This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Revelli, A. Articles by Tesarik, J. Search for Related Content PubMed PubMed Citation Articles by Revelli, A. Articles by Tesarik, J.

Nongenomic Actions of Steroid Hormones in Reproductive Tissues¹

Department of Obstetrical and Gynecological Sciences (A.R., M.M.), University of Torino, 10126 Torino, Italy; and Laboratoire d’Eylau (J.T.), 75116 Paris, France

	Abstract

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 

I. Introduction

II. Nongenomic Actions of Estrogens

A. Granulosa cells

B. Endometrial cells

C. Oocytes

D. Spermatozoa

III. Nongenomic Actions of P

A. Granulosa cells

B. Oocytes

C. Spermatozoa

IV. Nongenomic Actions of Androgens

A. Sertoli cells

B. Oocytes

V. Signal Transduction Pathways Involved in Nongenomic Steroid Effects

A. Ca²⁺ signal generation and amplification

B. Ca²⁺ signal transduction

C. The PTK system

D. The GABA_A-like receptor/Cl^- channel

VI. Cross-Talk Between the Nongenomic and Genomic Responses of Cells to Steroids

VII. Conclusions

	I. Introduction

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 
THE STEROID hormone receptor superfamily includes glucocorticoid, mineralocorticoid, progestin, estrogen, androgen, and 1,25-dihydroxyvitamin D₃ receptors. All members of this superfamily share similar structural and functional characteristics and have highly conserved structural domains involved in DNA and ligand binding. The activation of steroid hormone receptors regulates the transcriptional activity of specific genes, thus mediating the so-called classic, or genomic, actions of steroids.

A detailed description of the structure and function of the classic steroid receptors is beyond the scope of this review and has already been published (see reviews in Refs. 1–5). Briefly, within the intact cell the unoccupied form of the classic steroid receptor resides in the nucleus or in the nucleus and cytoplasm, dependent on the receptor and cell type. Steroid hormones act in the target tissues with a mechanism that may be summarized in the following steps: 1) free steroid enters the target cell by passive diffusion through the plasma membrane, passes through the nuclear membrane (in the case of nuclear receptors), and binds with high affinity to the receptor; 2) the steroid-receptor complex undergoes activation, a process that involves conformational changes (tertiary and quaternary structure) and enables the receptor to bind to selective sites on the chromatin; 3) the activated steroid-receptor complex interacts with specific DNA sequences referred to as steroid response elements, usually located upstream of the steroid-responsive genes; 4) the activated steroid-receptor complex acts as a transcription factor modulating the synthesis of specific mRNAs and proteins, in turn responsible for the final effect of the hormone.

A large body of experimental data obtained in several cell systems suggests the existence of steroid effects that cannot be explained by the classic model of steroid-target cell interaction. These effects may rather be explained by the existence of signal-generating steroid receptors on the cell surface and have been called nonclassic, nongenomic steroid effects. In general, nongenomic steroid effects have the following characteristics: 1) are too rapid (from seconds to a few minutes) to be compatible with the involvement of changes in mRNA and protein synthesis; 2) can be observed even in highly specialized cells that do not accomplish mRNA and protein synthesis (e.g., spermatozoa) or by cell clones lacking the steroid nuclear receptors; 3) can be elicited even by steroids coupled with high-molecular weight substances that do not pass across the plasma membrane and do not enter into the cell; 4) are not blocked by inhibitors of mRNA or protein synthesis; 5) may not be blocked by antagonists of the classic, genomic steroid receptors (unless these antagonists react with structurally related steroid-binding domains that may be common to both the genomic and the nongenomic steroid receptors); 6) are highly specific, as steroids with very similar, but not identical, chemical structure may show various degrees of potency in exerting them.

This review summarizes the current knowledge about the nongenomic actions of steroid hormones in the reproductive tissues, describes the intimate mechanisms by which these effects are accomplished, and analyzes the cross-talk pathways between nongenomic and genomic actions of steroids in the reproductive system (Table 1). All the effects that are dealt with in this review can be induced by physiological concentrations of steroid hormones (i.e., concentrations with which each respective cell can be confronted in the physiological context). Effects of steroids at supraphysiological (pharmacological) concentrations are not included in this paper.

	II. Nongenomic Actions of Estrogens

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

A. Granulosa cells
In chicken and pig granulosa cells taken from preovulatory follicles, measurements with the Ca²⁺-reporter dye Fura-2 show an immediate (<5 sec) 4- to 8-fold increase in intracellular Ca²⁺ ([Ca²⁺]i) after the addition of 10^-6 to 10^-10 M 17ß-estradiol (17ß-E₂) (6). This effect is specific of estrogenic compounds, as estrone, 17-E₂, and estriol are as effective as 17ß-E₂ in increasing [Ca²⁺]i, whereas progesterone (P), pregnenolone, testosterone (T), androstenedione, and 5-dihydrotestosterone (5-DHT) are uneffective up to a concentration of 10^-5 M. The estrogen-triggered [Ca²⁺]i increase is not affected by the classic estrogen receptor (ER) antagonist tamoxifen or by the RNA and protein synthesis blockers, such as actinomycin-D and cycloheximide. The rapid [Ca²⁺]i increase is not reduced by incubating these cells in Ca²⁺-free medium containing the Ca²⁺ chelator EGTA or by pretreating them with the Ca²⁺ channel blockers La³⁺, Co²⁺, methoxyverapamil, and nifedipine. The mechanism of the rapid estrogen action in this experimental model is therefore different from that displayed in other cell systems by other steroid hormones, such as vitamin D₃ and P, which stimulate the influx of external Ca²⁺ through membrane channels (see Section III). In granulosa cells, the estrogen-induced rise of [Ca²⁺]i is abolished by pretreating the cells with inhibitors of inositol phospholipid hydrolysis such as neomycin and U-73,122 (7). The 17ß-E₂-induced [Ca²⁺]i increase in rat granulosa cells, therefore, seems to be the result of the release of Ca²⁺ from intracellular stores triggered by inositol 1,4,5-triphosphate (InsP₃) generated by a steroid receptor-induced hydrolysis of membrane phosphatidylinositol 4,5-biphosphate. The very brief (5 sec) time elapsed between 17ß-E₂ addition and the [Ca²⁺]i peak probably represents the time needed for enough InsP₃ to accumulate from the breakdown of membrane phosphoinositides and bind to Ca²⁺-releasing receptors on intracellular storage vesicles.

In mammalian granulosa cells, the estrogen-induced [Ca²⁺]i rise could be an important regulating mechanism, possibly able to modulate the function of Ca²⁺-binding signal proteins (e.g., calmodulin), affect the secretion of Ca²⁺-sensitive autocrine and/or paracrine factors (e.g., transforming growth factor-), stimulate cell proliferation, and modify the cell-to-cell intraovarian communication (8). Moreover, since the classic ER is a substrate of the calmodulin-dependent protein kinase II (9), the [Ca²⁺]i transient could even mediate a cross-talk interaction between ERs located on the plasma membrane and the genomic, classic ER.

