| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute of Anatomy and Cell Biology (D.M., M.P., R.M.), Justus-Liebig-University, 35385 Giessen, Germany; Institute for Hormone and Fertility Research (D.M., L.C.-D., L.T.B., B.B.-S., A.K.M.), University of Hamburg, 20251 Hamburg, Germany; and Department of Pharmacology (R.C.S.), University of Mississippi, University, Mississippi 38677
Address all correspondence and requests for reprints to: Dr. Dieter Müller, Institute of Anatomy and Cell Biology, Justus-Liebig-University, Aulweg 123, 35385 Giessen, Germany. E-mail: hans-dieter.mueller{at}anatomie.med.uni-giessen.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ANP/GC-A/cGMP signaling plays an essential role in regulating blood pressure and fluid volume homeostasis, with vascular and kidney cells representing important peptide target sites (2). However, GC-A is widely distributed in many different tissues and cell types, and there is convincing evidence for additional functions (2). For example, GC-A is highly expressed in testis (3, 4), and ANP-mediated accumulations of cGMP in Leydig cells have been shown to increase the cellular production of the male sex hormone, testosterone (5, 6, 7, 8).
Representing a specific phenomenon, the biological activities of GC-A (and GC-B) are regulated by mechanisms leading to alterations in the extent of receptor phosphorylation at a number of serine/threonine residues that are localized intracellularly within a kinase homology domain (1). Dephosphorylation at these sites decreases ligand-stimulatable GC activity, designated as receptor desensitization (9). Such reactions can be elicited by chronic exposure of cells to natriuretic peptides themselves (referred to as homologous desensitization), functionally resembling desensitization processes in many other ligand/receptor systems in which, however, activity down-regulation occurs via receptor endocytosis/internalization (10). Desensitization of GC-A and GC-B can also be induced by signaling molecules other than natriuretic peptides. This phenomenon, referred to as heterologous desensitization, provides a basis for cross talk between different signaling systems. Agents shown to elicit natriuretic peptide receptor desensitization include hormones like angiotensin II, vasopressin, endothelin (1), and pituitary adenylate cyclase activating peptide (11). Representing a novel aspect of particular interest, recent findings have demonstrated that the biologically active lipid lysophosphatidic acid (LPA) can desensitize GC-B (12, 13). This lipid generates diverse and potent cellular effects via binding to at least three different G protein-coupled receptors (designated as LPA1 to LPA3), which use manifold downstream signaling pathways (14). In case of GC-B desensitization, the LPA receptor type(s) involved have not yet been defined.
Although the principle of natriuretic peptide receptor regulation has been well established by basic studies predominantly performed with receptor-transfected cell lines, important questions remain to be addressed. For example, the signaling pathways mediating homologous and heterologous desensitization and cell type-related physiological functions are still poorly understood.
In this study, we examined whether GC-A is desensitized by ANP and/or LPA in MA-10 Leydig cells. The MA-10 cell line (15) has been frequently used as a model system for studying Leydig cell physiology, in particular regarding the regulation of steroidogenesis by different factors (16). Like native mouse and rat Leydig cells, MA-10 cells highly express GC-A endogenously (8, 17) without detectable levels of GC-B and NPR-C, suggesting a major role for ANP in this cell type, and that effects of ANP are mediated at the receptor level exclusively by interactions with GC-A. In addition, and considering a proposed role of cGMP-dependent protein kinase (PKG) I for regulation of GC-A activity (18), our findings regarding the absence of this kinase in MA-10 and isolated Leydig cells require attention.
Because both ANP and LPA were found to elicit GC-A desensitization in MA-10 cells, this study allowed for the first time to characterize comparatively homologous and heterologous desensitization reactions. Whereas ANP/GC-A/cGMP signaling is known to be involved in steroidogenesis, we in addition recognized pronounced morphoregulatory effects of LPA, indicating that LPA may act as an important signaling molecule for Leydig cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T3-1 cell line (19) was kindly provided by Dr. P. Mellon (University of California, San Diego, La Jolla, CA). The protein kinase A (PKA) inhibitors, H 89 and 8-(4-chlorophenylthio)adenosine-3', 5'-cyclic monophosphorothioate, Rp-isomer (Rp-8-CPT-cAMPS), were obtained from Biomol (Hamburg, Germany) or BIOLOG (Bremen, Germany), respectively, the MAPK kinase (MEK) inhibitor PD 98059 from Calbiochem (Bad Soden, Germany). 1-Methyl-3-isobutyl xanthine (IBMX) was from Sigma, 8-bromo-cAMP (8-Br-cAMP) from BIOLOG, and protease inhibitors from Boehringer (Mannheim, Germany).
