Stability of transbilayer phospholipid asymmetry in viable ram sperm cells after cryotreatment

Transcription

1 Journal of Cell Science 112, (1999) Printed in Great Britain The Company of Biologists Limited 1998 JCS Stability of transbilayer phospholipid asymmetry in viable ram sperm cells after cryotreatment Karin Müller 1,2, Thomas Pomorski 3, Peter Müller 1 and Andreas Herrmann 1, * 1 Humboldt-Universität zu Berlin, Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Invalidenstr. 43, D Berlin, Germany 2 Institut für Fortpflanzung landwirtschaftlicher Nutztiere Schönow e.v., Bernauer Chaussee 10, D Schönow, Germany 3 Max-Planck-Institut für Infektionsbiologie, Abteilung Molekulare Biologie, Monbijoustrasse 2, D Berlin, Germany *Author for correspondence ( andreas=herrmann@rz.hu-berlin.de) Accepted 29 October; published on WWW 8 December 1998 SUMMARY The transbilayer dynamics of lipids in the plasma membrane of mammalian sperm cells is crucial for the fertilization process. Here, the transbilayer movement and distribution of phospholipids in the plasma membrane of fresh, ejaculated and cryopreserved ram spermatozoa was studied by labeling cells with fluorescent analogues of phosphatidylserine and phosphatidylcholine. By colabeling cells with the DNA-binding dye propidiumiodide as well as by employing fluorescence microscopy and flow cytometry we were able to determine the transbilayer redistribution of fluorescent phospholipid analogues in intact (propidiumiodide-negative) and in impaired (propidiumiodide-positive) spermatozoa. The transbilayer distribution of the fluorescent phosphatidylserine and phosphatidylcholine analogues was not perturbed in intact sperm cells after cryopreservation. In those cells, the phosphatidylserine analogue became rapidly enriched on the cytoplasmic leaflet by the activity of a putative aminophospholipid translocase similar to intact cells of fresh, ejaculated samples. However, upon cryopreservation the activity of the putative aminophospholipid translocase was significantly reduced in intact cells. Employing annexin V-FITC, we found that even after cryopreservation the sequestering of endogenous phosphatidylserine to the cytoplasmic leaflet is maintained in intact cells, but not in impaired cells. The phosphatidylcholine analogue redistributed very slowly remaining essentially confined to the exoplasmic leaflet of the plasma membrane of intact cells from both fresh, ejaculated and cryopreserved samples. The physiological consequences of a perturbed transbilayer asymmetry in sperm plasma membranes is discussed. Key words: Ram sperm, Plasma membrane, Phospholipid, Phosphatidylserine, Asymmetry, Aminophospholipid translocase, Cryopreservation INTRODUCTION The movement and distribution of phospholipids between the exoplasmic and cytoplasmic leaflet of the plasma membrane of ram spermatozoa is a highly dynamic, ATP-dependent process (Müller et al., 1994b, 1997). Employing fluorescent as well as spin-labeled phospholipid analogues, distinct lipid-specific mechanisms of transbilayer movement and an asymmetric phospholipid distribution between both halves of the cell membrane were detected (Müller et al., 1994b, 1997) which are already known for the plasma membrane of a number of other eukaryotic cells (for reviews see Op den Kamp, 1979; Zachowski, 1993; Roelofsen and Op den Kamp, 1994). Similar to those cells, the activity of the putative aminophospholipid translocase (Seigneuret and Devaux, 1984) mediates specifically a rapid ATP-dependent translocation of aminophospholipids from the exoplasmic to the cytoplasmic leaflet. The half-time for this transport of the fluorescent analogues of phosphatidylserine (C6-NBD-PS) and phosphatidylethanolamine (C6-NBD-PE) is about 3 and 5 minutes at 20 C, respectively, for ram sperm cells (Müller et al., 1994b, 1997). A similar picture of transbilayer movement and distribution of phospholipids emerged from investigations on the plasma membrane of bull spermatozoa (Nolan et al., 1995) as well as of a non-mammalian sperm cell, that of trouts (Müller et al., 1994a), using fluorescent and spin-labeled phospholipid analogues, respectively. Presumably, the aminophospholipid translocase is responsible for the asymmetric transbilayer distribution of phospholipids. Phosphatidylserine (PS) and phosphatidylethanolamine (PE) are enriched in the cytoplasmic membrane leaflet, the cholinecontaining lipids phosphatidylcholine (PC) and sphingomyelin (SM) in the outer one (Müller et al., 1994b). However, the nature of this aminophospholipid translocase has not been unequivocally identified so far (Schroit et al., 1990; Auland et al., 1994; Tang et al., 1996). Recently, comprehensive overviews on the lipid dynamics in the plasma membrane of mammalian sperm cells and its possible physiological relevance have been given (Martinez and Morros, 1996; Nolan and Hammerstedt, 1997). The efficiency

2 12 K. Müller and others of the ATP-dependent translocation of aminophospholipids in sperm cells (Müller et al., 1994b) implies that this process as well as the maintenance of the asymmetric transbilayer distribution of phospholipids in the plasma membrane may be essential for the sperm cell at earlier stages of its genesis, in the epididymal tract, as well as after ejaculation (Müller et al., 1997; Nolan and Hammerstedt, 1997). Furthermore, there is strong evidence that alterations of the lipid phase are crucial for capacitation and the acrosome reaction (Bearer and Friend, 1982; Langlais and Roberts, 1985; Wolf et al., 1986; Roldan and Fragio, 1993; Gadella et al., 1995). In their model, Nolan and Hammerstedt (1997) considered that for the sperm plasma membrane a transbilayer rearrangement of lipids and destabilization of the lipid phase are necessary to become potentially fusiogenic. Premature destabilization may impair viability and fertilizing potential of sperm cells. Cryopreservation exerts osmotic, mechanical and chemical stress on