B. Endometrial cells
In cultured rat endometrial cells, 17ß-E₂ elicits both functional and morphological changes that appear too rapid to be mediated via the classic genomic pathway. The addition of 10^-9 M 17ß-E₂ rapidly (<10 min) stimulates ⁴⁵Ca²⁺ uptake from endometrial cells (10), and this effect appears to be mediated via surface estrogen-binding sites, whose existence has been reported in rat endometrial and liver cells (11, 12, 13). Furthermore, the intravenous injection of 17ß-E₂ to ovariectomized rats rapidly (within 1 min) modifies cell morphology, increasing the number and clustering of microvilli in the endometrial cells and the luminal surface ciliation (14). Similar morphological changes may be obtained after exposure to diethylstilbestrol (14). Ultrastructural analysis confirms the rapid estrogen-induced increase in microvillar density and length and shows that the structural modifications induced by estrogens are biphasic, as they appear further enhanced 7 min after 17ß-E₂ injection. 17ß-E₂ is even able to elicit micropinocytotic activity in the endometrial cells within 2 min of its intravenous injection to ovariectomized rats (15). It is of interest that morphological effects similar to those induced by 17ß-E₂ in endometrial microvilli can be evoked in other cell types by peptidic growth factors, whose action is mediated by surface receptors. For instance, such effects are produced by the epidermal growth factor in cultured sympathetic neurons (16) and by the nerve growth factor in the human carcinoma cell line A-431 (17).

C. Oocytes
Human oocytes at the germinal vesicle stage develop a rapid (within a few seconds) Ca²⁺ influx in response to 17ß-E₂ (18). The addition to oocytes of 10^-6 M 17ß-E₂, either free or linked to a membrane-impermeant conjugate with BSA, induces Ca²⁺ influx followed by a series of oscillations in the free cytosolic Ca²⁺ concentration ([Ca²⁺]i) that are dependent on the continuous presence of Ca²⁺ in the incubation medium (18). The presence of 17ß-E₂ in the medium does not influence the progression of oocytes through meiotic maturation (to metaphase I and metaphase II stages) but improves the subsequent fertilization potential of the in vitro matured oocytes. In fact, oocytes that reach the metaphase II stage and are inseminated in vitro show better cleavage rate and development of the resulting embryos. Thus, 17ß-E₂ does not appear to affect the nuclear maturation of human oocytes but exerts a beneficial effect on their cytoplasmic maturation (18). A subthreshold concentration of 17ß-E₂ (10^-7 nM) also promotes Ca²⁺ oscillations in germinal vesicle-stage human oocytes, but with a delay of 1–3 h after hormone addition (19). This slow nongenomic response may be closer to the in vivo condition, when oocytes are exposed to slow progressive changes in ovarian follicular steroid concentrations rather than to an abrupt increase. It should be noted that spontaneous Ca²⁺ oscillations can be observed in mouse germinal vesicle oocytes (20) but never in human oocytes (18, 19). Ca²⁺ oscillations, however, represent a typical early reaction of oocytes to the fertilizing spermatozoon in various mammalian species (21) including humans (22, 23). It remains to be determined whether the priming of oocytes with 17ß-E₂ during their previous intrafollicular development is important for the process of fertilization. Both the spontaneous Ca²⁺ oscillations of mouse oocytes (20) and the 17ß-E₂-induced Ca²⁺ oscillations of human oocytes (19) stop after the germinal vesicle breakdown and do not reappear in metaphase I or metaphase II oocytes.

D. Spermatozoa
The incubation of human spermatozoa in the presence of 17ß-E₂ stimulates sperm motility (24, 25), [³H]tetracycline binding capacity (26), oxygen uptake, and lactate production (27, 28), as well as the metabolization of several substrates (27). Furthermore, the in vitro sperm-agglutinating property of selected sera is significantly decreased by estrogens (29). The addition of 17ß-E₂ (30 ng/ml) to the sperm incubation medium is even claimed to improve the results of the zona-free hamster ova penetration test by human spermatozoa of both fertile and infertile men (30). Since spermatozoa are known to have a densely packaged DNA, and protein synthesis in spermatozoa is very scarce and limited to mitochondria, a genomic action of steroids is not possible in this cell type. The above mentioned effects of estrogens are thus necessarily mediated by a nongenomic mechanism.

A study by Hyne and Boettcher (31) reported the existence of specific, low-affinity binding sites for 17ß-E₂ on human spermatozoa for which other steroids could compete. A subsequent autoradiographic study observed that the binding of [³H]17ß-E₂ to spermatozoa can be reduced by coincubation with other steroids, such as progestagens, which can strongly compete for the 17ß-E₂ receptor (32). The presence of high-affinity (K_D 6.6 x 10^-10 M) binding sites for 17ß-E₂ on human sperm has been further confirmed: after fractionation of 17ß-E₂-bound spermatozoa, they are detectable mainly in the membrane fraction (75–85%), whereas the nuclear fractions shows only about 10% of the total bound radioactivity, and no radioactivity is detected in the cytosolic fraction (33). Other autoradiographic data confirm that the plasma membrane is the site of receptors with specificity for 17ß-E₂, which are mostly concentrated in the central part of the sperm tail (34). In competition experiments, the binding sites for [³H]17ß-E₂ appear to be specific, the labeled steroid being displaced by cold 17ß-E₂. Specific binding sites for [³H]17ß-E₂ are not detectable in the sperm cytosol and in sperm nuclei isolated after homogenization, detergent treatment, and centrifugation (35).

	III. Nongenomic Actions of P

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 
A. Granulosa cells
Data on the nongenomic effect of P in granulosa cells are scarce. There is one study reporting the induction of a rapid Ca²⁺ response of pig granulosa cells by P (36). This response involves both Ca²⁺ influx from the extracellular space and Ca²⁺ release from intracellular stores. It appears to be mediated by a pertussis toxin-insensitive G protein.

B. Oocytes
Several data suggest that the physiological stimulus triggering the resumption of meiotic division in amphibian oocytes is P. In Xenopus laevis, P acts at the oocyte plasma membrane to reinitiate the first meiotic division (37). The photoaffinity labeling technique identifies a monomeric membrane binding site for P with an approximate mol wt of 110,000, whose activation by P or the synthetic progestin R5020 induces germinal vesicle breakdown (38, 39). In Xenopus oocyte, P is effective in increasing [Ca²⁺]i, leads to a sudden rise in intracellular pH, decreases membrane conductance, and finally regulates the activity of plasma membrane enzymes (40). Some of them (e.g., serine proteases) are activated by P addition; others (e.g., adenylate cyclase) undergo down-regulation. In amphibian oocytes, the binding of P to plasma membrane receptors activates several systems associated with intracellular signal transduction:

1. One system is linked to the activation of a phosphatidylinositol diphosphate-specific phospholipase C that generates InsP₃ and a smaller quantity of diacylglycerol (DAG): both metabolites are subsequently released into the cytosol and reach a maximum after approximately 15 min of oocyte exposure to P (37).

2. Both in Rana pipiens (41) and Xenopus laevis (42, 43) oocytes, P inhibits adenylate cyclase, thus reducing cAMP production and stimulating the Na⁺/H⁺ exchange as well as changes in Ca²⁺ and K⁺ conductance. In vitro addition of P elicits an almost immediate decrease in cAMP concentration inside the oocyte, even if the enzyme has previously been stimulated by pretreating the oocytes with cholera toxin (42). Furthermore, the addition of physiological concentrations (30 x 10^-9 M) of P to intact oocytes selectively inhibits membrane-bound adenylate cyclase activity, which represents about 30% of the total enzyme (44). The inhibition of X. laevis oocyte adenylate cyclase involves a G protein; in fact, the level of inhibition of plasma membrane adenylate cyclase by both P and 2',5'-dideoxyadenosine, which acts as a potent progesterone receptor (PR) agonist, correlates with slowing of guanine nucleotide exchange in photoaffinity labeling assays (43). The inhibitory action of P is abolished by the ion Mn²⁺ (43).