Cell culture
MA-10 cells were grown in culture medium containing 7.5% horse serum and 2.5% fetal calf serum as detailed before (20). Depending on the assays intended, cells were dispensed into 12-well plates (for whole-cell stimulations followed by determinations of cGMP content), 21.4 -m2 dishes (for assessments of cell lysates), or T75 culture flasks (for generation of membrane/cytosolic proteins or for RNA isolation). In each case, cells were washed once with PBS and incubated with serum-free medium containing 1% BSA overnight at 37 C in an atmosphere of 95% air-5% CO2 before specific treatments. T3 cells, used in certain control experiments, were cultured in DMEM, supplemented with 4.5 g/liter glucose, 110 mg/liter pyruvate, 549 mg/liter L-glutamine (Sigma, D6429), 10% fetal calf serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin sulfate (Sigma, P0781). Isolated Leydig cells were cultured as described before (7).
Cell treatments
Most assays were performed in 12-well plates using incubation volumes of 500 µl and approximately 50,000 cells/well. After overnight incubations (see above), media were aspirated, and cells were treated in serum-free medium containing 0.25 mM IBMX with the agents and for the time periods indicated. Reactions were stopped by addition of 1.5 ml ice-cold 100% (vol/vol) ethanol. After storing the plates for 1 h at –20 C, ethanol extracts were collected and centrifuged at 2000 x g for 30 min at 4 C. The supernatant fractions were withdrawn, vacuum evaporated to dryness, and resolved in 0.5 ml of assay buffer [0.1 M sodium phosphate, 0.15 M NaCl, 5 mM EDTA, 0.2% BSA, 0.01% thimerosal (pH 7.5)] before measurements of cGMP. When assays were performed in 21.4-cm2-dishes (to examine cell lysates) or T75 flasks (to facilitate guanylyl cyclase assays), cells were treated as indicated above.
Preparation of membrane and cytosolic proteins
The generation of membrane (particulate) and cytosolic (soluble) protein fractions used for affinity labeling (4) or immunoblotting (21, 22) experiments has been described before. To prepare membrane protein fractions from MA-10 cells for subsequent assessments of GC activities, cells were cultured in 75-cm2 flasks. After the treatments indicated, cells were washed once with PBS, scraped off into PBS using a rubber policeman, and centrifuged at 4 C for 3 min at 500 x g. Cell pellets derived from two flasks were pooled and homogenized at 4 C in 0.45 ml homogenization buffer [50 mM Tris-HCl (pH 7.5), containing 1 mM EDTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonylfluoride] by 20–30 strokes in a Potter-Elvehjem homogenizer. After centrifugation at 3000 x g for 8 min at 4 C to remove cell debris and nuclei, the supernatant fractions were centrifuged at 4 C for 30 min at 100,000 x g. The resulting supernatant fractions were aspirated, and the crude membrane pellets were resuspended in 70–100 µl of 50 mM Tris-HCl (pH 7.5). After determining protein concentrations by using a kit from Bio-Rad Laboratories (Munich, Germany) with BSA (fraction V) as standard, these membrane preparations were either used directly or after storage (for up to 1 wk) at –80 C (without any detectable alterations in enzyme activity, not shown) for assessments of GC activity.
Generation of cell lysates
After treatment, cells (in 21.5-cm2 dishes) were rinsed once with ice-cold PBS, scraped off into 300 µl ice-cold lysis buffer [20 mM Tris-Cl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM ß-glycerolphosphate, 1 mM Na3VO4, 1 µg/µl leupeptin, 1 mM phenyl methyl sulfonyl fluoride] and transferred to Eppendorf tubes. The tubes were kept on ice for 30 min and frequently vortexed. After centrifugation at 10,000 x g for 10 min at 4 C, the supernatant fractions, representing the cell lysates, were stored at –80 C until usage for Western blot analyses. Protein concentrations were determined as described above.
Assays of membrane GC activity
Determinations of membrane (particulate) GC activity were performed essentially as described (4). The only exceptions were that reactions contained 5 (instead of 10) µg of protein and that cGMP levels were measured by time-resolved fluorescence immunoassay (TRFIA) (see below). To assess maximal (ligand-independent) GC activities (23), incubations were carried out in the same buffer as described but with 5 mM MnCl2 per 1% (vol/vol) Triton X-100 instead of ANP and MgCl2.
Measurements of cGMP
Levels of cGMP were determined in most cases by means of a cGMP-specific TRFIA, representing a competitive solid-phase assay based on Eu3+-chelate-labeled cGMP as tracer. This commercially available (Bio
conomic$, Hamburg, Germany; www.ihf.de) assay was designed for measurements of cGMP in assay buffer and cell culture medium. Cross-reactivities of structurally related compounds are 0.00004 (cAMP), 0.00003 (guanosine), 0.00001 (GTP), and less than 0.00001 (other nucleotides/nucleosides). Assays were carried out according to the protocol of the supplier, and fluorescence was measured in a multilabel counter (Wallac 1420 Victor2) from EG&G Wallac (Turku, Finland). In some experiments, cGMP levels were determined by a commercial ELISA, as previously described (4).
