Human Preadipocytes-To precisely determine the contribution of cGKI in the ANP-mediated transduction pathway, experiments were further performed in an adipocyte model containing less lipid content, i.e. primary cultures of human preadipocytes differentiated in adipocytes. The presence of a large lipid vacuole in mature adipocytes limits the use of lipophilic pharmacological tools. Because the ontogenesis of the NP system was never described, the expression of the various components of the ANP transduction pathway, i.e. NP receptors, cGKI, and HSL, was studied during the adipocyte differentiation process (days 0, 5, 10, and 15). The amounts of NP receptor mRNAs, GC-A, GC-B, and NPr-C, were determined using real time RT-PCR analysis. As depicted in Fig. 5A, GC-A mRNAs levels were increased during the adipocyte differentiation with a maximal expression observed after 5 days (3-fold increase in mRNA levels in day 5 compared with day 0). GC-B or NPr-C mRNA levels were not statistically modified. The expression of cGKI, evidenced by Western blot analysis (Fig. 5B), decreased with differentiation (a decrease of 37 Ϯ 7% in the amount of cGKI protein was observed in day 15 versus day 0 (p Ͻ 0.05, n ϭ 5)); however, the protein level of HSL, assessed via Western blot analysis (Fig. 5C), progressively increased (15 Ϯ 2-fold increase in HSL protein amount in day 15 versus day 0). Finally, lipolysis challenges were performed during the differentiation process with 10 nmol/liter ANP (Fig.  5D). After 5 days, ANP significantly increased the release of glycerol in the extracellular medium (3-fold increase compared with basal value). The extent of the lipolytic response induced by ANP increased during the differentiation process (9-and 5-fold increases compared with basal value after 10 and 15 days, respectively). In parallel we performed lipolysis challenges in the presence of 10 mol/liter thiorphan, a neutral endopeptidase inhibitor because NPs are known to be degraded by neutral endopeptidase. In that context, no significant improvement of the NP-induced lipolysis was found (data not shown).

ANP Induces the Serine Phosphorylation of HSL Together with Perilipin A in Differentiated Human
Preadipocytes-HSL and perilipin are described to be concomitantly phosphorylated in response to lipolytic stimulation in intact adipocytes. Then we examined the effect of 10 nmol/liter ANP on the phosphorylation of HSL and perilipin in 10-day differentiated human preadipocytes. 10-day differentiated human preadipocytes were chosen because of their low lipid content but high cGKI and HSL protein levels and maximal ANPinduced lipolysis.
As shown in Fig. 6A, Western blot analysis using the antihuman HSL antibody, performed on immunoprecipitates obtained with the anti-phosphoserine antibody on protein extracts from cells treated with ANP for increasing amounts of time (0, 10, and 30 min, respectively), showed that ANP induced a time-dependent phosphorylation of HSL on serine residues with a time course similar to that described in mature human adipocytes (n ϭ 5). The ANP-mediated serine phosphorylation of HSL in human preadipocytes was confirmed by Western blot analysis using the anti-phosphoserine antibody performed on immunoprecipitates obtained with the antibody against the human HSL on protein extracts from cells treated for 30 min with ANP (Fig. 6B). In parallel, we show that a 10-min incubation period of human preadipocytes with 100 nmol/liter isoproterenol also led to increased serine phosphorylation of HSL.
Our preliminary experiments have shown that perilipin A is the main perilipin isoform expressed in differentiated human preadipocytes. To determine whether ANP could stimulate the phosphorylation of perilipin A, Western blot analysis was performed using an antibody against perilipin A on protein extracts from 10-day differentiated human preadipocytes, which were either treated or not treated for 30 min with 10 nmol/liter ANP or for 10 min with 100 nmol/liter isoproterenol. As depicted in Fig. 6C, ANP treatment induced a marked shift in the migration of perilipin A, similar to that observed under stimulation with isoproterenol (n ϭ 6).

Involvement of cGK and Not of MAP Kinase in the Control of ANP-induced Lipolysis and HSL Phosphorylation-To exam-
ine the involvement of cGKI in the control of ANP-induced lipolysis and HSL phosphorylation, 10-day differentiated human preadipocytes were preincubated for 40-min periods with increasing concentrations (25 and 50 mol/liter) of the cGK inhibitor Rp-8-pCPT-cGMPS and then stimulated with 10 nmol/liter ANP. As depicted in Fig. 7A, the pretreatment of the cells with Rp-8-pCPT-cGMPS led to a statistically significant decrease in the ANP-induced lipolysis (maximal decrease of 49 Ϯ 11% of ANP-induced lipolysis in the presence of 50 mol/ liter Rp-8-pCPT-cGMPS compared with ANP alone). In parallel, Western blot analysis using the anti-human HSL antibody, performed on immunoprecipitates obtained with the antiphosphoserine antibody, showed that Rp-8-pCPT-cGMPS pretreatment significantly diminished the ANP-induced HSL phosphorylation (Fig. 7B) (maximal decrease of 52 Ϯ 5% of ANP-induced HSL phosphorylation in the presence of 50 mol/ liter Rp-8-pCPT-cGMPS compared with ANP alone).
