| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Urology Research Laboratory, Royal Victoria Hospital and Faculty of Medicine, McGill University, Montréal, Québec, Canada H3A 1A1
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cyclic adenosine monophosphate, gamete biology, phosphodiesterases, sperm
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
PDEs differ in their primary structure, kinetics, and mechanisms of regulation, as well as substrate and inhibitor specificities. PDEs belong to at least 11 identified families [8, 9]. PDEs of each family are highly specific for the hydrolysis of either cAMP (PDE4, PDE7, and PDE8), cGMP (PDE5, PDE6, and PDE9), or have mixed specificities and hydrolyze both cAMP and cGMP (PDE1, PDE2, PDE3, PDE10, and PDE11). The activity of these PDEs is regulated by different mechanisms [8, 10]. For example, availability of cAMP stimulates the activity of PDE4, whereas cGMP inhibits cAMP hydrolysis by PDE3. Moreover, the interaction with calcium and calmodulin stimulates PDE1 activity, whereas allosteric binding of cGMP promotes cAMP breakdown by PDE2. Finally, PDE1, PDE3, and PDE4 activities are regulated by feedback mechanism involving a cAMP-dependent PKA phosphorylation.
In spermatozoa there is evidence for the presence of more than one isoform of PDEs. Measurement of sperm PDE activity in the presence of inhibitors for PDE1, PDE4, and PDE5 in humans [5, 7] or of stimulators for PDE1 (calcium and calmodulin) in mice [11] and bovine [12] suggested the presence of PDE1 and PDE4, but not of PDE5 in spermatozoa. The expression of PDE1 and PDE4 in mice [13, 14] and rat [15, 16] testis germ cells support the preceding data. Recently, mRNA transcripts of six types of PDE have also been identified in ejaculated human spermatozoa [17].
Bovine and murine sperm PDE activities are stimulated by calmodulin [11, 12]. Calmodulin is a calcium-binding protein present at high concentrations in spermatozoa [18, 19] and modulates the activity of a variety of key enzymes such as adenylyl cyclase [20], protein phosphatase 2B (calcineurin), [21] and other sperm proteins (calmodulin-binding proteins) of unknown role [22]. Calmodulin is localized on the sperm head and flagellum [18, 19] and may be involved in hamster and mouse sperm motility and capacitation [23, 24].
Very little is known about PDEs in ejaculated human spermatozoa, and only the effect of PDE inhibitors on sperm functions was investigated [3, 5, 7]. Lefièvre et al. [7] observed that the rate of cAMP hydrolysis by sperm PDEs is about 3-fold higher than that of cGMP, indicating that PDEs present in human sperm have higher affinity for cAMP. Because cAMP is an important second messenger in the control of sperm functions, it appeared important to identify PDEs such as PDE1 (calmodulin-dependent PDE) and PDE3 (cGMP-inhibited PDE) in these cells. The aims of the present study were to 1) ascertain the presence of PDE1 and PDE3 (enzyme activity, immunocytochemical localization, and immunoblotting) in human spermatozoa, 2) to examine whether PDE3 is involved in human sperm functions, 3) partially purify sperm PDEs, and 4) to study calmodulin regulation of PDE1 activity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Lysophosphatidylcholine (LPC), 3-isobutyl-1-methylxanthine (IBMX), BSA, Pisum sativum agglutinin conjugated to fluorescein isothiocyanate (PSA-FITC), 1,4-diazabicyclo-[2.2.2] (dabco), milrinone, trifluoperazine (TFP), Crotalus atrox venom, Dowex 1X8 (200–400 mesh), Tween-20, Triton X-100, Nonidet P-40 (NP-40), CHAPS, and poly-L-lysine were purchased from Sigma Chemical Company (St. Louis, MO). Monoclonal