|Home | About | Journals | Submit | Contact Us | Français|
Because sperm cannot synthesize new proteins as they journey to the egg, they use multiple mechanisms to modify the activity of existing proteins, including changes in the diffusion coefficient of some membrane proteins. Previously, we showed that during capacitation the guinea pig heterodimeric membrane protein ADAM1/ADAM2 (fertilin) transforms from a stationary state to one of rapid diffusion within the lipid bilayer. The cause for this biophysical change, however, was unknown. In this study we examined whether an increase in cAMP, such as occurs during capacitation, could trigger this change. We incubated guinea pig cauda sperm with the membrane-permeable cAMP analog dibutyryl cAMP (db-cAMP) and the phosphodiesterase inhibitor papaverine and first tested for indications of capacitation. We observed hypermotility and acrosome-reaction competence. We then used fluorescence redistribution after photobleaching (FRAP) to measure the lateral mobility of ADAM1/ADAM2 after the db-cAMP treatment. We observed that db-cAMP caused roughly a 12-fold increase in lateral mobility of ADAM1/ADAM2, yielding diffusion similar to that observed for sperm capacitated in vitro. When we repeated the FRAP on testicular sperm incubated in db-cAMP, we found only a modest increase in lateral mobility of ADAM1/ADAM2, which underwent little redistribution. Interestingly, testicular sperm also cannot be induced to undergo capacitation. Together, the data suggest that the release of ADAM1/ADAM2 from its diffusion constraints results from a cAMP-induced signaling pathway that, like others of capacitation, is established during epididymal sperm maturation..
Sperm must be able to adapt and function not only in the different environments within the male reproductive tract but also through the various milieus of the female reproductive tract. It is noteworthy that sperm accomplish these adaptations and changes in activity without synthesizing new proteins. Once sperm leave the testis, they shed their synthetic machinery to become the extremely streamlined gametes necessary to swim to and fertilize the egg. However, without the ability to produce new proteins, sperm must use other mechanisms to detect and respond to changing environments.
Three mechanisms sperm can use to affect functional changes on preexisting proteins are 1) posttranslational modifications (such as proteolytic cleavage or phosphorylation); 2) the relocalization of proteins from one domain on the sperm to another, thus allowing for new protein-protein and protein-lipid interactions; and 3) changes in a membrane protein's ability to diffuse within the lipid bilayer (such as the alteration of a protein from an immobile to a mobile state within the membrane) [1–12]. This last type of event can occur by a number of mechanisms, including loss of a binding interaction that restricts protein mobility (e.g., by posttranslational modification of the binding partner), a reduction or increase in cross-linking of a membrane associated network that restricts the free path of the protein, or a change in the surrounding lipid composition such as a change in liquid-ordered lipid domains that could potentially restrict long-range mobility of the protein.
ADAM1/ADAM2 (fertilin) is a heterodimeric, transmembrane protein, the exact function of which remains unclear but which appears important for producing functionally efficient sperm . ADAM1/ADAM2 may require all three of the previously mentioned mechanisms for it to become a fully functional protein. ADAM1 (fertilin α) is proteolytically processed when the sperm is in the testis, while ADAM2 (fertilin β) is proteolytically processed during the sperm's transit through the epididymis [6, 14–16]. Previously we showed that the proteolytic cleavage of ADAM2 coincides with the relocalization of ADAM1/ADAM2 from a whole head pattern to consolidation solely into the posterior head domain (PHD) . We demonstrated that the relocalization results from the redistribution of existing ADAM1/ADAM2 from the anterior head domain (AHD) to the PHD and is not due to the removal of ADAM1/ADAM2 from the AHD with concomitant exposure of a previously hidden population of ADAM1/ADAM2 within the PHD .
Using fluorescent redistribution after photobleaching (FRAP), we also determined that lateral fluidity of ADAM1/ADAM2 within the membrane exists in two different states depending on the functional stage of the sperm. ADAM1/ADAM2 is highly restricted from diffusing on testicular sperm (when it is found covering the whole sperm head) and in cauda epididymal sperm (when ADAM1/ADAM2 has consolidated into only the PHD). During capacitation, however, ADAM1/ADAM2 becomes highly mobile within the PHD membrane, and its diffusion increases even more after the sperm undergo the acrosome reaction (Fig. 1) .