Photoaffinity labeling of natriuretic peptide receptors
The protocols for photoaffinity labeling of GC-A by 125I-ANP (24) and GC-B by 125I-[Tyr0]CNP (25) have been described in detail before. In brief, membranes were incubated with either 125I-labeled ANP (0.5 nM) or CNP (2 nM), ligand/receptor cross-links were induced by UV light irradiation, and reaction products were analyzed by SDS-PAGE under reducing conditions in 7.0% acrylamide separation gels followed by autoradiography at –80 C using XAR-5 films (Kodak, Rochester, NY) and intensifying screens.
Immunoblotting
After separation by SDS-PAGE under reducing conditions in 9% acrylamide gels, proteins were transferred to nitrocellulose membranes, and blots were pretreated as described before (4) before probing with antibodies directed against cGMP-dependent protein kinase I (anti-GK I-CT, no. 539729, diluted 1:3000, Calbiochem, San Diego, CA), ERK1/2 (anti-ERK1/2-CT, no. 06-182, 1:5000, Biomol), phospho-ERK1/2 (no. 9106, 1:1000; Cell Signaling, Beverly, MA), PKA catalytic subunit (PKA
cat, sc-905, 0.5 µg/µl; Santa Cruz, Heidelberg, Germany) or PKA regulatory subunits RI (no. 610165) and RII (no. 610625), each 1:1000 (BD Biosciences, San Jose, CA). Either antimouse or antirabbit IgG, linked to peroxidase (Pierce, Rockford, IL), was used as secondary antibodies. Signals were detected as described before (4).
RT-PCR analysis of LPA receptor subtypes
The procedures for isolation of total RNA, cDNA synthesis, and the PCR conditions used have been described in detail recently (26). The LPA1-specific primers used were 5'-ATC TTT GGC TAT GTT CGC CA-3' (sense) and 5'-TTG CTG TGA ACT CCA GCC A-3' (antisense), yielding a PCR product of the expected size of 394 bp; the LPA2-specific primers were 5'-TGG CCT ACC TCT TCC TCA TGT TCC A-3' (sense) and 5'-GGG TCC AGC ACA CCA CAA ATG CC-3' (antisense), resulting, as expected, in the amplification of 516-bp fragments; the LPA3-specific primers, 5'-AGT GTC ACT ATG ACA AGC-3' (sense) and 5'-GAG ATG TTG CAG AGG C-3' (antisense), induced the generation of 513-bp fragments (26). Reaction products were resolved by electrophoresis in 1.5% agarose gels and ethidium bromide staining. Analogous assessments of glyceraldehyde-3-phosphate dehydrogenase gene expression served to control the cDNA syntheses (not shown).
Immunocytochemistry
MA-10 cells (12,000 cells/chamber) were grown on four-well thin borosilicate cover slides (Nunc, Wiesbaden, Germany). After incubation in serum-free medium in either the absence or presence of LPA (10 µM) or ANP (0.1 µM), cells were washed with PBS and fixed (3% paraformaldehyde for 30 min). After washing (3 times with PBS), cells were permeabilized for 2 min with 0.5% Triton X-100 and blocked for 60 min with 5% rabbit nonimmune serum (Sigma). The slides were then incubated for 60 min with the primary antibody (antismooth muscle -actin, Sigma, diluted 1:500) followed by treatment (60 min) with a Cy3-conjugated secondary antibody (antirabbit, 1:200; Jackson Immunochemicals, Dianova, Hamburg, Germany). To stain cell nuclei, the fluorescence groove-binding probe for DNA, 4'6'-diamidino-2-phenylindole (DAPI, diluted 1:200; Sigma), was applied for 5 min after the secondary antibody. After washing, the chamber slides were kept in dark at 4 C overnight before viewing on a fluorescence/laser-scanning microscope (Nikon, Dusseldorf, Germany). Different Nikon filter blocks were used for DAPI (330–380 nm) and Cy3-immunofluorescence (using fluorescein isothiocyanate filter, 450–490 nm). Micrographs were taken using a digital camera (Leica, Wetzlar, Germany).