anti-phosphotyrosine (clone 4G10) and anti-calmodulin (both from Upstate Biotechnology Inc., Lake Placid, NY) antibodies, polyclonal anti-PDE1A, and anti-PDE3A antibodies (a generous gift of Dr. J.A. Beavo; Seattle, WA), goat anti-mouse (Gibco BRL, Burlington, ON, Canada) and donkey anti-rabbit (Jackson Immunoresearch Laboratories, West Grove, PA) immunoglobulin G (IgG) conjugated with horseradish peroxidase, nitrocellulose (0.22 µm pore size; Osmonics Inc., Westborough, MA), an enhanced chemiluminescence kit (Lumi-Light; Roche Molecular Biochemicals, Laval, QC, Canada) and x-ray films (Fuji, Minami-Ashigara, Japan) were used for immunodetection of sperm proteins. Biotinylated goat anti-rabbit IgG (Pierce, Rockford, IL), avidin conjugated with Cy3 and mounting media (Prolong Antifade; Molecular Probes, Eugene, OR) were used for immunocytochemistry. 4',6-Diamidine-2'-phenylindole-6-dihydrochloride (Dapi), aprotinin, leupeptin, pepstatin A, benzamidine, and PMSF were purchased from Roche Molecular Biochemicals. Collodion was purchased from JBS-Supplies (Dorval, QC, Canada). Percoll, [8-3H]cAMP, [8-3H]cGMP, and the Mono Q anion-exchange column (HR 5/5) were purchased from Amersham Pharmacia Biotech (Baie d'Urfé, QC, Canada). Other chemicals used were of at least reagent grade. The specific antiserum against cAMP (CV27) was obtained through the National Hormone and Pituitary Program of the National Institute of Diabetes and Digestive and Kidney Diseases (Bethesda, MD), and Dr. A.F. Parlow (Torrance, CA).
IBMX and milrinone were dissolved in dimethyl sulfoxide (DMSO), then diluted in an incubation medium so that the concentration of DMSO did not exceed 0.1% (v/v), an amount that did not influence sperm motility or capacitation. Other chemicals were dissolved in distilled water.
Sperm Preparation and Treatments
Semen samples were obtained by masturbation from healthy volunteers after 3 days of sexual abstinence. After liquefaction, semen was layered on top of a gradient composed of 4 Percoll layers (20%, 40%, 65%, and 95%) made isotonic and buffered in Hepes balanced saline (HBS; 115 mM NaCl, 4 mM KCl, 0.5 mM MgCl2, 14 mM fructose, and 25 mM Hepes pH 8) as described previously [7]. The samples were centrifuged for 30 min at 2300 x g, and spermatozoa were recovered from the 65%–95% and within the 95% Percoll layer. Spermatozoa were pooled and diluted to 500 x 106 cells/ml with the 95% Percoll solution. Only samples with progressive sperm motility >70% were used.
Phosphodiesterase Assay
PDE activity was measured using the modified 2-stage radioisotopic method of Sonnenburg et al. [25]. Briefly, the standard incubation mixture (pH 7.5) contained 20 mM Tris-HCl, 20 mM imidazole, 3 mM MgCl2, 15 mM magnesium acetate, 0.2 mg/ml BSA, and 1 µM cAMP or 1 µM cGMP. For the determination of PDE1 activity, 2 mM EGTA, 1.5 mM CaCl2 + 4 µg/ml calmodulin or 45 µM TFP (a calmodulin antagonist) were added. For the determination of PDE3 activity 2 mM EGTA and 50 µM milrinone (a specific PDE3 inhibitor) or 1 µM cGMP was added. The reaction was started by the addition of 50 000 cpm of [8-3H]cAMP or [8-3H]cGMP and incubated for 10 min at 30°C. The reaction was stopped by heat denaturation at 100°C for 1 min, and the 5'-nucleotide produced was then converted to the corresponding nucleoside by a second incubation of 5 min at 30°C with 2.5 mg/ml of Crotalus atrox venom. The reaction products were separated by anion-exchange chromatography on Dowex 1X8, and the amount of unbound [3H] nucleoside was quantified by liquid scintillation counting. Results were reported as picomoles of cyclic nucleotides hydrolyzed per minute per 108 spermatozoa.
