| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Institute of Biomembranes, Department of Biochemistry and Cell Biology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
b Graduate School of Animal Health, Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
c Laboratory of Gamete Signaling, The Babraham Institute, Cambridge, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
acrosome reaction, apoptosis, calcium, cyclic adenosine monophosphate, sperm capacitation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A particular bicarbonate-induced alteration noted was an increase in membrane phospholipid packing disorder as detected by the sperms' ability to bind the fluorescent amphiphilic probe merocyanine 540 [9]. The increased disorder was found to be concomitant with a change in the transport and transverse membrane distribution of C6NBD-labeled phospholipids [7]. Unstimulated sperm cells (i.e., incubated in the absence of bicarbonate) showed a high degree of lipid asymmetry, added C6NBD-PE and C6NBD-PS rapidly translocating to the inner leaflet of the plasma membrane while C6NBD-PC and C6NBD-SM remained in the outer leaflet [10]; phospholipid analogues translocated to the inner leaflet did not reappear to any extent in the outer leaflet. Incubation in the presence of bicarbonate/CO2, however, led to a collapse in the asymmetric distribution, with a 10-fold reduction in the rate of C6NBD-PS and C6NBD-PE translocation to the inner leaflet and a marked increase in the proportion of C6NBD-PC and C6NBD-SM locating to the inner leaflet [7]. At the same time, phospholipids preloaded into the inner leaflet were translocated back to the outer leaflet. From these observations, we deduced that the bicarbonate-stimulated protein phosphorylation pathway was leading to an activation of phospholipid scramblase, the enzyme that translocates phospholipid species back and forth across the membrane lipid bilayer [11].
Scramblase activation is generally considered to lead to a collapse of asymmetry (i.e., equilibration of the various phospholipid species between the two lipid leaflets) and a resultant exposure of phosphatidylserine (PS) and phosphatidylethanolamine (PE) at the outer surface. While the use of C6NBD-phospholipid analogues can give information on phospholipid translocation dynamics, the technique cannot actually demonstrate the occurrence of phospholipid scrambling. It has important drawbacks compared with techniques that directly detect these exposed aminophospholipids (e.g., use of fluorescein-conjugated annexin V, which labels specifically surface-exposed PS [11, 12]), especially with respect to studies on spermatozoa. 1) C6NBD-phospholipids are not natural substrates and, due to their conjugated acyl fluorochrome moiety, may behave differently from endogenous phospholipids. 2) The technique only provides information about the mean proportional transverse distribution of the various C6NBD-phospholipid classes in the sperm plasma membrane within the whole sperm population. Direct labeling of exposed endogenous aminophospholipids, on the other hand, provides information about the size of the subpopulation of sperm cells showing active phospholipid scrambling. 3) C6NBD-phospholipid labeling is not suitable for detecting cellular locations of phospholipid scrambling. 4) Albumin, a regular component of in vitro fertilization (IVF) media, cannot be included in systems using C6NBD-phospholipid analogues to measure phospholipid asymmetry because of its high affinity for lipids. Labeling of exposed endogenous aminophospholipids, on the other hand, can be done in the presence of albumin. 5) Due to their amphiphilic properties, C6NBD-phospholipid analogues that have translocated to the inner leaflet of the plasma membrane can diffuse across the cytosol into intracellular membranes, thereby distorting evaluations of plasma membrane phospholipid asymmetry.
Using C6NBD-phospholipids, though we deduced that bicarbonate was inducing an activation of scramblase, we were unable to detect any change in steady-state aminophospholipid distribution. In the present study, we have used direct labeling with annexin V-FL (fluorescein-conjugated annexin V, which specifically detects PS [12, 13]) and Ro-SA-FL (a stable complex of biotinylated Ro09-0198 and fluorescein-conjugated streptavidin, which specifically detects PE [14]) to demonstrate that bicarbonate does indeed induce aminophospholipid exposure in boar spermatozoa as a result of phospholipid scrambling.
Changes in the transverse distribution of membrane phospholipids as a result of scrambling is observable in a variety of cellular phenomena, including cell adhesion [15] and exocytosis [16]; however, it is often considered to be an early indicator of cell apoptosis [17, 18]. While Weil et al. [19] were unable to find any evidence for the involvement of caspases (key apoptotic catalysts) in mouse sperm death, a number of other recent studies have used the detection of PS exposure in sperm samples as evidence of an apoptotic change. Because incubation in the presence of bicarbonate does appear to enhance sperm cell death [2, 7], we have examined the possibility that bicarbonate's effect on phospholipid scrambling is part of an apoptotic response. We found no evidence of either DNA fragmentation [20] or mitochondrial membrane depolarization [21], and we were unable to block the bicarbonate-induced aminophospholipid exposure by prior treatment of the spermatozoa with broad-spectrum caspase inhibitors.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Fluorescein-conjugated annexin (annexin V-FL) was kindly donated by Dr. Reutelingsperger (Academic Hospital, Maastricht, The Netherlands). Fluorescein-conjugated streptavidin/biotinylated Ro09-0198 complex (Ro-SA-FL) was kindly donated by Dr. Umeda (RINSHOKEN, Tokyo, Japan). R-phycoerythrin peanut agglutinin (PNA-RPE) was prepared as previously described [7]. Fluorescent-conjugated peanut agglutinin (PNA-FL) was obtained from EY Laboratories, Inc. (San Mateo, CA). Propidium iodide (PI), merocyanine 540 (MC540), Yo-Pro-1, Mitotracker Red, and acridine orange were from Molecular Probes Europe (Leiden, The Netherlands). The TUNEL apoptotic cell death fluorescein detection kit for flow cytometry was from Roche Diagnostics (product 1 684 795; Almere, The Netherlands). The broad-spectrum caspase inhibitor zVAD-fmk was from Promega Corporation (Madison, WI), Boc-D-fmk was from Alexis Biochemicals (Lausanne, Switzerland). Filipin, staurosporine, cycloheximide, and the protein phosphorylation effectors (see Table 1) were from Sigma-Aldrich Chemie BV (Zwijndrecht, The Netherlands). Bovine serum albumin (BSA) was obtained from Sigma (product A4503) and then defatted according to Chen [22].