Data presentation and statistical analyses
The data were graphed and analyzed using Prism 3.02 (GraphPad Software Inc., San Diego, CA). When error bars are not visible, they are contained within the data point. The significance of effects was assessed by unpaired t test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), performed with either 125I-ANP (label GC-A) (24) or 125I-[Tyr0]CNP (label GC-B) (25), indicate GC-A but not GC-B expression in MA-10 cells, whereas both receptors are detectable in a control (T3) cell line, previously proposed by RT-PCR analyses to coexpress GC-A and GC-B (27). The same assays concurrently revealed (not shown) the absence of detectable amounts of the 60-kDa natriuretic peptide receptor, NPR-C, which would be labeled in these experiments by both radioligands (4, 22). Consistent with the cross-linking data, all three natriuretic peptides induced massive accumulations of cGMP in T3 cells (not shown), whereas only ANP and BNP (the ligands for GC-A) were highly effective in stimulating cGMP production in MA-10 cells (Fig. 1B). Analogous experiments were performed with isolated rat Leydig cells. These studies showed increases (up to 300-fold) in cGMP in response to ANP but not CNP (not shown). In both, MA-10 and isolated rat Leydig cells, basal cGMP contents calculated per 50,000 cells were 0.1 pmol or less, and treatments with ANP (0.1 µM, 30 min) induced greater than 100-fold elevations, resulting in total amounts of up to 25 pmol.
|
). In addition, we also did not find any expression in Leydig cells of PKG II, using rat brain extracts as a positive control (data not shown). In contrast, expression of PKA (Fig. 1C shows the catalytic subunit; similar results were obtained for the regulatory subunits RI and RII) and other kinases examined (e.g. glycogen synthase kinase-3, protein kinase B, and phosphatidylinositol 3-kinase; data not shown) was well detectable in both MA-10 and isolated Leydig cells. Thus, with regard to investigations of GC-A desensitization, immortalized MA-10 Leydig cells represent an interesting model based on three notable features: 1) the stable expression of high levels of GC-A; 2) the absence (at least at significant levels) of the two other natriuretic peptide receptors that could hypothetically (GC-B) or in a well-established manner (NPR-C; e.g. Ref.28) interfere with ANP/GC-A signaling; and 3) the absence of PKG, representing a major cellular target for ANP-induced cGMP (29).
Homologous desensitization
Initial studies served to examine the time dependency of ANP-dependent cGMP production in MA-10 cells. Incubations with 0.1 µM ANP were carried out for different time periods before measurements of cGMP (Fig. 2A). ANP was found to rapidly stimulate GC-A activity, as indicated by 12-fold increases (from 0.05 to 0.6 pmol per 25,000 cells) in cGMP after 1 min incubation time. Highest values (7.6 ± 0.8 pmol per 25,000 cells) of cGMP were measured after incubation times of 30 min, whereas longer treatments failed to further enhance cGMP concentrations. Thus, in the following experiments, ANP incubation times of 30 min were used to generate maximum cGMP accumulations.
|
, preexposure of cells to ANP triggered, in a time-dependent manner, decreases in subsequent ANP-dependent accumulations of cGMP. A significant (P < 0.01) reduction by 26% already was found after a 10-min preincubation time, and pretreatments for 30 min resulted in about 40% less cGMP, compared with either untreated (0 min) or in the absence of ANP preincubated (–ANP) controls (Fig. 2B). No further attenuation in cGMP levels was observed after a 60-min preexposure, and preincubations with CNP instead of ANP failed to elicit any effects (not shown). Basal values of cGMP, measured in the absence of ANP stimulation after pretreatments with ANP and washing, ranged from 0.1 ± 0.05 (0 min ANP) up to 2.6 ± 0.4 (30 min ANP) pmol per 50.000 cells.
To assess whether the ANP-induced effects are indeed explained by alterations in hormone-dependent GC-A activity, GC assays were carried out with membranes prepared from cells after preincubations for 30 min in either the absence or presence of ANP. These experiments (Fig. 2C) revealed similar decreases (by 40%) in ANP-elicited cGMP production as observed with intact cells (Fig. 2B), indicating that alterations in membrane GC activity accounted for the effect.
To examine whether the reduced cGMP-generating capacity of GC-A after ANP preincubations could be due to decreases in molecular levels (rather than in intrinsic activity) of the receptor, GC assays were performed in the presence of MnCl2/Triton X-100. This treatment maximally activates membrane GCs ligand independently and represents an indicator of total receptor amount (12, 23). The results (Fig. 2D) show that ANP pretreatment does not reduce MnCl2/Triton-dependent GC activity.
Thus, these studies revealed that ANP can induce decreased GC-A activity in MA-10 Leydig cells, based on mechanisms directly affecting the hormone responsiveness of the receptor. Because cGMP is the primary signaling molecule generated during this process and taking into account that control experiments ruled out any ANP-induced elevations in cellular cAMP levels (data not shown), it is reasonable to assume that cGMP target proteins are implicated as secondary signal mediators. Considering the apparent absence of PKG I (Fig. 1C), a key mediator of cGMP effects, and the fact that PKA has been shown previously to act as a functional target for cGMP in Leydig cells (7) and other cell types (30, 31, 32), we next examined whether the PKA inhibitor, H89, can influence homologous desensitization of GC-A in MA-10 cells. These studies revealed that the presence of 1 µM H89 during preincubation with ANP blocked GC-A desensitization (Fig. 3), whereas pretreatments with the inhibitor alone failed to elicit any effects on GC-A activity (see Fig. 5). Thus, these results suggested that PKA mediates the ANP-induced desensitization of GC-A in MA-10 cells. To prove the involvement of PKA, analogous assays were performed with Rp-8-CPT-cAMPS, representing a structurally different inhibitor of PKA. The data obtained (Table 1) confirmed that PKA inhibition is capable of inhibiting the ANP-induced desensitization of GC-A. Moreover, receptor desensitization was elicited when cells were preexposed to 8-Br-cAMP (Table 1), showing that PKA activation by a membrane-permeable cAMP analog can mimic the effects of ANP pretreatment.