sperm cells. Damage of sperm cells by cryopreservation has been associated with alterations of the plasma membrane lipid phase. Changes of the lipid composition of sperm cell membranes have been made responsible for the rise in membrane fluidity upon cryotreatment (Hinkovska- Galcheva et al., 1988, 1989; Buhr et al., 1994). Alterations of lateral lipid organization caused by phase separation have been described as well (Holt and North, 1986; DeLeeuw et al., 1990; Gadella et al., 1996). Moreover, there is an experimental indication for an altered transbilayer lipid distribution of diphosphatidylglycerol in sperm cell membranes during cryotreatment (Hinkovska-Galcheva et al., 1989). However, a significant part of cells which maintained their fertilizing potential can be recovered after cryopreservation. Presumably, this population of sperm cells which in the following will be designated as intact cells survived the process of cryopreservation without significant disturbances, i.e. of the plasma membrane. In order to elucidate this notion, we have investigated here the redistribution of fluorescently labeled phospholipids, C6-NBD-PS and -PC, in the plasma membranes of fresh as well as of cryopreserved ram sperm cells. Particular emphasis has been given to differentiate between intact and impaired sperm cells. To this end, sperm cells were simultaneously labeled with fluorescent lipid analogues and the fluorescent propidiumiodide (PI) which has been shown to stain non-viable (impaired), but not viable (intact) cells (Riedy et al., 1991). By this double labeling approach and by employing flow cytometry and fluorescence microscopy, we were able to follow the transbilayer movement and distribution of lipid analogues specifically in intact sperm cells. We could show that the active translocation to the cytoplasmic membrane leaflet as well as the typical asymmetric transbilayer distribution of the fluorescent PS-analogue are maintained in intact spermatozoa of cryopreserved semen. Likewise, we found that the transbilayer distribution of endogenous phospholipids, at least of PS, is maintained in those cells. However, in impaired or even nonviable cells, significant defects of transbilayer arrangement of phospholipids were detected. MATERIALS AND METHODS Preparation of fresh and cryopreserved ram sperm cells Ejaculates were obtained from rams (Merino and Blackhead, IFN Schönow e.v.) who give semen continuously over the whole year. Cryopreservation was performed by a modified Aamdal method (Aamdal et al., 1982): fresh ejaculates were diluted with solution I (containing skim milk, 11 g/100 ml, 3.8 mm N-acetyl-cysteine and 5% egg yolk) to a final volume of 7.5 ml and stored at 4 C for 60 minutes. Subsequently, 7.5 ml of ice-cold solution II (solution I with 110 mm fructose and 14% glycerol) were added. After 120 minutes at 4 C the suspension was centrifuged (1000 g, 20 minutes, 4 C), and the supernatant was discharged to give cells/ml. Straws were filled with the cell suspension to volumes of 0.2 ml (corresponding to cells per straw) at 4 C, incubated for 10 minutes at 110 C and stored in fluid nitrogen. Thawing of spermatozoa was performed by incubating straws for 5 seconds in a water bath at 70 C and transferring the cell suspension to glass tubes kept at room temperature (RT). A tenfold volume of modified Beltsville thawing solution (mbts) containing 205 mm glucose, 20 mm Na-citrate, 15 mm NaHCO 3, 10 mm KCl, 1.1 mm MgCl 2 and 1 mm EGTA (ph 7.5) was added. To remove the extender sperm cells were washed twice (5 minutes at 800 g after medium acceleration; RT) with mbts. Fresh ram sperm cells were prepared after 10-fold dilution of the ejaculate with mbts and two successive centrifugations (8 minutes at 1200 g after medium acceleration; RT). All further manipulations were performed at 22 C. Labeling of sperm cells with fluorescent phospholipid analogues The fluorescent phospholipid analogues 1-acyl-2-[6-[(7-nitro-2,1,3- benzoxadiazol-4-yl)amino]caproyl]-sn-glycero-3-phosphatidylcholine (C6-NBD-PC) and -phosphatidylserine (C6-NBD-PS) were synthesized as described previously (Colleau et al., 1991) or purchased from Avanti Polar Lipids (Birmingham, AL, USA). Labeling of sperm cells was performed as described by Müller et al. (1994b). Briefly, appropriate amounts of analogues in chloroform/methanol (1:1) were transferred to a glass tube, dried under nitrogen, and vortexed with the desired volume of mbts. Washed sperm cells were preincubated with diisopropylfluorophosphate (DFP, Aldrich) at a final concentration of 5 mm to prevent hydrolysis of analogues (Colleau et al., 1991; Müller et al., 1994b). After 5 minutes two volumes of DFP incubated cell suspension (about cells/ml) were mixed with one volume of label suspension representing time zero for all kinetic measurements. The label concentration was about 4 mol% of endogenous membrane phospholipids. Determination of transbilayer distribution by fluorescence spectrometry The redistribution of C6-NBD-lipids between the two leaflets of the sperm cell membrane was determined by the selective reduction of analogues in the exoplasmic one using dithionite as described by Pomorski et al. (1994, 1996) and Müller et al. (1994b). At given time points 50 µl samples of labeled cells were transferred to 2 ml of mbts (22 C). The decrease of fluorescence intensity at 540 nm (excitation at 470 nm) was monitored (Shimadzu RF 5001 PC spectrometer, Darmstadt, Germany) in the presence of 25 mm dithionite added from a freshly prepared stock solution (1 M in 100 mm Tris buffer, ph 9.0). The difference between initial and remaining fluorescence intensity is a measure for the amount of label located in the outer leaflet of the cell membrane. Treatment of labeled cells with Triton X-100 (final concentration 0.5%) renders all label accessible to dithionite, abolishing the fluorescence totally. Determination of transbilayer distribution by flow cytometry One minute after addition of fluorescent analogues to the cell suspension (see above), an aliquot (20 µl) of cell suspension was diluted with 80 µl mbts and mixed with 10 µl PI from a stock solution (1 mg/ml in PBS) at 22 C. After an incubation of one minute, one ml mbts was added and surplus dye was removed by