To determine whether other serine/threonine kinases might be involved in the ANP-induced lipolysis, we examined the effect of ANP on the activation of the MAP kinases (ERK and p38). 10-day differentiated human preadipocytes were exposed to 10 nmol/liter ANP or to 100 nmol/liter isoproterenol for increasing periods of time (0, 2, 5, 10, and 30 min, respectively), and the activation of ERK and p38 MAP kinase were assessed by Western blot analysis performed using specific antibodies directed against the phosphorylated forms of ERK and p38. In parallel, Western blot analysis using anti-ERK or anti-p38 MAP kinase antibodies was performed on the same protein extracts. Irrespective of the time period of treatment, ANP or isoproterenol modified neither ERK nor p38 MAP kinase phosphorylation (n ϭ 3, data not shown). Moreover, the preincubation of 10-day differentiated human preadipocytes for 90 min either SB203580 (0.2 and 2 mol/liter), inhibitor of p38 MAP kinase or U0126 (5 mol/liter) inhibitor of ERK, did not affect the HSL phosphorylation induced by 10 nmol/liter ANP for 30 min or 100 nmol/liter isoproterenol for 10 min, as assessed by Western blot analysis using anti-human HSL antibody performed on anti-phosphoserine immunoprecipitates (n ϭ 3, data not shown). DISCUSSION We have already shown that NPs, ANP and brain NP but not C-type NP, are strong activators of lipolysis in human fat cells (1, 2). The present study was undertaken to further elucidate the mechanisms underlying ANP-induced lipolysis in human adipocytes. We demonstrate that ANP-induced lipolysis is mediated by cGMP with the activation of cGKI, identified for the first time in human fat cells, which leads to the phosphorylation of HSL and perilipin A. Moreover, we describe ontogenesis of the ANP-dependent transduction pathway during the differentiation of human preadipocytes into adipocytes.
In humans, the hormonal control of lipolysis has always been related to insulin (inhibition of lipolysis) and catecholamines

cGMP/cGKI Involvement in NP-mediated HSL Phosphorylation
(stimulation of lipolysis) (21,22). It is generally accepted that the modulation of intracellular cAMP levels (increased under ␤-adrenergic receptor stimulation or decreased by insulin) is the crucial point in the control of lipolysis because it determines the activation of PKA, which stimulates HSL activity by phosphorylation of HSL on serine residues. HSL catalyzes the first and rate-limiting step of hydrolysis of stored triglycerides and is thereby a key enzyme in the mobilization of free fatty acids from adipose tissue (23). The hallmark of HSL, which distinguishes this enzyme from all other known triacylglycerol lipases, is the control of its activity through phosphorylation (24). Thus, the ultimate rate of lipolysis in adipose tissue is dependent on the phosphorylation state of HSL (23). We demonstrated that the ANP-induced lipolysis is associated with an increase in the serine phosphorylation of HSL in human mature adipocytes but not in rat mature adipocytes, which cells are known to be from a nonresponsive species (2), thus confirming the necessity of an HSL phosphorylation to induce lipolysis and indicating the involvement of a serine protein kinase. Moreover, the different kinetic profiles of HSL phosphorylation induced by ANP and by the ␤-adrenergic receptor agonist isoproterenol strongly suggested that the ANP-induced lipolysis was indeed mediated by a different pathway than the classical well known sequence, cAMP/PKA/HSL. To further analyze the components of the ANP-stimulated transduction pathway, experiments were performed with the GC inhibitor LY-83583.
LY-83583 strongly inhibited the ANP-induced accumulation of cGMP, HSL phosphorylation, and lipolysis without a modification of isoproterenol-mediated lipolysis. This result clearly demonstrates that the signal transduction pathway stimulated by ANP to promote lipolysis in human adipocytes is strictly connected to an increase in intracellular cGMP concentrations.
