Response of goat sperm to hypoosmotic steps modelled probit analysis

Response of goat sperm to hypoosmotic steps modelled probit analysis

Animal Reproduction Science 91 (2006) 265–274 Response of goat sperm to hypoosmotic steps modelled probit analysis B. Leboeuf a,∗ , Y. Le Vern b , V...

128KB Sizes 0 Downloads 62 Views

Animal Reproduction Science 91 (2006) 265–274

Response of goat sperm to hypoosmotic steps modelled probit analysis B. Leboeuf a,∗ , Y. Le Vern b , V. Furstoss a , D. Kerboeuf b , P. Guillouet a , M. Magistrini c a

Unit´e Exp´erimentale d’Ins´emination Artificielle Caprine et Porcine, INRA, Centre Poitou-Charentes, 86480 Rouill´e, France b Service de Cytom´ etrie en Flux INRA-BASE, 37380 Nouzilly, France c Reproduction Equine INRA-PRC, 37380 Nouzilly, France

Received 31 January 2005; received in revised form 7 April 2005; accepted 19 April 2005 Available online 27 June 2005

Abstract Hypoosmotic swelling test (HOS) has been proposed by many authors to evaluate the functional integrity of the sperm membrane. Our approach in this experiment has consisted in exposing spermatozoa to a wide range of osmotic pressures then evaluating the reacted sperm cells by flow cytometry and finally modelling the sperm cell responses. Semen samples were diluted in skim milk or NPPC (native phosphocaseinate) extenders, and stored at 4 ◦ C for 3 days. At D0 and D3 aliquots from each ejaculate (n = 12) were submitted to seven hypoosmotic solutions varying from 230 to 10 mOsm/kg. Sperm samples were analyzed using flow cytometry to determine two populations of spermatozoa identified by propidium iodide (PI): PI+ (including PI, red fluorescence) and PI− (excluding PI, no fluorescence). Spermatozoa PI+ were considered as spermatozoa with membrane damages. PI+ exhibited a high variation from 230 to 10 mOsm/kg which was considered as a dose–response curve. Data were modelled using Mixed procedure and probit analysis to a sigmoid curve. Each model curve characterized the profile of response of the variable PI+ to the range of osmotic pressure from 230 to 10 mOsm/kg. The estimated parameters modelling the sigmoid curves are discussed in order to evaluate the effect of extender (skim milk versus NPPC) and duration of preservation (D0 versus D3).



Corresponding author. Tel.: +33 0 5 49 89 00 86; fax: +33 0 4 49 43 93 72. E-mail address: [email protected] (B. Leboeuf).

0378-4320/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.anireprosci.2005.04.014

266

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

Such modelling could help to differentiate storage method ejaculates within males or between male, contributing therefore to improve semen technology. © 2005 Elsevier B.V. All rights reserved. Keywords: Goat spermatozoa; Hypoosmotic steps; Flow cytometry; Probit analysis; Membrane integrity

1. Introduction The integrity of the sperm cell plasma membrane is crucial for the survival of spermatozoa. Osmotic stress has been proposed by many authors since Jeyendran et al. (1984) to evaluate the functional integrity of the sperm membrane. The high hydraulic conductivity of mammalian spermatozoa (Watson et al., 1992) ensures rapid osmotic equilibration across the plasma membrane within a few seconds. When placed in increased hypotonic solutions the plasma membrane of the sperm cells swells and stays intact until the maximum volume was exceeded, before lysis (Drevius and Ericksson, 1966). After hypoosmotic stress the higher the number of cells with swollen membranes, the better the plasma membrane quality as showed by Curry and Watson (1994), during long term sample storage at a positive temperature. The hypoosmotic swelling test (HOS) was originally developed for human sperm (Jeyendran et al., 1984) but has also been applied to several species of domestic animals: bull (Revell and Mrode, 1994), pig (Vasquez et al., 1997; Perez-Llano et al., 2001), horse (Pommer et al., 2002), dog (Rota et al., 1995). HOS evaluates the functional status of the sperm plasma membrane. The principle of the test is based on water transport across the sperm membrane under hypoosmotic conditions. It has previously been used to assess semen quality (Revell and Mrode, 1994) and to analyze fertilizing capacity (Rota et al., 2000; Perez-Llano et al., 2001). Most of the authors have used the HOS test at only one particular hypoosmotic pressure around 50–100 mOsm, and have analyzed low numbers of spermatozoa under a phasecontrast microscope to determine the percentage of total swelling patterns (Jeyendran et al., 1984). But considering the heterogeneity of the sperm population in an ejaculate, a large range of hypoosmotic pressure could be necessary to better discriminate the different subpopulations of spermatozoa. Our approach in this experiment consisted in the evaluation and the modelling of spermatozoa response across a wide range of osmotic pressures (from 230 to 10 mOsm/kg) using flow cytometry analysis after dilution in two different extenders and after 3 days storage. 2. Material and methods 2.1. Animals and semen The experiment was carried out on 12 ejaculates from three alpine and five saanen adult bucks which were submitted to photoperiodic treatments, alternatively 45 long days (16 h light per day) then 45 short days (8 h light per day) in order to reduce seasonal sexual