3 Lipid asymmetry in cryotreated sperm cells 13 centrifugation (1 minute, 600 g). The remaining, soft pellet (about 20 µl) was mixed with 20 µl mbts. 10 µl of this cell suspension were gently diluted with 2.5 ml mbts in a measuring tube and, within one minute, the pattern of green (NBD) and red (PI) fluorescence of 10,000 cells was recorded in a flow cytometer (Partec GmbH, Münster, Germany). Due to sample preparation, the earliest time point of measurement was 8 minutes after labeling with phospholipid analogues. Subsequently, to assess transbilayer distribution of C6- NBD-analogues, dithionite was added to the measuring tube to give a final concentration of 20 mm. In some cases we used a much lower concentration of dithionite (3 mm, see Results for explanation). After a short incubation of 30 seconds with dithionite, NBD and PI fluorescence were recorded again. After 48 minutes of label incorporation, an aliquot of labeled cells was stained with PI and flow cytometry was performed as described above. Using the software WinMDI 2.5 available from the Internet ( four polygon regions could be defined according to the fluorescence of cell populations before and after dithionite treatment: PI-negative with strong or weak C6-NBDfluorescence and PI-positive with strong or weak C6-NBDfluorescence. For each region, the percentage of cells and the geometric mean of the intensity of C6-NBD-fluorescence were determined. Fluorescence microscopy of C6-NBD-PS and annexin V- FITC labeled cells Labeling of sperm cells with C6-NBD-PS and, subsequently, after 30 minutes, staining with PI were performed as described for flow cytometry, with the exception that only 6 µl PI were used. After removal of surplus dye by washing, the soft pellet was resuspended in 100 µl mbts. Sperm cells before and after addition of dithionite (final concentration 20 mm) were observed on wet slides with an inverse standard microscope (Axiovert 100, Carl Zeiss, Germany) at a magnification of The microscope was equipped with a fluorescence facility, Planapo 100/1.3 numerical aperture and following barrier filter set: BP 450 to 490 excitation filter, FT 510 dichroic mirror and LP 515 emission filter (Carl Zeiss, Oberkochen, Germany). Photographs were taken on a Kodak EPH 1600 film pushed up to 3200 ASA. For annexin V-FITC binding washed sperm cells were resuspended in 500 µl mbts without EGTA. 2.5 µl FITC-labeled annexin V (Nexins Research B.V., Netherlands) were diluted in 235 µl binding buffer (Nexins Research B.V., Netherlands). Subsequently, 15 µl sperm suspension were added and incubated for 10 minutes at RT in the dark. After addition of 7.5 µl PI, incubation was continued for one minute. Unbound dye was removed by addition of 1 ml binding buffer and subsequent centrifugation (1 minute, 600 g). Fluorescence microscopy was done as described above. To visualize cells that were not stained (intact cells, see Results) phase-contrast optics were used. Assessment of sperm cell quality Cells which rejected the DNA binding dye PI were defined as intact cells, otherwise as impaired cells (PI-staining see above). Apart from PI-staining, two independent assays were used to assess the sperm cell quality. The normal acrosomal ridge (NAR) of an intact acrosome was characterized by phase-contrast optics at a magnification of In parallel, differentiation of cell status was performed by staining sperm cells with the fluorescent dye Hoechst which rapidly penetrates defective cells. Briefly, 5 µl samples of sperm cells were diluted 5- fold with mbts and subsequently fixed with 2% formol in mbts (1:1, v:v). After dilution with mbts cells were stained with Hoechst according to the method of De Leeuw et al. (1991) and were observed by fluorescence microscopy on wet slides (see above). Determination of cellular ATP ATP of sperm samples was extracted with trichloric acid/edta. The determination of cellular ATP content was performed on a BioOrbit- Luminometer using an ATP monitoring kit (Colora, Lorch, Germany). The ATP-values were normalized to the number of Hoechst-negative sperm cells with NAR, i.e. intact sperm cells, in the respective samples. Statistical analysis Results are expressed as mean ± standard deviation (s.d.), with n being the number of experiments. Differences were analyzed by Student s t-test with the indicated P-value. Data for the fast redistribution of C6- NBD-PS were fitted to a single exponential function (SigmaPlot, SPSS Inc., Erkrath, Germany). RESULTS Transbilayer redistribution of C6-NBD-labeled phospholipids We have determined the transbilayer distribution of C6-NBDlabeled phospholipid analogues in the plasma membrane of ram sperm cells by the selective reduction of analogues in the exoplasmic leaflet with dithionite. Immediately after labeling of membranes, either of fresh or cryopreserved sperm cells, the fluorescence is rapidly and completely destroyed upon addition of dithionite independent of the lipid head group (data not shown). This is consistent with the initial exclusive orientation of analogues to the exoplasmic leaflet (Müller et al., 1994b). However, after longer incubation at 22 C a different picture emerged. In Fig. 1 typical kinetics for dithionite mediated destruction of C6-NBD-analogues measured after 60 minutes of labeling are shown. Only for cryopreserved cells labeled with C6-NBD-PC a rapid and almost complete disappearance of fluorescence was found. For all other cases, the reaction of Fluorescence intensity [%] Time [sec] Fig. 1. Reduction of fluorescence intensity of C6-NBD-labeled phospholipid analogues in ram sperm cells by dithionite: C6-NBD- PC 60 minutes after labeling of cryopreserved (a) and fresh spermatozoa (b); C6-NBD-PS 60 minutes after labeling of cryopreserved (c) and fresh spermatozoa (d). Labeling was carried out as described. At t=0, dithionite was added to a final concentration of 25 mm and the fluorescence was recorded at 22 C. Fluorescence intensity before addition of dithionite was set to 100%. In order to access all analogues, membrane integrity was disrupted by addition of Triton X-100 with a final concentration of 0.5% after 3 minutes (arrow). d c b a