|
Sperm Cell Preparation
Sperm-rich fractions of boar semen were collected from highly fertile Dutch Landrace boars kept at the Collaborative Artificial Insemination Centre, Bunnik (The Netherlands). After collection, the semen samples were diluted and stored in Beltsville thawing solution as described previously [10].
For experimentation, spermatozoa were isolated by centrifugation through a two-step discontinuous gradient of 35% and 70% isotonic Percoll as previously described [9]. After removal of the supernatant layers, the resultant loose pellet was resuspended in residual 70% Percoll (final concentration about 6 x 108 cells/ml).
Incubations
The basal media used were the Hepes-buffered bicarbonate-free Tyrode medium (HBT) and its bicarbonate-supplemented form (HBT-Bic), both detailed by Gadella and Harrison [7]. For the present experiments, unless otherwise indicated, both were supplemented with 3 mg/ml BSA and with 2 mM CaCl2 (to allow binding of annexin V). Effectors and inhibitors were included in these media as described in the relevant Results sections.
Incubations were initiated by addition of aliquots of washed spermatozoa to the prewarmed media (final concentration 107 cells/ml). The sperm suspensions in HBT were incubated at 38°C for up to 4 h in air-tight closed tubes under normal humidified atmosphere, whereas sperm in HBT-Bic were incubated at 38°C under humidified 5% CO2 in air. At designated intervals, subsamples were removed for assay.
Annexin V Staining
In order to detect surface exposure of PS, sperm samples taken from incubating suspensions were labeled for 5 min at 38°C with a combination of 1 µg/ml annexin V-FL, 1 µg/ml PNA-RPE (to discriminate between acrosome-intact and acrosome-deteriorated cells) and 1 µg/ml PI (to discriminate between live and dead cells); suspensions were maintained under a humidified atmosphere in equilibrium with 5% CO2 where appropriate.
Flow cytometric analysis was performed on a FACScalibur flow cytometer equipped with a 100-mW argon laser exciting at 488 nm (Becton Dickinson, San Mateo, CA); the flow cell and sampling system was preequilibrated at 38°C [7]. Annexin V-FL fluorescence was detected on FL-1 (530/30-nm band-pass filter), while PNA-RPE and PI fluorescence was detected on FL-3 (620-nm long-pass filter). Fluorescence data were collected in logarithmic mode while light-scatter data were collected in linear mode. As the sample passed through the machine prior to data collection, nonsperm scatter events were gated out; two-dimensional dot plots were then set up with FL-1 data expressed on the x-axis and FL-3 data on the y-axis; because sperm cells with red fluorescence (PNA and/or PI-positive) were considered as degenerate and thus to be excluded [7], computer-generated boundaries were set to define the sperm subpopulation that did not expose PS (low staining with annexin V-FL) and the subpopulation exposing PS (high staining with annexin V-FL); finally, data events representing 10 000 live acrosome-intact sperm cells were collected, from which the relative proportions of the two subpopulations were calculated.
For confocal microscopy, sperm suspensions incubated and labeled as described above were transferred to a heated sample chamber attached to a Leica TCS-SP inverted spectral emission confocal laser-scanning microscope (Leica GmbH, Heidelberg, Germany). The sample chamber, which sat on the cross-table, consisted of a temperature-controlled block maintained at 38°C whose center contained a sealed coverslip on which the sample was mounted. The surface of the sample was flushed continually with warmed humidified air with or without 5% CO2 as appropriate; the microscope's objectives were also maintained at 38°C by means of a warming ring. Using single 488-nm excitation scans via a 63x APO-water immersion objective, annexin V-FL labeling was detected in PMT-1 (emission range 510–540 nm) while PNA-RPE and PI labeling were detected in PMT-2 (emission range 580–700 nm); each scan took less than 0.1 sec. Cells negative for PNA and/or PI were considered as live acrosome intact.
Ro-09-0198 Staining
To detect surface exposure of PE, sperm samples were stained in a similar manner to that described for annexin V staining except that 0.5 µg/ml Ro-SA-FL replaced annexin V-FL. Stained suspensions were analyzed using both flow cytometry and confocal microscopy as described above.
Merocyanine 540 Staining
To detect increases in plasma membrane lipid packing disorder, sperm samples were stained for 10 min with 2.7 µM merocyanine 540, 25 nM Yo-Pro-1, and 10 µg/ml PNA-FL and then subjected to flow cytometric analysis at 38°C as previously described [7, 9].