3. Another consequence of P action in the amphibian oocyte is the stimulation of massive phospholipid turnover in the oocyte plasma membrane (45) and the activation of plasma membrane-bound, Ca²⁺-dependent protein kinase C (PKC) (46). In fact, the addition of GTP to the isolated plasma membrane of R. pipiens oocytes increases both the P-induced protein phosphorylation (46) and the activation of a trypsin-like serine protease, which is released from the plasma membrane within 1 min after P binding to its receptor (45).

4. In R. pipiens oocytes, P induces N methylation of phosphatidylethanolamine within seconds, and this reaction is followed by an increased synthesis of phosphatidylcholine and further conversion into phosphatidylmonomethylethanolamine within the first minute of the steroid addition. The latter conversion is associated with a sharp, transient rise of DAG, an increase in the phosphorylation of plasma membrane proteins, and the first peak of a biphasic Ca²⁺ release from the oocyte surface (47). Since inhibitors of phospholipid N methylation are able to block P-induced meiosis, and exogenous phosphatidylmonomethylethanolamine alone is able to induce meiosis, N-methyltransferase is likely to be a downstream element of the signal transduction cascade mediating the P effect on oocyte maturation (47).

C. Spermatozoa
Follicular fluid (FF) mixed with the oviductal fluid or trapped in the extracellular matrix of the cumulus oophorus represents a source of P acting on spermatozoa at the fertilization site. P is able to stimulate the motility and acrosomal exocytosis of human sperm preserving sperm viability in vitro (48) and also enhances the acrosomal exocytotic response of mouse sperm to the subsequent exposition to the zona pellucida glycoprotein ZP3 (49). P has also been observed to increase the percentage of spermatozoa exhibiting hyperactivated motility at very low concentrations (3.1 ng/ml) and within 10 min of addition to the incubation medium (50), and to increase the velocity of spermatozoa when added to peritoneal fluid (51). A significant correlation between P levels in FF and the ability of FF to induce the acrosomal exocytosis of human sperm has been noticed after charcoal delipidation of FF and its subsequent supplementation with exogenous P (52). A final concentration of 10 µg P/ml almost completely restores (88%) the stimulating effect of FF on the acrosomal reactivity (52). Data obtained combining organic precipitation and HPLC suggest that P and 17-hydroxy-P are the major substances responsible for the acrosomal reaction-inducing activity of FF (53). Moreover, charcoal treatment, as well as preincubation of FF with P-specific antibodies, prevents the FF-induced Ca²⁺ influx in human spermatozoa, a phenomenon that is strictly related to acrosomal exocytosis (54).

The effect of P on acrosomal reaction is due to its capacity of inducing a rapid increase in free cytosolic [Ca²⁺]i, which is evident after a few seconds of exposure to the steroid (54). The increase in [Ca²⁺]i is entirely due to the influx of extracellular Ca²⁺, as it may be blocked by the Ca²⁺ channel-blocker La³⁺ or by the addition of the Ca²⁺-chelator EGTA to the incubation medium (54). The effect of P on Ca²⁺ influx is dose-dependent, with small effects being observed with 10^-8 to 10^-9 M and maximal effects with 10^-5 to 10^-6 M P (54). It has been calculated that P is able to initiate the acrosomal reaction in vitro at concentrations ranging from 1 µg/ml (approximately 3 x 10^-6 M) to 100 ng/ml (55). P stimulates the rapid influx of Ca²⁺ in parallel with a rapid hydrolysis of phosphatidylinositol 4,5-biphosphate (56); both phenomena are blocked by preventing the entry of extracellular Ca²⁺ by La³⁺ addition, showing that the latter is a sequela, and not the cause, of the former. Thapsigargin, a specific inhibitor of the endoplasmic reticulum Ca²⁺ ATPase Ca²⁺-pump, and thus a mobilizer of intracellular Ca²⁺, can initiate the acrosomal reaction in human sperm (57). Since there is no apparent endoplasmic reticulum in the cytoplasm of mature sperm, the Ca²⁺ contained in the acrosome, and possibly the mitochondria, is likely to be involved in these phenomena. Although some authors reported that the Ca²⁺ influx after P addition is even higher in the absence of Na⁺ (58), Na⁺ seems to play a role in these phenomena since the absence of Na⁺ in the medium causes inhibition of the P-mediated increase in human sperm [Ca²⁺]i and of the acrosomal exocytosis (59).

Trypsin-like activity in sperm might also play a role in the P-mediated [Ca²⁺]i increase. In fact, preincubation of capacitated sperm with either benzidamine hydrochloride or 4'-acetamidophenyl,4-guanidinobenzoate, two inhibitors of trypsin-like enzymes, is able to inhibit the P-induced acrosomal reaction by 68–85% by blocking the membrane fusion events of the acrosome reaction (60). Moreover, the action of P on human spermatozoa is enhanced in the presence of an acrosin activator from the human FF (61). The action of proteases, however, needs to be regulated by a yet unknown mechanism because the addition of trypsin to the incubation medium removes the P-binding activity from the sperm surface without affecting sperm motility, viability, and acrosomal reactivity (62). Sperm surface trypsin-like activity, represented by the acrosomal enzyme acrosin (63) or by an acrosin-like protease from the seminal vesicles (64), is claimed to be important in cleaving a P-binding protein (corticosteroid-binding globulin, CBG), finally producing a high local P concentration (65, 66).

The existence of specific receptors for P on the plasma membrane of spermatozoa is suggested by the observation that P conjugated with BSA [P-3-(O-carboxymethyl)oxime: BSA], a large molecule incapable of crossing the plasma membrane, is able to increase [Ca²⁺]i and to stimulate the acrosomal reaction (67, 68). Moreover, the [Ca²⁺]i-increasing effect of P is extremely rapid, dose-dependent, and not counteracted by RU486, a potent antagonist of the nuclear PR (69). RU486 is ineffective even in inhibiting the P-induced stimulation of hyperactivated motility (50). Further, neither RU486 nor ZK98299, another blocker of the classic genomic receptor, are able to avoid the P-mediated Ca²⁺ entry into spermatozoa, and this process may not be mimicked by synthetic progestins such as megestrol, norgestrel, medroxyprogesterone acetate, norethynodrel, norethindrone, and R5020, which interact with the nuclear PR (70). Direct evidence of the presence of specific binding sites for P on the sperm plasma membrane has been provided by a study with P coupled to fluorescein isothiocyanate-labeled BSA: the P-BSA complex binds tightly to the sperm plasma membrane (71). This study, however, does not demonstrate that the binding is specific, since competition tests showing the ability of unlabeled P to displace the bound labeled P have not been performed. Fluorescence microscopy and flow cytometry indicate that only about 10% of spermatozoa possess PRs on their surface, and thus are able to bind the steroid and undergo the acrosome reaction (71). The binding of P appears to be limited to the sperm acrosomal region of the plasma membrane, and a marked variability among individual spermatozoa may be observed (71). These observations are consistent with three other studies: the first reporting that only 30% of spermatozoa have detectable PRs on their heads (67), the second localizing PRs at the equatorial region of the sperm head by means of immunofluorescence with the monoclonal antibody MAb C26s, directed against the C-terminal P-binding domain of the intracellular PR (72), and the third showing that the P-mediated increase in [Ca²⁺]i begins in the midhead region and spreads rapidly (within 1 sec) anteriorly over the rest of the head (73). Some authors, however, hypothesize that a greater proportion of sperm in an ejaculate can possess PRs in a functionally inactive form; their activation could be asynchronous and stimulated by the increase of extracellular Ca²⁺ (74). More recent findings indicate that about 90% of spermatozoa may respond to P with an abrupt increase in [Ca²⁺]i (75, 76), and thus the size of the sperm population having PRs seems to be greater than originally believed.