Capacitation is the term given to the preparatory cellular events sperm undergo in the female reproductive tract before they are capable of fertilizing the egg [17–19]. Cellular changes in sperm at capacitation include but are not limited to hypermotility, increased protein tyrosine phosphorylation (pY), and membrane lipid reorganization, which culminate in the sperm becoming competent to undergo an acrosome reaction and fertilize the ovum . Capacitation can be induced in vitro for most sperm by incubating them in defined medium supplemented with a combination of calcium, bicarbonate, and BSA or 2-hydroxypropyl-β-cyclodextrin (βCD). BSA or βCD act by removing sterols from the membrane [21, 22], while one function of bicarbonate, enhanced by calcium, is to stimulate soluble adenylyl cyclases (sACs) within the sperm to increase the intracellular concentration of cAMP [23–28]. For sperm of many species, the bicarbonate requirement for capacitation can be bypassed if sperm are given a membrane-permeable cAMP analog along with a phosphodiesterase inhibitor (PDEi) to artificially increase their intracellular cAMP concentration [29–38]. The cAMP can then be used by various signaling pathways, one of which leads to the stimulation of protein tyrosine kinases (PTK) and results in enhanced pY formation [39, 40].
For this study, we sought to extend the understanding of the role of cAMP in capacitation and the biological events surrounding it that cause sperm to transform into fertilizing gametes. We initially tested whether exogenous cAMP can induce capacitation-associated changes in guinea pig sperm, as it does in other species, leading to hypermotility and competence to undergo an acrosome reaction. Guinea pig sperm, exceptionally, do not require exogenous calcium during in vitro capacitation . When calcium is present in the media, capacitated guinea pig sperm spontaneously undergo the acrosome reaction. We exploited this phenomenon to establish whether sperm incubated in exogenous cAMP achieve capacitation by determining the percentage of sperm in which acrosome reaction was induced on the addition of calcium. Because guinea pig sperm have exceptionally large acrosomes, distinguishing between acrosome-intact and acrosome-reacted sperm was easily done using light microscopy.
We then considered whether cAMP would also trigger the capacitation-associated conversion of ADAM1/ADAM2 from its restricted-diffusion state to its highly mobile state. We tested this directly by incubating cauda epididymal sperm with a cAMP analog and phosphodiesterase inhibitor and used FRAP to measure whether the diffusion properties of ADAM1/ADAM2 altered. Finally, we used testicular sperm, which are unable to undergo cAMP-stimulated capacitation until they have been epididymally matured, to test whether exogenous cAMP would affect the lateral mobility of ADAM1/ADAM2 in these less functionally mature cells.
All chemicals were purchased from Sigma (St. Louis, MO) unless otherwise noted.
Male Hartley guinea pigs (retired breeders) were purchased from Charles River Laboratories (Wilmington, MA) and housed in the Laboratory Animal Research Center at the Rockefeller University or at the animal facility at the University of Connecticut Health Center. The respective Institutional Animal Care and Usage Committees approved all animal protocols.
The mouse anti-guinea pig ADAM2 monoclonal antibody used has been described previously in Cowan and Myles . For the FRAP experiments, F(ab) fragments of this monoclonal antibody were directly conjugated with rhodamine and used as described in Cowan et al. .
The dissected epididymides were placed into either Hanks balanced salt solution lacking calcium (HBSS-Ca2+) for FRAP studies or Hepes-Mg2+ (1 M NaCl, 1 M KCl, 0.1 M Hepes, 0.1 M dextrose, 1 M MgCl2, pH 7.4) for the remaining studies. Sperm were collected by mincing the epididymides and allowing the sperm to swim free as well as being gently eased out of the lumen by applying slight pressure to the tubule fragments . The sperm were filtered through a 100-μm sieve to remove chunks of epididymal tissue, and the sperm concentration was determined using a hemocytometer.
