- Email: [email protected]
Nitric Oxide–Induced F-Actin Disassembly Is Mediated via cGMP, cAMP, and Protein Kinase A Activation in Rat Mesangial Cells
Experimental Cell Research 271, 329 –336 (2001) doi:10.1006/excr.2001.5378, available online at http://www.idealibrary.com on
Nitric Oxide–Induced F-Actin Disassembly Is Mediated via cGMP, cAMP, and Protein Kinase A Activation in Rat Mesangial Cells Katrin B. Sandau,* Florian Gantner,† and Bernhard Bru¨ne* ,1 *University of Kaiserslautern, Faculty of Biology, Department of Cell Biology, 67663 Kaiserslautern, Germany; †Byk Gulden, Department of Biochemistry, 78467 Konstanz, Germany
Glomerular mesangial cells contain actin and myosin, and in analogy to vascular smooth muscle cells, they can contract and relax to regulate the glomerular filtration rate. A key molecule that determines hemodynamic properties is nitric oxide, which is produced by nitric oxide synthase isoenzymes located in individual cells of the kidney. The contractility of mesangial cells is based on the interaction of actin microfilament bundles (F-actin) with myosin. We had the notion that nitric oxide influences the shape change of mesangial cells, so we analyzed the signal transduction involved. Chemically unrelated nitric oxide donors induced Factin dissolution, which was mediated by cGMP but was unrelated to protein kinase G activation. Actin disassembly was achieved with inhibitors of phosphodiesterase-3 and -4 or forskolin-evoked cAMP generation. We assumed that signal transmission involves activation of protein kinase A, and we went on to attenuate F-actin disassembly by protein kinase A inhibition. In conclusion, we found evidence that nitric oxide triggered F-actin dissolution via cGMP generation, inhibition of cAMP-hydrolyzing phosphodiesterase-3, and subsequent protein kinase A activation. © 2001 Elsevier Science
Key Words: Nitric oxide; cGMP; phosphodiesterases; protein kinase A; cytoskeleton; F-actin.
INTRODUCTION
The radical nitric oxide (NO) is produced by a family of NO synthase isoenzymes (NOS) which utilize L-arginine and oxygen to produce citrulline and NO [1]. Constitutive and inducible NOS isoforms can be distinguished. The constitutive NOS (cNOS) is regulated by Ca2⫹ and transiently generates NO in small concentrations, whereas the inducible form (iNOS) is activated by diverse agonists such as cytokines or lipopolysaccharide and releases high amounts of NO over an extended period of time. Once NO 1 To whom correspondence and reprint requests should be addressed at University of Kaiserslautern, Faculty of Biology, ErwinSchro¨dinger-Strasse, 67663 Kaiserslautern, Germany. E-mail: [email protected].
is produced, it may trigger physiological as well as pathophysiological mechanisms. In general, physiological processes are evoked by cGMP and protein kinase G (PKG) as NO binds and activates the heme subunit of soluble guanylyl cyclase (sGC) [2]. Pathophysiological functions of NO often provoke apoptosis or necrosis, but the mechanisms are more diverse, and no specific, cell-type–independent target has been identified so far. The ambivalent function of NO is demonstrated by its action in the kidney. During kidney-related diseases, such as diabetes or proliferative glomerulonephritis, increased NO production contributes to pathogenic hemodynamic changes and cell death of glomerular mesangial cells (MC) [3, 4], whereas chronic NO insufficiency causes hypertension [5]. These hemodynamic alterations are controlled by MC, which are specialized smooth muscle cells in the glomerulus of the kidney. MC contain the ability to contract and relax, thereby regulating the glomerular filtration rate and renal blood flow [5]. Under physiological conditions, these parameters are controlled by NO, generated by the endothelial NOS which is found in close proximity to MC. In the cardiovascular system, the action of endothelium-derived NO, which diffuses to smooth muscle cells and activates sGC, is well established [6]. However, it is controversially discussed how a cGMP signal leads to vasorelaxation and how NO affects mesangial contractility. Therefore, we were interested to analyze the signal transduction involved. To study the function of NO, we used different NO donors such as S-nitrosoglutathione (GSNO) or spermine-NO. These NO donors induced F-actin disassembly within 30 min, and by analyzing signal transduction, we established that the effect was mediated by cGMP production but was independent of PKG activation as described for other circumstances [7]. In contrast, we found protein kinase A activation, and as it is known that cGMP attenuates phosphodiesterase-3 (PDE), a cAMP and cGMP hydrolyzing enzyme, we used PDE inhibitors to mimic Factin dissolution. As F-actin dissolution was attenuated by PKA inhibition, we conclude that NO triggers F-actin disassembly in MC by cGMP production, inhi-
329
0014-4827/01 $35.00 © 2001 Elsevier Science All rights reserved.
330
¨ NE SANDAU, GANTNER, AND BRU
bition of cAMP-hydrolyzing PDEs, and finally executed by PKA activation. MATERIALS AND METHODS
FIG. 1. NO donor–induced F-actin disassembly. Rat mesangial cells (10 5 cells/well) were stimulated with vehicle (A), 250 M GSNO (B), or 250 M spermine-NO (C) for 30 min. After stimulation, cells were fixed, and F-actin was stained with phalloidinTRITC as outlined under Materials and Methods.
Reagents. Insulin, IBMX, erythro-9-(2-hydroxy-3-nonyl)-adenine, forskolin, phalloidin-TRITC, 8-bromo-cAMP, sulfanilamide, adenosine triphosphate (ATP), and N-naphthylethylenediamine were purchased from Sigma (Deisenhofen, Germany). Pefabloc was purchased from Boehringer Mannheim (Mannheim, Germany); [ 3H]cAMP and [ 3H]cGMP, and [␥- 32P]ATP were purchased from Amersham (Braunschweig, Germany); spermine-NO was purchased from Research Biochemicals International (Cologne, Germany); NS 2028 was purchased from Alexis (Gru¨nberg, Germany); and 8-chloroadenosine-3⬘,5⬘-cyclic monophosphorothioate Rp-isomer and Rp-8pCPT-cGMPS were purchased from BioLog (Bremen, Germany). RPMI 1640 and medium supplements were purchased from Biochrom (Berlin, Germany). Fetal calf serum (FCS) and protein kinase A assay system were purchased from Life Technologies (Berlin, Germany). GSNO was synthesized as described previously [8]. Motapizone was kindly provided by Nattermann (Ko¨ln, Germany); RP 73401 and sildenafil were provided by the Chemistry Department of Byk Gulden, Konstanz, Germany. All other chemicals were of the highest grade of purity and were commercially available. Culture of MC. Rat MC were cultured, cloned, and characterized as described previously [9]. Cells were grown in RPMI 1640 medium supplemented with 10% FCS, penicillin (100 U/ml), streptomycin (100 g/ml), and bovine insulin (5 g/ml). One day before and during the experiments, controls and stimulated cells were kept in medium with 0.5% FCS. For the experiments, 10 to 25 passages of MC were used. F-actin staining. MC were cultured and stimulated in eight-well chamber slides for F-actin staining. After stimulation, medium was discarded and cells were washed with phosphate-buffered saline (PBS) before fixing the cells with 3% paraformaldehyde for 30 min at room temperature. After a second step of