The apparent contradiction among these studies probably originates from the presence of different types of PRs in the sperm plasma membrane. Many features of the sperm response to P, in fact, suggest the implication of at least three types of surface receptors, by means of which P targets, respectively, 1) a plasma membrane Ca²⁺ channel (PR1), 2) a membrane-associated protein tyrosine kinase (PTK; PR2), and 3) a plasma membrane chloride channel (PR3) (77) (Fig. 1). The tyrosine kinase-associated PR (PR2) seems to be the one visualized by the hormone-binding assay because those spermatozoa that bind the BSA conjugate also increase their phosphotyrosine content and undergo the acrosome reaction (78). This PTK-associated PR is probably responsible both for the effect of P on the acrosome reaction and on hyperactivated motility (79). It appears to be sensitized during the capacitation process, during which an initial phosphorylation event takes place under the control of redox- (80) and cAMP-regulated (81) mechanisms. When tyrosine phosphorylation is stimulated, human spermatozoa become particularly sensitive to physiological concentrations of P, responding to the P-generated Ca²⁺ transient with high rates of sperm-oocyte fusion occurring after 5 min of exposure to the steroid (80). On the other side, when P is administered to uncapacitated spermatozoa, they rapidly develop an insensitivity to further stimulation that is due both to the down-regulation of the cellular mechanisms necessary to generate a Ca²⁺ transient, and to interferences with the downstream events coupling Ca²⁺ influx with a biological response (80). Tyrosine phosphorylation may be the site at which premature exposure to P exerts its suppressive effect; as in other cell types, cell-specific phosphorylation events may influence steroid receptor interaction with the transcription apparatus, altering receptor-mediated induction mechanisms (82, 83). The PR responsible for the rapid opening of the Ca²⁺ channel (PR1) is active in a higher percentage (more than 90%) of spermatozoa (75, 76), but has different ligand-binding properties than the receptor responsible for PTK activation. Moreover, it is not capable of initiating the P-induced acrosome reaction. A pharmacological dissection of the P effects on Ca²⁺ influx and on tyrosine phosphorylation may be obtained in peculiar experimental conditions (77). The third PR (PR3) is likely to be a -aminobutyric acid (GABA)_A-receptor/chloride channel complex, and probably mediates the Cl^- fluxes occurring during acrosomal exocytosis (84). It does not appear to be involved in the P-induced Ca²⁺ influx (85). Thus, P effects on spermatozoa, and perhaps the nongenomic effects of steroids on cells in general, appear to be mediated by multireceptor systems rather than by a single type of receptor. When one member of this system is not operational, the biological response may fail even if the other types of receptor function normally. P receptors in spermatozoa are probably relevant to their fertilizing ability: in fact, an impaired P-binding capacity with subsequent failure in generating the typical Ca²⁺ influx wave and the acrosomal reaction has been observed in some cases of unexplained infertility (86), and spermatozoa from oligospermic men seem to be less prompt to respond to P with an increase in [Ca²⁺]i, probably for a lack of PRs on the sperm surface (87).

View larger version (23K): [in this window] [in a new window]   Figure 1. PRs at the plasma membrane of human spermatozoa. Experimental data suggest that three types of PR exist in the plasma membrane of human spermatozoa. P-corticosteroid-binding globulin (CBG) complexes contained in FF are dissociated by an acrosine-like enzyme of the sperm plasma membrane; as a consequence, a high P concentration is obtained locally. Also, free P molecules may be found in FF and can bind to the P surface receptors. P appears to act directly at a voltage-independent Ca2+ channel after binding to type 1 P receptor (PR1), a membrane surface receptor expressed in a high percentage of sperm cells. The P-PR1 interaction triggers a rapid (seconds) Ca2+ influx, but is not sufficient to initiate the acrosome reaction. At the same time P reacts with a second type of surface receptor (PR2), which may be found only in about 10% of spermatozoa. After P binding, PR2 aggregates and stimulates a PTK which leads to the opening of a PTK-dependent Ca2+ channel, both directly and via a signaling cascade involving a series of intracellular messengers (indicated in square brackets). The activation of both types of Ca2+ channels finally initiates exocytosis of the acrosome. A third type of surface P receptor (PR3), coupled to a Cl- channel and resembling a GABAA/Cl- channel complex, mediates the Cl- influx which takes place during the acrosome reaction.

After binding to the PTK-associated surface receptor, P produces receptor aggregation on the sperm surface (62), which may be instrumental in at least some of the known P effects. Artificial aggregation of PR complexes, as generated by adding anti-P antibodies to spermatozoa previously incubated with subthreshold concentrations of P, is able to trigger a Ca²⁺ influx into the sperm and is followed by the acrosomal reaction within a few minutes (89). Under appropriate conditions (inhibition of immediate exocytosis), a spontaneous aggregation of ligand-occupied P-binding sites on the sperm surface can be demonstrated (62). P is also able to increase the fluidity of human sperm plasma membrane, to aggregate membrane vesicles, to induce the fusion of these vesicles, and to render the membrane permeable to hydrophilic molecules such as carboxyfluorescein (90). The increase in membrane fluidity, which facilitates receptor aggregation, is a common finding when spermatozoa are incubated in capacitating conditions: the decrease in cholesterol/phospholipid ratio inside sperm membranes occurs during capacitation, and is a prerequisite for acrosomal reaction (91).

Several mechanisms mediating the transduction of the P signal inside the sperm cell have been studied (see also Section V). The P-induced onset of hyperactivated motility seems to be linked to an increased 3',5'-cAMP concentration inside the spermatozoon (79), which in turn is responsible for cAMP-dependent kinase (protein kinase A; PKA) activation. The stimulating effect of human FF on acrosomal reactivity, which has been thoroughly studied (92, 93, 94, 95, 96), is inhibited by KT5720, a PKA inhibitor (97). Human spermatozoa incubated for 2 h with increasing P concentrations show a significant, dose-dependent increase in intracellular cAMP that occurs biphasically, with one peak after 30 min and another peak after 120 min incubation. These effects of P on cAMP levels are paralleled by a significant rise in the percentage of hyperactivated spermatozoa, occurring at the same times. The effect of P on intracellular cAMP concentration is Ca²⁺-dependent, is mimicked by the Ca²⁺-ionophore A23187, and is inhibited by the PTK inhibitor genistein, but not by the genomic PR antagonist RU486. Since phosphodiesterase inhibitors, such as pentoxyfilline, have been shown to increase hyperactivation in human spermatozoa (98, 99), the P-induced increase in cAMP could also depend on the inhibition of phosphodiesterase.

The intracellular mechanisms by which P elicits the acrosome exocytosis and prepares the subsequent sperm-oocyte fusion are complex. They are schematically represented in Fig. 1.