Testes were removed from euthanized guinea pigs and decapsulated, and the freed seminiferous tubules were minced into fine pieces using razor blades and placed into HBSS-Ca2+ medium containing 0.5 μg/ml DNase I. Cells were pelleted through 7% Percoll in HBSS-Ca2+, resuspended in 60% Percoll in HBSS-Ca2+, and centrifuged at 12000 × g for 10 min. The band of testicular sperm was removed, diluted 10-fold in Hepes-buffered saline (pH 7.4), pelleted, and resuspended at 108/ml. For photobleaching experiments, sperm are subjected to several additional washes during antibody staining to ensure removal of Percoll, which has been reported to have effects on membrane structure . No significant difference was observed in lateral mobility of ADAM1/ADAM2 between testicular sperm and in elongating spermatids isolated without the use of Percoll , indicating that Percoll is not affecting the results.
Caudal epididymal or testicular sperm (0.5–1.0 × 107 sperm/ml) were washed into modified Tyrodes medium (mT) (112 mM NaCl, 2.8 mM KCl, 0.5 mM MgCl2, 25 mM NaHCO3, 5.6 mM glucose, 10 mM sodium lactate, 1 mM sodium pyruvate; pH 7.5); 1 mM dibutyryl cAMP (db-cAMP) was added from a 100-mM stock solution made fresh in 10 mM Pipes, pH 7.2. The working concentration of the PDEi papaverine (0.1 mM) was diluted from a 10-mM stock solution freshly made using a newly opened ampoule of DMSO. Sperm were incubated for 30–90 min at 37°C, depending on the experiment. For negative controls, parallel sets of sperm were incubated in mT alone or with vehicle solvents.
After 30 or 90 min of incubation with db-cAMP and papaverine, sperm samples were combined with an equal volume of mT medium supplemented with 4 mM Ca2+ (2 mM Ca2+ final) and incubated an additional 15 min at 37°C to induce the acrosome reaction. Vehicle-treated negative control sperm were similarly sampled and supplemented with calcium and incubated alongside the test samples. At the end of the incubation, samples were fixed in 3% formaldehyde, and the percent of acrosome-reacted sperm was quantified by microscopic examination. At least 200 sperm were counted from each sample.
Following 60 min of incubation with db-cAMP and papaverine, sperm prelabeled with rhodamine-conjugated anti-ADAM2 Fab antibody fragments  were prepared for FRAP analysis by mixing with agarose that had been heated and cooled to just above solidification temperature and then quickly spreading in a thin layer on a slide. For the epididymal sperm, only cells with embedded heads but freely moving tails were chosen for FRAP examination to eliminate the risk of examining dead cells . Testicular sperm exhibit little mobility, although in most cases some infrequent twitching spasms were observed. Cells for analysis, therefore, were selected on the basis of exhibiting normal morphology with intact flagellum and acrosome.
Fluorescence redistribution after photobleaching experiments were performed using a Zeiss LSM510 laser scanning confocal microscope [44, 45]. Images were collected using a 63× 1.4 N.A. planapochromat objective with the pinhole fully opened to ensure collection of fluorescence through the entire depth of the specimen. The scan axis was rotated such that the x-axis of the scan was parallel to the long axis of the sperm head, and a region consisting of approximately one-third of the middle section of the sperm head was photobleached using 100% laser power. The first postphotobleach image (t = 0 sec) was collected within 3 msec of the end of the photobleach, and subsequent postphotobleach images were collected immediately thereafter. For epididymal sperm, the time per image ranged from 1.1 to 1.9 sec; for testicular sperm, the time per image ranged from 2.3 to 3.0 sec. Image analysis was performed essentially as described in Cowan et al. . Background values were subtracted from all images, and average intensity values were calculated perpendicular to a line drawn down the long axis of the sperm head.
The analysis of the fluorescence redistribution was performed with the software system MLAB, constructed by Civilized Software, Inc. (Bethesda, MD). The diffusion coefficient (D) and percent mobile fraction (%R) were calculated on the basis of a normal-mode analysis as previously described . Briefly, we treat the data as a normal-mode analysis on a spherical surface of radius r. For an initial condition with bleaching symmetric about the scanning axis and a monodisperse redistribution function, we have
where c(x, t) is the concentration distribution of the fluorophore at time t, D is the diffusion coefficient of the fluorophore, and P0(x) and P2(x) are the zero- and second-order Legendre polynomials, respectively . For a polydisperse redistribution function, we have 
where . For a simple case of a diffusion coefficient and an immobile fraction, we add a constant to μ2(t) ,
where R is the so-called mobile fraction.