Visualization of Cholesterol Distribution by Filipin
Filipin is an antibiotic that forms complexes with nonesterified membrane cholesterol ([23] and references therein). Samples from incubating sperm suspensions were fixed by dilution with an equal volume of 4% (w/v) glutaraldehyde in 0.15 M Na-cacodylate/HCl buffer, pH 7.4. After 30 min of gentle shaking, the cell suspensions were washed twice (500 x g, 15 min) in 0.15 M cacodylate buffer and then resuspended in cacodylate buffer containing 25 µM filipin (added from a 10 mM stock solution in dimethyl sulfoxide [DMSO]); as controls, parallel samples were treated with similar DMSO concentrations without filipin. Labeling was performed in the dark under gentle shaking for 30 min, after which the cells were washed once in cacodylate buffer. Finally, the cell pellets were suspended in a mixture of 3:7 electron microscopical-grade glycerol:cacodylate buffer under gentle shaking for 1 h. Filipin fluorescence was observed on a Leica DMRE fluorescence microscope equipped with a 100-W Hg lamp and a UV filter block (340–380-nm excitation filter, 400-nm dichroic mirror, and 425-nm long-pass emission filter). Specimens were mounted under coverslips sealed with nail polish. For each sample, 200 cells were counted in triplicate.
Detection of DNA Degradation
Sperm cells were incubated at 38°C for 2 h in HBT, HBT-Bic, or HBT with UV illumination using a UVP high 2 x 2 Amp transilluminator (UVP, San Gabriel, CA). The cells were then centrifuged for 3 min at 1000 x g and the pellets resuspended in PBS (137 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4, 1.5 mM KH2PO4, pH 7.4). The ensuing procedure was based on that of Guthrie et al. [24]. Aliquots of 100 µl (
106 cells) were vortexed at full speed for 30 sec and placed on ice. Nine hundred microliters of a precooled (0°C) mixture of 7:3 ethanol:PBS was carefully added on top of the sperm suspension and mixed in. At this point, the suspension could be stored at 2°C for up to 1 wk before being processed further. Subsequently, the suspension was centrifuged for 3 min at 1000 x g and the pellet washed once in PBS, resuspended in a mixture of 1:2 PBS:extraction buffer (192 mM Na2HPO4, 4 mM citric acid, pH 7.8), vortexed, kept at room temperature for 10 min, and revortexed. The suspension was recentrifuged and the pellet resuspended in 1 ml PBS containing 50 µg/ml propidium iodide and 50 µg/ml RNase I, after which it was incubated for 30 min in the dark at room temperature. The mixture was then passed through a 100-µm filter (Millipore, Etten Leur, The Netherlands) to remove clumped material and analyzed by flow cytometry. The PI fluorescence of nonaggregated nuclear events was detected on FL-3 (630-nm long-pass filter) in linear mode. As a positive control, similar analyses were made of preparations of granulosa cells obtained from atretic bovine follicles (i.e., under conditions where massive regression of granulosa cells takes place by apoptosis [24]); granulosa cells from growing follicles were used as a negative control (absence of apoptosis).
Acridine Orange Detection of Single-Stranded DNA
Staining of spermatozoa that had been incubated in HBT, HBT-Bic, or in HBT with UV illumination, as described in the previous section, was performed according to the method described by Evenson et al. [25]. The cells were centrifuged for 3 min at 1000 x g and the pellet washed once and resuspended in TNE buffer (0.15 M NaCl, 1 mM EDTA, and 10 mM Tris, pH 7.2) at 4°C to a final concentration of 5 x 106 cells/ml. A 200-µl aliquot of this suspension was mixed with 400 µl of a detergent/acid solution consisting of 0.1% v/v Triton X-100 in 0.08 M HCl, 0.15 M NaCl. After 30 sec, 1.2 ml of a solution containing 6 µg/ml acridine orange in 0.15 M NaCl, 1 mM EDTA, 0.2 M Na2HPO4, 0.1 M citric acid (pH 6.0) was added to the sample. The cells were subjected to flow cytometry after 30 min incubation at room temperature. After gating out nonsperm and aggregated events, the fluorescence detected by the FL-1 detector (530/30-nm band-pass filter: green fluorescence due to acridine orange binding to intact DNA) was compared with that detected by the FL-3 detector (620-nm long-pass filter: red fluorescence due to acridine orange binding to single-stranded DNA).
TUNEL Detection of DNA Strand Breaks
DNA strand breaks in the nuclei of spermatozoa that had been incubated in HBT, HBT-Bic, or in HBT with UV illumination were detected using a kit supplied by Roche Diagnostics. The incubated cells were pelleted (3 min at 1000 x g) and then washed, resuspended in the TUNEL reaction mixture, and processed, all according to the supplier's protocol. The stained cells were analyzed by flow cytometry. After gating out nonsperm and aggregated events, the incorporation of fluorescein-conjugated dUTP into the sperm DNA was quantitated on the FL-1 detector (530/30-nm band-pass filter). Similarly processed granulosa cells from atretic and growing-phase follicles were used to validate the methodology.