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

267

variations (Delgadillo et al., 1992). Bucks were individually housed and maintained on a diet of mixed grass hay and 400 g of concentrate. The bucks were previously trained to mount a teaser. The semen from each buck was routinely collected 2 days per week, from 3 weeks before the beginning and until the end of the experiment. Semen was collected using a teaser doe with the aid of an artificial vagina at 37 ◦ C. Only ejaculates containing more than 2 × 109 spermatozoa were used for the experiment. Samples were kept in a collection tube in a water bath at 30 ◦ C until volume and sperm concentration could be determined. 2.2. Media Two preservation media were used: the NPPC extender composed of NPPC (native phosphocaseinate) (Leboeuf et al., 2003) diluted in BesKOH saline solution (Dunner and Impastato, 1993) supplemented with glucose (0.01 M), and a skim milk extender. Skim milk extender is used routinely for goat semen (Corteel, 1974; Leboeuf et al., 2000). 2.3. Preparation of hypoosmotic solutions The hypoosmotic solutions (n = 7) used in this experiment were as follows: 230, 190, 150, 90, 55, 30 and 10 mOsm/kg. Kreps ringer phosphate glucose solution (KRPG; 280 mOsm/kg) (Corteel, 1974) was used to prepare the different hypoosmotic solutions from 230 to 10 mOsm/kg by adding distilled water to the 280 mOsm/kg solution. Each hypoosmotic solution was prepared and kept frozen for all the experiments. 2.4. Semen processing The 12 ejaculates were handled according to a standard procedure (Corteel, 1974). Briefly and immediately after collection and determination of volume and concentration of spermatozoa, ejaculates were split into two parts before dilution at 37 ◦ C based on a concentration of 500 × 106 spermatozoa/mL. One part was diluted with skim milk, the other part with the NPPC extender. The diluted sample from each aliquot was packed at room temperature into 0.25 mL straw (IMV, L’Aigle, France), for a final concentration of 100 × 106 spermatozoa. The straws were chilled from 37 to 4 ◦ C over 90 min. The straws containing the diluted sperm were stored horizontally at 4 ◦ C for 3 days. Motility parameters were evaluated before splitting each ejaculate. The percentage of motile sperm and progressive motility were scored by microscopic observations. All ejaculates with more than 66% motile spermatozoa and progressive motility higher than 3 on a scale from 0 to 5 (by subjective estimation) were used (Bishop et al., 1954). 2.5. Hypoosmotic steps On collection day (D0) and then after 3 days of storage (D3), two straws per extender (skim milk versus NPPC extender) were chosen at random. Each straw was immediately diluted at 37 ◦ C into 1 mL of KRPG (280 mOsm/kg) and then centrifuged at room