As was previously found, lateral mobility of ADAM1/ADAM2 in testicular sperm was highly restricted, and because there can be some motion in these cell preparations, we were not able to follow the fluorescence redistribution for a sufficiently long time to obtain accurate values for the fractional recovery. In this case we used a method of cumulants  to compute a value for mobility, , equivalent to the average diffusion coefficient of a polydisperse distribution of all labeled molecules of that cell. We use the method of cumulants, where we define the cumulant generating function ,
where the mth cumulant of Γ, Km(Γ), is the coefficient of (−3t)m/m! in the MacLaurin expansion of , which thus takes the form of
For a single exponential function, K(−3t; Γ) is linear in t. Deviations from “exponentiality” of μ2(t) thus appear as higher order terms in t.
We fit μ2(t) to an approximate function, that is, A[RX exp(−3Γt) + 1 – R)], where Γ = 2D′/r2 and R is a “mobile” fraction. From there, we then calculate
which leads to
Statistical significance between groups was determined using the nonparametric Mann-Whitney U-test. Standard deviations are reported for all the FRAP and acrosome reaction data presented.
We first determined whether cAMP could induce capacitation-associated competence for the acrosome reaction in guinea pig sperm as it does in several other species. Unlike sperm of many other species, guinea pig sperm do not require an exogenous calcium supply during in vitro capacitation. The addition of calcium to the media causes capacitated guinea pig sperm to spontaneously undergo the acrosome reaction and can therefore be used as a bioassay to determine the proportion of sperm that have completed capacitation at any given time point.
For each experiment, cauda epididymal sperm were isolated and incubated with the membrane-permeable cAMP analog, dibutyryl cAMP (db-cAMP), and the PDEi papaverine to determine if this caused the sperm to capacitate. Negative control sperm were incubated with vehicle alone. Based on concentrations of db-cAMP and papaverine used to trigger signaling events of capacitation and/or fertilization in gametes of other species [29, 37, 49–54], we incubated 107 cauda sperm/ml in 1 mM db-cAMP and 100 μM papaverine at 37°C. After 30 or 90 min of incubation, calcium was added to the samples to induce the fully capacitated sperm to undergo the acrosome reaction. The sperm were then fixed and the percentage of acrosome-reacted sperm determined. We found on average 60%–67% of sperm to be acrosome-reacted regardless of whether the sperm were treated with db-cAMP for 30 or 90 min (Table 1). To determine if we were working at suboptimal concentrations, we tested the db-cAMP at concentrations of 2, 3, and 10 mM. None of these concentrations improved the level of acrosome reactivity within the population. In fact, the acrosome reactivity averages obtained for the 2- and 3-mM db-cAMP were slightly lower (46% and 53%, respectively, at 90 min). The 10-mM concentration of db-cAMP resulted in lowered sperm motility. Decreasing the concentration of db-cAMP to 0.1 mM resulted in levels of acrosome reaction that were only slightly higher than background. Sperm motility was also assessed by microscopic examination during cAMP incubation. By the 30-min time point, the 1-mM db-cAMP treatment had induced hypermotility, evidenced by vigorous tail thrashing. No hypermotility was observed in the vehicle control samples.
Previously we found, using fluorescence redistribution after photobleaching, that the lateral diffusion of ADAM1/ADAM2 within the membrane is highly restricted prior to capacitation, with the majority of ADAM1/ADAM2 showing little or no movement . However, ADAM1/ADAM2 becomes highly diffusible within the membrane during capacitation and remains exceedingly mobile even after the acrosome reaction (Fig. 1) . We thus asked if this increased diffusion for ADAM1/ADAM2 could be induced in cauda sperm by artificially increasing the intracellular levels of cAMP. To test this, we incubated cauda sperm with db-cAMP and papaverine and then determined the D for ADAM1/ADAM2 using FRAP analysis.