Solubilization of Sperm PDE
Percoll-separated spermatozoa were washed twice (5 min at 1000 x g) with HBS to remove Percoll. Pellets were resuspended to a concentration of 80 x 106 cells/ml and incubated in a hypotonic buffer (50 mM Tris-HCl pH 7.5, 2 mM EDTA, 1 mM dithiothreitol [DTT], 5 µg/ml leupeptin, 5 µg/ml pepstatin, 10 µg/ml aprotinin, and 200 µM PMSF). The suspension was then either 1) homogenized (30 strokes) in a Dounce homogenizer on ice, 2) sonicated for 1 min at 50 V on ice, 3) extracted with 0.1% (v/v) Triton X-100, (v/v) NP-40 or (w/v) CHAPS, or 4) extracted twice with 0.01% (v/v) Triton X-100. The different resulting preparations were centrifuged at 100 000 x g at 4°C for 30 min. The supernatants (soluble fractions) were collected and the pellets (particulate fractions) were resuspended in an equal volume of hypotonic buffer containing protease inhibitors, and both fractions were assayed for PDE activity as described earlier or diluted in solubilization buffer (final concentrations: 2% SDS, 10% glycerol, 1.4% DTT, 62.5 mM Tris-HCl pH 6.8, 5 µg/ml leupeptin, 5 µg/ml pepstatin, 10 µg/ml aprotinin, 200 µM PMSF, and 1 mM EDTA) for SDS-PAGE and immunoblotting analysis.
SDS-PAGE and Immunoblotting
Proteins were separated on SDS-PAGE and electrotransferred onto nitrocellulose membrane and immunodetected as described before [7]. For detection of phosphotyrosine-containing proteins, spermatozoa at 20 x 106 cells/ml were incubated in Biggers, Whitten, and Whittingham medium (BWW; pH 8) [26] modified as described in Lefièvre et al., [7] alone, or supplemented with 500 µM IBMX or 50 µM milrinone for 2.5 h at 37°C. PDE1A and PDE3A isoforms present in fractionated spermatozoa as well as in anion-exchange chromatography fractions were detected by polyclonal anti-PDE1A and anti-PDE3A antibodies. In some cases, electrophoresis was also performed under nonreducing and nondenaturating conditions (no DTT, no SDS). Positive immunoreactive bands were detected by enhanced chemiluminescence after incubation of the blots with horseradish peroxidase-conjugated goat anti-mouse or donkey anti-rabbit antibodies. Silver staining of the proteins transferred on the nitrocellulose membrane was performed after the immunodetection [27].
Immunocytochemistry
Percoll-separated spermatozoa were washed in PBS (128 mM NaCl, 2 mM KCl, 8 mM Na2HPO4, 2 mM KH2PO4, and 1 mM sodium azide pH 7.5) and dispersed on microscope slides previously treated with poly-L-lysine and air-dried for 5–10 min. Spermatozoa were then fixed in methanol for 6 min at -20°C, rehydrated with 50% (v/v) PBS-methanol for 5 min, and washed 3 times (5 min each) with PBS containing 0.1% Triton X-100 (TPBS). To minimize nonspecific binding, sperm were preincubated with 5% (v/v) goat serum in TPBS for 30 min before incubation with the primary anti-PDE1A and anti-PDE3A antibodies. After extensive washing with TPBS, sperm were further incubated for 1 h with a biotinylated goat anti-rabbit antibody and then for 45 min with avidin conjugated to Cy3 (0.9 µg/ml in TPBS). Following extensive washing with TPBS, slides were finally mounted in Prolong Antifade containing 0.3 µg/ml of Dapi to stain nuclei. The specificity of the antibodies was determined in an additional experiment in which antibodies were preadsorbed with the GST-fusion proteins [25] corresponding to the C-terminal epitopes of PDE1A and PDE3A isoforms (about 45 amino acids). Spermatozoa showed no immunoreactive staining (Fig. 2F).
|
Evaluation of Sperm Capacitation and Hyperactivation
Spermatozoa were resuspended at 20 x 106 cells/ml in BWW. Spermatozoa were incubated in BWW alone (control) or supplemented with 500 µM IBMX or 50 µM milrinone. Sperm capacitation was evaluated after 3.5 h of incubation by the LPC-induced acrosome reaction (3 mg/ml BSA and 100 µM LPC) as previously described [28]. The acrosomal status was evaluated using PSA-FITC [7]. At least 200 spermatozoa were evaluated in each sample. The level of spontaneous capacitation occurring in spermatozoa incubated in BWW alone was subtracted from those obtained in spermatozoa treated with IBMX or milrinone.