4 14 K. Müller and others NBD-PL non-accessible [%] Time [min] Fig. 2. Fraction of fluorescent-labeled phospholipid analogues in ram sperm cells non-accessible to dithionite. At time t=0 the exoplasmic leaflet of the plasma membrane of cryopreserved (open symbols) and fresh (filled symbols) ram sperm cells was labeled with the respective analogue and incubated at 22 C for various times. The fraction of non-accessible analogues was determined by the dithionite-assay as described and shown in Fig. 1:, C6-NBD-PS;, C6-NBD-PC. In the case of the aminophospholipid analogue C6-NBD-PS data and their fit to a single exponential function of one typical experiment out of three are shown. For C6-NBD-PC the mean of three independent experiments is shown. S.d. did not exceed the size of the symbols. Differences between cryopreserved and fresh sperm cell fractions are significant (P<0.012) at 20, 30 and 60 minutes. dithionite with labeled sperm cells was biphasic (Fig. 1): a steep drop followed by a slow diminution of the fluorescence intensity was observed. We have previously shown for fresh ram sperm cells that the initial steep drop characterizes the fast reduction of the analogues on the exoplasmic leaflet by dithionite (Müller et al., 1994b). It is completed within <30 seconds after addition of dithionite and can be clearly distinguished from the subsequent slower phase. The latter reflects the destruction of the fluorescence of analogues on the cytoplasmic leaflet by dithionite upon its slow permeation (Müller et al., 1994b). However, we cannot preclude that a slow inward-to-outward redistribution (flop) of analogues and their subsequent destruction on the exoplasmic leaflet contributes to the phase of slow fluorescence quenching (Nolan et al., 1995). Although the kinetics of the slower phase of C6-NBD-PS reduction by dithionite is very similar, its amplitude is lower for cryopreserved cells in comparison to fresh cells. The amount of analogues rapidly accessible to dithionite can be deduced either by extrapolating the slow phase of fluorescence quenching to time zero or by non-linear regression using two exponentials (not shown). Both procedures gave identical results. To follow the transbilayer dynamics of C6-NBD-analogues, we have measured the accessibility of C6-NBD-analogues to dithionite after various times of labeling. Fig. 2 shows the amount of C6-NBD-analogues non-accessible to dithionite as a function of time. Characteristic parameters of the kinetics are summarized in Table 1 (only given for C6-NBD-PS). We have to take into account that sperm cells with an impaired plasma membrane integrity do not allow a determination of transbilayer distribution of C6-NBD-analogues due to the rapid penetration of dithionite (see below, microscopic observation and flow cytometry). In that case rapid reduction of analogues localized on the cytoplasmic leaflet would contribute to the initial steep drop of fluorescence intensity (see Fig. 1). This would give rise to an underestimation of the amount of analogues on the cytoplasmic leaflet. Therefore, in order to characterize the transbilayer redistribution in intact sperm cells, we have normalized parameters of kinetics in Table 1 to the number of PI-negative cells (see Table 2). C6-NBD-PS rapidly distributed to the cytoplasmic leaflet of PI-negative cells from fresh samples with an initial velocity of about 5.5 pmol analogues/minute per 10 6 cells (Table 1). The half-time of redistribution was approximately three minutes in agreement with that reported previously (Müller et al., 1994b). The plateau of analogues protected from a rapid destruction by dithionite corresponding to analogues on the cytoplasmic side was about 98%. A similar plateau was observed for C6-NBD- PS in cryopreserved sperm cells. However, the initial velocity was reduced by a factor of seven in comparison to intact sperm cells of fresh samples (Table 1). The time course of C6-NBD-PC accessibility to dithionite is clearly different from that of the PS analogue in both fresh and cryopreserved ram sperm cells (Fig. 2). Only 10% of C6- NBD-PC became non-accessible to dithionite after 60 minutes of labeling for fresh sperm cells. This is consistent with the slow passive movement of PC-analogues across the plasma membrane of sperm cells (Müller et al., 1994b; Nolan et al., 1995). The amount of non-accessible C6-NBD-PC was even lower in cryopreserved sperm cells (Fig. 2). The analogue remained almost completely accessible to dithionite during the experimental time course (2.9% after 60 minutes). However, if Table 1. Redistribution of C6-NBD-PS in the plasma membrane of ram sperm cells C6-NBD-PS on the cytoplasmic leaflet at steady state (%) v i (pmol analogue* minute 1 * 10 6 cells) PI-negative and Normalized to PI-negative and Normalized to Sperm cell sample -positive cells PI-negative cells -positive cells PI-negative cells Cryopreserved 31.5± ± Fresh 64.5± ± The initial velocity v i of transbilayer redistribution and the steady-state distribution of the fluorescent aminophospholipid analogue C6-NBD-PS in cryopreserved and freshly ejaculated, ram spermatozoa at 22 C (n=3). Both parameters were calculated from monoexponential fitting of redistribution kinetics as those shown in Fig. 2. Values were given without (left columns; mean ± s.d.) or with correction to PI-negative cells (right columns). Data for PI-negative cells were taken from Table 2 (48 minutes after addition of dithionite).