PBS washing, cells were treated with 0.2% Triton X-100/PBS for 5 min, followed by PBS washing and F-actin staining with 10 g/ml phalloidin-TRITC for 1 h at 37°C. Afterward, chamber slides were washed precisely with PBS, dried, and embedded in glycerol. Stained cells were visualized with a Leica fluorescence microscope, and photographs were taken with a Leica camera. All photographs are of similar magnification (40⫻). Phosphodiesterase activity assay. MC were harvested, resuspended in 200 l homogenization buffer (10 mM Tris–HCl, pH 8.2, 1 mM -mercaptoethanol, 1 mM MgCl 2, 1 mM EGTA, 5 M pepstatin A, 10 M leupeptin, 400 M pefablock, 10 M soybean trypsin inhibitor, and 2 mM benzamidine) and disrupted by sonication (Branson 250 sonifier, output control 20%, duty cycle 1). The complete disruption of the cells (⬎98%) was checked via trypan blue exclusion. The homogenate was spun at 1000 ⫻ g for 5 min to remove viable cells (1–2%) and nuclei. The supernatant was decanted and stored at ⫺80°C until further use. PDE activity was determined as described by Thompson and Appleman [10] with some modifications [11]. In brief, 5 l of the 1000 ⫻ g supernatant was assayed in a final volume of 200 l containing 60 mM Tris HCl, pH 7.4, 5 mM MgCl 2, 0.5 M cAMP or cGMP (28,000 cpm [ 3H]cAMP or [ 3H]cGMP) and were incubated in the presence or absence of activators or inhibitors for 30 min at 37°C. The reaction was terminated by the addition of 50 l 0.2 N HCl, and the assay mixture was left on ice for further 15 min. Crotalus atrox snake venom (0.5 mg/ml) was added for 15 min at 37°C, and afterward, the assay mixture was loaded onto QAE-Sephadex A-25 columns (1-ml bed volume) and eluted with 2 ml ammonium formiate (30 mM, pH 6.0). The radioactivity in the eluate was counted in a liquid scintillation counter (Beckman). PDE isoenzyme activity calculations were performed by using activators (5 M cGMP for
NITRIC OXIDE–INDUCED F-ACTIN DISASSEMBLY
331
FIG. 2. Inhibition of sGC-attenuated NO-initiated F-actin disassembly in contrast to protein kinase G inhibition. Mesangial cells (10 5 cells/well) were stimulated with vehicle (A), 5 M NS 2028 (B), 250 M GSNO (C), 5 M NS 2028/250 M GSNO (D), 1 mM CPT-cGMPS (E), or the combination of 1 mM CPT-cGMPS/250 M GSNO (F). The specific sGC inhibitor NS 2028 and the PKG inhibitor CPT-cGMPS were preincubated for 15 min. After a 30-min incubation period, cells were fixed, and F-actin was visualized by phalloidin-TRITC staining as outlined under Materials and Methods.
PDE-2; 1 mM Ca 2⫹ and 100 nM calmodulin for PDE-1) or PDE isoenzyme–selective inhibitors (100 M erythro-9-(2-hydroxy-3nonyl)-adenine for PDE-2, 1 M Motapizone for PDE-3, 1 M RP73401 for PDE-4, and 100 nM sildenafil for PDE-5). The residual cAMP hydrolyzing activity in the presence of 1 M Motapizone and 1 M RP 73401, respectively, was designed as PDE res and theoretically could be composed of the novel cAMP-specific isoforms PDE-7, PDE-8, and PDE-10, or some combination of these. PKA activity assay. cAMP-dependent protein kinase activity was determined by in vitro phosphorylation of Kemptide by using an assay system from Life Technologies. Briefly, cells were washed twice with PBS and resuspended in 150 l extraction buffer (5 mM
EDTA, 50 mM Tris pH 7.5, 1% NP-40, 1 mM Na 3VO 4, 1 g/ml leupeptin, and 1 mM PMSF) followed by a 15-s sonication (Branson 250 sonifier, output control 20%, duty cycle 1). Cell debris were kept on ice for 30 min and centrifuged (17,000 ⫻ g, 15 min), and the protein content in the supernatant was analyzed. Afterward, 10 g protein of each sample was mixed with 4⫻ substrate solution with a final concentration of 50 M Kemptide, 100 M ATP, 10 mM MgCl 2, 0.25 mg/ml bovine serum albumin, 12.5 mM Tris pH 7.5, and 1 Ci [␥- 32P]ATP, and incubated for 5 min at 30°C. Reactions were terminated by spotting the reaction mix onto P81 phosphocellulose discs. The P81 paper sheets were immersed two times in 1% (v/v) orthophosphoric acid and washed three to five times with distilled
332
¨ NE SANDAU, GANTNER, AND BRU
TABLE 1 PDE Profile of Rat Mesangial Cells PDE isoform
PDE activity (pmol/mg ⫻ min)
1 2 3 4 5 Res.
45 ⫾ 16 0 38 ⫾ 9 6⫾ 6 5⫾ 2 2⫾ 1
Note. Unstimulated mesangial cells (10 7 cells/assay) were harvested and prepared for PDE activity determination as described under Materials and Methods. Data are expressed as pmol/mg protein ⫻ min mean total activities (soluble plus particulate) ⫾ standard deviation from five independent preparations.
water. Incorporated radioactivity was determined by scintillation counting. We ensured that the assay was linear with respect to time. Values are expressed as the percentage of increase of PKA activity vs control. Nitrite determination. Nitrite, a stable NO oxidation product, was determined by using the Griess reaction. Cell-free culture supernatants were collected (200 l), adjusted to 4°C, and mixed with 20 l sulfanilamide (dissolved in 1.2 M HCl) and 20 l N-naphthylethylenediamine-dihydrochloride. After 5 min at room temperature, the absorbance was measured at 560 nm with a reference wavelength at 690 nm. Nitrite concentrations were calculated by using a NaNO 2 standard. Statistical analysis. Each experiment was performed at least three times, and representative illustrations are shown.
RESULTS
FIG. 3. Inhibition of phosphodiesterases leads to F-actin dissolution. Mesangial cells (105 cells/well) were incubated with vehicle (A), the unselective PDE inhibitor IBMX (500 M) (B), or the combination of the PDE-3 inhibitor Motapizone (1 M) and the PDE-4 inhibitor RP 73401 (1 M) (C). After 30 min, cells were fixed and F-actin was visualized by phalloidin-TRITC staining as outlined under Materials and Methods.
In the first set of experiments, we used two chemically unrelated NO donors to analyze the role and signal transduction of NO leading to F-actin disassembly in rat MC. We stimulated MC for 30 min with GSNO (Fig. 1B) or spermine-NO (Fig. 1C) in a concentration of 250 M (Fig. 1). For control reasons, we incubated MC with glutathione and spermine, decomposed end products of GSNO and spermine-NO, but no F-actin changes were detectable (data not shown). Therefore, NO donor–induced F-actin disassembly is based on the release of NO. Generally, F-actin resolution appeared 10 min after NO addition and was completely reversible after 4 h. Dose-response studies showed visible F-actin changes with 500 nM and more pronounced effects with 1 M GSNO (data not shown); the alterations were sharpest at 250 M. Next, we used the sGC inhibitor NS 2028 to block sGC activation and cGMP-dependent signaling. NS 2028 at a concentration of 5 M revealed no cytotoxic effect or F-actin changes (Fig. 2B). In combination with 250 M GSNO (Fig. 2D), NS 2028 attenuated F-actin disassembly induced by NO after a 30-min incubation time, when NS 2028 was preincubated for 15 min. F-actin staining of these cells resembled control cells.
NITRIC OXIDE–INDUCED F-ACTIN DISASSEMBLY
FIG. 4. Lipophilic cAMP analogs and forskolin evoke F-actin disrupture. Mesangial cells (105 cells/well) were treated for 30 min with vehicle (A), 1 mM 8-bromo-cAMP (B), or 10 M forskolin (C). Cells were fixed and stained for F-actin as described under Materials and Methods.