The aggregation of the PTK-associated type of P receptors stimulates the tyrosine phosphorylation of a 94-kDa phosphoprotein, which, in turn, is able to activate intracellular effectors leading to acrosomal exocytosis (78). This protein is a unique, germ cell-specific receptor PTK localized at the sperm plasma membrane in the acrosomal region; it may enhance its kinase activity in response to both P and solubilized zona pellucida, behaving as a receptor PTK with regard to the zona glycoprotein ZP3 and as a receptor-coupled PTK with regard to P (100). The PR responsible for tyrosine phosphorylation (PR2) is expressed in only 10% of spermatozoa, aggregates within minutes, and stimulates the PTK which, in turn, leads to the opening of voltage-dependent, PTK-dependent Ca²⁺ channels. The P-induced increase in tyrosine phosphorylation is detectable after at least 5–10 min after P addition to spermatozoa, contrasting with the rapidity of the P-induced Ca²⁺ influx occurring within a few seconds after P treatment (100). This rapid Ca²⁺ influx depends on the direct effect of P on voltage-independent Ca²⁺ channels, probably coupled with a different type of PR (PR1). The result of the direct P-induced, rapid activation of the voltage-independent Ca²⁺ channels first, and of the second messenger-operated activation of voltage-dependent Ca²⁺ channels in a second time, is a biphasic Ca²⁺ influx into sperm. The P effect on tyrosine phosphorylation, in any case, is not a simple consequence of the P-induced Ca²⁺ influx because a similar Ca²⁺ influx can be obtained with the use of a Ca²⁺ ionophore without producing the PTK-mediated phenomena leading to the acrosomal exocytosis (77, 100).

A unique steroid receptor/Cl^- channel complex (resembling but not identical to a GABA_A receptor/Cl^- channel complex) is possibly involved in the P-initiated acrosomal reaction of sperm. Both GABA and the GABA_A-receptor agonist muscimol induce acrosomal exocytosis in murine spermatozoa (101), but this effect cannot be observed with human sperm (80). In the mouse, P can also potentiate the ability of GABA agonists to stimulate the acrosome reaction (101). The GABA_A receptor/Cl^- channel complex blockers picrotoxin or pregnenolone sulfate and the GABA_A receptor antagonist bicuculline are effective in reducing the P-initiated increase in [Ca²⁺]i (69) and the acrosomal reaction of human and porcine sperm (84, 102). The absence of extracellular Cl^- in the culture medium does not prevent P from stimulating the Ca²⁺ entry and the rise in [Ca²⁺]i, but the acrosomal reaction occurs only in the presence of Cl^- (103). Such observations may be explained by the existence of a third type of plasma membrane PR (PR3), different from that initiating the rapid increase in [Ca²⁺]i and from the one associated with PTK activation. PR3 (similar to the GABA_A/Cl^- channel complex) is thought to mediate an increased Cl^- influx leading to a subsequent reversal of HCO₃^-/Cl^- exchange (85, 103). Another possible model is that a chloride influx induces an initial hyperpolarization of the cell, as happens when GABA activates neuronal GABA_A-receptor/Cl^- channel complexes (84), and this hyperpolarization is then followed by a proton efflux, depolarization, and activation of voltage-sensitive Ca²⁺ channels. The hypothesis of the existence of PR3 is supported by the observation that steroids that are very active on the GABA_A/Cl^- channel are poor stimulators of Ca²⁺ influx (85). Moreover, known modulators of GABA_A/Cl^- channels, such as picrotoxin, diazepam, GABA, muscimol, and pentobarbital, are ineffective in influencing basal [Ca²⁺]i or the Ca²⁺ flux induced by P (80, 85).

Several other substances appear to play a role in coupling P binding to its receptors on the sperm surface and Ca²⁺ fluxes through the sperm membrane (Fig. 1, in square brackets). The P-PR complex is thought to exert the following effects:

1. Activation of phospholipase D, which in turn elevates the intracellular concentration of phosphatidic acid, a substance able to promote Ca²⁺ influx (104).

2. Modulation of polyamine biosynthesis increasing spermidine production, in turn promoting Ca²⁺ influx (105). Inhibitors of polyamine synthesis, such as -difluoromethylornithine, an inhibitor of putrescine synthesis, or MDL73811, an inhibitor of S-adenosylmethionine decarboxylase (required for spermine and spermidine synthesis), are effective in blocking the P-induced Ca²⁺ influx and the acrosomal reaction, but preincubation with putrescine or spermidine reverses such a block (105).

3. Stimulation of the production and release of platelet-activating factor (PAF), a phospholipid able to induce acrosomal exocytosis (106, 107, 108, 109, 110). Treatment of human spermatozoa with P results in the activation of phospholipase A₂ (PLA₂) and an increase of [³H]acetate incorporation into PAF is observed (111, 112); such newly synthesized PAF could act autocrinally on spermatozoa and play a role in mediating P-induced Ca²⁺ influx and acrosomal reaction.

4. Stimulation of protease-driven mechanisms, which have been hypothesized in the P-induced acrosomal reaction: the acrosomal reaction of human spermatozoa, in fact, is impaired by protease inhibitors (113).

5. Activation of phospholipase C, thus increasing the intracellular content in InsP₃ and DAG, in turn mediating the stimulation of PKC (56, 114). However, while in most hormonally responsive cells the phospholipase C-mediated generation of InsP₃ is responsible for the release of Ca²⁺ from intracellular stores and consequently for the increase in [Ca²⁺]i, in spermatozoa the synthesis of InsP₃ appears to be a consequence rather than a cause of the Ca²⁺ influx induced by P. In fact, the phospholipase C inhibitor U73122 is uneffective in disrupting the P-induced Ca²⁺ transient in human spermatozoa (80).

The involvement of the G protein system in the P-induced stimulation of sperm motility and acrosomal reactivity has never been demonstrated. Pertussis toxin, which catalyzes the ADP ribosylation of G proteins and thereby inhibits their function, does not influence the P-mediated Ca²⁺ influx and the acrosome reaction (80, 115). Moreover, although G proteins insensitive to this toxin have been identified in mouse spermatozoa (116), they have not been detected in spermatozoa from mammalian species including the human (117).

	IV. Nongenomic Actions of Androgens

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 
A. Sertoli cells
In freshly isolated rat Sertoli cells, both T and 5-DHT rapidly (20–40 sec) increase the intracellular Ca²⁺ concentration (118). This effect is inhibited by preincubation with either the nonsteroidal antiandrogen hydroxyflutamide or the 5-reductase inhibitor finasteride (118), indicating that T acts via a receptor-mediated mechanism and, at least partially, after conversion to 5-DHT. Aromatization of T to 17ß-E₂ consistently reduces the T-induced Ca²⁺ entry into the cell (118), suggesting that T effect is mediated by specific androgen binding sites, but not by ERs. Other steroids, like P, are completely inactive in Sertoli cells (118). The abrupt increase of [Ca²⁺]i can be prevented by the removal of extracellular Ca²⁺ or by the pharmacological blockade of voltage-dependent or voltage-independent membrane Ca²⁺ channels (118), indicating that the androgen effect on Sertoli cells is accomplished by transplasmalemmal influx of extracellular Ca²⁺. Interestingly, in avian and porcine (but not rat) granulosa cells (the female analog of Sertoli cells), 17ß-E₂ increases the cytosolic [Ca²⁺]i independently of extracellular Ca²⁺, releasing the ion from the intracellular stores via the activation of phosphoinositide pathways (6), as accomplished by peptides (119). BSA-conjugated T, which is unable to cross the plasmalemma of the Sertoli cell, is still effective in rapidly increasing cytosolic Ca²⁺ (118). This observation, together with the rapidity of the androgen effect and its dependence on extracellular Ca²⁺, suggests that the effect of T and its 5-reduced metabolite 5-DHT may be mediated directly at the plasma membrane. Since T and FSH have synergistic effects in initiating spermatogenesis (120), and the increase of [Ca²⁺]i is one of the signal transduction mechanisms of FSH action in Sertoli cells (121), it may be postulated that the interaction between T and FSH in these cells is mediated via common effects on Ca²⁺ fluxes.

