Cryopreservation of human spermatozoa. IV. The effects of cooling rate and warming rate on the maintenance of motility, plasma membrane integrity, and mitochondrial function.

OBJECTIVE: To test the hypotheses that there is a two-factor aspect of cellular damage during cryopreservation that occurs in human sperm (osmotic effects versus intracellular ice formation) and that there is a cooling rate by warming rate interaction related to this damage. DESIGN: Ejaculates from healthy men were cooled at 0.1, 1.0, 10, 175, or 800 degrees C/min to -80 degrees C in a solution of 0.85 M glycerol and plunged into liquid nitrogen. Samples were warmed at 400 degrees C/min (experiment 1) or either 1 degrees C or 400 degrees C/min (experiment 2). After warming, sperm were assessed for survival using motility as the endpoint in experiment 1 and motility, plasma membrane integrity, and mitochondrial function in experiment 2. RESULTS: In experiment 1, over the various cooling rates with a standard 400 degrees C/min warming rate, a plot of motility versus cooling rate produced a classical inverted U-shaped curve (n = 6) with maximum motility at the 10 degrees C/min cooling rate. In experiment 2, over the various cooling rates, both 1 and 400 degrees C/min warming rates produced similar but shifted plots of motility, plasma membrane integrity, and mitochondrial function versus cooling rate, which also produced inverted U-shaped patterns (n = 11). Maximal survival for each of the three endpoints occurred at 10 degrees C/min cooling rate for the rapidly warmed sperm and at 1 degree C/min for the slowly warmed sperm. CONCLUSIONS: These data support the hypotheses that a two-factor hypothesis of cryodamage applies to human spermatozoa and that an interaction exists between cooling rate and warming rate. These data also suggest that motility, plasma membrane integrity, and mitochondrial function are not differently affected by cooling and warming during cryopreservation.

Mechanism of superoxide dismutase loss from human sperm cells during cryopreservation.

Earlier studies on human sperm cryodamage have shown that plasma membrane stress is the primary process and that phospholipid peroxidation in cryopreserved samples is not inhibited by addition of antioxidants. One consistent effect of cryopreservation is loss of enzymatic activity of the peroxidation defense enzyme, superoxide dismutase (SOD). To clarify this aspect of the freeze-thaw process and to develop a more complete resolution of the reactions leading to cryodamage, we sought to identify which of the two most probable mechanisms, loss of enzyme protein from the cells of denaturation of the protein, operates. If the first operates, cellular enzymatic activity and enzyme protein as identified by immunocytochemistry should give a linear correlation. If the second operates, there should be no correlation. In this study, five individual samples were analyzed before and after cryopreservation for immunoreactive Cu/Zn SOD and cell intactness by flow cytometry, for SOD enzymatic activity by a highly sensitive fluorimetric method, and for motility characteristics by Hamilton-Thorn motility analyzer. Fresh samples were obtained by the "swim-up" method and had > 95% intact cells with > 78% motile cells. After freeze-thaw, about half the cells were intact. SOD enzymatic activity was determined on Triton X-100 cell extracts, a method that removes all enzymatic activity from the cell structure, and compared with immunoreactive SOD in the cells as determined by indirect immunofluorescence mean intensities. Residual immunofluorescence was observed in the cells after Triton X-100 treatment; if this was taken into account, a close linear correlation between SOD enzyme activity and SOD immunoreactivity was obtained (r = 0.90; P = 0.00014). There was no correlation between SOD enzyme activity ratios for cryopreserved and fresh cells and fraction of intact cells after freeze-thaw. We conclude that loss of SOD protein from the subset of cells undergoing acute membrane damage is the most probable primary mechanism of SOD enzymatic activity loss from the sample and that resistance to cryodamage and SOD activity in any given cell are quite independent of one another.

Prediction of daily sperm output in stallions.

Ten sexually inexperienced Thoroughbred stallions, ranging in age from 4 to 7 yr, were used to determine the relationship between available extra-gonadal reserves (AEGR) and daily sperm output (DSO). Ejaculates were obtained during the months of June through December. Each stallion was ejaculated at 0700, 0800, and 1700 h on Day 1 and 0700, 1200, and 1700 h on Days 2 and 3. A single ejaculate was collected at 0700 h on Days 4 through 7. DSO was calculated by averaging the total spermatozoa obtained on Days 5, 6, and 7. A minimum of 14 d was allotted to each stallion between trials to allow replenishment of AEGR. Weekly trials were classified as 1) primary: ejaculates taken from sexually inexperienced stallions; 2) normal: all weekly ejaculated in which the total spermatozoa in the second ejaculate (0800 h on Day 1) divided by the total spermatozoa in the first ejaculate (0700 h on Day 1) yielded percentages >20 and <70%; and 3) all: all weekly trials. Regression analysis yielded an equation of estimated DSO=0.18 (total spermatozoa in first ejaculate) + 0.93. Measured DSO, mean of total spermatozoa collected on Days 5, 6, and 7, divided by total spermatozoa in the first ejaculate of normal weekly trials averaged 27.5+/-1.9%. When 27.5% was multiplied by the total spermatozoa found in the first ejaculate (0700 h on Day 1) in primary, normal, and all trials, correlation coefficients between measured and estimated DSOs of 0.95, 0.95, and 0.92, respectively, were obtained. In conclusion, there appears to be a relatively stable relationship between AEGR and DSO in Thoroughbred stallions aged 4 to 7 yr. This relationship allowed a famrly accurate (82%) estimation of DSO when the total number of spermatozoa found in the first ejaculate of sexually rested stallions is multiplied by 27.5%.