Sperm hyperactivation was measured after 90 min of incubation at 37°C using the CellSoft computer-assisted digital image analysis system (Cryo Resources, Montgomery, NY). Spermatozoa were placed between Collodion-coated slides and coverslips and maintained at 37°C. Motility criteria for sperm hyperactivation were as follows: curvilinear velocity of at least 100 µm/sec, linearity lower than 0.6, and amplitude of the lateral head displacement at least 6.5 µm [7]. Parameters for at least 200 motile spermatozoa per sample were assessed. The percentage of hyperactivation obtained for spermatozoa incubated in BWW alone was subtracted from that obtained in spermatozoa treated with IBMX or milrinone.
Cyclic AMP Measurements
Spermatozoa were incubated for 30 min at 37°C in BWW alone or supplemented with 500 µM IBMX or 50 µM milrinone. Intracellular cAMP was extracted with ice-cold 90% ethanol and measured by radioimmunoassay using [125I]cAMP as described previously [7]. The level of cAMP in spermatozoa incubated in BWW alone was subtracted from that obtained in spermatozoa treated with IBMX or milrinone.
High-Performance Anion-Exchange Chromatography
The combined sperm soluble fractions of 3 or 4 semen samples extracted twice with 0.01% Triton X-100 (see above) were loaded at 0.5 ml/min on a Mono Q anion-exchange column HR 5/5, previously equilibrated with buffer A (50 mM Tris-HCl pH 7.5, 2 mM EDTA, 2 mM EGTA, 1 mM DTT, 0.01% Triton X-100, 1 mM benzamidine, 5 mM leupeptin, and 100 µM PMSF) [25]. Under these conditions, more than 95% of the total PDE activity was bound to the column. The column was washed with buffer A for 20 min or until the absorbance at 280 nm returned to the baseline value. The bound PDE activity was eluted at a flow of 0.5 ml/min with a linear gradient (20%–50%) of buffer B (buffer A + 0.8 M NaCl) over 40 min. Fractions of 0.25 ml were collected and kept at 4°C. Each fraction was assayed for PDE activity and some samples were boiled in sample buffer for immunoblotting.
Data Analysis
ANOVA (two-tailed, paired values) was used to test the differences in PDE activity, levels of capacitation, cAMP, and hyperactivation between control and treated spermatozoa. A difference was considered statistically significant at P 
0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sonication, homogenization, and detergent treatments were investigated in an attempt to solubilize sperm PDE activity. The 100 000 x g soluble and particulate fractions of spermatozoa obtained after these treatments were tested both with cAMP and cGMP as substrates (Table 1). As previously described [7], sperm PDE activity was 1.5-fold to 2-fold greater when using cAMP rather than cGMP. Very low PDE activity was recovered in the soluble fraction of human spermatozoa after homogenization or sonication.
|
The total PDE activity recovered after addition of detergents to spermatozoa was significantly higher than that obtained after homogenization or sonication for both substrates (Table 1). Triton X-100 and NP-40 (both at 0.1%), two nonionic detergents, extracted 50%–65% of sperm PDE activity, leaving 30%–50% in the particulate fraction (Table 1). The presence of CHAPS, a zwitterionic detergent, caused a similar increase in total PDE activity but did not solubilize as much PDE (10%–25%) as nonionic detergents. Two consecutive extractions with 0.01% Triton X-100 were as efficient as one with 0.1% Triton X-100 or NP-40 to solubilize PDE activity toward both cAMP and cGMP (Table 1). Because NP-40 has a 3-fold to 5-fold lower critical micelle concentration than Triton X-100, we extracted sperm PDE twice with 0.01% Triton X-100 for subsequent experiments, particularly for purification of sperm PDEs by high-performance chromatography (see below).