In summary, inhibition of sGC blocked NO-induced cytoskeleton changes, which is indicative of a cGMPmediated pathway. In several cases, cGMP-mediated
333
processes are triggered by PKG activation. To test the involvement of PKG, we employed 1 mM of the specific PKG inhibitor Rp-8-pCPT-cGMPS (CPT-cGMPS) [12] alone (Fig. 2E) or in combination with 250 M GSNO (Fig. 2F), but CPT-cGMPS failed to modulate F-actin disassembly induced by NO. Similar results were achieved with another PKG inhibitor KT-5823 (data not shown), supporting the conclusion that PKG activation is not involved. Cyclic nucleotide PDE isoforms are currently classified into 10 families. Two of them are regulated by cGMP (PDE-2, the so-called cGMP-stimulated PDE activity, and PDE-3, classified as cGMP-inhibited PDE family). Therefore, we used different PDE inhibitors alone or in combination to assess their effect on F-actin organization. First, we used the nonspecific PDE inhibitor IBMX at a concentration of 500 M. After 30 min of incubation, F-actin disrupture resembled NOtreated cells (Fig. 3B). In addition, we supplied more specific PDE inhibitors such as erythro-9-(hydroxy-3nonyl)-adenine for PDE-2, Motapizone for PDE-3, RP 73401 for PDE-4, and sildenafil for PDE-5. These inhibitors, added individually, displayed no effect on Factin (data not shown). However, the combination of Motapizone and RP 73401 induced stress fiber dissolution after 30 min (Fig. 3C), and alterations became reversible after 24 h (data not shown). In order to obtain information on PDE isoforms present in our system, we determined a PDE profile in MC. We measured the activity of PDE-1 to PDE-5 according to the protocol described under Materials and Methods (Table 1). We detected a high activity for PDE-1 and PDE-3, and some activity of PDE-4 and PDE-5. PDE-1, depending on the subtype or subtypes expressed, preferably degrades cGMP, whereas PDE-4 and to a lesser extent PDE-3 are specific for cAMP. Obviously, PDE-3 is the most prominent cAMP-hydrolyzing PDE in MC, and as it is defined as cGMP inhibitable, an endogenous cGMP increase as generated by NO donors would result in a cAMP increase. Next, 1 mM lipophilic 8-bromo-cAMP was applied or the adenylyl cyclase was stimulated by forskolin (10 M) to test whether cAMP leads to F-actin resolution. F-actin disassembly occurred after 30 min, similar to NO when 8-bromo-cAMP or forskolin were added (Fig. 4). This process was again fully reversible within 24 h (data not shown). In addition, we used NS 2028 in combination with Motapizone/RP 73401, forskolin, or 8-bromo-cAMP (data not shown). As NS 2028 did not interfere with F-actin disassembly in response to 8-bromo-cAMP, we conclude that cAMP action is downstream of sGC. We measured PKA activity after stimulating MC with 250 M GSNO, 1 mM 8-bromo-cAMP, which served as a positive control, or with 1 M Motapizone/1 M RP 73401. Reactions were terminated after 10 min.
334
¨ NE SANDAU, GANTNER, AND BRU
FIG. 5. F-actin disassembly is attenuated by PKA inhibition. Mesangial cells (10 5 cells/well) were stimulated for 30 min with vehicle (A), 1 mM Cl-AMPS (B), 1 M Motapizone/1 M RP 73401 (C), 1 mM Cl-AMPS/1 M Motapizone/1 M RP 73401 (D), 1 M GSNO (E), and 1 mM Cl-AMPS/1 M GSNO (F). Cl-AMPS was preincubated for 15 min. Cells were fixed and stained for F-actin as described under Materials and Methods.
As expected, 8-bromo-cAMP induced strong PKA activation (69 ⫾ 8% increase vs control) after 10 min. GSNO stimulation achieved 24 ⫾ 4% PKA activation vs control, which is similar to Motapizone/RP 73401 (24 ⫾ 2% activation vs control), measured after 30 min.
Apparently, the PKA activation can occur in the same time frame as actin disassembly. To examine the role of an active PKA for F-actin disassembly, we used a specific PKA type I inhibitor, 8-chloroadenosine-3⬘,5⬘-cyclic monophosphorothiotae
NITRIC OXIDE–INDUCED F-ACTIN DISASSEMBLY
TABLE 2 Increased PKA Activity after NO Stimulation
Stimulation
Time (min)
Increase of PKA activity vs control (%)
Control GSNO (250 M) 8-bromo-cAMP (1 mM) Motapizone (1 M) ⫹ RP 73401 (1 M)
10 10 10 30
0 24 ⫾ 4 69 ⫾ 8 24 ⫾ 2
Note. Mesangial cells (10 7 cells/assay) were stimulated with vehicle (control), 250 M GSNO, 1 mM 8-bromo-cAMP, or 1 M Motapizone in combination with 1 M RP 73401 for the times indicated. Afterward, PKA activity was determined as described under Materials and Methods. Data are expressed as the percentage increase of PKA substrate phosphorylation as a marker of PKA activity vs unstimulated controls. The experiments were performed at least five times and mean values ⫾ standard deviation are given.
Rp-isomer (Cl-AMPS) [13], which was effective as seen in Fig. 5. Cl-AMPS by itself did not change the F-actin structure (Fig. 5B). To attenuate F-actin changes induced by Motapizone/RP 73401 (Fig. 5C) or GSNO (Fig. 5E), we preincubated Cl-AMPS for 15 min. Cells stimulated with an agonist plus the PKA inhibitor displayed normal F-actin structure (Figs. 5D, 5F). Cytoskeleton changes achieved with lipophilic cAMP analogs, as shown in Fig. 4, were blocked with ClAMPS as well (data not shown). These results point to the involvement of PKA type I, leading to F-actin disassembly in MC. As a PKA inhibitor attenuates the effect of NO, PDE inhibitors, and cAMP, we predict that PKA activation appears downstream of cGMP and cAMP production. DISCUSSION
Herein, we have shown that different NO donors provoke F-actin dissolution in glomerular MC (Fig. 1). The process can be attenuated by inhibition of either sGC or type I PKA (Figs. 2, 5). In addition, inhibition of the predominant cAMP-hydrolyzing enzymes present in MC such as PDE-3 and PDE-4 (Fig. 3), activation of the adenylyl cyclase by forskolin, or addition of lipophilic cAMP analogs (Fig. 4) achieved F-actin disassembly. All stimuli induced a transient, fully reversible effect which was sensitive toward PKA inhibition (Fig. 5), while the sGC inhibitor NS 2028 was effective after NO stimulation only (Fig. 2). This allowed us to postulate the position of the cGMP increase as upstream of the PDE inhibition and PKA activation. In contrast to the findings of an earlier report [14] that described the presence of PDE-1, PDE-2, and PDE-4 in rat MC, but in line with the findings of Dousa [15], the MC in our study displayed high PDE-3 and PDE-1 activity and a lower but functionally important
335
PDE-4, whereas PDE-2 was completely lacking (Table 1). Inhibitors against PDE-1, PDE-2, or PDE-5 were ineffective in inducing F-actin changes, whereas selective inhibition of PDE-3 and PDE-4 with the combination Motapizone/RP 73401 achieved F-actin disassembly. Such a PDE-3/4 synergism is well established in a variety of functional immune cell responses [16] and has also been described in experimental settings where PDE-3 inhibition due to NO-mediated cGMP elevation was potentiated by a PDE-4 inhibitor [17]. In the context of renin secretion, a similar scenario for NO has been suggested by Kurtz et al. [18]. They measured renin secretion of isolated perfused rat kidneys and found increased amounts of renin after stimulation with the NO donor SNP after activation of adenylyl cyclase or supplementation of different PDE inhibitors. The NO effect was suppressed by sGC inhibition, while renin secretion was drastically enhanced by attenuating PDE-3. In this study, the authors did not analyze the involvement of PKA. In our case, Factin disassembly is mediated by PKA activation (Table 2), and as expected, Cl-AMPS attenuated stress fiber relaxation (Fig. 5). Therefore, we assumed a cAMP increase, but as we have not obtained cAMP data, we cannot ultimately exclude the possibility that cGMP activates PKA directly, a pathway suggested by Cornwell et al. [19] who analyzed the role of NO for smooth muscle proliferation. They reported an increase in cGMP as a result of NO addition, activation of PKA, and inhibition of smooth muscle cell growth. PKA inhibitors attenuated growth inhibition, but cAMP accumulation was not found. The effect of PKA on the cytoskeleton machinery (e.g., myosin light chain [MLC] kinase, MLC, actin, or other kinases or phosphatases) were not in the focus of this study, but a recent article analyzed the effect of dibutyryl cGMP on MLC phosphorylation in MC [20]. The dynamic phosphorylation and dephosphorylation cycle of MLC influences the assembly and disassembly of the actin microfilament network responsible for stress fiber cell contraction and relaxation. The authors report that dibutyryl cGMP attenuated MLC phosphorylation elicited by angiotensin II or phorbol ester. As this was blocked by the phosphatase inhibitor calyculin A, it was concluded that cGMP activates MLC phosphatase. However, data on MLC phosphatase activation were not presented, and in light of our results, we instead suggest that cGMP acts via PKA, a pathway which cannot be excluded on the basis of the study of Torrecillas et al. [20]. The importance of NO for physiological or pathophysiological actions in the glomerulus of the kidney is well established. MC itself contain a NO-generating system with the implication that NO may have autocrine as well as paracrine functions [21–23]. We have shown that exogenously (Fig. 1) as well as endog-
¨ NE SANDAU, GANTNER, AND BRU
336
enously generated NO influenced mesangial contraction and relaxation in vitro. It can be assumed that NO will act in a similar way in vivo, thus provoking hemodynamic changes in the glomerulus as described before. This now can be explained by NO formation and its signaling properties involving guanylyl cyclase activation, cGMP generation, PDE-3 and PDE-4 inhibition, and PKA activation. We thank C. Blechner and S. Liehner for excellent technical assistance and Drs. H. Tenor and A. Hatzelmann (Byk Gulden, Konstanz) for helpful discussions. This work was supported by the Deutsche Forschungsgemeinschaft (Br 999 and SFB 423, A5).