Detection of Inner Mitochondrial Transmembrane Potential
Washed spermatozoa (5 x 106 cells/ml) were incubated for 1 h in HBT, HBT-Bic, or HBT with UV illumination (as described in the previous sections). Samples were stained for 18 min at 38°C with a combination of 2.5 nM Mitotracker Red (chloromethyldihydroX-rosamine), 25 nM Yo-Pro-1, and 10 µg/ml PNA-FL in a humidified atmosphere in equilibrium with 5% CO2 as appropriate; a parallel set of samples were stained similarly but with the combination 0.5 µg/ml Ro-SA-FL, 1 µg/ml PI, and 1 µg/ml PNA-RPE. All stained samples were then analyzed by flow cytometry. After gating out nonsperm and aggregated events, Mitotracker Red fluorescence was detected on FL-3 (630-nm long-pass filter) while Yo-Pro fluorescence was detected on FL-1 (530/30-nm band-pass filter). Mitotracker-positive sperm events indicated cells with active membrane-polarized mitochondria, whereas cells with reduced Mitotracker fluorescence were those whose mitochondria had lost their transmembrane potential. Viable (plasma membrane-intact) cells were Yo-Pro negative whereas degenerate cells were Yo-Pro positive. Ro-SA-FL and PI fluorescences were detected and interpreted as described in an earlier section.
For visualization of Mitotracker staining, cells incubated for 2 h in HBT were stained with 2.5 nM final concentration of Mitotracker Red together with either 0.5 µg/ml Ro-SA-FL or 1 µg/ml annexin V-FL or with 25 nM Yo-Pro-1. The staining was carried out for 10 min at 38°C in an appropriately humidified atmosphere, and confocal microscopy was carried out as described for annexin V-FL labeling (see section above). Annexin V-FL, Ro-SA-FL, and Yo-Pro fluorescence were detected in PMT-1 (emission range 510–540 nm), while Mitotracker Red labeling was detected in PMT-3 (emission >600 nm). The alternative staining procedure enabled comparative visualization of Mitotracker Red staining either in the sperm subpopulation that showed apical PS or PE exposure or in the intact population (Yo-Pro negative).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bicarbonate induction of aminophospholipid exposure was tested using triple-stain flow cytometry. Spermatozoa were incubated in HBT or HBT-Bic for up to 2 h and stained either with annexin V-FL to detect PS exposure or with Ro-SA-FL to detect PE exposure. The cells were counterstained with PI to detect damaged plasma membranes and with PNA-RPE to detect reacted or deteriorated acrosomes [7]. In the absence of bicarbonate, very few intact spermatozoa (those negative for both PI and PNA-RPE) were stained with either annexin V-FL or Ro-SA-FL (Fig. 1, A and C, respectively); in other words, there was no exposed PS or PE on the great majority of intact cells. On the other hand, incubation in HBT-Bic (i.e., in the presence of 15 mM bicarbonate) induced exposure of PS and PE in a substantial subpopulation of intact cells (Fig. 1, B and D). Because annexin V-FL and Ro-SA-FL cannot pass through an intact plasma membrane, PS and PE exposure in these cells must have taken place at the external surface of the plasma membrane and must have stemmed from a collapse (or scrambling) of the plasma membrane's phospholipid asymmetry.
|
The kinetics of PE and PS exposure differed considerably (Fig. 2). Maximal proportions of intact spermatozoa with surface-exposed PE were detected within 15 min (Fig. 2A; t½ = approximately 5 min), whereas sperm exposing PS only reached maximal proportions after 60 min (Fig. 2B; t½ = approximately 20 min). However, after 60 min incubation, similar proportions of the intact sperm population had exposed PS as had exposed PE.
|
Because annexin V-FL and Ro-SA-FL can be used in the presence of albumin (unlike C6NBD-phospholipid analogues, for which albumin has a high affinity), we were able to test the effect of albumin on aminophospholipid exposure. In fact, the protein, though a frequent component of mammalian in vitro fertilization media, had only a very slight effect on rates of exposure and did not affect overall levels (compare triangles with circles in Fig. 2).
Confocal Microscopic Localization of Bicarbonate-Induced Aminophospholipid Exposure
Intact (PI-negative) spermatozoa in samples that had been incubated for up to 2 h in HBT did not show aminophospholipid exposure at their surface. However, after incubation in HBT-Bic, a subpopulation of intact cells showed exposed PS (stained with annexin V-FL) and PE (stained with Ro-SA-FL) over the anterior acrosomal region of their head plasma membrane (Fig. 3, A and C, respectively). In some degenerated (PI-positive) cells, the midpiece was labeled (Fig. 3B), but this labeling was observed regardless of whether the incubation medium contained bicarbonate. In some immature spermatozoa that were intact (PI negative), the attached cytoplasmic droplet was labeled (Fig. 3D); this labeling too was independent of the presence of bicarbonate during incubation. Neither the degenerated cells nor the intact immature sperm cells showed any bicarbonate-dependent PS and PE exposure over the anterior acrosomal region of their heads.