5 Lipid asymmetry in cryotreated sperm cells 15 Fresh sperm cells 10 4 PI 10 4 PI before dithionite after dithionite NBD-PS NBD-PS Cryopreserved sperm cells PI 10 4 PI before dithionite after dithionite NBD-PS NBD-PS Fig. 3. Flow cytometry of cryopreserved and fresh spermatozoa after 8 minutes of labeling with C6-NBD-PS and propidiumiodide (PI). Recording was done before and 30 seconds after addition of dithionite (3 mm). Two subpopulations could be identified. Before addition of dithionite both exhibited almost the same intensity of C6- NBD-PS fluorescence, but differ in the degree of PI labeling. After addition of dithionite the C6-NBD-PS fluorescence of the subpopulation with PI fluorescence (high red channels) diminished (shift to the left), while that of the subpopulation without PI fluorescence (low red channels) became only slightly reduced (see Table 3). Note the lower dithionite concentration in comparison to that used in Figs 1 and 2. A typical experiment out of three is shown. we normalize the amount of C6-NBD-PC non-accessible to dithionite to the number of PI-negative cells (not shown for experiments with C6-NBD-PC), about 15% and 12% of the analogue redistributed to the cytoplasmic leaflet of PI-negative cells of fresh and cryopreserved samples, respectively, within 60 minutes. Transbilayer distribution of C6-NBD-analogues detected by flow cytometry To distinguish between transbilayer redistribution of analogues in intact sperm cells (PI-negative) and impaired cells (PIpositive), we have employed flow cytometry as well as fluorescence microscopy. Staining of cells by PI was performed in parallel to labeling with phospholipid analogues (see Materials and Methods). Fig. 3 shows a typical flow cytometric experiment with C6-NBD-PS labeled fresh and cryopreserved sperm cells, respectively. Table 2 summarizes the results of the flow cytometry experiments on a quantitative basis (only shown for C6-NBD-PS). Staining with PI results in two different cell populations essentially with similar intensity of fluorescent phospholipid analogues (left part of Fig. 3). Only a minor part of the PI-positive cells exhibits a higher degree of labeling with respect to the PI-negative cells. After addition of dithionite, significant reduction of C6-NBD-PS fluorescence occurred only for PI-positive spermatozoa. Concomitant with that we observed for those cells also a shift in red fluorescence channels detecting the PI fluorescence. This is caused by the decline of the NBD-fluorescence overlapping with the red channel. No fluorescence decrease in the red channel upon dithionite addition was detected for PI-positive cells not labeled with C6-NBD-analogues*. In order to visualize the decrease, and thus, the shift of the NBD-fluorescence for PIpositive cells, we have used a lower concentration of dithionite (3 mm). Longer preincubation of cells or using a higher concentration of dithionite (see above) resulted in an almost complete disappearance of the NBD-fluorescence in PIpositive cells. In contrast, the fluorescence of C6-NBD-PS of PI-negative cells decreased only slightly in the presence of dithionite (Table 2) even at concentrations of 25 mm dithionite. No significant differences between fresh and cryopreserved samples were found. About 87% of C6-NBD-PS in PI-negative cells became non-accessible to dithionite already after 8 minutes of labeling. After longer incubation (48 minutes) the non-accessible amount of C6-NBD-PS increased to about 94% (Table 2). In the case of C6-NBD-PC (data not shown), we observed a similar disappearance of the NBD-fluorescence for PIpositive cells upon addition of dithionite as for the PS *It must be noticed that although the emission spectrum of the NBD group has its maximum in the green range, it extends with a much lower intensity to the red region. Thus, NBD fluorescence may contribute to the intensity in the red fluorescence channel which is mainly determined by the PI-emission.

6 16 K. Müller and others Table 2. Resistance towards PI-staining of fresh and cryopreserved ram sperm cells labeled with C6-NBD-PS Remaining fluorescence intensity of C6-NBD-PS PI-negative cells (%) in PI-negative cells (%) Before dithionite After dithionite After dithionite Sperm cell sample 8 minutes 48 minutes 8 minutes 48 minutes 8 minutes 48 minutes Cryopreserved 39±7 40±5 27±5 32±3 86.5± ±5.0 Fresh 72±9 75±12 64±16 66± ± ±12.5 PI-negative cells correspond to the subpopulation of cells with a lower fluorescence intensity in the red (PI) channel (see Fig. 3). The amount of PI-negative cells was recorded by flow cytometry before and after addition of dithionite and normalized to the total amount of cells (n=3). Recording was performed 8 and 48 minutes after labeling cells with C6-NBD-PS (incubation at 22 C). The fluorescence intensity of C6-NBD-PS of PI-negative cells remaining after addition of dithionite is presented. NBD- and PI-fluorescence intensity was recorded in parallel by flow cytometry. Values are mean ± s.d. The remaining fluorescence intensity of C6-NBD-PS between 8 and 48 minutes was significantly different for cryopreserved cells (P<0.05; Student s paired t-test). analogue. However, we observed a strong decline of the NBD-fluorescence for PI-negative sperm cells, too. But different to PI-positive cells, a residual NBD-intensity remained indicating that a small part of C6-NBD-PC became non-accessible to dithionite after longer incubation (48 minutes at 22 C). Transbilayer distribution of C6-NBD-analogues and endogenous PS detected by fluorescence microscopy In order to sustain results of flow cytometry, in particular the relationship between complete quenching of NBDfluorescence by dithionite and PI-positive staining of sperm A B C D Fig. 4. Fluorescence microscopy of cryopreserved ram spermatozoa. (A and B) Thirty minutes after labeling with C6-NBD-PS (green fluorescence) and PI (red fluorescence) before (A) and after (B) addition of dithionite (final concentration 20 mm). Fluorescence of C6-NBD- PS remained preserved in the presence of dithionite only in those cells which were not stained by PI. Note the motile green tails in some spermatozoa. (C and D) Cells labeled with annexin V-FITC (green fluorescence) and PI were visualized by transmission light (C) and fluorescence microscopy (D). Annexin V-FITC binding was only seen to those cells which were stained by PI. Magnification was identical for all photographs. Bar, 20 µm (see C).