PDE1 and PDE3 Activity in Triton X-100 Soluble and Particulate Fractions of Human Spermatozoa
PDE1 and PDE3 activities were present in both Triton X-100 soluble and particulate fractions of human spermatozoa (Fig. 1). PDE1 activity is activated by calcium + calmodulin and inhibited by EGTA [25]. Calcium + calmodulin increased PDE hydrolysis of cAMP in the soluble fraction by about 20%, but no stimulation was observed when cGMP was used as substrate (Fig. 1, A and B). In the particulate fraction, the presence of calcium + calmodulin increased the hydrolysis of both cAMP and cGMP by about 20% and 60%, respectively (Fig. 1, A and B). TFP, an antagonist of calmodulin, was used to determine whether the EGTA added in the assay mixture was sufficient to inhibit all PDE1 activity. PDE activity toward cAMP was not affected by TFP, whereas PDE activity toward cGMP contained in both soluble and particulate fractions were inhibited by an additional 33% and 22%, respectively, as compared to basal (EGTA-insensitive) PDE activity (Fig. 1, A and B). Moreover, addition of EGTA did not inhibit total sperm PDE activity (no addition of EGTA or inhibitors), and addition of calcium alone (0.2 mM to 5 mM of calcium) did not stimulate the activity (data not shown).
|
Inhibition of PDE hydrolysis by milrinone is indicative of the presence of PDE3. Milrinone inhibited about 10% and 25% of PDE hydrolysis of cAMP, measured in the presence of EGTA, from soluble and particulate fractions, respectively. However, cGMP, used as a competitive inhibitor for PDE3, had no effect on PDE activity toward cAMP (Fig. 1C).
Cellular Localization and Identification of PDE1A and PDE3A in Human Spermatozoa
Immunocytochemistry experiments indicated that PDE1A was localized in the mid and principal pieces of the flagellum as well as in the equatorial segment of the sperm head (Fig. 2D). PDE3A was localized in the postacrosomal segment of the sperm head (Fig. 2E). Spermatozoa incubated with only the secondary biotinylated goat anti-rabbit antibody or with antibodies preadsorbed with the GST-fusion proteins showed no immunoreactive staining (Fig. 2F).
Immunoblotting experiments indicated that anti-PDE1A antibody recognized a band at 67 kDa as well as other protein bands (Fig. 2G) that could result from partial proteolysis because the sperm acrosome is a rich source of proteases. The 67-kDa band was found in both soluble and particulate sperm fractions. The anti-PDE3A antibody recognized a 97-kDa protein band (Fig. 2G), which was present mostly in the soluble fraction. The 97-kDa band identified as PDE3A was also recognized by the anti-PDE1A antibody (Fig. 2G), possibly because PDEs are similar between each other, the antibodies are polyclonal, and they were raised against mouse PDE.
Effect of Milrinone on Sperm Capacitation and Associated Events
Because cAMP is a key second messenger in sperm capacitation, protein tyrosine phosphorylation and motility [1, 28], the involvement of PDE3 in these sperm functions was evaluated with the use of milrinone (Fig. 3). IBMX, a nonspecific PDE inhibitor, was used as a positive control. This inhibitor consistently increased levels of cAMP, capacitation, hyperactivation, and protein tyrosine phosphorylation (Fig. 3). However, whereas milrinone increased the level of cAMP (Fig. 3A) it did not affect sperm functions. The increase in cAMP observed with milrinone was lower than that obtained with IBMX. On the other hand, the effects of PDE1 on sperm capacitation and associated events could not be evaluated because specific PDE1 inhibitors are not available [25, 29].