It is now recognized that there are four major classes of cGMP-regulated proteins: cGKs, cGMP-stimulated/cGMP-inhibited PDEs, cGMP-gated cation channels, and PKA (11,12). Our previous study established that neither cAMP nor the PDE-3B (a cGMP-inhibited PDE), were involved in the NPinduced lipolysis. Because PKA could be activated by cGMP (25), we examined whether ANP could modify PKA activity in human adipocytes and found that ANP did not stimulate the activity of PKA. Moreover, the inhibition of PKA by H-89 did not affect the ANP-induced lipolysis, but it did decrease the isoproterenol-stimulated lipolysis. Taken together, both results demonstrate that ANP-mediated lipolysis did not involve crosstalk between cGMP and PKA. NPs are known to modulate some cGMP-gated channels like the L-type calcium channels, inducing a Ca 2ϩ increase in cardiac muscles and vascular smooth muscle cells (26 -28). Previous observations point to a putative positive regulatory role of calcium in adipocyte lipolytic cascade (29 -31). However, recent reports have brought some contradictory results showing the association of the increase in intracellular Ca 2ϩ concentrations with inhibition of FIG. 6. ANP and isoproterenol-induced HSL and perilipin A phosphorylation in 10-day differentiated human preadipocytes. A, 10-day differentiated human preadipocytes were exposed to 10 nmol/liter ANP, and HSL phosphorylation profile was analyzed by immunoprecipitation using an antibody directed against phosphoserine followed by a Western blot analysis using an antibody directed against human HSL. A representative blot of five independent experiments for ANP is shown. B, 10-day differentiated human preadipocytes were exposed to 10 nmol/liter ANP or 100 nmol/ liter isoproterenol (ISO). Western blot analysis using the anti-phosphoserine antibody were performed on immunoprecipitates obtained with the antibody against the human HSL on protein extracts from cells treated for 30 min with ANP and 10 min with isoproterenol. A representative blot of three independent experiments is shown. The blots were subsequently stripped and probed with antibodies directed against human HSL to assess the level of HSL in each lane. A representative blot of three independent experiments is presented. C, 10-day differentiated human preadipocytes were exposed for 30 min to 10 nmol/liter ANP or for 10 min to 100 nmol/liter isoproterenol (ISO), and homogenates were immunoblotted with anti-perilipin. The similar characteristic upward shift observed with ANP and ISO reveals the phosphorylation of perilipin A in stimulated cells (six and two independent experiments for ANP and ISO, respectively). IP, immunoprecipitation; IB, immunoblot.
lipolysis (32,33). Therefore, the involvement of Ca 2ϩ in the NP-induced lipolysis may be ruled out. Finally, the potential involvement of a cGK, the most relevant target of cGMP, was considered. Two main forms of cGKs, encoded by distinct genes (34 -36), have been identified and defined as type I and type II cGK. Type I is a cytosolic homodimeric protein, whereas the type II is a membrane-bound homodimeric protein. Smooth muscle cells, platelets, and cerebellum contain high concentrations of cGKI, whereas cGKII is highly concentrated in brain, lung, kidney, bone, and intestinal mucosa (12,37). Although early observations pointed to the presence of a cGMP-binding protein at very low levels in rat adipose tissue (38), to our knowledge no information concerning the cGK isoform and its physiological role have ever been provided. The present study identified mRNAs of cGKI in human adipocyte but not of cGKII. To evaluate the contribution of cGKI in the NP-induced lipolysis, we used several specific cGK inhibitors such as KT5823, H8, or Rp-8-pCPT-cGMPS. Although these agents are known to inhibit the cGK-induced activation in many cell-free systems, we did not observe any effect on the NP-induced lipolysis in mature human adipocytes. The reason for the failure of such tools in adipose tissue may arise from the following problems: (i) difficulty of getting access to cGK in intact fat cells because of compartmentalization of the drug; (ii) requirement of longer incubation times (in comparison with broken cell systems) that are necessary to enable diffusion/transport of the drug through the cellular membrane, which is impossible for mature fat cells; and (iii) hydrophobicity of the compounds such that they are immediately trapped into the large lipid vacuole. This kind of problem has also been reported by groups working FIG. 7. Effect of the inhibition of cGK on the ANP-induced lipolysis and HSL phosphorylation in human preadipocytes. A, 10-day differentiated human preadipocytes were preincubated for 40 min with increasing concentrations of Rp-8-pCPT-cGMP (25 and 50 mol/liter) and stimulated with 10 nmol/liter ANP. The results are the means Ϯ S.E. from four separate experiments. ***, p Ͻ 0.001 versus ANP alone. B, 10-day differentiated preadipocytes were pre-exposed to increasing concentrations of Rp-8-pCPT-cGMP (25 and 50 mol/liter) and then stimulated with 10 nmol/liter ANP. The HSL phosphorylation profile was analyzed by immunoprecipitation using an antibody directed against phosphoserine followed by a Western blot analysis using an antibody directed against human HSL. A representative blot is shown. Densitometric analysis of the blots were performed. The results are expressed as percentages of HSL phosphorylation and are expressed as the means