7 Lipid asymmetry in cryotreated sperm cells 17 Table 3. Sperm cell quality NAR and ATP-content NAR* (%) Hoechst-negative (%) (nmol 10 8 intact sperm.) Incubation time (minutes) Incubation time (minutes) Incubation time (minutes) Sperm cell sample and C6-NBD-analogue Cryopreserved PS 63±3 55±10 23±4 17±5 17±6 15±9 PC 63±7 53±7 23±6 13±3 18±6 12±6 Control 64±9 68±3 30±9 26±6 24±4 24±6 Fresh PS 80±5 89±5 71±7 76±5 22±5 10±1 PC 83±3 84±7 72±2 80±13 20±5 11±2 Control 85±2 88±9 68±6 77±24 22±4 19±5 Morphological parameters and ATP-content of cryopreserved and freshly ejaculated ram sperm populations at various times without (control) and after incorporation of C6-NBD-phospholipid analogues of PS and PC, respectively (n=3). The ATP-content was normalized to the number of intact cells. Values are mean ± s.d. For control and labeled cells, values between cryopreserved and fresh samples are significantly different (P<0.05). Cryopreserved cells: sixty minutes after incorporation of C6-NBD-PC, values of labeled cells differ significantly from the corresponding control (P<0.05). NAR of labeled cells decreases significantly between 5 and 60 minutes (P<0.05). Fresh cells: sixty minutes after incorporation of C6-NBD-analogues, the ATP-content of intact cells is significantly reduced, and, in the case of C6-NBD-PS, it differs significantly from the corresponding control value (P<0.05). *Percentage of cells with a normal acrosomal ridge (determined by phase-contrast microscopy, magnification 1000). Percentage of cells with a normal acrosomal ridge, not stained by Hoechst The ATP-content of each sample was normalized to the number of NAR and Hoechst-negative cells (i.e. intact cells). cells, we studied cells of fresh (data not shown) and cryopreserved samples by fluorescence microscopy (Fig. 4). In Fig. 4A cells preincubated with C6-NBD-PS and PI (see Materials and Methods) are shown. In agreement with the results above most of the cryopreserved sperm cells were stained by PI as evident from the intense red fluorescence of sperm heads. All cells became labeled with the PS-analogue (green fluorescence). Labeling of PI-stained cells with the PSanalogue was obvious by the green fluorescence of the tails. For the head of those cells, the strong PI-fluorescence was superior to the NBD-fluorescence. However, a clear labeling of heads by C6-NBD-PS was seen for PI-negative sperm cells (see cell upper right, Fig. 4A). Addition of dithionite completely abolished fluorescence of C6-NBD-PS in PIpositive cells, but not in PI-negative cells (Fig. 4B). Intense green fluorescence was maintained for the later cells indicating strongly the non-accessibility of the phospholipid analogue to dithionite. The diffuse appearance of the tails of PI-negative cells is caused by their rapid motion during exposure of the film. We found essentially the same results for sperm cells of fresh samples, but with a much higher amount of PI-negative cells (not shown). To probe for exposure of endogenous PS on the exoplasmic leaflet of the sperm plasma membrane, we studied binding of annexin V-FITC to ram sperm cells (data shown only for cryopreserved cells). Labeling with annexin V-FITC (green fluorescence) was only seen for PI-positive spermatozoa (Fig. 4D). Binding of annexin V-FITC was observed in particular to tails and to the postacrosomal region of the head, but not to intact acrosomal membrane regions. Remarkably, those cells which did not become stained by PI did not exhibit any green fluorescence (compare transmission light microscopy, Fig. 4C, and fluorescence microscopy, Fig. 4D; cells in the lower middle). Influence of cryopreservation on plasma membrane integrity To characterize the quality of ram sperm cells, in particular the integrity of their plasma membrane, we determined the number of cells which possess an intact acrosome (NAR) and reject the Hoechst dye (Table 3). Based on these indicators, we found that about 70% of fresh cells had an intact plasma membrane (i.e. NAR- and Hoechst-negative). Neither labeling with C6-NBD-analogues nor incubation for 60 minutes at 22 C caused a significant deterioration of sperm cell quality. The number of NAR- and Hoechst-negative cells is very similar to that determined by the resistance of sperm cells towards labeling with PI using flow cytometry (Fig. 3 and Table 2). About 70 to 75% of C6-NBD-PS labeled cells were resistant to PI staining even after longer incubation at 22 C. We note that a slight, but not significant reduction of the amount of PInegative cells was observed after treatment with dithionite. Cryotreatment impaired dramatically the integrity of the plasma membrane*. Only about 30% of cells had a normal acrosomal ridge and were resistant to staining with Hoechst (Table 3). Both, labeling with C6-NBD-analogues as well as incubation for 60 minutes at 22 C caused a slight reduction of the numbers of cells with an intact plasma membrane. The influence of cryopreservation is confirmed by measuring the amount of PI-negative cells. About 40% of C6- NBD-PS labeled cells did not become stained by PI (Table 2). This value did not change significantly upon incubation at 22 C. However, we note that dithionite treatment reduced the amount of PI-negative cells by about 10%. To characterize further the physiological state of fresh and cryopreserved ram sperm cells, we have measured the intracellular ATP-content. Data were normalized to the number of intact cells (Table 3). As can be seen the ATP-content of intact cells was similar for fresh and cryopreserved samples. For intact cells labeled with C6-NBD-phospholipid analogues *We note that centrifugation contributed to the high number of impaired sperm cells. In comparison to fresh semen, cryopreserved ram sperm cells were found to be more susceptible to the mechanical stress caused by centrifugation which was necessary in order to remove the cryo-extender. Components of the cryo-extender could bind C6-NBDanalogues and, thus, interfere with their incorporation into the plasma membrane. Immediately after thawing (without further centrifugation), we found 67±6% of sperm cells with NAR, 49±5% simultaneously rejected the Hoechst-dye (n=3). For fresh sperm cells, removal of the seminal plasma by centrifugation did not markedly influence the acrosome integrity or Hoechst-dye penetration (data not shown).