Figure 2 (top row) shows four images from a time series of a representative FRAP experiment (the entire times series can be viewed as a movie in Supplemental Movie 1, available online at www.biolreprod.org ). A stripe running anterior-posterior in the posterior head area was photobleached, as indicated in the first postphotobleach image at t = 0 sec. The fluorescence within the photobleached area quickly began to return, and by 15 sec redistribution was essentially complete. An analysis of this experiment is shown in Figure 3 (top row). The leftmost panel (A) shows the fluorescence intensity of ADAM1/ADAM2 averaged along the anterior-posterior axis of the posterior head domain at each point in a line perpendicular to the anterior-posterior axis at each time point. Panel B shows at higher resolution three individual scans from panel A at three selected time points: the initial fluorescence intensity in the PHD prior to bleaching (circles), the loss of fluorescence intensity in the region of the PHD immediately after bleaching (squares), and the final scan after bleaching (triangles) showing the recovery of fluorescence intensity in the bleach region concomitant with a loss of fluorescence intensity in the nonbleached regions. By the 78-sec time point, the fluorescence intensity within the bleached area has equilibrated essentially to its initial distribution, revealing rapid diffusion within the PHD, although the overall level of fluorescence has been reduced because of the photobleaching. Fluorescence redistribution was analyzed on the basis of fitting the data to the change in the second moment of the distribution function over time (panel C) and from which both D as well as the %R (the fraction of the ADAM1/ADAM2 population exhibiting diffusion) were determined .
The average lateral diffusion of ADAM1/ADAM2 on cauda sperm treated with db-cAMP and papaverine was (9.8 ± 4.2) × 10−10 cm2/sec and had an averaged %R of 62 ± 16% (n = 22). Thus, the majority of the ADAM1/ADAM2 population (62%) on a sperm was found to be highly diffusible after cAMP treatment, while the remaining 38% of the population was immobile under these conditions. This increase in D was an order of a magnitude larger than our previous measurements of ADAM1/ADAM2 on untreated cauda sperm (0.8 × 10−10 cm2/sec) with a larger portion of the ADAM1/ADAM2 population diffusing than what we had seen previously (%R = 45) as well . Parallel samples of these treated sperm were incubated in calcium to determine the percent acrosome reaction for each experiment and to prove that the drugs added were efficacious. The average acrosome reactivity in these experiments was 64%.
On testicular sperm, ADAM1/ADAM2 is localized over the entire sperm head and shows little to no lateral diffusion within the membrane (Fig. 1) . Until sperm released from the testis undergo epididymal maturation, they cannot be induced to capacitate or undergo the acrosome reaction . It is hypothesized that this is because the cAMP signaling pathway(s) of capacitation require epididymal maturation before they are fully established. We thus tested whether cAMP would induce the rapid diffusion of ADAM1/ADAM2 in testicular sperm as it did on epididymal sperm. We performed FRAP analysis of ADAM1/ADAM2 on testicular sperm treated with db-cAMP and papaverine as described previously. Although ADAM1/ADAM2 is found over the entire head of testicular sperm, we analyzed its mobility specifically in the posterior head region, where it is located on cauda sperm, although a photobleached stripe was performed along the entire long axis of the sperm head (see Fig. 2 and Supplemental Movie 2, available online at www.biolreprod.org ). In comparison to the findings for cauda sperm, we found that db-cAMP caused only a small increase in the lateral mobility of ADAM1/ADAM2 within the membrane. As shown in Figure 2 (bottom panels), redistribution into the bleach region was very slow and incomplete during the experiment; after more than 100 sec, the bleached region was still somewhat apparent in the fluorescence image. These results are quite similar to redistribution times of control testicular sperm, as the example in panels G and H of Figure 3 show.
The lateral mobility of ADAM1/ADAM2 was at such a low level on the testicular sperm with or without cAMP treatment that redistribution was often not complete during the time scale of the experiment, as indicated in panels F and I of Figure 3, where we see no evidence of a plateau in the distribution function during the time course of the experiment. As a result, we were unable to calculate the %R or D of ADAM1/ADAM2 for these cells directly. To directly compare these cells with the db-cAMP-treated cauda sperm as well as with our previously reported FRAP analyses of ADAM1/ADAM2 on sperm stages , we used a previously described method of cumulants  to assess lateral mobility by measuring the parameter (see Materials and Methods) for all experiments. These quantified averages showed that ADAM1/ADAM2 on cAMP-treated testicular sperm ([1.1 ± 1.3] × 10−10 cm2/sec) did not undergo the dramatic increase in mobility that was seen on cAMP-treated cauda sperm ([5.9 ± 2.7] × 10−10 cm2/sec) and that this difference was statistically highly significant (P < 0.0001). The addition of cAMP to testicular sperm resulted in modest yet significant (P < 0.03) increase in the of ADAM1/ADAM2 compared to controls ([0.29 ± 0.28] × 10−10 cm2/sec); however, the overall mobility of ADAM1/ADAM2 remained highly restricted, and the was not statistically different (P > 0.3) from ADAM1/ADAM2′s highly restricted state measured on untreated cauda sperm ([0.5 ± 0.41] × 10−10 cm2/sec). Finally, because testicular sperm cannot acrosome react, we verified that the db-cAMP and papaverine added in these experiment were biologically active by incubating parallel samples of cauda sperm, collected from the same animals, with the same drug preparations followed by calcium to induce the acrosome reaction. In all experiments, this bioassay showed that the db-cAMP and papaverine preparations used were biologically effective.