B. Oocytes
Though incapable of inducing a Ca²⁺ response, androstenedione (A), the main intrafollicular androgen, influences the Ca²⁺ response of human germinal-vesicle oocytes to 17ß-E₂ (see Section II). When 10^-6 M A is added to oocytes during ongoing Ca²⁺ oscillations induced by previous addition of 10^-6 M 17ß-E₂, the oscillations stop; the 17ß-E₂-induced Ca²⁺ oscillations of human germinal-vesicle oocytes are also disabled to a variable degree by previous exposure of the oocytes to A, even when the oocytes are washed from the hormone before the treatment with 17ß-E₂ (19). Because 17ß-E₂ appears to support oocyte cytoplasmic maturation, acting by a nongenomic, Ca²⁺-mediated mechanism (18), these observations may be interpreted to suggest that the quality of mature human oocytes recovered from large ovarian follicles is conditioned by the estrogen-to-androgen ratio to which the oocytes have been temporarily exposed in the midfollicular phase (19).

	V. Signal Transduction Pathways Involved in Nongenomic Steroid Effects

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 
Most studies aimed at the identification of the signal transduction mechanisms employed in nongenomic steroid effects used spermatozoa as a model. The reason for this is the ease with which spermatozoa can be prepared and maintained in vitro in highly reproducible conditions and the absence of genomic response mechanisms whose superimposition in other cell types may complicate interpretations. Hence, the information presented in this chapter is essentially based on data obtained with spermatozoa although results obtained with other cell types are also taken into account.

Most nongenomic actions of steroids yet described involve Ca²⁺ as a second messenger. Thus, we will first focus on the fundamental phenomena underlying the steroid-induced [Ca²⁺]i increase and subsequently show what parallel signaling mechanisms steroids use, in addition to the Ca²⁺ signal, and what downstream elements are involved in the transduction of all these signals to corresponding effectors of the cellular biological response (Fig. 2).

View larger version (34K): [in this window] [in a new window]   Figure 2. Schematic representation of signal transduction pathways involved in nongenomic steroid effects on cells and their possible impact on the classic genomic steroid response (compiled from data obtained with different cell types). The steroid ligand reacts, alternatively, with a voltage-independent, poorly selective cation channel (upper left) and with a PTK-coupled receptor (at 6 and 12 o’clock). The early Ca2+ response, mediated by Ca2+ influx through the cation channel, is further amplified by a positive feedback loop consisting of Ca2+-dependent phosphoinositidase C (PIC) cleaving phosphatidylinositol 4,5-biphosphate (PIP2) to generate InsP3 and DAG. The former opens the InsP3 receptor/Ca2+ channel (InsP3-R) in the membrane of intracellular Ca2+ stores leading to Ca2+ release. The latter, after its own autoamplification through the activation of phosphatidylcholine-specific phospholipase C (PLC), activates PKC which, in its turn, opens a voltage-dependent (VD) Ca2+ channel (lower left) in the plasma membrane. Cytoplasmic alkalinization, due to the Na+ influx through the poorly selective cation channel (upper left) and to the Cl- efflux (see below), may be important for the changes in the plasma membrane polarity required for the optimal function of this VD channel. The autoamplified Ca2+ signal is then transduced to different effector systems, namely those involving the effects of activated protein kinase A (PKA) and PKC on cytoplasmic and nuclear targets, such as the classical steroid receptor (SR) and general transcription factors (TF), as well as the DAG-activatable phospholipase A2 (PLA2) producing fatty acids (FA) and lysophospholipids (LPL) involved in membrane fusion reactions. Finally, the PTK-associated pathway involves the PTK aggregation with the ligand-activated receptor leading, in addition to further stimulation of Ca2+ influx through plasma membrane Ca2+ channels and Cl- efflux through a GABAA-like receptor/Cl- channel (lower right), to the coactivation (together with Ca2+) of PIC and to the activation of a cascade involving p21ras protein and MAP-kinase with the probable participation of common PTK substrates, such as Shc, and the mSos/GRB2 complex. MAP kinase can transduce the signal to the nucleus and thus represents another potential participant in the genomic-nongenomic response cross-talk.

The initial Ca²⁺ influx has been shown to stimulate several bifurcating signal transduction pathways (Fig. 2). One of these pathways employs cAMP whose intracellular concentration in human spermatozoa undergoes a Ca²⁺-dependent increase in response to P (79), although another study has failed to detect any P-induced cAMP response in the same system (125). This discrepancy may be explained by a particular kinetics of the P-induced cAMP increase reported by Parinaud and Milhet (79) and characterized by two peaks separated by a period during which no increase is detectable. It is not known whether the observed P-dependent cAMP increase is due to stimulation of adenylate cyclase or to inhibition of phosphodiesterase.

Another pathway activated downstream of the Ca²⁺ influx (Fig. 2) involves the stimulation of hydrolysis of plasma membrane phosphoinositides by phosphoinositide-specific phospholipase C (PIC) leading to the formation of two important second messengers, DAG and InsP₃ (56, 114, 126, 127). The DAG produced by this pathway amplifies its own production by stimulating the activity of phosphatidylcholine-specific phospholipase C, another DAG-generating enzyme in mammalian spermatozoa (126). In many biological systems, DAG exerts its signaling role by activating PKC. However, the actual presence of PKC in spermatozoa, claimed by some workers (128, 129), was questioned by others (130). It now seems clear that PKC plays a role in transducing steroid-generated signals in mammalian spermatozoa (131, 132) although its activity in these cells is extremely low (133, 134). When activated by DAG, the two PKC isoforms found in bovine spermatozoa, PKC and PKCßI, rapidly translocate from the cytosol to the membrane fraction, presumably with the participation of membrane PKC-binding proteins, collectively termed RACKs (Receptors for Activated C-Kinase) that have been identified in the bovine sperm membrane fraction (132).

The pathways leading, respectively, to the increase in the intracellular cAMP content and to the stimulation of PKC appear to play an important role in the amplification of the initial steroid-induced [Ca^2+]i increase (Fig. 2). This amplification is mediated by opening, through phosphorylation by PKC, of a plasma membrane Ca²⁺ channel leading to additional Ca²⁺ influx, on the one hand, and by the release of Ca²⁺ stored in the sperm acrosome through cAMP-gated channels in the acrosomal membrane (135) on the other hand. Unlike the plasma membrane Ca²⁺ channel responsible for the initial steroid-induced Ca²⁺ influx, both the PKC-dependent plasma membrane channel and the cAMP-gated acrosomal channel are voltage-dependent (135). This may be important because the channel involved in the initial Ca²⁺ influx shows a poor cation selectivity and its opening also leads to Na⁺ influx (58). Together with Cl^- efflux via GABA_A-like receptors/Cl^- channels, which is also directly activatable by P and related steroids in mammalian spermatozoa (84, 85, 101, 136, 137), this Na⁺ influx seems to be responsible for membrane depolarization events necessary for the optimal function of these voltage-dependent channels. Sperm plasma membrane depolarization in response to P has been experimentally demonstrated (58). A testis-specific L-type voltage-dependent Ca²⁺ channel has been cloned and suggested to be the voltage-dependent plasma membrane Ca²⁺ channel employed in the cell signaling by P in mammalian spermatozoa (124).