|
Correlation of PE and PS Exposure with Increase in Merocyanine 540 Binding in Cells Stimulated by Bicarbonate
The merocyanine stainability of intact cells reflects the degree of lipid disorder in the outer leaflet of their plasma membranes (see [9] for references), and an increase in merocyanine stainability has been associated with phospholipid scrambling [26]. To see whether bicarbonate-induced merocyanine stainability in boar spermatozoa did indeed correlate with direct measures of phospholipid scrambling in sperm, PE and PS exposure was measured in parallel with merocyanine binding in spermatozoa that had been incubated for either 10 or 60 min in HBT-Bic; samples from 13 ejaculates were analyzed (Fig. 4). There was much variability in response between the 13 ejaculates, which enabled the correlation between the parameters to be observed clearly. After both 10 and 60 min of bicarbonate stimulation, the proportion of cells showing PE and PS exposure was closely correlated with the proportion highly stained by merocyanine (regression analysis showed that R ranged from 0.92 to 0.98 for the four associations illustrated in Fig. 4). However, the relationship between aminophospholipid exposure and merocyanine stainability was different for the two phospholipid species. The proportion of intact cells with exposed PE was closely similar to the proportion highly stained by merocyanine after both 10 and 60 min of bicarbonate stimulation (Fig. 4A); the ratio of these proportions was 0.98 ± 0.08 after 10 min and 1.01 ± 0.07 after 60 min (means ± SD). In all cases, longer incubation led to only slight increases in both parameters, which were not statistically significant (P > 0.05, Wilcoxon matched-pairs signed-ranks test).
|
Given the much slower kinetics of PS exposure in response to bicarbonate (see Fig. 2), the relationship between PS exposure and high merocyanine staining was, as expected, different (Fig. 4B). After 10 min exposure to bicarbonate, there were many fewer PS-exposing cells than high merocyanine-staining cells (ratio of proportions 0.41 ± 0.07), whereas after 60 min, the two parameters were close (ratio of proportions 1.08 ± 0.15), almost identical to the PS:merocyanine relationship; by this later time point, the high merocyanine-stained cell population had increased only slightly but the proportion of PS-exposing cells had increased by a highly significant factor of approximately 2.5 (P = 0.0002, Wilcoxon matched-pairs signed-ranks test).
Effect of Calcium on Detection of Changes in Plasma Membrane Phospholipid Architecture
Calcium (2 mM) was routinely included in HBT and HBT-Bic because annexin V requires a substantial level of this ion to bind to PS [27]; however, the bicarbonate-induced increases in merocyanine binding take place independently of external calcium [9], while Ro09-0198 does not apparently require this ion for binding [28]. We tested to what extent our three parameters for detecting phospholipid scrambling depended on millimolar external Ca2+ by incubating cells in Ca2+-supplemented and unsupplemented HBT and HBT-Bic. As expected, in the absence of millimolar Ca2+, there was no binding of annexin V-FL to sperm regardless of incubation conditions. However, detection of surface PE with Ro-SA-FL did not require Ca2+, enabling us to demonstrate that PE exposure was also not dependent on this ion. After 10 min incubation in HBT-Bic containing 2 mM Ca2+, 48.7% ± 11.0% of the intact cells had exposed PE, while in unsupplemented HBT-Bic, this value was 49.1% ± 9.9% (mean values ± SD for 8 individual ejaculates, each measured in triplicate). After 60 min, the respective values were 62.0% ± 17.4% and 60.5% ± 18.1%. The proportions of intact spermatozoa highly stained by merocyanine were closely similar to those exposing PE, regardless of medium, and the increases in merocyanine staining induced by bicarbonate were closely similar to the increases in PE exposure regardless of the level of extracellular Ca2+.
Role of cAMP-Dependent Protein Phosphorylation in the Control of Membrane Lipid Architecture
In our previous study [7], we had demonstrated that changes in the transbilayer movements of C6NBD-phospholipid analogues were controlled through a cAMP-dependent phosphorylation pathway in the same way as increases in merocyanine stainability [6]. Accordingly, we tested various modulators of cAMP-dependent phosphorylation for their effects on the induction of aminophospholipid exposure, comparing the results of staining with annexin V-FL, Ro-SA-FL, and merocyanine 540. Also, using filipin as a probe, we studied the lateral distribution of cholesterol because we have recently found that bicarbonate induces a lateral redistribution of cholesterol to the apical surface of the sperm head in the cells that display high merocyanine binding [8]. Incubation of sperm cells for 2 h in HBT with the phosphodiesterase inhibitor papaverine resulted in PE and PS exposure similar to that induced in HBT-Bic (Table 1); the merocyanine and filipin responses were also similar to those in HBT-Bic. The cAMP analogue 8Br-cAMP provoked similar responses, whereas the cGMP analogue 8Br-cGMP did not induce aminophospholipid exposure, merocyanine stainability, or cholesterol redistribution. Forskolin also did not induce the plasma membrane alterations even at high concentrations (the bicarbonate-dependent form of adenylyl cyclase is forskolin insensitive [5]). On the other hand, the protein phosphatase inhibitors okadaic acid and calyculin induced aminophospholipid exposure as well as the merocyanine- and filipin-detectable membrane changes. All of these results indicated that bicarbonate was exerting its effect by increasing cAMP levels, thereby stimulating protein kinase A to raise protein phosphorylation states. To test this hypothesis further, we incubated spermatozoa in HBT-Bic in the presence of the protein kinase A inhibitors H89 and Rp-cAMPS; both compounds indeed blocked the effect of bicarbonate (Table 1).
Apoptotic Indicators Are Not Associated with Bicarbonate-Induced Phospholipid Scrambling
As pointed out in the Introduction, surface exposure of aminophospholipids can be an early indicator of cell apoptosis. Two other hallmarks of apoptosis are deterioration of the nuclear DNA and mitochondrial degeneration: cytochrome C leaked into the cytosol is a key component of the apoptosome that initiates activation of destructive caspases [21, 29].