8 18 K. Müller and others a reduction of the ATP-content was revealed after longer incubation (60 minutes at 22 C) with no difference between fresh and cryopreserved samples. DISCUSSION The capacitation as well as the acrosome reaction requires a rebuilding of the lipid phase in order to destabilize and to prime the plasma membrane for membrane fusion events including the final merger of the sperm cell with the oocyte (Martinez and Morros, 1996; Nolan and Hammerstedt, 1997). This resembles a time and spatially defined process in the female genital tract. Any untimed destabilization of the lipid phase may abolish irreversibly the fertilizing potential of sperm cells. For example, the problem of a decreased fertilizing potential of mammalian sperm cells upon cryopreservation is well known. Very likely, this decrease is caused by a perturbation of the plasma membrane by an intense stress during the process of cryopreservation (Hammerstedt et al., 1990; Watson, 1995), too. One characteristic feature of the plasma membrane of mammalian sperm cells is the asymmetric transbilayer distribution of lipids (Hinkovska-Galcheva et al., 1989; Müller et al., 1994b; Nolan et al., 1995), which has been suggested to play an important role in the fertilization process of spermatozoa (Martinez and Morros, 1996; Nolan and Hammerstedt, 1997). Here, we have investigated the transbilayer redistribution of fluorescent phospholipid analogues in the membrane of frozen-thawed ram spermatozoa to analyze a possible disturbance of the recently characterized transbilayer asymmetry and movement of phospholipids in fresh, ejaculated cells (Müller et al., 1994b, 1997). In order to differentiate between intact and impaired cells, we characterized the quality of the sperm cell population by independent assays. Cells which rejected DNA-binding dye PI were defined as intact cells, otherwise as impaired cells. Using this assay, in fresh samples about 70% of cells were identified as intact cells. Upon cryopreservation the relative amount of intact cells decreased by a factor of two. The relation between intact and impaired cells of fresh and cryopreserved samples, respectively, was confirmed when intact cells were identified by a normal acrosomal ridge and a resistance towards the penetration of the Hoechst dye In agreement with previous results, we observed a rapid inward movement of the PS-analogue C6-NBD-PS in the plasma membrane of ram sperm cells. Choline containing phospholipids as SM and PC redistribute only slowly to the cytoplasmic leaflet (Müller et al., 1994b; Nolan et al., 1995) as shown here for C6-NBD-PC. The PS-analogue became rapidly non-accessible to dithionite which quenches irreversibly the fluorescence upon chemical reaction with the NBD-moiety. The rapid transport to the cytoplasmic leaflet is specific for aminophospholipids and consistent with the existence of a putative ATP-dependent aminophospholipid translocase (Müller et al., 1994b, 1997) as found also for bull sperm cells (Nolan et al., 1995). However, different from previous reports we normalized the amount of analogue non-accessible to dithionite to the number of intact (PI-negative) cells. Since the molecular sizes of PI and dithionite are of the same order and differ only by a factor of about three, it is very likely that permeation of PI parallels that of dithionite. This normalization was also justified by our observations from fluorescence microscopy and flow cytometry showing that even after longer incubation subsequent to labeling with phospholipid analogues, the NBD-fluorescence became rapidly and completely abolished upon dithionite addition in PI-positive, but not in PI-negative cells. Thus, even if PI-positive cells with a residual activity of the putative aminophospholipid translocase would exist, intracellularly localized C6-NBD-PS would not contribute to the fluorescence intensity remaining after dithionite addition. By that normalization, we found a steady-state plateau with 95% (or even more) of PS-analogues on the cytoplasmic leaflet of intact cells. This is consistent with the data obtained by flow cytometry. After 48 minutes, about 95% of C6-NBD-PS were localized on the cytoplasmic leaflet. Remarkably, we found the same steady-state plateau for C6- NBD-PS in intact cells of cryopreserved samples. The almost complete absence of PS-analogues from the exoplasmic leaflet of intact cells in the steady-state is consistent with the lack of any binding of annexin V-FITC to those cells. Binding of annexin V (-FITC) is known to be very sensitive to exposure of PS on the exoplasmic leaflet of plasma membranes of mammalian cells (Kuypers et al., 1996). We observed no binding of annexin V to PI-negative cells neither for fresh nor for cryopreserved samples. Binding was only detected for PIpositive cells supporting the normalization procedure discussed above. Taken together, this implies that PS is almost completely oriented to the cytoplasmic leaflet of intact ram sperm cells, and that cryopreservation does not perturb this pronounced transbilayer asymmetry in those cells. Moreover, we have no indication that the transbilayer distribution of other phospholipids is affected in intact sperm cells upon cryopreservation. We found no significant difference of the redistribution of C6-NBD-PC between fresh and cryopreserved samples. However, we observed a significant decline of the activity of the putative aminophospholipid translocase in intact cells of cryopreserved samples. This is indicated by a sevenfold lower initial velocity of the inward redistribution of C6-NBD-PS in those cells compared with intact cells from fresh samples. At the present stage of investigation, we have no explanation for this reduced activity of the putative aminophospholipid translocase. Presumably, a loss of intracellular ATP content cannot be made responsible for that. After correcting the measured ATP-values to the number of intact sperm cells, we found no difference for the ATP-concentration of intact sperm cells between fresh and cryopreserved samples. Thus, other reasons have to account for the reduced activity of the putative aminophospholipid translocase. For example, an enhanced concentration of cytoplasmic free Ca 2+ which may inhibit the putative aminophospholipid translocase (Bitbol et al., 1987) has to be considered. Although not known for ram sperm cells, an accumulation of calcium in cryopreserved bovine spermatozoa has been reported (Bailey and Buhr, 1994). Furthermore, altered protein-lipid interactions in the postfreezing membrane as discussed by Hammerstedt et al. (1990) could impair the function of membrane enzymes. Several independent lines of evidence show that the putative aminophospholipid translocase maintains the phospholipid asymmetry in plasma membranes of mammalian cells (Calvez et al., 1988; Gudi et al., 1990; Kuypers et al., 1993; Heinrich