As to the intracellular Ca²⁺ release from the sperm acrosome, the involvement of the InsP₃ released from membrane phosphoinositides by PIC stimulated through the steroid-induced Ca²⁺ influx is also possible as InsP₃ receptors have been detected in spermatozoa of several mammalian species and localized to the outer acrosomal membrane of mouse spermatozoa (138). The need for these Ca²⁺ signal amplification events for the generation of biologically efficient response may explain why the P-induced acrosome reaction can be mimicked (57) or potentiated (139) by thapsigargin, a mobilizer of InsP₃-sensitive intracellular Ca²⁺ stores. The observation of the P-induced Ca²⁺ waves and Ca²⁺ oscillations in human spermatozoa (137) also argues in favor of the existence of a complex Ca²⁺ exchange mechanism in these cells because such phenomena would be hardly compatible with the presence of only one mechanism of Ca²⁺ influx through a single type of plasma membrane channel. The reason why the inhibitors of voltage-dependent Ca²⁺ channels nifedipine and verapamil do not inhibit the P-induced [Ca²⁺]i increase in human spermatozoa (58) is not clear. A recent study suggests that a prolonged incubation with nifedipine is required for an efficient inhibition of the P-induced Ca²⁺ influx and acrosome reaction in human spermatozoa (124). There may also be interspecies differences relating to the relative importance of the Ca²⁺ influx and the Ca²⁺ release mechanisms in the amplification of the initial steroid-induced Ca²⁺ signal.

The above mechanisms of Ca²⁺ signal generation and amplification were essentially discovered and analyzed using spermatozoa of several mammalian species as a model. However, fragmentary data that are available from studies in other systems reveal the presence of similar events, although the order in which individual mechanisms are employed is sometimes different. Therefore, in human oocytes the initial Ca²⁺ signal induced by 17ß-E₂ is also generated by the action of the steroid at the outer surface of the plasma membrane and relies on Ca²⁺ influx through a yet unknown type of channel; this influx is only secondarily amplified by Ca²⁺ release from intracellular stores (18). In contrast, the 17ß-E₂-induced [Ca²⁺]i increase in chicken granulosa cells appears to be primarily due to Ca²⁺ release from InsP₃-sensitive stores after the generation of InsP₃ from plasma membrane phosphatidylinositol 4,5-biphosphate by PIC stimulated by the steroid (6). However, measurable [Ca²⁺]i rises in human granulosa cells in response to androstenedione are precluded both by voltage-dependent Ca²⁺ channel blockers and thapsigargin (V. Machelon, F. Nomé, and J. Tesarik, unpublished observation), suggesting either that the initial limited Ca²⁺ influx (below the sensitivity threshold of the technique employed) is subsequently amplified by Ca²⁺-induced Ca²⁺ release from internal stores or that an initial limited Ca²⁺ release is amplified by capacitative Ca²⁺ entry (store depletion-induced opening of plasma membrane Ca²⁺ channels) from the extracellular space. Thus, the principle of cross-amplification of the initial steroid-induced Ca²⁺ signal by a combination of Ca²⁺ influx and Ca²⁺ release appears to be a widely spread characteristic of nongenomic steroid effects although different cell types may differ as to the source of the initial [Ca²⁺]i increase triggering the amplification reaction.

B. Ca²⁺ signal transduction
A key molecule realizing the downstream transduction of the steroid-generated Ca²⁺ signal (Fig. 2) is DAG resulting from the Ca²⁺-dependent activation of PIC and of phosphatidylcholine-specific phospholipase C (reviewed in Ref.127). In addition to the activation of PKC (see above), a major role of DAG in spermatozoa is to activate phospholipase A₂ (140) with an ensuing generation of fatty acids and lysophospholipids (141), which then act as direct effectors of the biological response (membrane fusion during the acrosome reaction).

Another mechanism by which the Ca²⁺ signal is transduced to effectors of membrane fusion is the hydrolysis of phosphatidylinositol-4,5-bisphosphate by the action of Ca²⁺-dependent PIC: the removal of phosphatidylinositol-4,5-bisphosphate alleviates its inhibitory effect on actin-severing proteins, which thus become active and depolymerize F-actin localized between the plasma membrane and the outer acrosomal membrane and representing a physical barrier to membrane fusion (135, 142). Last, but not least, the Ca²⁺-dependent, steroid-induced increase in sperm cAMP (79) is likely to activate cAMP-dependent PKA. In fact, the effect of FF (with P as the main active component) on the human sperm acrosome reaction can be inhibited by KT5720, a PKA inhibitor (97). Both of the latter two signal transduction pathways appear to require, in addition to Ca²⁺, a simultaneous stimulation through other, Ca²⁺-independent steroid response mechanisms, namely the protein tyrosine kinase system and the GABA_A-like receptor/Cl^- channel (see below).

C. The PTK system
The implication of protein tyrosine phosphorylation in transducing the P-generated signal in spermatozoa has been suspected since the observation that artificial, antibody-mediated cross-linking of steroid-receptor complexes on the sperm surface leads to an additional Ca²⁺ influx and induces the acrosome reaction in those spermatozoa that fail to respond to the addition of the hormone alone (89). A subsequent study confirmed the working hypothesis that, as in other systems in which ligand-receptor binding leads to receptor aggregation, P induces tyrosine phosphorylation in human spermatozoa; this study also pointed out a 94-kDa sperm phosphoprotein as a major Triton-soluble substrate for tyrosine phosphorylation in human spermatozoa (78). This phosphorylation was later shown not to be required for the initial P-induced Ca²⁺ influx (77, 143) and, inversely, external Ca²⁺ was not required for the P-induced stimulation of tyrosine phosphorylation (78).

A deeper analysis of the dynamics of the P effect on protein tyrosine phosphorylation suggests that this effect is less rapid than the P-induced Ca²⁺ response. This conclusion is based on the following observations. First, ligand-induced aggregation of a human sperm P receptor, visualized with the use of P conjugated with fluorescein-labeled BSA, needed several minutes to develop (62) as opposed to the several seconds that are required for the beginning of the P-induced Ca²⁺ response (see above). Second, the initial, PTK-independent P-induced [Ca²⁺]i increase, monitored in single sperm cells, is followed, after 2–10 min, by a second major [Ca²⁺]i increase, this time a PTK-dependent one (100). Interestingly, this delayed, PTK-dependent [Ca²⁺]i increase is only seen in about 10% of spermatozoa, whereas the initial, PTK-independent [Ca²⁺]i increase may be observed in 35% of the cells (100). Depending on the sensitivity setting of the assay system, this initial P-induced [Ca²⁺]i increase can be detected in up to 100% of viable spermatozoa exposed to this hormone (75, 137). Because so many spermatozoa can generate a Ca²⁺ response to P but so few (10%) actually bind the conjugate of P with fluorescein-labeled BSA (71), activate protein tyrosine phosphorylation (78, 100), and develop the biological response (71, 89), it has been suggested that spermatozoa possess at least two types of plasma membrane steroid receptor, one responsible for the Ca²⁺ influx, ubiquitous and undetectable by fluorescent steroid conjugate binding, and the other responsible for PTK activation, restricted to a small sperm subpopulation and detectable by fluorescent steroid conjugate binding (144). The stimulation of the former receptor may functionally activate the latter because a gradual increase in sperm Ca²⁺ load was related to the recruitment of new cohorts of spermatozoa capable of steroid conjugate binding followed by the acrosome reaction (74). The relatively slow response of the PTK-coupled steroid receptor may also partly explain the reported failure to observe an increase in tyrosine phosphorylation after a short (5 min) exposure of human spermatozoa to P (76).