Sperm DNA integrity was assessed in three different ways in cells that had been incubated for 2 h either in the absence or the presence of bicarbonate or subjected to UV illumination during incubation in the absence of bicarbonate (UV is a well-known inducer of apoptosis in somatic cells [17]). 1) We investigated DNA fragmentation, using the procedure of Guthrie et al. [24], to treat the incubated sperm nuclei briefly with RNase and then stain them with PI before flow cytometric analysis. Normal haploid nuclei give a uniform PI fluorescence signal because they all have essentially the same amount of DNA (ignoring sex differences), whereas fragmentation of DNA can lead to subnuclear particles, containing subhaploid amounts of DNA and thus less PI fluorescence compared with the intact nuclear events. No such subhaploid nuclear events were detected after normal incubation, whether or not bicarbonate was included (Fig. 5, A and B); however, illumination of sperm cells with UV light resulted in a significant subhaploid peak (Fig. 5C). 2) We used the method of Evenson et al. [25] to detect single-strand DNA structures in the sperm nuclei by means of acridine orange staining. This intercalating DNA probe fluoresces green when interacting with intact DNA but fluoresces red when it interacts with single-stranded DNA or RNA; it can be used to detect DNA damage directly in sperm cells because RNA is more or less absent. Spermatozoa incubated normally showed only minimal DNA damage regardless of the presence or absence of bicarbonate (Fig. 6, A and B); however, substantial DNA damage (increase in red-fluorescing events) was observed in UV-illuminated cells (Fig. 6C). 3) We used a flow cytometric TUNEL assay [30] to detect DNA strand breaks in the incubated spermatozoa. The TUNEL assay detects such breaks through labeling the free 3'-OH termini of the broken strands quantitatively with fluorescein-conjugated dUTP; the greater the number of strand breaks, the greater the incorporated fluorescence. Normal incubation in the presence of bicarbonate did not cause any changes in TUNEL labeling (compare Fig. 7, A and B), whereas illumination with UV resulted in a substantial TUNEL-positive subpopulation (Fig. 7C).
|
|
|
To investigate mitochondrial functionality, we measured depolarization of the mitochondrial inner membrane using a rosamine derivative, Mitotracker Red. This cell-permeant organelle-selective dye is actively sequestered by functioning mitochondria with polarized inner membranes but is not accumulated by depolarized (nonactive) mitochondria. After 2-h incubation of spermatozoa in HBT-Bic, addition of the probe resulted in bright staining of the midpiece of all cells showing PS or PE exposure in the anterior acrosomal region of their heads (as in Fig. 3, A and B); in such incubated samples, almost all the Yo-Pro-negative (plasma membrane-intact) cells showed similar bright staining of the midpiece. Sperm samples were also incubated for 1 h in either HBT or HBT-Bic or HBT with UV illumination and then stained either with Mitotracker Red in the presence of Yo-Pro and PNA-FL or with Ro-SA-FL in the presence of PI and PNA-RPE before being analyzed by flow cytometry (Fig. 8). By this means, the staining characteristics of the intact population could be quantified. In HBT and HBT-Bic, almost all the intact population (>99.5%) showed high Mitotracker Red accumulation, whether or not that population contained many scrambled cells (compare Fig. 8, A and B). Notably, although UV illumination resulted in considerable cell death, the very great majority (>98.5%) of the residual intact population possessed fully functional mitochondria and did not show exposed PE (Fig. 8C).
|
In a final series of experiments, we examined the effect of caspase inhibitors on bicarbonate induction of scrambling. Caspases, a specific group of cysteine proteases with strong preference for aspartyl residues, are considered to lie at the heart of apoptotic degenerative processes no matter what inductive pathway is involved [29]. Preincubation of spermatozoa for 3 h with 100 µM concentrations of the broad-spectrum caspase inhibitors zVAD-fmk and Boc-D-fmk did not in any way inhibit subsequent bicarbonate induction of scrambling (Table 2); other preincubation periods and lower concentrations similarly had no effect (data not shown). We also tested the effect of preincubating spermatozoa with the apoptosis inducers staurosporine and cycloheximide [17]. In the absence of bicarbonate, these did not of themselves induce scrambling; moreover, in the presence of bicarbonate, as previously shown [6], higher concentrations of staurosporine (a potent protein kinase inhibitor) actually blocked bicarbonate-induced scrambling (see Table 2).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
One of the earliest detectable changes is the increase in plasma membrane lipid disorder, as measured by increased merocyanine binding: this is essentially complete within about 5 min of bicarbonate exposure [9]. In other cell types, enhanced merocyanine binding has been associated with a collapse in the asymmetric transverse distribution of plasma membrane phospholipids brought about by increased scramblase activity [26]. In the present study, we have used the recently developed probes Ro-SA-FL and annexin V-FL to monitor directly exposure of PE and PS, respectively, on the sperm cell surface. We have been able to show that bicarbonate does indeed induce the exposure of PE and PS in live acrosome-intact spermatozoa. The exposure of PE takes place very rapidly, being clearly detectable within 5 min and largely complete within 10 min; a similar population of sperm cells expose PS (detected with annexin V-FL), albeit with slower kinetics. Notably, aminophospholipid exposure takes place only over the apical region of the head plasma membrane.