In Figure 4 we compiled all the results for ADAM1/ADAM2 from our current FRAP experiments and plotted them in context with past FRAP experiments after recalculating those data using the method of cumulants. The figure shows that ADAM1/ADAM2 remains highly restricted from diffusing until the sperm undergo capacitation, but if cauda sperm are treated with exogenous cAMP, the restriction of lateral movement is lifted, and the heterodimeric protein increases its rate of diffusion over 10-fold. This dramatic enhancement of diffusion by cAMP, however, is not inducible prior to epididymal maturation, as evidenced by the low mobility of ADAM1/ADAM2 on testicular sperm even after their exposure to cAMP.
Previously, we showed that ADAM1/ADAM2 undergoes a change in its diffusion state during capacitation . Because intracellular cAMP rises sharply during the early stages of capacitation, we asked in this study if cAMP signals this increased diffusion rate of ADAM1/ADAM2. Our results show that increasing cAMP levels causes ADAM1/ADAM2 on cauda sperm to switch from its relatively immobile diffusion state to its freely diffusing state. There is precedent for cAMP control of the diffusion state of a membrane protein. Through FRAP analysis, the diffusion coefficient (D) for the membrane protein aquaporin 2 (AQP2) was determined in LLC-PK1 (porcine kidney) cells stably transfected with GFP-AQP2. D was found to be 10-fold fewer cells when treated with the cAMP agonist, forskolin, versus untreated controls . The hypothesized mechanism of action in that case was a cAMP-generated phosphorylation of AQP2, leading to increased AQP2 interaction with cytoskeletal proteins, resulting in the impedance of its diffusion . Although we see the opposite effect on ADAM1/ADAM2 on cauda sperm (cAMP causes an increase in the diffusion coefficient of ADAM1/ADAM2), the compelling aspect of the data is that cAMP in two different systems appears to act as the switching mechanism that alters a protein's membrane diffusion.
The molecular mechanism by which cAMP effects a change in lateral mobility of ADAM1/ADAM2 is unclear. Capacitation is known to entail changes in lipid composition and to affect lipid order in the sperm plasma membrane of a variety of species [9, 50, 53, 56–63]. It has been previously reported that during boar sperm capacitation, the plasma membrane architecture is modified through a translocation of phospholipids to the outer leaflet and that cAMP can trigger this lipid scrambling [53, 58, 59, 64]. We do not know if a similar alteration in lipid architecture occurs in guinea pig sperm. However, it is unlikely that such an increase in lipid disorder would be sufficient to account for the changes in the lateral mobility of ADAM1/ADAM2 observed in this report. On testicular sperm and epididymal sperm, ADAM1/ADAM2 is essentially immobile. In the absence of another mechanism, this would mean a complete lack of membrane fluidity. Furthermore, previous studies in guinea pig [2, 65], as well as other species , revealed no large increases in global fluidity of the sperm plasma membrane either in different domains or following capacitation. In fact, lipid probes, such as C14diI, have actually shown a moderate, 2- to 3-fold decrease in diffusion coefficient between testicular sperm and cauda epididymal sperm  and a further 2-fold decrease in D on acrosome-reacted cells .