The mechanism by which the steroid-generated signal is transduced downstream of PTK activation is largely unknown for the time being. The delayed, PTK-dependent [Ca²⁺]i increase observed in human spermatozoa (100) may be involved. If this is the case, this [Ca²⁺]i increase is likely to be due to Ca²⁺ release from internal stores rather than to Ca²⁺ influx because P exerts a genistein-sensitive stimulatory effect on the human sperm acrosome reaction even in the actual absence of extracellular Ca²⁺ provided that spermatozoa have been previously loaded with Ca²⁺ using a Ca²⁺ ionophore (77). Activation of an InsP₃-gated Ca²⁺ channel by tyrosine phosphorylation was reported in mouse T cells and was suggested to modulate intracellular Ca²⁺ release during T cell activation (145), and a similar mechanism may be involved in the opening of the InsP₃ receptor regulating Ca²⁺ release from the sperm acrosome. Furthermore, the P-activated PTK may be required, together with Ca²⁺, for the activation of PIC (135, 142), which is responsible for the generation of DAG that subsequently acts to activate phospholipase A₂, resulting in the generation of fatty acids and lysophopholipids needed for membrane fusion (see above).

There is also evidence for a cross-talk between the PTK and PKA signaling pathways because A-kinase anchor proteins (AKAPs), responsible for the targeting of PKA to specific subcellular locations, are major substrates of PTK in the Triton-insoluble sperm protein fraction (146). The phosphorylation status of AKAPs is controlled by Ca²⁺ via calmodulin-dependent protein tyrosine dephosphorylation (146). It remains to be determined whether AKAPs are also substrates of the P-activatable sperm PTK.

D. The GABA_A-like receptor/Cl^- channel
The involvement of Cl^- efflux (136) through the P-activated GABA_A-like receptor/Cl^- channel represents a well documented feature of the P-induced acrosome reaction (72, 85, 101). However, Cl^- fluxes may not explain all the effects supposedly mediated by the GABA_A-like receptor in spermatozoa, such as the increase in the intracellular concentration of cAMP (79) or the stimulation of DAG generation (127). In fact the GABA_A-like receptor of mammalian spermatozoa appears to be linked to the PTK system because the inhibition of PTK inhibits the GABA-induced generation of DAG in mouse spermatozoa (127) and the P-induced Cl^- efflux in human spermatozoa (147). It is possible that the ligand-bound GABA_A-like receptor comigrates with PTK, forming functionally active complexes in which PTK is activated and activates, in its turn, PIC responsible for DAG production and adenylate cyclase responsible for cAMP synthesis or inactivates phosphodiesterase, thus slowing down cAMP degradation. Alternatively, phosphorylation by PTK may be important for the function of the GABA_A-like receptor/Cl^- channel. Such a modulation has been reported for GABA_A receptors in neurons (148). These possible relationships still remain to be experimentally assessed.

	VI. Cross-Talk Between the Nongenomic and Genomic Responses of Cells to Steroids

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 
Unlike spermatozoa and oocytes, an active genomic steroid response mechanism has been found to coexist with the nongenomic one in some other cell types. This is the case, for instance, with granulosa, endometrial, and mammary gland cells. Data suggesting the possible cross-talk between genomic responses to steroid hormones and cell surface receptor-mediated signal transduction pathways are emerging for the mammary gland (149), breast cancer (150, 151, 152), and uterine cells (153, 154, 155). In these cell types, the cross-talk is realized by interactions between steroid-responsive genes and the cAMP system. Moreover, studies dealing with the mechanism of the cell proliferation-promoting effect of 17ß-E₂ in different types of cells reveal the implication of protein tyrosine phosphorylation (156, 157, 158, 159). Interestingly, 17ß-E₂ has been shown to activate the tyrosine kinase/p21^ras/mitogen-activated protein (MAP)-kinase pathway in a human mammary cancer cell line by acting at a cell surface-located variant of the classic ER (157). Thus, the same steroid ligand can activate, simultaneously or consecutively, a membrane-associated receptor and the classic nuclear steroid receptor (Fig. 2). In this scenario, the resultant global steroid effect on the cell will be a superimposition of the two distinct receptor-mediated events of which one may condition or modulate the other. In fact, the PKC pathway, known to be involved in some nongenomic steroid effects, potentiates androgen-mediated gene expression of a mouse vas deferens-specific protein (160), and the 17ß-E₂-binding mechanism of the human nuclear ER is regulated by tyrosine phosphorylation (161). It is not known whether, inversely, events set in motion by the genomic response mechanism can influence the nongenomic response. It remains to be determined how different cell types regulate the relative contributions of both kinds of response to steroids in concrete physiological and pathological situations.

	VII. Conclusions

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 
Several animal and human cell systems appear to be influenced in their function, and sometimes even in their morphology, by nonclassic steroid actions that occur after interaction between the steroid hormone and specific binding sites at the cell plasma membrane. Multiple-site receptor complexes showing binding domains for steroids, located closely to binding sites for peptidic substances, have been identified. They are probably far more diffused in nature than currently believed. Nongenomic surface receptors represent a portion of a complex system of cell regulation in which steroids and peptides not only act on the same target cell, but also involve the same intracellular messengers, modulating the same signaling pathways. Steroids with similar structure can sometimes modulate in a different way a given signal transduction system, finally exerting different, and often opposite, effects. Nongenomic and genomic actions may even synergize, leading to the appearance of biphasic effects that have both a rapid onset and a long-lasting persistence. A better knowledge of nongenomic actions of steroids is likely to open new perspectives in the understanding of animal and human physiology, as well as in the pharmacological treatment of some pathological conditions.

	Acknowledgments

 
The authors wish to thank Dr. Fausto De Vecchi and Dr. Gianluca Gennarelli for their valuable help and constructive criticism. A special thanks to Dr. Daniela Guidetti for drawing the figures.

	Footnotes

 
Address reprint requests to: Alberto Revelli, M.D., Department of Obstetrical and Gynecological Sciences, University of Torino, Via Ventimiglia 3, 10126 Torino, Italy.

¹ The writing of this review was supported by C.E.G. (Study Group in Gynecological Chronoendocrinology), Torino, Italy.

	References

Top Abstract I. Introduction II. Nongenomic Actions of... III. Nongenomic Actions of... IV. Nongenomic Actions of... V. Signal Transduction Pathways... VI. Cross-Talk Between the... VII. Conclusions References

 

1. Bagchi MK, Tsai MJ, O’Malley BW, Tsai SY 1992 Analysis of the mechanism of steroid hormone receptor-dependent gene activation in cell-free systems. Endocr Rev 13:525–535[CrossRef][Medline]
2. Orti E, Bodwell JE, Munck A 1992 Phosphorylation of steroid hormone receptors. Endocr Rev 13:105–128[Abstract]
3. Schuchard M, Landers JP, Sandhu NP, Spelsberg TC 1993 Steroid hormone regulation of nuclear proto-oncogenes. Endocr Rev 14:659–669[Abstract]
4. Truss M, Beato M 1993 Steroid hormone receptors: interaction with deoxyribonucleic acid and transcription factors. Endocr Rev 14:459–479[Abstract]
5. Ciocca DR, Roig LM 1995 Estrogen receptors in human nontarget tissues: biological and clinical implications. Endocr Rev 16:35–62[CrossRef][Medline]
6. Morley P, Whitfield JF, Vanderhyden BC, Tsang BK, Schwartz JL 1992 A new, nongenomic estrogen action: the rapid release of intracellular calcium. Endocrinology 131:1305–1312[Abstract]
7. Shears SB 1991 Regulation of the metabolism of 1,2-diacylglycerols and inositol phosphates that respond to receptor activation. Pharmacol Ther 49:79–104[CrossRef][Medline]
8. Eppig JJ 1991 Intercommunication between mammalian oocyte and companion somatic cells. Bioessays 13:569–574[CrossRef][Medline]
9. Colbran RJ, Soderling TR 1990 Calcium/calmodulin-dependent protein kinase II. Curr Top Cell Regul 31:181–221