Effect of controlled exercise on semen characteristics in two-year-old stallions.

A two-phase study was conducted to evaluate the effect of controlled daily exercise on semen characteristics of 2-yr-old stallions. Eight nonexercised, sexually inexperienced 2-year-old Morgan stallions were randomly assigned to one of two groups. In Phase 1, four stallions were randomly assigned to receive daily controlled exercise for 16 wk. In Phase 2, these stallions were confined to box stalls for 16 wk. The remaining four stallions were confined to box stalls for 16 wk during Phase 1 and received daily controlled exercise for 16 wk during Phase 2. Individual semen collections were obtained from all stallions every 14 days. Daily sperm output (DSO) was determined by daily semen collections for seven consecutive days on Wk 0, 8, 16, 24, and 32. Mean DSO (10(9)) in Phase 1 was 3.52+/-.62 and 2.79+/-.75 for Wk O and 3.12+/-.39 and 2.61+/-.68 for Wk 16 for exercised and nonexercised stallions, respectively. Mean DSO (10(9)) in Phase 2 was 2.61+/-.68 and 3.12+/-.39 for Wk 16 and 2.48+/-.69 and 3.95+/-.51 for Wk 32 for exercised and nonexercised stallions, respectively. No effect (P>.05) of treatment was observed in either phase when parameters of spermatozoa concentration, total spermatozoa per ejaculate, DSO, total volume, gel-free volume, gel volume, percentage motility, percentage live, pH, and percentage normal were measured and analyzed.

Ovarian activity in progesterone treated mares following administration of pregnant mare serum gonadotropin.

Twelve non-pregnant, non-lactating mares were randomly assigned to four treatment groups using a 2x2 factorial arrangement with three replicates per group. Mares were administered PGF(2alpha) (10 mg, IM) on days -14 and 0, followed by HCG (3000 IU, IM) on day 5. The following treatments were administered: Group A received PMSG on days 2 (4000 IU, IM) and 5 (1000 IU, IV); Group B received PMSG (4000 IU, IM) on day 2; Group C received PMSG (1000 IU, IV) on day; Group D received no PMSG. Mares received progesterone (25 mg, IM) on days 1 through 4. Reproductive tracts were recovered at necropsy on day 16 (10 days post-ovulation). Ovaries were weighed, CL number and weight determined, follicles counted and measured, and volume of follicular fluid quantified. Mean ovarian weight (g) and number of CL per mare, respectively were: Group A, 100.0+/-15.6, 1.7+/-.7; Group B, 128.6+/-40.4, 1.3+/-.7; Group C, 92.4+/-21.0, 2.0+/-.0; Group D, 93.3+/-12.3; .3+/-.3. Mean number of follicles >10 mm and total volume (ml) of follicular fluid per mare, respectively, were: Group A, 9.4+/-2.0, 21.8+/-10.9; Group B, 1.3+/-.3, 32.2+/-28.9; Group C, 4.3+/-1.8, 5.4+/-2.3; Group D, 6.0+/-4.5, 24.0+/-10.3. There was no difference (P>.05) in mean ovarian weight, CL number, CL weight, follicular fluid volume, number of follicles, or size of follicles between treatment groups. These results show no significant effect on ovarian activity in progesterone treated mares following administration of exogenous PMSG.

Effect of progesterone impregnated vaginal sponges and PMSG administration on estrus synchronization in mares.