268

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

temperature (600 × g, 5 min) to discard the preservation media. The pellet was resuspended in KRPG (280 mOsm/kg) at 100 × 106 sperm/mL. Aliquots from each diluted straw were submitted to osmotic stress: 10 ␮L from each straw were diluted into 990 ␮L of each hypoosmotic solution. After 5 min of incubation at 37 ◦ C the spermatozoa were stained with 5 ␮L/mL of propidium iodide (PI) (final concentration 2.5 ␮g/mL) and incubated for 5 more minutes (Garner and Johnson, 1995) before flow cytometry analysis. 2.6. Flow cytometry analysis Sperm samples were analysed at a rate of 1000 events per second in a sheath fluid using a FACStarPlus Analyser (Becton-Dickinson, San Jos´e, CA, USA) equipped with an argon laser operating at 488 nm with a power of 80 mW. The two light-scattering parameters (forward scatter and side scatter) were used to define a gate excluding debris and aggregates from all fluorescence analyses. PI fluorescence was measured through a 610-nm longpass filter. For each sample a total number of 10,000 events were analysed for the two populations of spermatozoa: PI+ (including PI, red fluorescence) and PI− (excluding PI, no fluorescence). PI+ spermatozoa were considered as spermatozoa with membrane damages. 2.7. Statistical analysis The percentage of spermatozoa with membrane damages after exposure to the different osmotic pressures was presented as PI+ %. Data were submitted to a factorial design with repeated measurements using the SAS Mixed procedure (SAS Institute Inc., Cary, NC, USA; 1990 Version 6). The characteristics of the model were as follows: ejaculate as statistical unit, studied factors were storage duration (D0 and D3), and extender (NPPC and skim milk), controlled factor was ejaculate (n = 12). Two straws per ejaculate were chosen by drawing lots as repetition. Factors of variations of PI+ % were tested using the SAS Mixed procedure. The mixed procedure took into account the correlations between PI+ % at different osmotic pressure within one straw per ejaculate to test studied effects. In accordance with Littell et al. (1998) we tested different covariance structures to take into account the repeated measures. The unstructured covariance was chosen according to the best value of Schwartz Bayesian Criteria. 2.8. Modelling Among the possible models, we held the one whose parameters had a biological meaning. The model of the dose–response curves was performed using the SAS probit procedure (SAS Institute Inc., Cary, NC, USA; 1990 Version 6). The main purpose of the modelling was to synthesize the responses of spermatozoa to hypotonic steps (PI+ %) with a limited number of parameters to facilitate the interpretation of the data. The biological hypothesis of the modelling was that the individual tolerance of spermatozoa to osmotic pressure is normally distributed. The parameters of distribution estimated by this procedure were (1) C which is the natural percentage of spermatozoa with membrane damages (PI+ %) at 230 mOsm. The lower the C parameter, the lower the PI+ %. (2) The mean (µ) which is the osmotic

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

269

pressure calculated to get 50% of PI+ from the PI− population at 230 mOsm considered as the inflexion point of the curve. The lower the µ, the higher the quality of the sperm plasma membrane. (3) Sigma (σ) which is the inverse of the slope at the inflexion point. A strong slope at the inflexion point, characterized by a low value of σ, means a more homogeneous sperm population at the inflexion point. Consequently the lower the σ parameter, the higher the homogeneity of the sperm population. 3. Results 3.1. Data distribution The median of the distribution of PI+ % (as percentage of spermatozoa with membrane damages) showed a high variation from 230 to 10 mOsm/kg (Fig. 1). The PI+ % was lower at D0 than at D3. Whatever the duration of storage (D0 or D3), the median of PI+ % increased slowly from 230 to 90 mOsm/kg, then increased dramatically from 90 to 10 mOsm/kg to reach 100% of PI+ (Fig. 1). PI+ response to the osmotic pressures (230–10 mOsm) was considered as a sigmoid dose–response curve. 3.2. Factors of variations The results presented in Table 1 indicate a highly significant effect of the different studied factors on PI%: duration of semen storage (D0 versus D3; p < 0.0001), and extenders (NPPC versus skim milk; p < 0.0001). A significant effect was also observed on the interaction osmotic pressure × duration of storage (p < 0.0001) and on the interaction osmotic pressure × extenders (NPPC versus skim milk; p < 0.0005). The duration × extender × osmotic pressure interaction was slightly significant (p < 0.04).

Fig. 1. Boxplot distribution of PI+ % by value of osmotic pressure. (Boxplots illustrate median, first and third quartile. Whiskers are drawn to nearest value not beyond the standard span which is 1.5 × inter quartile range. Horizontal lines are outliners.)

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

270

Table 1 Tests of studied effects computed from the mixed procedure analysis Sources

n degrees of freedom

Type III F

Pr > F

Ejaculate Duration Extender Osmotic pressure (OP) Duration × extender Duration × OP Extender × OP Duration × extender × OP