Ϯ S.E. from six independent experiments. **, p Ͻ 0.01 versus ANP alone. IP, immunoprecipitation; IB, immunoblot. on cGKs (12,39). Please note that the quantitative analysis of cGK activation in intact cells is difficult because of the relative low expression of cGKs in most cell types, especially when compared with expression of other protein kinase such as PKA (12). Therefore, we shifted toward another adipocyte model and used the adipocytes that originate from primary culture of human preadipocytes and accumulate less lipid. Most preadipocyte cell lines currently available originate from mice and are not relevant for purposes of this study because we have demonstrated that the NP-induced lipolysis is unique to the primate fat cell (2). The ontogenesis of the different components involved in the transduction pathway of the NP-induced lipolysis was first investigated in the human preadipocytes during the differentiation process. As for mature adipocytes, human preadipocytes expressed the mRNAs of the three NPs receptor subtypes (GC-A, GC-B, and NPr-C). However, we found only GC-A to be up-regulated during the differentiation process. The cGKI protein was found to be expressed in nondifferentiated human preadipocytes, but its expression was slightly decreased with differentiation. However, prolonged cell culture is often associated with a down-regulation or a loss of cGK expression in many cell types (12). HSL expression was increased during the differentiation process, confirming its expression as a late marker of adipocyte differentiation (23). Finally, as for the mature adipocyte, ANP was able to trigger a lipolytic effect that was markedly enhanced all along the differentiation process. The ANP-induced lipolysis was associated with an increase in the serine phosphorylation of HSL assessed by Western blot analysis using either HSL or phosphoserine antibodies on immunoprecipitates obtained with phosphoserine or HSL antibodies, respectively. The hydrolysis of triglycerides is the result of an HSL activation by phosphorylation on serine residues followed by a translocation of HSL from the cytosol to the lipid droplets. The binding of phospho-HSL to the lipid droplet is only allowed when the perilipins (40), which are proteins that coat the surface of intracellular lipid droplets, are phosphorylated (41). Three perilipin isoforms (perilipins A, B, and C) have been described, and perilipin A is the most abundant isoform in adipose tissue (42). In differentiated human preadipocytes, we show that ANP-dependent lipolysis was associated with an increase perilipin A phosphorylation. Indeed, isoproterenol as previously described (40) and ANP induced a similar shift in perilipin A migration. These data confirm that the activation of lipolysis in adipocytes is the result of concerted cross-talk between phospho-HSL and phospho-perilipin. Finally, we demonstrate the involvement of cGKI in the ANP-induced lipolysis, using the inhibitor Rp-8-pCPT-cGMPS. Indeed, Rp-8-pCPT-cGMPS dramatically diminished the ANP-induced lipolysis and decreased the ANP-mediated phosphorylation of HSL. This result is consistent and provides a biologically relevant role to early data demonstrating, on a cell-free system, the phosphorylation and activation of HSL by cGK (43,44). Further studies should be performed to identify which phosphorylated serine residues could be involved in the HSL and perilipin phosphorylation during ANP stimulation of human adipocytes. This point was considered to be out of the scope of the present study.
To rule out the potential involvement of other serine/threonine kinases in the ANP-dependent pathway, we examined the effect of ANP on the activation of MAP kinases. Indeed, recent studies have shown a role for the MAP kinases in the control of the lipolysis induced by tumor necrosis factor ␣ in human adipocytes (45,46). Our results show that both ERK and p38 MAP kinase were not involved in the ANP-mediated HSL phosphorylation because ANP did not modulate the phosphorylation of ERK and p38 MAP kinase, and the MAP kinase inhib-itors did not affect the ANP-induced HSL phosphorylation.
Until now it has been generally accepted that lipolysis was mainly controlled by the activity of the sympathetic nervous system and by plasma insulin levels (22,47). We propose in the present work a new additional mechanism involving a peptide system, the NPs, able to activate lipolysis, to an extent similar to that induced by catecholamines, via the increase of intracellular cGMP, activation of cGKI, and phosphorylation of HSL and of perilipin A. Importantly, we observed that the lipolytic effect of NPs is independent of the control by insulin (1). Moreover, the NP-induced lipolysis has a primate fat cell specificity, and it is not depot-, age-, or gender-dependent (2,13,14). The present findings demonstrating a new control of HSL activity and lipolysis by natriuretic peptides, cGMP and especially cGKI, in human fat cell, raises a number of questions about the physiological role of this lipolytic pathway in comparison with the ␤-adrenergic control of human adipose tissue metabolism. A putative involvement in the development and in the pathogenesis of obesity cannot be excluded. Moreover, because elevated circulating NPs levels are found in several pathological states and are often associated to disease severity, this new pathway could play a decisive role in cachexia (48) and wasting diseases leading to adipose tissue regression.