12.
Butt, E., Van Bemmelen, M. X. L., Fischer, L., Walter, U., Jastorff, B. (1990) Inhibition of cGMP-dependent protein kinase by Rp-fuanosine 3⬘, 5⬘-monophosphorothioates. FEBS Lett. 263, 47–50.
13.
Gjertsen, P. J., Mellgren, G., Otten, A., Maronde, E., Genieser, J. G., Jastroff, B., Vintermyr, O. K., McKnight, G. S., and Doskeland, S. O. (1995). Novel (Rp)-cAMPS analogs as tools for inhibition of cAMP-kinase in cell culture. Basal cAMP-kinase activity modulates interleukin-1 action. J. Biol. Chem. 270, 20599 –20607.
14.
Ahn, H. S., Foster, M., Arik, L., Bykow, G., and Foster, C. (1995). Cyclic nucleotide phosphodiesterase isoenzymes in rat mesangial cells. Eur. J. Pharmacol. 289, 49 –57.
15.
Dousa, T. P. (1999). Cyclic-3⬘,5⬘-nucleotide phosphodiesterase isozymes in cell biology and pathophysiology of the kidney. Kidney Intl. 55, 29 – 65.
16.
Schudt, C., Gantner, F., Tenor, H., and Hatzelmann, A. (1999). Therapeutic potential of selective PDE inhibitors in asthma. Pulm. Pharmacol. Ther. 12, 123–129, doi:10.10061pupt. 1999.0182.
17.
Komas, N., Lugnier, C., and Stoclet, J. C. (1991). Endotheliumdependent and independent relaxation of the rat aorta by caclic nucleotide phosphodiesterase inhibitors. Br. J. Pharmacol. 104, 495–503.
18.
Kurtz, A., Go¨tz, K. H., Hamann, M., and Wagner, C. (1998). Stimulation of renin secretion by nitric oxide is mediated by phosphodiesterase. Proc. Natl. Acad. Sci. USA 95, 4743– 4747.
19.
Cornwell, T. L., Arnold, W., Boerth, N. J., and Lincoln, T. M. (1994). Inhibition of smooth muscle cell growth by nitric oxide and activation of cAMP-dependent protein kinase by cGMP. Am. J. Physiol. 267, C1405–C1413.
20.
Torrecillas, G., Diez-Marques, M. L., Garcia-Escribano, C., Bosch, R. J., Rodriguez-Puyol, D., and Rodriguez-Puyol, M. (2000). Mechanisms of cGMP-dependent mesangial-cell relaxation: A role for myosin light-chain phosphatase activation. Biochem. J. 346, 217–222.
21.
Pfeilschifter, J., Rob, P., Mu¨lsch, A., Fandrey, J., Vosbeck, K., and Busse, R. (1992). Interleukin 1 beta and tumour necrosis factor alpha induce a macrophage-type of nitric oxide synthase in rat renal mesangial cells. Eur. J. Biochem. 203, 251–255.
22.
Shultz, P. J., Tayeh, M. A., Marletta, M. A., and Raij, L. (1991). Synthesis and action of nitric oxide in rat glomerular mesangial cells. Am. J. Physiol. 261, F600 –F606.
23.
Marsden, P. A., and Ballermann, B. J. (1990). Tumor necrosis factor alpha activates soluble guanylate cyclase in bovine glomerular mesangial cells via an L-arginine– dependent mechanism. J. Exp. Med. 172, 1843–1852.
REFERENCES 1. 2. 3.
4.
5. 6.
7.
8.
9.
10.
11.
Marletta, M. A. (1994). Nitric oxide synthase: Aspects concerning structure and catalysis. Cell 78, 927–930. Nathan, C. (1992). Nitric oxide as a secretory product of mammalian cells. FASEB J. 6, 3051–3064. Cattell, V., and Cook, T. (1995). The nitric oxide pathway in glomerulonephritis. Curr. Opin. Nephrol. Hypertens. 4, 359 – 364. Sandau, K., Pfeilschifter, J., and Bru¨ne, B. (1997). The balance between nitric oxide and superoxide determines apoptotic and necrotic death of rat mesangial cells. J. Immunol. 158, 4938 – 4946. Raij, L., and Baylis, C. (1995). Glomerular actions of nitric oxide. Kidney Intl. 48, 20 –32. Moncada, S., Palmer, R. M., and Higgs, E. A. (1991). Nitric oxide: Physiology, pathophysiology, and pharmacology. Pharmacol. Rev. 43, 109 –142. Ruth, P. (1999). Cyclic GMP-dependent protein kinases: Understanding in vivo functions by gene targeting. Pharmacol. Ther. 82, 355–372. Hart, T. W. (1997). Some observation concerning the S-nitroso and S-phenylsulfonyl derivatives of L-cysteine and glutathione. Tetrahedron Lett. 26, 2013–2016. Pfeilschifter, J., and Vosbeck, K. (1991). Transforming growth factor inhibits interleukin 1 and tumor necrosis factor induced of nitric oxide synthase in rat renal mesangial cells. Biochem. Biophys. Res. Commun. 175, 372–379. Thompson, W. J., and Appleman, M. M. (1979). Assay of cyclic nucleotide phosphodiesterase and resolution of multiple molecular forms of the enzyme. Adv. Cycl. Nucl. 10, 69 –92. Bauer, A. C., and Schwabe, U. (1980). An improved assay of cyclic 3⬘,5⬘-nucleotide phosphodiesterase with QAE Sephadex A-25. Naunyn-Schmiedebergs Arch. Pharmacol. 311, 193–198.