11 1 1 6 1 6 6 6

11.13 80.07 20.17 210.37 2.72 20.41 5.02 2.57

0.0001 0.0001 0.0001 0.0001 0.1054 0.0001 0.0005 0.0306

3.3. Correlation between the different osmotic pressures Table 2 shows highly significant correlation (p < 0.001) between the different osmotic pressures except for 55 (0.01 < p < 0.05), 30 or 10 mOsm (non significant). At 10 mOsm the estimated variance is not significantly different from 0, which explains the non significant correlations between 10 mOsm and the other osmotic pressures. These results mean that, whatever the ejaculate, there is no linear relation between 55 and 30 mOsm steps and the other steps of the curve: 230, 190, 150 and 90 mOsm. 3.4. Modelling the sigmoid curves The estimated means (lsmeans) of the interaction between studied factors and osmotic pressures showed a sigmoid curve (Fig. 2). According to Collett (1991) this dose–response curve can be modelled by the probit procedure of SAS. The modelling was performed as follows: (1) probit as link function and (2) the lsmeans from the Mixed procedure as response to the osmotic steps. The lsmeans were adjusted according to the interaction extender (skim milk versus NPPC) × duration of storage (D0 versus D3) × osmotic pressure (range of osmotic pressure). Each model curve characterized the profile of response of the variable PI+ % to the range of osmotic pressure from 230 to 10 mOsm/kg (Fig. 2). Table 2 Correlations between the different osmotic pressures estimated by mixed procedure Osmotic pressure 230 190 150 90 55 30 10 * ** ***

p < 0.05. p < 0.01. p < 0.001.

230

190

150

90

55

30

10

1.00

0.92***

0.91***

0.81***

1.00

0.95*** 1.00

0.90*** 0.91*** 1.00

0.20 0.32** 0.25* 0.33** 1.00

0.05 0.01 0.05 0.07 0.19 1.00

0.69 0.59 0.67 0.55 −0.16 −0.05 1.00

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

271

Fig. 2. Effect of the duration of storage × extender interactions to a range of osmotic pressures and the corresponding probit regression curves. () D0 NPPC; () D3 NPPC; () D0 skim milk; () D3 skim milk. Table 3 Estimated parameters of probit procedure on lsmeans of studied factors after mixed procedure analysis C (PI+ %)

µ (mOsm)

σ (mOsm)

Extender Skim milk NPPC

59.5 45.1

56.5 59.7

24.4 26.3

Duration of semen storage D0 D3

41.4 63.1

54.2 65.4

22.4 30.2

Interactions D0 × skim milk D0 × NPPC D3 × skim milk D3 × NPPC

46.1 36.8 72.9 53.4

52.0 55.9 66.2 65.1

20.1 24.2 32.7 28.9

3.5. Estimated parameters modelling the sigmoid curves The results of probit procedure are presented in Table 3. The C parameter increased from D0 (41%) to D3 (63%). This parameter was lower for the NPPC extender (45%) than for the skim milk extender (59%) The variation of the µ parameter was lower between skim milk and NPPC extenders (56.5 mOsm versus 60 mOsm) than between D0 and D3 (54 mOsm versus 65 mOsm). Finally, the σ parameter was higher at D3 (30.2 mOsm) than at D0 (22.4 mOsm). The confident interval for the probit parameters cannot be estimated because this modelling is made on estimated means (lsmeans). Thus we can only make an indicative comparison between the values of each parameter for the different factor levels.

4. Discussion The hypoosmotic swelling test has been shown to be a simple assay for evaluating sperm membrane integrity and therefore, in part, the functional competence of the plasma