|
|
|
). Near maximum effects were obtained at 10 µM LPA, and we selected this concentration for further experiments. Preincubations with LPA for different time periods demonstrated that LPA effects are rapidly evident (within few minutes) and further increase up to preexposure times of 30–60 min (Fig. 4B). More prolonged pretreatments were found to be less effective (not shown), in agreement with observations in case of LPA-induced desensitization of GC-B in NIH3T3 fibroblasts (12). When cells were preincubated for 30 min in either the absence or presence of LPA and then exposed to different levels of ANP (Fig. 4C), the desensitizing effect of LPA was detectable at all ANP concentrations (10–6 to 10–10 M) examined.
|
). As revealed by further assays carried out in the presence of Triton X-100/MnCl2 (see above), the LPA-elicited decline in ANP-dependent GC activity was not accounted for by a decrease in total receptor amount (Fig. 4E). Moreover, when membranes, prepared from untreated MA-10 cells, were incubated for 30 min in either the absence or presence (10 µM) of LPA before measuring ANP-dependent GC activity, we did not find any LPA-induced effects, ruling out a direct inhibitory interaction between LPA and GC-A (data not shown). Thus, these results demonstrate that: 1) LPA induces desensitization of GC-A; 2) Leydig cells possess molecular pathways capable of receiving and specifically mediating LPA signals; and 3) both ANP and LPA elicit equal decreases in GC-A hormone responsiveness.
Taking into account that the kinase inhibitor H89 completely blocked the ANP-induced desensitization of GC-A, we next examined whether H89 also affects the LPA-induced desensitization. These experiments clearly demonstrated that H89 does not inhibit the LPA-elicited desensitization (Fig. 5), indicating that PKA is not involved in this case.
Because ERK 1 and ERK 2 are known potential targets for LPA-mediated cellular effects (26, 35, 36), we next investigated whether LPA exposure to MA-10 cells alters the activity of these enzymes. LPA was found to induce ERK phosphorylation in a time-dependent manner (Fig. 6), and reactions were as rapidly (after 5 min) detectable as in case of LPA-induced GC-A desensitization (Fig. 4B). In contrast, ANP exposure for 30 min (Fig. 6) or other time periods examined (15, 60 min; not shown) did not induce any effects, compatible with findings that cGMP-dependent ERK phosphorylation requires the presence of PKG (37). To examine whether LPA-induced ERK phosphorylation may be functionally linked to GC-A desensitization, we performed experiments in the presence (40 µM) of the MEK inhibitor, PD 98059, previously shown to efficiently prevent LPA-induced ERK phosphorylation (26). These studies revealed that MEK inhibition blocks ERK phosphorylation (not shown) but does not protect against the LPA-mediated decrease in GC-A hormone sensitivity (Fig. 7). During this investigation, we noticed a certain inhibitory effect of PD 98059 alone on ANP-dependent cGMP production. The mechanism(s) underlying this effect remained unclear. However, the specific effect of LPA on GC-A activity was in the same range (inhibition by 40%) when incubations were performed in either the absence or presence of the MEK inhibitor (Fig. 7). Thus, the LPA-elicited desensitization of GC-A is not mediated by the MEK/ERK pathway.
|
|
T3-1) cell line (Fig. 8). However, and in contrast to T3-1 cells, MA-10 cells apparently do not express the LPA1 gene (Fig. 8). These findings indicate that LPA transduces signals into MA-10 cells by interactions with LPA2 and/or LPA3 but not through LPA1.