Received January 30, 3001 Revised version received August 29, 2001 Published online October 29, 2001
Nitric Oxide–Induced F-Actin Disassembly Is Mediated via cGMP, cAMP, and Protein Kinase A Activation in Rat Mesangial Cells Katrin B. Sandau,* Florian Gantner,† and Bernhard Bru¨ne* ,1 *University of Kaiserslautern, Faculty of Biology, Department of Cell Biology, 67663 Kaiserslautern, Germany; †Byk Gulden, Department of Biochemistry, 78467 Konstanz, Germany
Glomerular mesangial cells contain actin and myosin, and in analogy to vascular smooth muscle cells, they can contract and relax to regulate the glomerular filtration rate. A key molecule that determines hemodynamic properties is nitric oxide, which is produced by nitric oxide synthase isoenzymes located in individual cells of the kidney. The contractility of mesangial cells is based on the interaction of actin microfilament bundles (F-actin) with myosin. We had the notion that nitric oxide influences the shape change of mesangial cells, so we analyzed the signal transduction involved. Chemically unrelated nitric oxide donors induced Factin dissolution, which was mediated by cGMP but was unrelated to protein kinase G activation. Actin disassembly was achieved with inhibitors of phosphodiesterase-3 and -4 or forskolin-evoked cAMP generation. We assumed that signal transmission involves activation of protein kinase A, and we went on to attenuate F-actin disassembly by protein kinase A inhibition. In conclusion, we found evidence that nitric oxide triggered F-actin dissolution via cGMP generation, inhibition of cAMP-hydrolyzing phosphodiesterase-3, and subsequent protein kinase A activation. © 2001 Elsevier Science
Key Words: Nitric oxide; cGMP; phosphodiesterases; protein kinase A; cytoskeleton; F-actin.
INTRODUCTION
The radical nitric oxide (NO) is produced by a family of NO synthase isoenzymes (NOS) which utilize L-arginine and oxygen to produce citrulline and NO [1]. Constitutive and inducible NOS isoforms can be distinguished. The constitutive NOS (cNOS) is regulated by Ca2⫹ and transiently generates NO in small concentrations, whereas the inducible form (iNOS) is activated by diverse agonists such as cytokines or lipopolysaccharide and releases high amounts of NO over an extended period of time. Once NO 1 To whom correspondence and reprint requests should be addressed at University of Kaiserslautern, Faculty of Biology, ErwinSchro¨dinger-Strasse, 67663 Kaiserslautern, Germany. E-mail: [email protected].
is produced, it may trigger physiological as well as pathophysiological mechanisms. In general, physiological processes are evoked by cGMP and protein kinase G (PKG) as NO binds and activates the heme subunit of soluble guanylyl cyclase (sGC) [2]. Pathophysiological functions of NO often provoke apoptosis or necrosis, but the mechanisms are more diverse, and no specific, cell-type–independent target has been identified so far. The ambivalent function of NO is demonstrated by its action in the kidney. During kidney-related diseases, such as diabetes or proliferative glomerulonephritis, increased NO production contributes to pathogenic hemodynamic changes and cell death of glomerular mesangial cells (MC) [3, 4], whereas chronic NO insufficiency causes hypertension [5]. These hemodynamic alterations are controlled by MC, which are specialized smooth muscle cells in the glomerulus of the kidney. MC contain the ability to contract and relax, thereby regulating the glomerular filtration rate and renal blood flow [5]. Under physiological conditions, these parameters are controlled by NO, generated by the endothelial NOS which is found in close proximity to MC. In the cardiovascular system, the action of endothelium-derived NO, which diffuses to smooth muscle cells and activates sGC, is well established [6]. However, it is controversially discussed how a cGMP signal leads to vasorelaxation and how NO affects mesangial contractility. Therefore, we were interested to analyze the signal transduction involved. To study the function of NO, we used different NO donors such as S-nitrosoglutathione (GSNO) or spermine-NO. These NO donors induced F-actin disassembly within 30 min, and by analyzing signal transduction, we established that the effect was mediated by cGMP production but was independent of PKG activation as described for other circumstances [7]. In contrast, we found protein kinase A activation, and as it is known that cGMP attenuates phosphodiesterase-3 (PDE), a cAMP and cGMP hydrolyzing enzyme, we used PDE inhibitors to mimic Factin dissolution. As F-actin dissolution was attenuated by PKA inhibition, we conclude that NO triggers F-actin disassembly in MC by cGMP production, inhi-
329
0014-4827/01 $35.00 © 2001 Elsevier Science All rights reserved.
330
¨ NE SANDAU, GANTNER, AND BRU
bition of cAMP-hydrolyzing PDEs, and finally executed by PKA activation. MATERIALS AND METHODS
FIG. 1. NO donor–induced F-actin disassembly. Rat mesangial cells (10 5 cells/well) were stimulated with vehicle (A), 250 M GSNO (B), or 250 M spermine-NO (C) for 30 min. After stimulation, cells were fixed, and F-actin was stained with phalloidinTRITC as outlined under Materials and Methods.
Reagents. Insulin, IBMX, erythro-9-(2-hydroxy-3-nonyl)-adenine, forskolin, phalloidin-TRITC, 8-bromo-cAMP, sulfanilamide, adenosine triphosphate (ATP), and N-naphthylethylenediamine were purchased from Sigma (Deisenhofen, Germany). Pefabloc was purchased from Boehringer Mannheim (Mannheim, Germany); [ 3H]cAMP and [ 3H]cGMP, and [␥- 32P]ATP were purchased from Amersham (Braunschweig, Germany); spermine-NO was purchased from Research Biochemicals International (Cologne, Germany); NS 2028 was purchased from Alexis (Gru¨nberg, Germany); and 8-chloroadenosine-3⬘,5⬘-cyclic monophosphorothioate Rp-isomer and Rp-8pCPT-cGMPS were purchased from BioLog (Bremen, Germany). RPMI 1640 and medium supplements were purchased from Biochrom (Berlin, Germany). Fetal calf serum (FCS) and protein kinase A assay system were purchased from Life Technologies (Berlin, Germany). GSNO was synthesized as described previously [8]. Motapizone was kindly provided by Nattermann (Ko¨ln, Germany); RP 73401 and sildenafil were provided by the Chemistry Department of Byk Gulden, Konstanz, Germany. All other chemicals were of the highest grade of purity and were commercially available. Culture of MC. Rat MC were cultured, cloned, and characterized as described previously [9]. Cells were grown in RPMI 1640 medium supplemented with 10% FCS, penicillin (100 U/ml), streptomycin (100 g/ml), and bovine insulin (5 g/ml). One day before and during the experiments, controls and stimulated cells were kept in medium with 0.5% FCS. For the experiments, 10 to 25 passages of MC were used. F-actin staining. MC were cultured and stimulated in eight-well chamber slides for F-actin staining. After stimulation, medium was discarded and cells were washed with phosphate-buffered saline (PBS) before fixing the cells with 3% paraformaldehyde for 30 min at room temperature. After a second step of PBS washing, cells were treated with 0.2% Triton X-100/PBS for 5 min, followed by PBS washing and F-actin staining with 10 g/ml phalloidin-TRITC for 1 h at 37°C. Afterward, chamber slides were washed precisely with PBS, dried, and embedded in glycerol. Stained cells were visualized with a Leica fluorescence microscope, and photographs were taken with a Leica camera. All photographs are of similar magnification (40⫻). Phosphodiesterase activity assay. MC were harvested, resuspended in 200 l homogenization buffer (10 mM Tris–HCl, pH 8.2, 1 mM -mercaptoethanol, 1 mM MgCl 2, 1 mM EGTA, 5 M pepstatin A, 10 M leupeptin, 400 M pefablock, 10 M soybean trypsin inhibitor, and 2 mM benzamidine) and disrupted by sonication (Branson 250 sonifier, output control 20%, duty cycle 1). The complete disruption of the cells (⬎98%) was checked via trypan blue exclusion. The homogenate was spun at 1000 ⫻ g for 5 min to remove viable cells (1–2%) and nuclei. The supernatant was decanted and stored at ⫺80°C until further use. PDE activity was determined as described by Thompson and Appleman [10] with some modifications [11]. In brief, 5 l of the 1000 ⫻ g supernatant was assayed in a final volume of 200 l containing 60 mM Tris HCl, pH 7.4, 5 mM MgCl 2, 0.5 M cAMP or cGMP (28,000 cpm [ 3H]cAMP or [ 3H]cGMP) and were incubated in the presence or absence of activators or inhibitors for 30 min at 37°C. The reaction was terminated by the addition of 50 l 0.2 N HCl, and the assay mixture was left on ice for further 15 min. Crotalus atrox snake venom (0.5 mg/ml) was added for 15 min at 37°C, and afterward, the assay mixture was loaded onto QAE-Sephadex A-25 columns (1-ml bed volume) and eluted with 2 ml ammonium formiate (30 mM, pH 6.0). The radioactivity in the eluate was counted in a liquid scintillation counter (Beckman). PDE isoenzyme activity calculations were performed by using activators (5 M cGMP for
NITRIC OXIDE–INDUCED F-ACTIN DISASSEMBLY
331
FIG. 2. Inhibition of sGC-attenuated NO-initiated F-actin disassembly in contrast to protein kinase G inhibition. Mesangial cells (10 5 cells/well) were stimulated with vehicle (A), 5 M NS 2028 (B), 250 M GSNO (C), 5 M NS 2028/250 M GSNO (D), 1 mM CPT-cGMPS (E), or the combination of 1 mM CPT-cGMPS/250 M GSNO (F). The specific sGC inhibitor NS 2028 and the PKG inhibitor CPT-cGMPS were preincubated for 15 min. After a 30-min incubation period, cells were fixed, and F-actin was visualized by phalloidin-TRITC staining as outlined under Materials and Methods.