Estrus synchronization trials with mares were carried out using progesterone impregnated vaginal sponges and pregnant mare serum gonadotropin (PMSG) injections. In Phase 1, 10 non-pregnant, non-lactating mares were administered 1 g progesterone via vaginal sponges (5 x 6 cm) without regard to stage of estrous cycle. Sponges were replaced on day 7 of trial for an additional seven days. On day 12, PMSG (1000 IU, IM) was administered to five mares (Group A); five control mares (Group B) received no injections. There was no difference (P>.05) in estrus synchronization between Group A and Group B. Total sponge retention was 75%. In Phase 2, 11 non-pregnant, non-lactating mares were administered 2 g progesterone via vaginal sponges (10 x 6 cm) without regard to stage of estrous cycle. Sponges were replaced on day 7 of trial for an additional seven days. Estrus behavior was exhibited in 54.5% of mares by day 19. Total sponge retention was 95.4%. There was no difference (P>.05) in estrus synchronization or sponge retention between Phase 1 and Phase 2. The larger Phase 2 sponges showed less (P<.01) posterior movement within the vagina than the smaller Phase 1 sponges.

Effect of controlled exercise on libido in 2-yr-old stallions.

Eight sexually inexperienced, 2-yr-old Morgan stallions were used in a consecutive two-phase design with two groups of four stallions each. Each phase lasted 16 wk, with semen collections every 14 d. Libido scores were assigned to stallions during each semen collection. Scores ranged from zero to four, with zero indicating minimum and four representing maximum libido. In Phase 1, four stallions received daily forced exercise for 16 wk, and the remaining four stallions were confined to box stalls. In Phase 2, the previously exercised stallions were confined to box stalls, and the non-exercised stallions of Phase 1 received daily forced exercise. No week X treatment effect (P greater than .05) was found in Phase 1. Exercised stallions, however, tended to have lower libido values than non-exercised stallions from wk 10 through wk 16. A week X treatment effect (P less than .01) was found in Phase 2. Libido scores were lower (P less than .05) over time among exercised stallions, whereas scores of non-exercised stallions tended to remain stable or rise slightly over time. Mean libido scores for exercised and non-exercised stallions were different (P less than .05) at the end of Phase 1 (2.06 +/- .37 and 3.5 +/- .37, respectively). By wk 26, mean libido scores were similar (exercised: 2.62 +/- .34; non-exercised: 2.52 +/- .34). However, by wk 32, libido values for exercised stallions were lower (P less than .05) than non-exercised stallions (1.87 +/- .34 and 2.81 +/- .34, respectively). In general, mean libido scores of the non-exercised group were higher than exercised stallions after 12 wk of forced daily exercise.

The temperature dependence in the hydraulic conductivity, Lp, of the mouse sperm plasma membrane shows a discontinuity between 4 and 0 degrees C.

Cryopreservation of sperm from transgenic mice would make possible the economical perpetuation of these genetic models. In order to design a protocol for this process, it was first necessary to determine the hydraulic conductivity, Lp, or water permeability, of the plasma membrane as a function of temperature to calculate its activation energy, Ea. This was done with sperm from the caudae epididymides of CD-1 mice, using the hyposmotic cell lysis method, in which the critical osmolality, Osmcrit, defined as the osmolality at which 50% of the cells lyse, was first determined, and then the time, t, to lysis of 50% of the cells in a medium of Osm < Osmcrit was obtained. Values (mean +/- SEM, n = 10) of Osmcrit at 37, 22, 4, and 0 degrees C were 32.1 +/- 2.7, 33.7 +/- 4.1, 30.8 +/- 3.4, and 101.0 +/- 10.8 mOsm, and of t in 17 mOsm medium were 40.6 +/- 2.4, 33.8 +/- 5.7, 25.0 +/- 4.4, and 4.7 +/- 1.2 s, respectively. Values of Lp were calculated from Osmcrit and t by two different algorithms, one based on the high-amplitude swelling model used with sperm from other mammalian species and the other based on a low-amplitude swelling model. Values of Lp in micron.min-1.atm-1 from the high-amplitude swelling model at 37, 22, 4, and 0 degrees C were 1.41 +/- 0.08, 1.88 +/- 0.35, 2.12 +/- 0.19, and 1.13 +/- 0.1, respectively. From the low-amplitude swelling model, the Lp values were 0.025 +/- 0.001, 0.040 +/- 0.008, 0.082 +/- 0.029, and 1.66 +/- 0.23. Examination of mouse sperm in hyposmotic media by microscopy revealed little swelling of the cells, indicating that the low-amplitude swelling model may be the one more applicable to these cells. The temperature dependence, and hence Ea, of Lp shows a marked discontinuity between 4 and 0 degrees C with values calculated from either model. This suggests a membrane phase transition to a more brittle structure in this temperature range, consistent with the observed marked increase in Osmcrit (P < 0.0003) and decrease in t (P < 0.0001) at 0 degrees C compared to the other temperatures. In the presence of 1 M glycerol, there was no discontinuity between 4 and 0 degrees C in the values of Osmcrit, consistent with the ability of glycerol to fluidize the membrane. Low hydraulic conductivity and low-temperature embrittlement of the plasma membrane are proposed as two factors leading to mouse sperm hypersensitivity to cryodamage.