A more likely explanation for the large (one to two orders of magnitude) increase in lateral mobility of ADAM1/ADAM2 is an alteration in a relatively high-affinity protein-protein or protein-carbohydrate interaction. This hypothesis is strengthened by findings involving the protein PH-20, which is known to occupy the same membrane domains at the same times as ADAM1/ADAM2 . PH-20 is largely immobile on testicular sperm, but on cauda epididymal sperm, it exhibits lateral mobility typical of many membrane proteins (D = 1.8 × 10−10 cm2/sec and %R = 73%), with a further increase in D after the acrosome reaction [2, 67]. Given that two proteins that occupy the same plasma membrane domain exhibit distinctly different temporal patterns for large changes in lateral mobility, we suspect that these are specific effects related to each protein individually rather than a global change in the lipid milieu. The most straightforward hypothesis, therefore, is that ADAM1/ADAM2 complexes with an immobile sperm component and that this binding interaction ceases to occur following cAMP signaling events.
Using the standard bicarbonate and BSA method of in vitro-induced capacitation, we previously observed two subpopulations of sperm (Fig. 4, bracketed data). In one population, ADAM1/ADAM2 was highly mobile, with average D similar to those for ADAM1/ADAM2 on acrosome-reacted sperm. In the second population, ADAM1/ADAM2 showed little or no diffusion, with an average D similar to the constrained mobility of ADAM1/ADAM2 measured for untreated cauda sperm . We hypothesized that these two sperm populations reflected the inefficiency of the in vitro method at inducing capacitation in 100% of the sperm. This was supported by the fact that on average only 40%–50% of these in vitro-capacitated sperm underwent the acrosome reaction when subsequently incubated with calcium chloride . Because FRAP analysis requires that we immobilize swimming sperm in an agarose matrix, we could not use hypermotility as an indicator of capacitation. Therefore, without a means of distinguishing between capacitated and noncapacitated sperm, our results likely contained a mixture of sperm from both states. We speculated that sperm with ADAM1/ADAM2 exhibiting restricted diffusion were sperm that had not capacitated, while the sperm with freely diffusing ADAM1/ADAM2 were those that had capacitated.
Interestingly, we did not observe two such subpopulations in the cAMP-treated cauda sperm; only one population was detected, and it corresponded to the previously mentioned group with freely diffusing ADAM1/ADAM2. The lateral mobility for ADAM1/ADAM2 after cAMP treatment, when calculated as D or as , was 12 times greater than the average mobility measured on untreated cauda sperm. We posit that the release of ADAM1/ADAM2 from some constraint is a cAMP-modulated event and that a threshold of intracellular cAMP must be reached before ADAM1/ADAM2 is released. During capacitation, sAC is stimulated in each sperm to produce cAMP. However, under conventional in vitro-capacitating conditions, each sperm reaches this putative cAMP threshold concentration at a different time because of its individual response to the stimulation of sAC. As a result, at any given sampling, two populations of sperm could be seen: 1) those sperm that have achieved the cAMP threshold and have released ADAM1/ADAM2 to diffuse readily and 2) those sperm whose intracellular levels of cAMP are still below this threshold and so maintain restricted diffusion of ADAM1/ADAM2. When membrane-permeable exogenous cAMP is given to the sperm, all the sperm simultaneously attain the same intracellular level of cAMP. If this concentration of cAMP is above the required threshold in the sperm, then ADAM1/ADAM2 would be converted on all the sperm from the constrained state to the freely diffusing state. This scenario could potentially explain why we detected only one population of sperm with freely diffusing ADAM1/ADAM2. However, even though a threshold may have been reached to convert ADAM1/ADAM2 to a highly mobile state in all the sperm examined, it was not enough to make all the sperm ready to acrosome react. When this population was challenged with calcium chloride, only about 60% of the sperm underwent the acrosome reaction. Therefore, the release of ADAM1/ADAM2 from its restraints may be an event that precedes final acrosome-reactive competence.
We think it is unlikely that cAMP directly releases ADAM1/ADAM2 from its constraints; instead, we postulate that it more likely stimulates a pathway that leads to this event. Cyclic AMP might trigger changes in phosphorylation, as has been suggested for AQP2, or might cause changes in intracellular calcium levels. Treating cauda sperm with the calcium ionophore A23187 and exogenous calcium induces the acrosome reaction in noncapacitated guinea pig cauda sperm [68, 69]. ADAM1/ADAM2 on such sperm is found to be highly mobile within the membrane. It remains unclear precisely how calcium and A23187 enhance ADAM1/ADAM2 diffusion, but collectively the data suggest that both cAMP and calcium have roles in this event.