PDE-2; 1 mM Ca 2⫹ and 100 nM calmodulin for PDE-1) or PDE isoenzyme–selective inhibitors (100 M erythro-9-(2-hydroxy-3nonyl)-adenine for PDE-2, 1 M Motapizone for PDE-3, 1 M RP73401 for PDE-4, and 100 nM sildenafil for PDE-5). The residual cAMP hydrolyzing activity in the presence of 1 M Motapizone and 1 M RP 73401, respectively, was designed as PDE res and theoretically could be composed of the novel cAMP-specific isoforms PDE-7, PDE-8, and PDE-10, or some combination of these. PKA activity assay. cAMP-dependent protein kinase activity was determined by in vitro phosphorylation of Kemptide by using an assay system from Life Technologies. Briefly, cells were washed twice with PBS and resuspended in 150 l extraction buffer (5 mM
EDTA, 50 mM Tris pH 7.5, 1% NP-40, 1 mM Na 3VO 4, 1 g/ml leupeptin, and 1 mM PMSF) followed by a 15-s sonication (Branson 250 sonifier, output control 20%, duty cycle 1). Cell debris were kept on ice for 30 min and centrifuged (17,000 ⫻ g, 15 min), and the protein content in the supernatant was analyzed. Afterward, 10 g protein of each sample was mixed with 4⫻ substrate solution with a final concentration of 50 M Kemptide, 100 M ATP, 10 mM MgCl 2, 0.25 mg/ml bovine serum albumin, 12.5 mM Tris pH 7.5, and 1 Ci [␥- 32P]ATP, and incubated for 5 min at 30°C. Reactions were terminated by spotting the reaction mix onto P81 phosphocellulose discs. The P81 paper sheets were immersed two times in 1% (v/v) orthophosphoric acid and washed three to five times with distilled
332
¨ NE SANDAU, GANTNER, AND BRU
TABLE 1 PDE Profile of Rat Mesangial Cells PDE isoform
PDE activity (pmol/mg ⫻ min)
1 2 3 4 5 Res.
45 ⫾ 16 0 38 ⫾ 9 6⫾ 6 5⫾ 2 2⫾ 1
Note. Unstimulated mesangial cells (10 7 cells/assay) were harvested and prepared for PDE activity determination as described under Materials and Methods. Data are expressed as pmol/mg protein ⫻ min mean total activities (soluble plus particulate) ⫾ standard deviation from five independent preparations.
water. Incorporated radioactivity was determined by scintillation counting. We ensured that the assay was linear with respect to time. Values are expressed as the percentage of increase of PKA activity vs control. Nitrite determination. Nitrite, a stable NO oxidation product, was determined by using the Griess reaction. Cell-free culture supernatants were collected (200 l), adjusted to 4°C, and mixed with 20 l sulfanilamide (dissolved in 1.2 M HCl) and 20 l N-naphthylethylenediamine-dihydrochloride. After 5 min at room temperature, the absorbance was measured at 560 nm with a reference wavelength at 690 nm. Nitrite concentrations were calculated by using a NaNO 2 standard. Statistical analysis. Each experiment was performed at least three times, and representative illustrations are shown.
RESULTS
FIG. 3. Inhibition of phosphodiesterases leads to F-actin dissolution. Mesangial cells (105 cells/well) were incubated with vehicle (A), the unselective PDE inhibitor IBMX (500 M) (B), or the combination of the PDE-3 inhibitor Motapizone (1 M) and the PDE-4 inhibitor RP 73401 (1 M) (C). After 30 min, cells were fixed and F-actin was visualized by phalloidin-TRITC staining as outlined under Materials and Methods.
In the first set of experiments, we used two chemically unrelated NO donors to analyze the role and signal transduction of NO leading to F-actin disassembly in rat MC. We stimulated MC for 30 min with GSNO (Fig. 1B) or spermine-NO (Fig. 1C) in a concentration of 250 M (Fig. 1). For control reasons, we incubated MC with glutathione and spermine, decomposed end products of GSNO and spermine-NO, but no F-actin changes were detectable (data not shown). Therefore, NO donor–induced F-actin disassembly is based on the release of NO. Generally, F-actin resolution appeared 10 min after NO addition and was completely reversible after 4 h. Dose-response studies showed visible F-actin changes with 500 nM and more pronounced effects with 1 M GSNO (data not shown); the alterations were sharpest at 250 M. Next, we used the sGC inhibitor NS 2028 to block sGC activation and cGMP-dependent signaling. NS 2028 at a concentration of 5 M revealed no cytotoxic effect or F-actin changes (Fig. 2B). In combination with 250 M GSNO (Fig. 2D), NS 2028 attenuated F-actin disassembly induced by NO after a 30-min incubation time, when NS 2028 was preincubated for 15 min. F-actin staining of these cells resembled control cells.
NITRIC OXIDE–INDUCED F-ACTIN DISASSEMBLY
FIG. 4. Lipophilic cAMP analogs and forskolin evoke F-actin disrupture. Mesangial cells (105 cells/well) were treated for 30 min with vehicle (A), 1 mM 8-bromo-cAMP (B), or 10 M forskolin (C). Cells were fixed and stained for F-actin as described under Materials and Methods.