|
). The effects, evident within 20–30 min during LPA exposure, were found to be reversed after withdrawal of the agonist (data not shown), indicating dynamic processes. Because cell rounding belongs to the biological effects of LPA, which are predominantly observed in the nervous system (38), these findings were particularly remarkable and seem to support recently recognized neural characteristics of Leydig cells (40).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Data provided in the present and a previous (21) report also showed the absence of PKG I in isolated human Leydig cells. In contrast and representing apparently conflicting results, significant levels of PKG I, associated with human Leydig cells, have been detected in one earlier study (41). To address this important issue, we reexamined the original data of the latter study and recognized that both the immunohistochemical and immunoblot analyses of this investigation have been carried out with testes of elderly men exhibiting pronounced Leydig cell hyperplasia. This is of potential interest regarding recent findings that Leydig cells derive by transdifferentiation from vascular smooth muscle cells, which are characterized by high levels of PKG I expression (21). Because newly generated Leydig cells still display immunoreactivity for a number of marker proteins of their vascular progenitors (40), it is reasonable to assume that this also holds for PKG I. Therefore, the detection of PKG I immunoreactivity in this specific study could be explained by the existence of significant amounts of newly formed Leydig cells under conditions of hyperplasia. Like normal human adult Leydig cell populations (i.e. in the absence of hyperplasia), Leydig cells isolated from adult mice and rats did not show any expression of PKG I by immunoblot analyses performed in our laboratory (data not shown). Consistently a recent careful immunohistochemical analysis of cell type-specific expression of PKG I in the rat testis demonstrated the absence of this protein in Leydig cells (42). Thus, there is considerable evidence that PKG I deficiency is not specific for the MA-10 cell line but rather may represent a general characteristic of mammalian postmitotic Leydig cells.
In this context, we examined whether PKA rather than PKG could act as the functional target for ANP-induced cGMP in MA-10 Leydig cells. Such a possibility was strongly supported by previous studies in mouse Leydig cells, which revealed that: 1) ANP and cGMP, like gonadotropins, can induce testosterone production; 2) both gonadotropin- and ANP-induced steroidogenesis is inhibited by a cAMP antagonist; 3) this antagonist and cGMP competitively reduced [3H]cAMP binding to Leydig cell proteins; and 4) PKA is activated in lysates prepared from either ANP- or gonadotropin-stimulated cells (7). Our findings that the PKA inhibitor H89 completely abolished the ANP-induced desensitization of GC-A in MA-10 cells suggested that the receptor desensitization process was also mediated by a direct interaction between cGMP and PKA. Consistent with and confirming such a pathway, the desensitization reaction was blocked by a different PKA-specific inhibitor, Rp-8-CPT-cAMPS, and could be triggered by treatments with 8-Br-cAMP. Note in addition that the LPA-induced (i.e. cGMP independent) desensitization of GC-A was insensitive to PKA inhibition. Thus, the present findings substantially support the idea (reviewed in ref.29) that cGMP can signal via activation of PKA in vivo.
Theoretically, this may occur not necessarily via direct stimulation of PKA by cGMP (32) but could be relied also on cGMP-mediated inhibition of cAMP-hydrolyzing phosphodiesterases (31). However, considering that all our assays have been performed in the presence of IBMX to inhibit phosphodiesterase activity and that ANP did not induce increases in cAMP levels, the homologous desensitization of GC-A in MA-10 cells is thought to be mediated via cross-activation of PKA by high concentrations of cGMP. Several studies have demonstrated that ANP-induced cGMP can mimic cAMP-elicited effects in isolated Leydig cells (e.g.5, 7, 43) and in vivo (8), indicating that cross-activation of PKA by cGMP occurs in not only MA-10 but also native Leydig cells. Whether such cross talk between cGMP and cAMP pathways depend on (or at least are favored by) a cellular PKG deficiency (32) remains an interesting question. Recent findings, demonstrating that ANP/cGMP-elicited effects in liver cells are mediated by PKA in the absence of gene expression of both PKG isoforms (30), suggest that Leydig cells are not unique in this regard.
Because compartmentation of cyclic nucleotide signaling is now generally well established (44), it is reasonable to assume that homologous (i.e. cGMP mediated) desensitization of GC-A occurs in a compartmentalized manner. Such spatially confined functional units should comprise the ANP receptor, the cGMP/cAMP-activated kinase (PKA) and at least one PKA target protein finally implicated in GC-A desensitization/dephosphorylation. Signaling in discrete microdomains would also allow for the relatively high local cGMP levels required for cross-activation of PKA. In Leydig cells, but also in transfected 293 cells (45), PKG I is apparently not involved in regulating GC-A desensitization.
LPA-induced desensitization of GC-A
This study demonstrates for the first time that LPA can induce desensitization of GC-A. Considering that previous publications have exclusively reported on GC-B desensitization by LPA (12, 33) or sphingosine-1-phosphate (13, 34), the present results establish that interactions between phospholipid and natriuretic peptide signaling are not a GC-B-specific phenomenon and occur in other cell types than those (fibroblasts, aorta smooth muscle cells) described so far.