272

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

membrane of spermatozoa in different species. However, the hypoosmotic pressure of the test varies among the species. Therefore the percentage of cells with intact membrane was markedly reduced near 75 mOsm for frozen-thawed bovine spermatozoa (Correa and Zavos, 1994), 90 mOsm for Human and Ram spermatozoa (Curry and Watson, 1994), 150 mOsm for boar sperm (Vasquez et al., 1997; Perez-Llano et al., 2001). It seems that the membrane permeability to water can greatly influence the response of sperm cells to hypotonic stress (Curry and Watson, 1994). According to Muldrew and Mc Gann (1994), several reasons for osmotic injury have been proposed, including (1) mechanical rupture of the cell membrane in hypoosmotic conditions and (2) flux of water through the membrane which can cause cell membrane damage. In order to focus on the evolution of in vitro survival based on the integrity of the plasma membrane, buck spermatozoa were exposed to decreasing osmolarity from 230 to 10 mOsm after dilution in two different extenders and 3 days of storage. We postulated that the differences in the dose–response curves (combined effect extenders × duration of storage) reflect the capacity of the extender to preserve the membrane integrity of spermatozoa. The results indicate that (1) the profile of spermatozoa responses to hypoosmotic solution is a sigmoid curve and (2) the natural percentage of sperm cells with membrane damages (PI+ %) (C), the osmotic pressure that results in 50% of spermatozoa with membrane damages (PI+ %) (µ), and the homogeneity of the population of sperm at the inflexion point (σ) vary with the extender (NPPC versus skim milk) and the duration of storage (D0 versus D3). The biological significance of the variations of the three defined parameters are discussed below. Based on natural percentage of sperm cells with membrane damages (PI+ %) (C), the protection afforded by the NPPC extender was higher than that of the skim milk extender at D0 as well as at D3. Likewise, the NPPC extender provided a higher homogeneity of the sperm population after 3 days of in vitro storage (σ parameter), whereas at D0 the motility of spermatozoa diluted in skim milk was higher than in the NPPC extender (results not shown). The µ parameter was more dependent on the duration of storage (D0, µ = 54.2 mOsm versus D3, µ = 65.4 mOsm) than on extender composition (NPPC, µ = 59.7 mOsm versus skim milk, µ = 56.5 mOsm). Our results show that the plasma membrane of spermatozoa is more resistant to hypotonic stress at D0 than at D3 and that the integrity of the plasma membrane of spermatozoa is better preserved when stored in NPPC extender than in skim milk extender. These results confirm that the NPPC extender, compared to skim milk extender, enhanced in vitro survival as demonstrated previously on motility parameters (Leboeuf et al., 2003). The efficiency of the NPPC extender could be explained by the presence of purified caseins (native phosphocaseinate), which provide a higher protection to sperm cells than raw milk. The native phosphocaseinate could limit the loss of intracellular components attributed to the dilution effect observed by Blackshaw (1953) for ram and bull spermatozoa. The higher resistance of the membrane to hypotonic stress at D0 compared to D3 and after dilution in NPPC extender compared to skim milk, can be related to modifications of the membrane composition and/or structure during survival. Native phosphocaseinate can be implicated in the protection of the sperm cells against the dramatic effect of reactive oxygen species. Mammalian spermatozoa are susceptible to oxidative stress from the relative abundance of unsaturated fatty acids which undergo lipid peroxidation; moreover the limited

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

273

availability of antioxidant defensive enzymes increases the exposure of cells to this attack. Indeed, preliminary results have demonstrated that lipid peroxidation can be limited when stallion spermatozoa are stored in INRA96® extender (IMV, L’Aigle, France) containing native phosphocaseinate (Bosc et al., 2002). Further experiments are necessary to confirm this hypothesis on buck spermatozoa. The σ parameter is higher at D3 than at D0 denoting a higher variability of the sperm response at the µ inflexion point at D3 and consequently a higher heterogeneity of the sperm population. Considering the σ parameter, our results demonstrate that NPPC extender maintains a lower heterogeneity of the sperm population after 3 days of storage than milk extender. The heterogeneity of the sperm responses to hypotonic steps can depend on several factors including the modification of the plasma membrane composition during storage which is induced by cell ageing and especially by lipid peroxidation (Chatterjee and Gagnon, 2001). NPPC extender could provide to sperm cells an antioxidant effect to sperm cells, thus limiting in that way the heterogeneity of the sperm population. 5. Conclusion The interest of our approach based on a wide range of hypoosmotic solutions and flow cytometry analysis is to model the sperm response to those solutions. Such modelling could help to differentiate storage method, ejaculates within males or between male, contributing therefore to improve semen technology. Furthermore this method evidences the heterogeneity of the sperm population by the σ parameter at the mean point of membrane injury (µ parameter). In order to reduce both the number of osmotic steps measured and to maintain the ability to show studied factor effect, it seems to be possible to select two osmotic pressures (55 and 30 mOsm) and a third one among the following: 230, 190, 150, 90, and 10 mOsm. Indeed there is no linear relation between those osmotic steps within an ejaculate as shown in Table 2. Such an hypothesis has to be confirmed before routine application. Acknowledgements The authors would like to thank Dr. J.L. Maubois and J. Fauquant for their kind supply of native phosphocaseinate and Jean Louis Pougnard, Yvonnick Forgerit and Daniel Bernelas for their technical assistance. We are very grateful to Robin Dolan for linguistic assistance. References Bishop, M.N.H., Campbell, R.C., Hancock, J.L., Walton, H., 1954. Semen characteristics and fertility in the bull. J. Agric. Sci. 44, 227. Blackshaw, A.W., 1953. The motility of ram and bull spermatozoa in dilute suspension. J. Gen. Physiol. 36, 449–462. Bosc, D., Retaillaud, F., Stradaioli, G., Giubini, C., Magistrini, M., 2002. Influence of temperature and extender on sperm motility and peroxidation process in the stallion. Proceedings of the Ninth International Symposium on Spermatology. Capetown, October 6–11th, P2-01, p. 59.