|
Anion-Exchange Chromatography of Triton X-100 Soluble Human Sperm Extract
A partial purification of PDE contained in the Triton X-100 soluble fraction of human spermatozoa was achieved by high-performance anion-exchange chromatography (Fig. 4). More than 95% of the total PDE activity was bound to the column in the initial loading. The elution profile showed a single peak of PDE activity that was greater for hydrolysis of cAMP than of cGMP (Fig. 4). Moreover, this peak of PDE activity corresponded to only 3% of the proteins present in the Triton X-100 soluble sperm extract. Analysis of the eluted fractions for PDE activity and immunoblotting confirmed the presence of PDE1A and PDE3A isoforms. No stimulation was observed in the presence of calcium + calmodulin with either cAMP or cGMP as substrates, and as previously described, addition of EGTA did not inhibit total sperm PDE activity (data not shown), but PDE activity was inhibited by about 10% by milrinone (Fig. 4). In addition, the 67-kDa band, PDE1A, and the 97-kDa band, PDE3A, were the only proteins that were immunodetected in the fractions obtained from the anion-exchange chromatography.
|
PDE Interaction with Calmodulin
PDE1A activity present in fractions from the anion-exchange chromatography (Fig. 4) was not stimulated by calcium + calmodulin, as expected for this PDE family, suggesting that calmodulin could stay bound to PDE1A throughout the separation process. To test this hypothesis, chromatographic fractions were submitted to nonreducing and nondenaturating electrophoresis, allowing the proteins to remain under their native forms. Western blots were probed with an anti-calmodulin antibody (Fig. 5) to identify calmodulin bound on proteins. The anti-calmodulin antibody recognized a doublet of protein bands of 67 and 68 kDa (Fig. 5) as well as unbound calmodulin at 17 kDa (data not shown). The 67-kDa protein band corresponded to that detected with the anti-PDE1A antibody (Fig. 5). The same doublet of proteins was also detected by the anti-calmodulin antibody when partial purification of PDEs was performed by gel filtration chromatography (Superdex-200; data not shown) instead of a anion-exchange chromatography.
|
Attempts were made to dissociate the complex PDE1A-calmodulin. Fractions 3 to 7 of the anion-exchange chromatography were pooled, concentrated, and treated with either EGTA + EDTA, urea, or TFP and subjected to SDS-PAGE (Fig. 6). These different treatments did not affect the association of PDE1A with calmodulin (Fig. 6), the doublet of proteins (67 and 68 kDa) recognized by the anti-calmodulin antibody was still being observed with the same intensity. As expected, PDE1A immunodetection (Fig. 6) was unaffected by these treatments. In another experiment, the anti-calmodulin antibody was adsorbed with calmodulin and then used for immunoblotting. No immunoreactive band was detected either at 17 kDa (data not shown) nor at 67 and 68 kDa (Fig. 6), demonstrating the specificity of the anti-calmodulin antibody for calmodulin and the doublet of proteins, 67 and 68 kDa, existing in a complex containing calmodulin.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Lefièvre et al. [7] observed that the rate of cAMP hydrolysis in sperm soluble and particulate fractions was greater than that of cGMP, which is probably due to the greater affinity for cAMP as substrate than cGMP in human spermatozoa. We previously reported that PDE hydrolysis of cAMP after sperm homogenization was recovered more in the cytosol (60%) than in membrane (30%) and insoluble (10%) fractions [7]. In the present study, we obtained a low recovery of PDE activity in soluble fractions (cytosol) of homogenized and sonicated spermatozoa (Table 1). This discrepancy between current and past results could be explained by the absence of BSA in our homogenization buffer. BSA is considered as a cholesterol-binding molecule [30] that could remove membrane-bound proteins, such as PDEs, while releasing this lipid from plasma membrane. Therefore, the higher PDE activity found in the cytosol fraction in our previous report could be due to a partial solubilization of PDEs due to the presence of BSA. When spermatozoa were treated with detergents, the total PDE activity recovered was higher than that measured from homogenized or sonicated spermatozoa (Table 1), suggesting that detergent is needed to either activate PDEs or release them to be available to the substrates. When spermatozoa were treated with CHAPS, 10%–25% of PDE activity was solubilized, whereas NP-40 and Triton X-100 extracted 50%–65% of PDE activity (mixture of cytosol and plasma membrane; Table 1). Zwitterionic detergents, such as CHAPS, extract peripheral membrane proteins, whereas nonionic detergents, such as NP-40 and Triton X-100, extract integral membrane proteins. These results strongly indicated that a detergent is needed to solubilize sperm PDEs, that most of the PDE activity is associated with the plasma membrane and that most of membrane PDEs are integral proteins.