LPA is a major constituent of serum, has diverse and potent activities on many different mammalian cell types, and is thought to elicit its effects in a membrane receptor-dependent manner (14). The elucidation of ERK phosphorylation and cell rounding in response to LPA exposure hence provided evidence for LPA receptor-mediated activities (26, 36) in MA-10 cells, consistent with the identification of cellular LPA receptor gene expression. The LPA dose dependency of the effects on GC-A activity (Fig. 4A) is also compatible with a receptor-mediated pathway. In addition, LPA-induced desensitization of GC-A was as rapidly initiated as LPA-induced ERK phosphorylation (both reactions were detectable already after 5 min of LPA treatment), and GC assays with cell membranes demonstrated the absence of a direct functional interaction between GC-A and LPA. In addition, it should be noted that the LPA-elicited decrease in GC-A hormone responsiveness is not explainable by a general loss of cellular activities under LPA treatment because agonist-stimulated steroid synthesis in MA-10 cells is not inhibited by LPA (46).
Experiments performed in the presence of kinase inhibitors revealed that neither PKA nor the MAPK/ERK pathway is functionally involved in the process of LPA-induced GC-A desensitization. Based on known LPA receptor-mediated signaling routes (14), all three cellular effects observed (ERK phosphorylation, cell rounding, receptor desensitization) may rely on different pathways, involving stimulation of the mitogenic Ras/MAPK cascade (ERK), activation of RhoA (cell rounding), or stimulation of phospholipase C with subsequent calcium mobilization (receptor desensitization). Whereas a contribution of protein kinase C in the process of receptor desensitization seems dispensable (1), intracellular calcium elevations have been clearly linked to LPA-induced desensitization/dephosphorylation of GC-B and proposed to act as a universal mechanism for heterologous desensitization of both GC-A and GC-B (33). Such a mechanism, although not specifically addressed in the present study, would be compatible with the data for LPA-induced desensitization of GC-A in MA-10 cells. Generally calcium increase is a well-established cellular response mediated by LPA receptors, and, unlike other LPA receptor-mediated pathways, it can be elicited by each one of the three receptor subtypes (38). Based on the RT-PCR data, both LPA2 and LPA3 are therefore candidate receptors for mediating the LPA-induced GC-A desensitization in MA-10 cells. More importantly, the present findings indicate that the MA-10 cell line might represent a useful model to study distinct cellular effects of LPA in the absence of LPA1.
As observed with ANP-induced desensitization, the LPA-elicited decrease in hormone-dependent GC-A activity was verified to an equal extent each in experiments with intact MA-10 cells and GC assays with membrane preparations, whereas basal activities and receptor levels remained unaltered. Thus, these findings are consistent with other investigations (47, 48), indicating that GC-A desensitization depends on reactions directly affecting the receptor hormone responsiveness and is not due to receptor endocytosis and/or degradation.
Comparison between homologous and heterologous desensitization
Both the ANP- and LPA-elicited effects observed in this study are based on endogenously expressed signaling components, allowing comparative characterizations in an in vivo-like context. Importantly, the selective inhibition of ANP- but not LPA-induced desensitization by H89 provided unequivocal evidence for distinct pathways. On the other hand, the time dependency of desensitization in response to both agents was remarkably similar. Of particular interest, however, was the observation that both agents finally caused equal decreases (by 40% each) in ANP-dependent GC-A activity. Although not proved experimentally in the present study, there is now, however, convincing evidence that time-dependent reduction in ligand-dependent GC-A activity is based on receptor dephosphorylation (1). Considering that the degree of receptor phosphorylation directly correlates with its hormone responsiveness (49), this would imply that ANP and LPA evoked an equal extent of dephosphorylation each. In this regard, the present data seem to be incompatible with previous findings (reviewed in ref.1) that homologous desensitization is based on a broad (global) dephosphorylation of the receptor, whereas the heterologous process results in the loss of only a single or few phosphate residue(s).
The observation that different signaling pathways resulted in a similar extent of GC-A desensitization raised interesting questions. In particular, we asked whether the same phosphatase or distinct phosphatases could be finally activated in these processes. However, experiments carried out with phosphatase inhibitors failed to provide conclusive results. Desensitization assays in the presence of okadaic acid (data not shown) did not demonstrate inhibitory effects on either ANP- or LPA-induced GC-A desensitization at concentrations of 0.1 nM (inhibits protein phosphatase (PP)2A and PP4, ref.50) or 20 nM (inhibits additionally PP1 and PP5) (50). Analogous experiments performed in the presence of microcystin-LR (effective on the same four phosphatases) (50) consistently failed to reveal any inhibition of homologous or LPA-induced desensitization. Thus, these results suggest that PP1, PP2A, PP4, or PP5 are not functionally involved as GC-A-dephosphorylating enzymes during both (homologous and heterologous) desensitization processes. However, clear-cut results require the identification of phosphatases, activated in response to ANP and/or LPA treatment. In addition, one has to consider the alternative possibility that the desensitization process primarily relies on an inhibition of GC-A phosphorylation rather than an activation of GC-A dephosphorylation. In such a scenario, the final target enzyme within the signaling cascade would be a (not yet identified) kinase acting on GC-A as substrate in vivo. As can be inferred from our findings that treatments with 8-Br-cAMP induce desensitization, this kinase should not be PKA.