274

B. Leboeuf et al. / Animal Reproduction Science 91 (2006) 265–274

Chatterjee, S., Gagnon, C., 2001. Production of reactive oxygen species by spermatozoa undergoing cooling, freezing and thawing. Mol. Reprod. Dev. 59, 451–458. Collett, D., 1991. Modelling binary data. In: Bioassay and Some Other Applications. Chapman & Hall, London, Chapter 4, pp. 93–119. Correa, J.R., Zavos, P.M., 1994. The hypoosmotic swelling test: its employment as an assay to evaluate the functional integrity of the frozen-thawed bovine sperm membrane. Theriogenology 42, 351–360. Corteel, J.M., 1974. Viabilit´e des spermatozo¨ıdes de boucs conserv´es et congel´es avec ou sans leur plasma s´eminal: effet du glucose. Ann. Biol. Anim. Biochim. Biophys. 14, 741–745. Curry, M.R., Watson, P.F., 1994. Osmotic effects on ram and human sperm membranes in relation to thawing injury. Cryobiology 31, 39–46. Delgadillo, J.A., Leboeuf, B., Chemineau, P., 1992. Abolition of seasonal variations in semen quality and maintenance of sperm fertilizing ability by photoperiodic cycles in goats bucks. Small Rumin. Res. 9, 47–59. Drevius, L.O., Ericksson, H., 1966. Osmotic swelling of mammalian spermatozoa. Exp. Cell Res. 42, 136–156. Dunner, S., Impastato, M., 1993. Effects of amine organic buffers for semen dilution on fertility of does. Small Rumin. Res. 11, 57–64. Garner, Duane L., Johnson, Lawrence A., 1995. Viability assessment of mammalian sperm using SYBR14 and propidium iodide. Biol. Reprod. 53, 276–284. Jeyendran, R.S., Van der Ven, H.H., Perez-Pelaez, M., Grabo, B.G., Zaneveld, L.J.D., 1984. Development of an assay to assess the functional integrity of the human sperm membrane and its relationship to other semen characteristics. J. Reprod. Fertil. 70, 219–225. Leboeuf, B., Restall, B., Salamon, S., 2000. Production and storage of goat semen for artificial insemination. Anim. Reprod. Sci. 62, 113–141. Leboeuf, B., Guillouet, P., Batellier, F., Bernelas, D., Bonn´e, J.L., Forgerit, Y., Renaud, G., Magistrini, M., 2003. Effect of native phosphocaseinate on the in vitro preservation of fresh semen. Theriogenology 60, 867–877. Littell, R.C., Henry, P.R., Ammerman, C.B., 1998. Statistical analysis of repeated measures data using SAS procedures. J. Anim. Sci. 76, 1216–1231. Muldrew, K., Mc Gann, L.E., 1994. The osmotic rupture hypothesis of intracellular freezing injury. Biophys. J. 66, 532–541. Perez-Llano, B., Lorenzo, J.L., Yennes, P., Treja, A., Garcia-Casado, P., 2001. A short hypoosmotic swelling test for the prediction of boar sperm fertility. Theriogenology 56, 387–398. Pommer Angela, C., Rutlland, J., Meyers Stuart, A., 2002. The role of osmotic resistance on equine spermatozoa function. Theriogenology 58, 1373–1384. Revell, S.G., Mrode, R.A., 1994. An osmotic resistance test for bovine semen. Anim. Reprod. Sci. 36, 77–86. Rota, A., Str¨om, B., Linde-Forsberg, C., 1995. Effects of seminal plasma and three extenders on canine semen stored at 4 ◦ C. Theriogenology 44, 885–900. Rota, A., Penzo, N., Vincenti, L., Mantovani, R., 2000. Hypoosmotic swelling (HOS) as a screening assay for testing in vitro fertility of bovine spermatozoa. Theriogenology 53, 415–1420. Vasquez, J.M., Martinez, E.A., Martinez, P., Garcia-Artiga, C., Roca, J., 1997. Hypoosmotic swelling of boar spermatozoa compared to other methods for analysing the sperm membrane. Theriogenology 47, 913–922. Watson, P.F., Kunze, E., Cramer, P., Hammerstedt, R.H., 1992. Comparison of critical osmolarity and hydraulic conductivity and its activation energy in fowl and bull spermatozoa. J. Androl. 13, 131–138.