PDE1 and PDE3 activities were detected in both the soluble and particulate fractions of human spermatozoa after Triton X-100 extraction (Fig. 1). Immunoblotting experiments indicated that PDE1A was distributed equally between the supernatant and particulate fractions, whereas PDE3A was mainly detected in the soluble fraction (Fig. 2G) of spermatozoa. The subcellular immunolocalization of PDE1A and PDE3A in the equatorial and postacrosomal region of the sperm head, respectively (Fig. 2, D and E), is consistent with a role in the regulation of sperm membrane changes, which are important for capacitation, the acrosome reaction, or both. PDE1A location in the mid and principal pieces of the flagellum (Fig. 2D) is also consistent for a role in sperm motility, hyperactivation, or both.
It was reported that an inhibitor of PDE1, 8-methoxy-IBMX, promoted the acrosome reaction in human spermatozoa [5]. However, because the specificity of this PDE1 inhibitor and of others is questioned [25, 29], their effect on sperm functions was not evaluated in the present study. However, the involvement of PDE3 in human sperm functions was tested using milrinone, a second generation of PDE inhibitors that is selective for PDE3 [31]. Milrinone has a IC50 value of 0.5 µM, a value that is 10-fold to 100-fold lower than for other PDE families [31]. Milrinone increased sperm cAMP levels in spermatozoa but this increase was probably not sufficient to induce capacitation, hyperactivation, or protein tyrosine phosphorylation of two fibrous sheath proteins (p105/81; Fig. 3) [3]. PDE3 activity corresponded to only 10% and 25% of total PDE activity of Triton X-100 soluble and particulate fractions, respectively (Fig. 1C). In addition, the increases in cAMP obtained with milrinone were lower than those caused by IBMX treatment (Fig. 3). One explanation could be that a higher level of cAMP than that caused by milrinone is needed to induce changes in sperm functions. This would suggest that inhibition of other PDEs than PDE3 is necessary to obtain a sufficient level of cAMP to induce an effect on sperm capacitation, motility, and protein tyrosine phosphorylation. However, because PDE3 is bound to a specific sperm structure (the equatorial segment; Fig. 2E), it is also conceivable that the increase in cAMP resulting from its inhibition would be only local and targeted for other actions. Spermatozoa are highly differentiated cells, and PDEs may contribute to signal compartmentalization by controlling the diffusion of the second messenger. The association of PDE4 with A-kinase anchoring proteins in cardiomyocytes [32], and the presence of structural hydrophobic membrane association domains on the N-terminal half of PDE3 [33] support this hypothesis.
Partial purification of PDEs from the Triton X-100 soluble fraction of human spermatozoa was achieved by anion-exchange chromatography. A single peak of PDE activity, which correspond to only 3% of the Triton X-100 soluble protein extract, was eluted, suggesting that these PDEs have similar charges. Moreover, these PDEs preferentially hydrolyzed cAMP (Fig. 4). Immunoblotting experiments indicated the presence of PDE1A even though no stimulation by calcium + calmodulin (Fig. 4), inhibition by EGTA, or activation by calcium (data not shown) was observed. Furthermore, both imunoblotting and inhibition of PDE activity by milrinone confirmed the presence of PDE3 (Fig. 4). PDE activity was inhibited by about 15% by milrinone, which corresponds to the inhibition observed for Triton X-100 soluble fraction from the initial preparation (Fig. 1C).
The presence of calmodulin-dependent PDE, PDE1, in spermatozoa was, for a long time, controversial. Tests were performed in spermatozoa from different species and failed to indicate the activation of PDE with calcium + calmodulin [34–36]. Some possible explanations for the lack of activation by calcium + calmodulin in spermatozoa are that 1) mild proteolysis renders PDE1 independent to calcium + calmodulin [22]; 2) detergents used to solubilize PDEs from membranes could also inhibit the calcium + calmodulin activation [37, 38]; and 3) calmodulin is tightly bound to plasma membrane [39], PDE1 [40], or its concentration is in excess to that required for activation of PDE1 in spermatozoa. The first two possibilities are not likely because a cocktail of protease inhibitors was used for these experiments and because detergents appeared to increase rather than decrease PDE activity (Table 1). The observation that TFP decreased PDE hydrolysis toward cGMP in Triton X-100 soluble fraction to a level even lower than that detected in the presence of EGTA (EGTA-insensitive PDE activity; Fig. 1, A and B) suggested that the last hypothesis is the most probable.