9 Lipid asymmetry in cryotreated sperm cells 19 et al., 1997) including sperm cells (Müller et al., 1994b; Nolan et al., 1995; Nolan and Hammerstedt, 1997). Thus, one may wonder why the transbilayer distribution of both the PS analogue as well as endogenous PS can be maintained even though the activity of the putative aminophospholipid translocase is significantly reduced. We have previously shown that the aminophospholipid transport mediated by the translocase in the ram sperm cell membrane is very efficient. Only after a dramatic reduction of the cellular ATPconcentration to about 5%, the fast inward transport of aminophospholipid analogues was significantly decreased. However, the asymmetric steady state distribution of PS and PE was not affected (Müller et al., 1994b). Based on this finding, we assume that in intact cells of cryopreserved samples the activity of the putative aminophospholipid translocase is still sufficient to maintain the asymmetric distribution of the aminophospholipids. We have previously suggested, that the high activity of the putative aminophospholipid translocase is required to avoid any appearance of aminophospholipids in the exoplasmic leaflet (Müller et al., 1994b). Exposure of PS in the exoplasmic leaflet may cause an interaction of sperm cells with the epithelia of the genital tract leading to a suppression of the motility of spermatozoa. It is known from various mammalian cells that the exposure of PS in the exoplasmic leaflet of the plasma membrane enhances cell-cell interactions (for reviews see Williamson and Schlegel, 1994; Bevers et al., 1996). For example, lipid-symmetric erythrocytes are more adherent to monolayers of endothelial cells than are erythrocytes having the typical lipid asymmetry with PS almost exclusively oriented to the cytoplasmic leaflet (Schlegel et al., 1985). Likewise, several studies have shown that red blood cells upon incorporation of exogenous PS or exposure of endogenous PS in their outer leaflet bind more readily to macrophages (Tanaka and Schroit, 1983; McEvoy et al., 1986). The appearance of PS on the surface of apoptotic T lymphocytes also provokes their recognition by macrophages (Verhoven et al., 1995). Presumably, a similar mechanism mediates phagocytosis of apoptotic spermatogenic cells with PS exposed on their surface by Sertoli cells (Shiratsuchi et al., 1997). In conclusion, the presence of aminophospholipids in the exoplasmic layer may impair the ability of sperm cells to move within the genital tract and, thus, eventually to fertilize the egg. A high activity of the putative aminophospholipid translocase allows a rapid elimination of a disturbance of aminophospholipid transbilayer distribution. However, restoration of a disturbed transbilayer asymmetry in intact cells of cryopreserved samples might be delayed by the reduced activity of the aminophospholipid translocase. Although the physiological consequences of such a delay are unknown, we may surmise, that their fertilizing potential may be more sensitive towards perturbations of the transbilayer asymmetry in comparison to intact cells from fresh samples. Remarkably, Parrish and Foote (1986) found a reduction in rabbit spermatozoa movement in the cervix after cryotreatment. Evidence that cryopreservation can alter transbilayer asymmetry of endogenous phospholipids in the plasma membrane of ram sperm cells is given by the exposure of endogenous PS to the exoplasmic leaflet of impaired cells (PIpositive). For those cells we found a significant binding of annexin V-FITC. Our data do not allow one to characterize the transbilayer distribution of phospholipids in impaired cells on a quantitative basis. Flow cytometry as well as fluorescence microscopy showed that upon addition of dithionite all NBDfluorescence was quenched in PI-positive cells. This can be explained by the absence of any redistribution of phospholipid analogues and/or a rapid permeation of dithionite into the intracellular lumen of sperm cells. PI penetration into those cells provides a strong indication that dithionite is able to have access to the cytosol of impaired sperm cells. Thus, any C6- NBD-phospholipid analogue on the cytoplasmic leaflet could become rapidly quenched by dithionite. By that, we cannot distinguish between analogues on the exoplasmic and on the cytoplasmic leaflet, respectively. Additionally, it is reasonable to assume that in impaired cells the activity of the putative aminophospholipid translocase suffers by the loss of intracellular ATP. Conclusions The transbilayer distribution of phospholipids, at least of PC and PS, is not perturbed in intact (PI-negative) cells after cryopreservation. As for intact cells of fresh samples, endogenous PS as well as a fluorescent PS-analogue are sequestered to the cytoplasmic leaflet. However, the activity of the aminophospholipid translocase is significantly reduced in intact cells of cryopreserved samples. In contrast to intact cells, endogenous PS becomes exposed in impaired (PI-positive) cells. Our study underlines that significant differences in plasma membrane organization exist between various subpopulations of sperm cell samples. To characterize adequately the plasma membrane of sperm (and other) cells, methods and assays are necessary which allow one to distinguish between subpopulations. Approaches which do not allow one to discriminate between subpopulations, could lead to invalid conclusions and may be, at best, only of indicative value. This work was supported by grants from the Deutsche Forschungsgemeinschaft (K.M., A.H., P.M.). REFERENCES Aamdal, J., Johansen, K. E. and Graffer, T. (1982). Künstliche Besamung von Muttern mit Gefriersperma. Proc. 14th Nordic Vet. Congr., Copenhagen, 5-9 July 1982, pp Auland, M. E., Roufogalis, B. D., Devaux, P. F. and Zachowski, A. (1994). Reconstitution of ATP-dependent aminophospholipid translocation in proteoliposomes. Proc. Nat. Acad. Sci. USA 91, Bailey, J. L. and Buhr, M. M. (1994). Cryopreservation alters the Ca2+ flux of bovine spermatozoa. Can. J. Anim. Sci. 74, Bearer, E. L. and Friend, D. S. (1982). Modifications of anionic-lipid domains preceding membrane fusion in guinea pig sperm. J. Cell Biol. 92, Bevers, E. M., Comfurius, P. and Zwaal, R. F. (1996). Regulatory mechanisms in maintenance and modulation of transmembrane lipid asymmetry: pathophysiological implications. Lupus 5, Buhr, M. M., Curtis, E. F. and Somnapan Kakuda, N. (1994). Composition and behavior of head membrane lipids of fresh and cryopreserved boar sperm. Cryobiology 31, Bitbol, M., Fellmann, P., Zachowski, A. and Devaux, P. F. (1987). Ion regulation of phosphatidylserine and phosphatidylethanolamine outsideinside translocation in human erythrocytes. Biochim. Biophys. Acta 904, Calvez, J. Y., Zachowski, A., Herrmann, A., Morrot, G. and Devaux, P. F. (1988). Asymmetric distribution of phospholipids in spectrin-poor erythrocyte vesicles Biochemistry 27, Colleau, M., Herve, P., Fellmann, P. and Devaux, P. F. (1991).