In an endeavor to elucidate the exact mechanism involved, we examined guinea pig sperm for enhanced protein tyrosine phosphorylation during capacitation but found that guinea pig sperm, unlike mouse and several other species of sperm, showed little or no increase in pY formation during capacitation even though on the same blot, mouse sperm showed an intense increase in pY formation. This does not rule out a role for pY formation during capacitation in guinea pig sperm, but it does indicate yet another difference between these two sperm species. We also tried 2-D gel immunoblotting of acrosome-reacted and acrosome-intact sperm, looking for subtle difference in ADAM2 that might correspond with the change in the diffusion state of ADAM1/ADAM2, but our results were inconclusive. Finally, we have performed yeast two-cell hybrid analysis using the cytoplasmic tail of ADAM2 to try to identify a binding partner that may be anchoring ADAM1/ADAM2 in its constrained state. Although no clear partner has been identified, a few candidates and several unknown proteins warrant further study. Interestingly, among the potential positive and numerous false positives identified, we found an abundance of divalent-cation-binding proteins, often specifically for calcium (unpublished data). This seems to further suggest that ADAM1/ADAM2 may be regulated through an interaction with a calcium-binding protein that may be cAMP regulated.
Regardless of the exact nature of the release event, the cAMP-induced increase in ADAM1/ADAM2 diffusion occurs only on sperm that have completed epididymal maturation. Treatment of testicular sperm (which cannot undergo capacitation or an acrosome reaction—even through calcium ionophore) with db-cAMP and papaverine did not result in significant alteration of the lateral mobility of ADAM1/ADAM2. It is possible that the mechanism that immobilizes ADAM1/ADAM2 within the membrane of testicular sperm is different than the one that keeps ADAM1/ADAM2 immobile on cauda sperm. ADAM1/ADAM2 on testicular sperm is found over the whole head, while on cauda sperm the protein is confined to the PHD . Although two separate mechanisms might be involved in confining ADAM1/ADAM2 in testicular and cauda sperm, the simplest explanation would involve just one tethering mechanism, the release from which is not possible until sperm have completed epididymal maturation. This requirement may be a means to prevent premature release and early activation of the ADAM1/ADAM2, a protein that may have a late-acting function, when sperm are in the female reproductive tract.
The functional consequences of modulating the diffusion of ADAM1/ADAM2 are not clear. The ADAM proteins represent a large family of proteins that are present in sperm but whose roles in fertilization events are not well understood. Mice lacking functional genes for ADAMS 1b, 2, or 3 are infertile, but given that multiple ADAMs appear to interact in ways that stabilize/destabilize each other's surface expression, there remain multiple functional mechanisms that could account for the infertility of the knockout mice. Guinea pig sperm do not appear to express ADAM3, so there are also likely differences in the function of these proteins between mouse and guinea pig. Nevertheless, ADAM1/ADAM2 is a major membrane protein of the sperm surface, and the dramatic change in lateral mobility we have observed is a significant event in capacitation. The magnitude of the change in lateral mobility of ADAM1/ADAM2 is very large. An individual protein that exhibits D in the range of 10−9 cm2/sec would be expected to traverse an area the size of one-half the posterior head domain (~5 μm2) in approximately 40 sec; for D in the range of 10−10 cm2/sec, it would take 7 min, and for D in the range of 10−11 cm2/sec, it would take 70 min. In testicular and cauda epididymal sperm, most of the ADAM1/ADAM2 molecules are even less mobile than 10−11 cm2/sec. Thus, if ADAM1/ADAM2 were involved in a binding event that required recruitment of multiple proteins to the site of interaction, the recruitment event would be unlikely to occur in sperm prior to capacitation but could occur on a physiological time scale after capacitation and/or the acrosome reaction, when D is in the range of 10−9 cm2/sec with a high mobile fraction .
These results have positioned us to explore thoroughly the steps in the cAMP-signaling pathway leading to the release and increased diffusion of ADAM1/ADAM2 during capacitation. Further studies may shed light on the operative advantage that this increase in lateral movement provides for ADAM1/ADAM2. Also, more generally, they would provide information regarding previously unknown mechanisms by which synthetically inert sperm can modulate activity of surface proteins and subsequently sperm function.
We would like to thank Mr. Evan Read for his help with the manuscript illustrations.
1Supported by the National Institute of Health research grant HD038807.