Although the proportion of cells that responded to bicarbonate varied greatly between ejaculates, the proportion exposing PE matched very closely the proportion acquiring high merocyanine-binding ability. Moreover, both parameters showed closely similar bicarbonate-response kinetics (compare Fig. 2A with data in [7] and [9]). Thus, it is very likely that the two probes are in fact signaling the same event. Merocyanine 540 is believed to detect decreased packing order of phospholipids in the outer leaflet of the plasma membrane lipid bilayer [35, 36]; the decrease enables deeper intercalation of the amphiphilic probe into the bilayer, whence the increase in hydrophobicity of its environment renders the probe 5–10 times more fluorescent. Because PE is a phospholipid with a relatively small head group, its translocation into the outer lipid leaflet will disturb the latter's packing order, thereby enabling merocyanine's deeper intercalation. The discrepancy between PE and PS with respect to the kinetics of detectable aminophospholipid exposure is intriguing. Although the two probes we used differ in their binding mechanism and stoichiometry (Ro-SA-FL binds to PE in a Ca2+-independent manner with a stoichiometry of 1:1 [14, 28] whereas annexin V-FL requires Ca2+ and binds to PS with a stoichiometry of 1:4 annexin:PS [27]), such differences would lead only to a difference in amount of probe bound; the kinetics of detection would not be affected significantly. Rather, we suspect that the discrepancy between PE and PS exposure may be due to differential rates of processing of these phospholipid species by the combined activities of the externalizing scramblase and the internalizing aminophospholipid translocase [11, 26]. In support of this explanation, our earlier study [7] revealed that, in bicarbonate-stimulated boar spermatozoa, outward translocation of a fluorescent analog of PE across the sperm plasma membrane lipid bilayer was considerably faster than the PS analog, whereas the opposite was true with respect to inward translocation. A discrepancy between PE exposure (measured by TNBS labeling) and PS exposure (measured with annexin V-FL) has also been demonstrated for cardiac muscle cells under ischemic conditions [37], and Zweifach [38] has reported delayed annexin V-FL detection of induced phospholipid scrambling as compared with FM1-43 (a fluorescent membrane probe structurally related to merocyanine).
Exposure of aminophospholipids at the cell surface, though associated with other cellular events such as exocytosis and cell adhesion, is now recognized as an early indicator of apoptosis in many cell types. However, although bicarbonate does induce instability and enhanced cell death in sperm populations (e.g., even cooling stimulated samples from 38 to 20°C is sufficient to cause significant damage [7]; see also [9]), we could find no evidence that the bicarbonate-induced exposure of PS and PE in our boar sperm samples was part of an apoptotic mechanism. 1) Apoptotic cells display genome damage [20], which can be detected either as nuclear fragmentation or as an increase in single-stranded DNA moieties or as the appearance of DNA strand breaks. However, a 2-h incubation of the sperm cells with bicarbonate did not induce DNA or nuclear damage by any of these criteria, although damage was clearly detectable in spermatozoa that had been illuminated with UV, a well-established apoptotic agent. 2) The mitochondria of cells undergoing apoptosis frequently show depolarization of their inner membranes as cytochrome c escapes to initiate caspase activation and thence cellular degeneration [21]. However, intact sperm cells, whether in the presence or absence of bicarbonate and regardless of exposure of aminophospholipids, always contained functional mitochondria as detected with Mitotracker Red. Although UV illumination caused plasma membrane disruption in a large proportion of the sperm population, the remaining intact spermatozoa displayed fully functional mitochondria and showed no evidence of scrambling. 3) The action of the proteases known as caspases is now known to lie at the heart of the apoptotic process as execution agents common to almost all cell types investigated [29]. However, prolonged preincubation of spermatozoa with very high doses of the broad-spectrum caspase inhibitors Boc-D-fmk and zVAD-fmk did not inhibit bicarbonate induction of phospholipid scrambling. 4) Prolonged incubation with the apoptosis inducers staurosporine and cycloheximide did not induce scrambling; on the contrary, as previously reported [6], staurosporine blocked the bicarbonate-induced scrambling in accord with its well-established role as a potent protein kinase inhibitor. Two further pieces of evidence also point against bicarbonate-induced aminophospholipid exposure in boar spermatozoa representing an apoptotic process. Aminophospholipid exposure in apoptosing cells is not detected for many minutes after treatment with apoptotic agents [17], whereas we observed PE exposure in boar spermatozoa within a few minutes of bicarbonate addition. Moreover, while apoptotic scrambling of the membrane phospholipids is considered to be an irreversible process, stemming as it does from intracellular caspase-catalyzed proteolysis, bicarbonate-induced acquisition of high merocyanine binding ability in boar spermatozoa (which we have shown above to be concomitant with PE exposure) is reversible [9].