To verify this hypothesis we submitted partially purified PDEs to a nondenaturating and nonreducing electrophoresis in order to preserve the native form of the proteins. The anti-calmodulin antibody cross-reacted with the same 67-kDa protein band detected with the anti-PDE1A antibody (Fig. 5). Similar results were obtained with proteins separated by gel filtration chromatography (data not shown). Moreover, the association of PDE1A and calmodulin in spermatozoa appeared very strong because no dissociation of this complex was observed with either EGTA + EDTA, TFP, or urea treatment (Fig. 6). Furthermore, the 67-kDa band immunodetected with the anti-calmodulin antibody preadsorbed with calmodulin disappeared, confirming that the calmodulin was really bound to PDE1A. These results indicate that human sperm PDE1A is tightly bound to calmodulin in plasma membrane and form a stable complex, which suggest that these enzymes are in their active form in ejaculated spermatozoa. A similar case was described for PDEs from rabbit and bovine lung [40] and the myosin light-chain kinase [41], which appear to be tightly bound to calmodulin even in the absence of calcium.
The anti-calmodulin antibody recognized a doublet of proteins at 67 and 68 kDa (Fig. 5). The 67-kDa band is PDE1A and the 68-kDa band could be similar to that previously identified in rat spermatozoa as a calmodulin binding protein with still unknown function [22]. This second protein did not cross-react with the anti-PDE1A antibody but reacted with the anti-calmodulin antibody, indicating that it is a calmodulin-binding protein or that it could also be a protein with some antigenic relationship with calmodulin. This protein of 68 kDa detected could also correspond to an alternative spliced variant of either PDE1A, PDE1B, or PDE1C isoforms, because their molecular masses are very similar [8]. Alternatively, the small difference in mass may result from posttranslational modifications (e.g., phosphorylation).
In summary, we clearly identified PDE1A and PDE3A isoforms in ejaculated human spermatozoa. Cellular localization suggested that these two PDEs could play a role in the acquisition of sperm fertilization ability. Moreover, sperm PDE1A is tightly bound to calmodulin, even in the absence of calcium, which explains its lack of response to calcium + calmodulin. These results suggest that PDE1A is present in an activated form in ejaculated human spermatozoa and that a net increase in the level of cAMP could be due to its synthesis by adenylyl cyclase rather than by inhibition of PDEs.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Grant support for this work was received from the Canadian Institutes of Health Research. 
2 Correspondence: Claude Gagnon, Urology Research Laboratory, Room H6.47, Royal Victoria Hospital, 687 Avenue des Pins ouest, Montréal, QC, Canada H3A 1A1. FAX: 514 843 1457; claude.gagnon{at}muhc.mcgill.ca
Accepted: February 22, 2002.
Received: January 7, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Bajpai, S. E. Fiedler, Z. Huang, S. Vijayaraghavan, G. E. Olson, G. Livera, M. Conti, and D. W. Carr AKAP3 Selectively Binds PDE4A Isoforms in Bovine Spermatozoa Biol Reprod, January 1, 2006; 74(1): 109 - 118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-H. Su and V. D. Vacquier Cyclic GMP-specific Phosphodiesterase-5 Regulates Motility of Sea Urchin Spermatozoa Mol. Biol. Cell, January 1, 2006; 17(1): 114 - 121. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Vasta, W. K. Sonnenburg, C. Yan, S. H. Soderling, M. Shimizu-Albergine, and J. A. Beavo Identification of a New Variant of PDE1A Calmodulin-Stimulated Cyclic Nucleotide Phosphodiesterase Expressed in Mouse Sperm Biol Reprod, October 1, 2005; 73(4): 598 - 609. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