In summary, inhibition of sGC blocked NO-induced cytoskeleton changes, which is indicative of a cGMPmediated pathway. In several cases, cGMP-mediated
333
processes are triggered by PKG activation. To test the involvement of PKG, we employed 1 mM of the specific PKG inhibitor Rp-8-pCPT-cGMPS (CPT-cGMPS) [12] alone (Fig. 2E) or in combination with 250 M GSNO (Fig. 2F), but CPT-cGMPS failed to modulate F-actin disassembly induced by NO. Similar results were achieved with another PKG inhibitor KT-5823 (data not shown), supporting the conclusion that PKG activation is not involved. Cyclic nucleotide PDE isoforms are currently classified into 10 families. Two of them are regulated by cGMP (PDE-2, the so-called cGMP-stimulated PDE activity, and PDE-3, classified as cGMP-inhibited PDE family). Therefore, we used different PDE inhibitors alone or in combination to assess their effect on F-actin organization. First, we used the nonspecific PDE inhibitor IBMX at a concentration of 500 M. After 30 min of incubation, F-actin disrupture resembled NOtreated cells (Fig. 3B). In addition, we supplied more specific PDE inhibitors such as erythro-9-(hydroxy-3nonyl)-adenine for PDE-2, Motapizone for PDE-3, RP 73401 for PDE-4, and sildenafil for PDE-5. These inhibitors, added individually, displayed no effect on Factin (data not shown). However, the combination of Motapizone and RP 73401 induced stress fiber dissolution after 30 min (Fig. 3C), and alterations became reversible after 24 h (data not shown). In order to obtain information on PDE isoforms present in our system, we determined a PDE profile in MC. We measured the activity of PDE-1 to PDE-5 according to the protocol described under Materials and Methods (Table 1). We detected a high activity for PDE-1 and PDE-3, and some activity of PDE-4 and PDE-5. PDE-1, depending on the subtype or subtypes expressed, preferably degrades cGMP, whereas PDE-4 and to a lesser extent PDE-3 are specific for cAMP. Obviously, PDE-3 is the most prominent cAMP-hydrolyzing PDE in MC, and as it is defined as cGMP inhibitable, an endogenous cGMP increase as generated by NO donors would result in a cAMP increase. Next, 1 mM lipophilic 8-bromo-cAMP was applied or the adenylyl cyclase was stimulated by forskolin (10 M) to test whether cAMP leads to F-actin resolution. F-actin disassembly occurred after 30 min, similar to NO when 8-bromo-cAMP or forskolin were added (Fig. 4). This process was again fully reversible within 24 h (data not shown). In addition, we used NS 2028 in combination with Motapizone/RP 73401, forskolin, or 8-bromo-cAMP (data not shown). As NS 2028 did not interfere with F-actin disassembly in response to 8-bromo-cAMP, we conclude that cAMP action is downstream of sGC. We measured PKA activity after stimulating MC with 250 M GSNO, 1 mM 8-bromo-cAMP, which served as a positive control, or with 1 M Motapizone/1 M RP 73401. Reactions were terminated after 10 min.
334
¨ NE SANDAU, GANTNER, AND BRU
FIG. 5. F-actin disassembly is attenuated by PKA inhibition. Mesangial cells (10 5 cells/well) were stimulated for 30 min with vehicle (A), 1 mM Cl-AMPS (B), 1 M Motapizone/1 M RP 73401 (C), 1 mM Cl-AMPS/1 M Motapizone/1 M RP 73401 (D), 1 M GSNO (E), and 1 mM Cl-AMPS/1 M GSNO (F). Cl-AMPS was preincubated for 15 min. Cells were fixed and stained for F-actin as described under Materials and Methods.
As expected, 8-bromo-cAMP induced strong PKA activation (69 ⫾ 8% increase vs control) after 10 min. GSNO stimulation achieved 24 ⫾ 4% PKA activation vs control, which is similar to Motapizone/RP 73401 (24 ⫾ 2% activation vs control), measured after 30 min.
Apparently, the PKA activation can occur in the same time frame as actin disassembly. To examine the role of an active PKA for F-actin disassembly, we used a specific PKA type I inhibitor, 8-chloroadenosine-3⬘,5⬘-cyclic monophosphorothiotae
NITRIC OXIDE–INDUCED F-ACTIN DISASSEMBLY
TABLE 2 Increased PKA Activity after NO Stimulation
Stimulation
Time (min)
Increase of PKA activity vs control (%)
Control GSNO (250 M) 8-bromo-cAMP (1 mM) Motapizone (1 M) ⫹ RP 73401 (1 M)
10 10 10 30
0 24 ⫾ 4 69 ⫾ 8 24 ⫾ 2
Note. Mesangial cells (10 7 cells/assay) were stimulated with vehicle (control), 250 M GSNO, 1 mM 8-bromo-cAMP, or 1 M Motapizone in combination with 1 M RP 73401 for the times indicated. Afterward, PKA activity was determined as described under Materials and Methods. Data are expressed as the percentage increase of PKA substrate phosphorylation as a marker of PKA activity vs unstimulated controls. The experiments were performed at least five times and mean values ⫾ standard deviation are given.
Rp-isomer (Cl-AMPS) [13], which was effective as seen in Fig. 5. Cl-AMPS by itself did not change the F-actin structure (Fig. 5B). To attenuate F-actin changes induced by Motapizone/RP 73401 (Fig. 5C) or GSNO (Fig. 5E), we preincubated Cl-AMPS for 15 min. Cells stimulated with an agonist plus the PKA inhibitor displayed normal F-actin structure (Figs. 5D, 5F). Cytoskeleton changes achieved with lipophilic cAMP analogs, as shown in Fig. 4, were blocked with ClAMPS as well (data not shown). These results point to the involvement of PKA type I, leading to F-actin disassembly in MC. As a PKA inhibitor attenuates the effect of NO, PDE inhibitors, and cAMP, we predict that PKA activation appears downstream of cGMP and cAMP production. DISCUSSION
Herein, we have shown that different NO donors provoke F-actin dissolution in glomerular MC (Fig. 1). The process can be attenuated by inhibition of either sGC or type I PKA (Figs. 2, 5). In addition, inhibition of the predominant cAMP-hydrolyzing enzymes present in MC such as PDE-3 and PDE-4 (Fig. 3), activation of the adenylyl cyclase by forskolin, or addition of lipophilic cAMP analogs (Fig. 4) achieved F-actin disassembly. All stimuli induced a transient, fully reversible effect which was sensitive toward PKA inhibition (Fig. 5), while the sGC inhibitor NS 2028 was effective after NO stimulation only (Fig. 2). This allowed us to postulate the position of the cGMP increase as upstream of the PDE inhibition and PKA activation. In contrast to the findings of an earlier report [14] that described the presence of PDE-1, PDE-2, and PDE-4 in rat MC, but in line with the findings of Dousa [15], the MC in our study displayed high PDE-3 and PDE-1 activity and a lower but functionally important
335
PDE-4, whereas PDE-2 was completely lacking (Table 1). Inhibitors against PDE-1, PDE-2, or PDE-5 were ineffective in inducing F-actin changes, whereas selective inhibition of PDE-3 and PDE-4 with the combination Motapizone/RP 73401 achieved F-actin disassembly. Such a PDE-3/4 synergism is well established in a variety of functional immune cell responses [16] and has also been described in experimental settings where PDE-3 inhibition due to NO-mediated cGMP elevation was potentiated by a PDE-4 inhibitor [17]. In the context of renin secretion, a similar scenario for NO has been suggested by Kurtz et al. [18]. They measured renin secretion of isolated perfused rat kidneys and found increased amounts of renin after stimulation with the NO donor SNP after activation of adenylyl cyclase or supplementation of different PDE inhibitors. The NO effect was suppressed by sGC inhibition, while renin secretion was drastically enhanced by attenuating PDE-3. In this study, the authors did not analyze the involvement of PKA. In our case, Factin disassembly is mediated by PKA activation (Table 2), and as expected, Cl-AMPS attenuated stress fiber relaxation (Fig. 5). Therefore, we assumed a cAMP increase, but as we have not obtained cAMP data, we cannot ultimately exclude the possibility that cGMP activates PKA directly, a pathway suggested by Cornwell et al. [19] who analyzed the role of NO for smooth muscle proliferation. They reported an increase in cGMP as a result of NO addition, activation of PKA, and inhibition of smooth muscle cell growth. PKA inhibitors attenuated growth inhibition, but cAMP accumulation was not found. The effect of PKA on the cytoskeleton machinery (e.g., myosin light chain [MLC] kinase, MLC, actin, or other kinases or phosphatases) were not in the focus of this study, but a recent article analyzed the effect of dibutyryl cGMP on MLC phosphorylation in MC [20]. The dynamic phosphorylation and dephosphorylation cycle of MLC influences the assembly and disassembly of the actin microfilament network responsible for stress fiber cell contraction and relaxation. The authors report that dibutyryl cGMP attenuated MLC phosphorylation elicited by angiotensin II or phorbol ester. As this was blocked by the phosphatase inhibitor calyculin A, it was concluded that cGMP activates MLC phosphatase. However, data on MLC phosphatase activation were not presented, and in light of our results, we instead suggest that cGMP acts via PKA, a pathway which cannot be excluded on the basis of the study of Torrecillas et al. [20]. The importance of NO for physiological or pathophysiological actions in the glomerulus of the kidney is well established. MC itself contain a NO-generating system with the implication that NO may have autocrine as well as paracrine functions [21–23]. We have shown that exogenously (Fig. 1) as well as endog-
¨ NE SANDAU, GANTNER, AND BRU
336
enously generated NO influenced mesangial contraction and relaxation in vitro. It can be assumed that NO will act in a similar way in vivo, thus provoking hemodynamic changes in the glomerulus as described before. This now can be explained by NO formation and its signaling properties involving guanylyl cyclase activation, cGMP generation, PDE-3 and PDE-4 inhibition, and PKA activation. We thank C. Blechner and S. Liehner for excellent technical assistance and Drs. H. Tenor and A. Hatzelmann (Byk Gulden, Konstanz) for helpful discussions. This work was supported by the Deutsche Forschungsgemeinschaft (Br 999 and SFB 423, A5).