Potential relevance for Leydig cell physiology
The Leydig cells of the testis act to produce testosterone and are distinguished by remarkable neuroendocrine properties (40). Immortalized mouse Leydig tumor MA-10 cells are gonadotropin responsive and well established as an experimental model in studies of Leydig cell physiology (16). Gonadotropins, using cAMP-mediated pathways, are the principal regulators of Leydig cell steroidogenesis. However, ANP-induced stimulation of GC-A in Leydig cells has been reported to increase testosterone levels in vitro (e.g.5, 7) and in vivo (8), indicating an additional, modulatory role. In this context, our findings that GC-A is desensitized during prolonged exposure to ANP identify a control mechanism by which signal intensity and cellular response can be attenuated. Such a mechanism could play a regulatory role not only in adulthood but also during postnatal developmental stages (4) and in the fetal (51) testis.
ANP failed to increase ERK phosphorylation in MA-10 cells. On the other hand, our data cannot exclude inhibitory rather than activating effects. In fibroblasts, CNP-induced accumulations of cGMP were found to block activation of the MAPK cascade (52), and several studies revealed analogous effects generated by ANP (39, 53). Physiologically these natriuretic peptide/cGMP-induced activities contribute to the control of cell growth (39, 53), and analogous functions, in addition to those implicated in regulation of steroidogenesis, are conceivable also for ANP signaling in Leydig cells. In this regard, it remains to be addressed whether Leydig cell hyperplasia, representing a frequently observed phenomenon in testes of elderly men (40), might be related to increases in ANP plasma levels during aging (54).
LPA was found to elicit diverse responses in MA-10 cells. To our knowledge, this is the first report on activities of LPA in this cell type, revealing that Leydig cells must be regarded as targets for phospholipid signaling in vivo. Consistently, the LPA concentrations effective in our experiments were similar to those (1–5 µM) estimated in human serum (55).
Whether the specific effects of LPA on GC-A activity play a role in the context of regulation of steroidogenesis or are related to other, not-yet-recognized functions for ANP-induced cGMP in Leydig cells remains to be elucidated. In contrast to ovarian luteal cells, LPA fails to inhibit cAMP-stimulated steroid production in MA-10 cells (46). Most remarkably, the morphoregulatory activities of LPA in MA-10 cells suggest a role in controlling the phenotype/differentiation of these cells. In this context, it has to be noted that Leydig cells are characterized by the occurrence during postnatal development of three distinct phenotypes, including hypertrophied and so-called spindle-shaped forms. They all derive from the same (vascular) progenitor cell type but in a characteristic spatiotemporal pattern (40), indicating local environmental effects. Based on the activities observed in this study, it is conceivable that LPA signaling is functionally involved in regulating such processes.
In conclusion, this study demonstrates for the first time that two desensitization pathways for GC-A are operative within the same cell system, that ANP-induced desensitization does not require PKG but is mediated by PKA, and that LPA can elicit desensitization of this natriuretic peptide receptor subtype. The identification of both homologous and heterologous desensitization pathways for its receptor, GC-A, supports a role of ANP in Leydig cells. This investigation in addition provides evidence that LPA has to be considered as a novel and important signaling molecule for this cell type.
|
Acknowledgments |
|---|
T3-1 cell line. We in addition thank Professor Freimut Leidenberger for his outstanding support of scientific activities in Hamburg. The excellent technical assistance by Sabine Tasch is gratefully acknowledged. This publication is dedicated to Professor Michail S. Davidoff on the occasion of his 65th birthday. |
Footnotes |
|---|
Current address for A.K.M.: ELGA Biotech, Elbgaustrasse 71, D-22523 Hamburg, Germany.
D.M., L.C.-D., L.T.B., B.B.-S., M.P., R.C.S., A.K.M., and R.M. have nothing to declare.
First Published Online March 9, 2006
Abbreviations: ANP, Atrial natriuretic peptide; BNP, brain natriuretic peptide; 8-Br-cAMP, 8-bromo-cAMP; CNP, C-type natriuretic peptide; DAPI, 4'6'-diamidino-2-phenylindole; GC, guanylyl cyclase; IBMX, 1-methyl-3-isobutyl xanthine; LPA, lysophosphatidic acid; MEK, MAPK kinase; NPR, natriuretic peptide receptor; PKA, cAMP-dependent protein kinase; PKG, cGMP-dependent protein kinase; PP, protein phosphatase; Rp-8-CPT-cAMPS, 8-(4-chlorophenylthio)adenosine-3', 5'-cyclic monophosphorothioate, Rp-isomer; TRFIA, time-resolved fluorescence immunoassay.
Received January 25, 2006.
Accepted for publication March 2, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T3-1 cell line. Endocrine 2:849–856