Our demonstration of the independence of bicarbonate-induced scrambling from apoptotic processes in boar sperm suggests that interpretation of detectable aminophospholipid exposure in spermatozoa must be made with care. While apoptosis clearly takes place during spermatogenesis and germ cells showing evidence of apoptosis are to be found in semen of various species [39], these apoptotic indicators are largely associated with samples of poor quality [40, 41]. In our experiments, using sperm samples from highly fertile boars from commercial breeding herds, we detected very few cells with damaged DNA. However, we noted that a few spermatozoa showed bicarbonate-independent aminophospholipid exposure in the midpiece; these cells were PI positive (indicating that the plasma membrane was dysfunctional) and were not stained by Mitotracker Red (indicating that the mitochondrial inner membrane was depolarized). We also noted that some cells classified as immature by virtue of their bearing an attached cytoplasmic droplet showed aminophospholipid exposure over the droplet. It is noteworthy that Weil et al. [19] found no evidence for the involvement of caspases in mouse sperm death during normal incubation in vitro, but from parallel studies on chicken erythrocytes of differing maturational states, they put forward the hypothesis that caspases might be progressively lost during maturation of transcriptionally inactive cells. Perhaps caspases in spermatids [42] become sequestered within the cytoplasmic droplet to be discarded when the latter is shed during passage of the spermatozoa through the epididymis. From a theoretical point of view, there is no reason or advantage for mature spermatozoa, as terminal nondividing cells, to retain a mechanism for apoptosis.
In other cell types, phospholipid scramblase shows strong Ca2+ dependency, and the scrambling process is believed to be a response to (local) increases in intracellular Ca2+ [11, 18, 26]. Our previous studies indicated, however, that, in boar spermatozoa, bicarbonate induces both high merocyanine-binding ability and increased bidirectional transverse movement of C6NBD-phospholipid analogues across the plasma membrane through an adenylyl cyclase/cAMP/protein kinase A signaling pathway [6, 7, 9]; external Ca2+ is not required. In the present study, we have confirmed that the same is true of bicarbonate induction of aminophospholipid exposure, thus finally tying together the three phenomena as manifestations of the same process: phospholipid scrambling. As far as we are aware, ours is the first demonstration that phospholipid scrambling can be controlled by a cyclic AMP-dependent protein phosphorylation signaling pathway. We have also shown that the same signaling pathway is responsible for inducing lateral concentration of cholesterol in the apical region of the head plasma membrane; this bicarbonate-dependent process was reported recently by our laboratory [8] when it was found that the proportion of sperm cells showing cholesterol redistribution correlated well with the proportion of high merocyanine-binding cells. The redistribution of cholesterol is relatively slow (t½ > 45 min; unpublished data) compared with the aminophospholipid exposure. Therefore, we believe that cholesterol redistribution is a process consequent on phospholipid scrambling. The apical concentration of cholesterol seems to be an intermediate capacitation state permissive for albumin-mediated cholesterol extraction, a further key process in capacitation [8]. A similar link between phospholipid scrambling and cholesterol extraction in apoptotic Jurkat cells was recently reported by Tepper et al. [43].
Two further points concerning bicarbonate-induced scrambling should be mentioned. Sperm cells bearing a cytoplasmic droplet (sign of immaturity) not only failed to expose PE or PS at the apical sperm head area in response to bicarbonate (see Fig. 3D) but also failed to induce cholesterol redistribution and albumin-mediated cholesterol depletion [8]. One may deduce that the cyclic AMP-dependent signaling pathway leading to phospholipid scrambling was in some way ineffective in these cells. While the molecular details of the lesion remain to be elucidated, a link between different maturational states and functional heterogeneity within the sperm population has been made by others [44]. The varying proportions of apparently normal cells that fail to express phospholipid scrambling in response to bicarbonate may therefore represent a subpopulation that has not attained full maturity.
The surface region of the spermatozoon in which bicarbonate-induced PE and PS exposure can be seen (i.e., the apical head region) is that in which the plasma membrane and the outer acrosomal membrane fuse at multiple foci during the zona pellucida-induced acrosome reaction. Notably, aminophospholipid exposure is not seen in the equatorial region. Fusion does not extend to this latter region during the zona-induced acrosome reaction, a restriction thought to be essential for subsequent gamete fusion [1, 3]. Thus, we hypothesize that induction of phospholipid scrambling and subsequent cholesterol redistribution in the apical head region of the sperm plasma membrane are primary events in the development of fusogenic potential, whereby the spermatozoon is enabled to undergo the acrosome reaction in response to the zona pellucida. Bicarbonate has also been shown to induce tyrosine phosphorylation of a range of sperm proteins via a cAMP-dependent pathway (for a review, see [45]); among these is a panel of apical plasma membrane proteins [46], some of which are involved in zona pellucida binding [33]. Hence, the early membrane change imposed by bicarbonate that we have described is the first cellular modification in a train of events that takes place in a restricted cellular locality and that is clearly of wide importance to several aspects of the mammalian fertilization process.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 The close collaboration between R.A.P.H. and B.M.G., from which this study has stemmed, was initiated in 1994–1995 through generous funding by the Human Mobility and Capability program of the European Community. B.M.G. is a senior fellow of the Royal Dutch Academy of Sciences and Arts (KNAW) and R.A.P.H. is supported by the U.K. Biotechnology and Biological Sciences Research Council. 
2 Correspondence: B.M. Gadella, Department of Biochemistry and Cell Biology, Faculty of Veterinary Medicine, Utrecht University, Yalelaan 2, 3584 CM Utrecht, The Netherlands. FAX: 31 30 2535492; b.gadella{at}vet.uu.nl
Accepted: February 7, 2002.
Received: December 12, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