12.
Butt, E., Van Bemmelen, M. X. L., Fischer, L., Walter, U., Jastorff, B. (1990) Inhibition of cGMP-dependent protein kinase by Rp-fuanosine 3⬘, 5⬘-monophosphorothioates. FEBS Lett. 263, 47–50.
13.
Gjertsen, P. J., Mellgren, G., Otten, A., Maronde, E., Genieser, J. G., Jastroff, B., Vintermyr, O. K., McKnight, G. S., and Doskeland, S. O. (1995). Novel (Rp)-cAMPS analogs as tools for inhibition of cAMP-kinase in cell culture. Basal cAMP-kinase activity modulates interleukin-1 action. J. Biol. Chem. 270, 20599 –20607.
14.
Ahn, H. S., Foster, M., Arik, L., Bykow, G., and Foster, C. (1995). Cyclic nucleotide phosphodiesterase isoenzymes in rat mesangial cells. Eur. J. Pharmacol. 289, 49 –57.
15.
Dousa, T. P. (1999). Cyclic-3⬘,5⬘-nucleotide phosphodiesterase isozymes in cell biology and pathophysiology of the kidney. Kidney Intl. 55, 29 – 65.
16.
Schudt, C., Gantner, F., Tenor, H., and Hatzelmann, A. (1999). Therapeutic potential of selective PDE inhibitors in asthma. Pulm. Pharmacol. Ther. 12, 123–129, doi:10.10061pupt. 1999.0182.
17.
Komas, N., Lugnier, C., and Stoclet, J. C. (1991). Endotheliumdependent and independent relaxation of the rat aorta by caclic nucleotide phosphodiesterase inhibitors. Br. J. Pharmacol. 104, 495–503.
18.
Kurtz, A., Go¨tz, K. H., Hamann, M., and Wagner, C. (1998). Stimulation of renin secretion by nitric oxide is mediated by phosphodiesterase. Proc. Natl. Acad. Sci. USA 95, 4743– 4747.
19.
Cornwell, T. L., Arnold, W., Boerth, N. J., and Lincoln, T. M. (1994). Inhibition of smooth muscle cell growth by nitric oxide and activation of cAMP-dependent protein kinase by cGMP. Am. J. Physiol. 267, C1405–C1413.
20.
Torrecillas, G., Diez-Marques, M. L., Garcia-Escribano, C., Bosch, R. J., Rodriguez-Puyol, D., and Rodriguez-Puyol, M. (2000). Mechanisms of cGMP-dependent mesangial-cell relaxation: A role for myosin light-chain phosphatase activation. Biochem. J. 346, 217–222.
21.
Pfeilschifter, J., Rob, P., Mu¨lsch, A., Fandrey, J., Vosbeck, K., and Busse, R. (1992). Interleukin 1 beta and tumour necrosis factor alpha induce a macrophage-type of nitric oxide synthase in rat renal mesangial cells. Eur. J. Biochem. 203, 251–255.
22.
Shultz, P. J., Tayeh, M. A., Marletta, M. A., and Raij, L. (1991). Synthesis and action of nitric oxide in rat glomerular mesangial cells. Am. J. Physiol. 261, F600 –F606.
23.
Marsden, P. A., and Ballermann, B. J. (1990). Tumor necrosis factor alpha activates soluble guanylate cyclase in bovine glomerular mesangial cells via an L-arginine– dependent mechanism. J. Exp. Med. 172, 1843–1852.
REFERENCES 1. 2. 3.
4.
5. 6.
7.
8.
9.
10.
11.
Marletta, M. A. (1994). Nitric oxide synthase: Aspects concerning structure and catalysis. Cell 78, 927–930. Nathan, C. (1992). Nitric oxide as a secretory product of mammalian cells. FASEB J. 6, 3051–3064. Cattell, V., and Cook, T. (1995). The nitric oxide pathway in glomerulonephritis. Curr. Opin. Nephrol. Hypertens. 4, 359 – 364. Sandau, K., Pfeilschifter, J., and Bru¨ne, B. (1997). The balance between nitric oxide and superoxide determines apoptotic and necrotic death of rat mesangial cells. J. Immunol. 158, 4938 – 4946. Raij, L., and Baylis, C. (1995). Glomerular actions of nitric oxide. Kidney Intl. 48, 20 –32. Moncada, S., Palmer, R. M., and Higgs, E. A. (1991). Nitric oxide: Physiology, pathophysiology, and pharmacology. Pharmacol. Rev. 43, 109 –142. Ruth, P. (1999). Cyclic GMP-dependent protein kinases: Understanding in vivo functions by gene targeting. Pharmacol. Ther. 82, 355–372. Hart, T. W. (1997). Some observation concerning the S-nitroso and S-phenylsulfonyl derivatives of L-cysteine and glutathione. Tetrahedron Lett. 26, 2013–2016. Pfeilschifter, J., and Vosbeck, K. (1991). Transforming growth factor inhibits interleukin 1 and tumor necrosis factor induced of nitric oxide synthase in rat renal mesangial cells. Biochem. Biophys. Res. Commun. 175, 372–379. Thompson, W. J., and Appleman, M. M. (1979). Assay of cyclic nucleotide phosphodiesterase and resolution of multiple molecular forms of the enzyme. Adv. Cycl. Nucl. 10, 69 –92. Bauer, A. C., and Schwabe, U. (1980). An improved assay of cyclic 3⬘,5⬘-nucleotide phosphodiesterase with QAE Sephadex A-25. Naunyn-Schmiedebergs Arch. Pharmacol. 311, 193–198.
Received January 30, 3001 Revised version received August 29, 2001 Published online October 29, 2001