Comparison of glycerol, other polyols, trehalose, and raffinose to provide a defined cryoprotectant medium for mouse sperm cryopreservation.

Most procedures for mouse sperm cryopreservation have utilized raffinose to provide hypertonicity for cell desiccation prior to freezing and glycerol to block intracellular ice formation. Trehalose has been shown in other cell systems to provide positive protection to the plasma membrane and so was examined as a replacement for raffinose. Comparison of 3 and 6% glycerol and 7.5 and 20% sugar showed that 6% glycerol and 7.5% sugar gave maximal protection consistently and so were adopted as standard. Comparison of raffinose and trehalose at this concentration showed trehalose to give significantly better recovery of intact cells: 48 +/- 6% for trehalose, 36 +/- 9% for raffinose (+/- SE, n = 5; arc sine transformed data; P < 0.01). Less hydrophilic polyols should prove more permeant to the membrane than glycerol, enter the cell rapidly, and so possibly inhibit lethal intracellular ice formation effectively. We hypothesized that one of these polyols plus glycerol would be a more effective cryoprotectant than glycerol alone. The polyols tested as supplements to 6% glycerol were propane-1,2-diol, propane-1,3-diol, 1,1,1-tris-(hydroxymethyl)ethane (THME), and 2-ethyl-2-(hydroxymethyl)-propane-1,3-diol (EHMP). With 6% glycerol and 7.5% raffinose or trehalose, the two diols and THME gave less cryoprotection than with glycerol alone, and EHMP reduced postthaw membrane integrity to nil, thus invalidating the hypothesis. Comparison of bicarbonate-containing medium MJB to bicarbonate-free medium NTP, both with 6% glycerol/7.5% trehalose, showed no difference in recovery of membrane-intact cells. For ease of pH maintenance, NTP was chosen for studies of addition prefreeze and removal postthaw of 6% glycerol/7.5% trehalose cryoprotectant with in vitro fertilization as endpoint. Three protocols for cryoprotectant handling were tested: serial addition/dilution; dialysis addition and removal; and dialysis addition and direct insemination without cryoprotectant removal. The last proved significantly superior (P < 0.01), giving 62% fertilized eggs, normalized to controls, compared to 21% for dialysis addition and removal and 32% for serial addition and dilution. The glycerol/trehalose combination thus provides a defined cryoprotectant which, when used with addition by dialysis prefreeze and direct insemination postthaw, yields a satisfactory yield of fertilized eggs in an in vitro fertilization system.

Water permeability, Lp, of the mouse sperm plasma membrane and its activation energy are strongly dependent on interaction of the plasma membrane with the sperm cytoskeleton.

Two parameters fundamental to cell cryobiology are the water permeability (hydraulic conductivity), Lp, and its activation energy, EA. The Lp can be calculated from two experimental determinations: the critical osmolality, Osmcrit, at which 50% of the cells lyse, and the time, tcrit, to 50% lysis in a highly hyposmotic medium, based on the assumption that the cells swell to lysis with minimal resistance to swelling. We have reported [Cryobiology 32, 220-238 (1995)] that mouse sperm in hyposmotic medium show minimal swelling and so fail to meet this assumption. The concept that resistance to swelling was due to anchoring of the plasma membrane through cytoskeletal interaction was examined by treating mouse sperm with 5 microM cytochalasin D to depolymerize the cytoskeletal filamentous actin (f-actin), whose presence was established by staining with fluorescently labeled phalloidin. Diminution of fluorescence due to loss of f-actin induced by cytochalasin D was shown by flow cytometry. Mouse sperm treated with cytochalasin D showed tail curling in hyposmotic medium, similar to that observed with bovine and human sperm, indicating that the standard swelling model was applicable to these cells. Two sets of Lp values were calculated from tcrit: one using individual means of Osmcrit and one using the mean of means of Osmcrit between 37 and 4 degrees C, as these individual means were not significantly different. Values (micron.min-1.atm-1), respectively, were 9.95, 7.15 (37 degrees C); 1.51, 0.91 (22 degrees C); 0.54, 0.78 (12 degrees C); 0.47, 0.50 (4 degrees C); 0.33 (0 degree C); and 0.36 (-3 degrees C). Arrhenius plots gave EA = 13.7 and 11.7 kcal/mol, respectively. Values of t1/2 were calculated from the first-order rate constants characterizing the kinetics of cell lysis at the higher four temperatures; Lp values calculated from these, and the two sets of Osmcrit values described were 5.70, 4.09 (37 degrees C); 1.18, 0.71 (22 degrees C); 0.62, 0.90 (12 degrees C); and 0.34, 0.37 (4 degrees C). Arrhenius plots gave EA = 14.2 and 11.0 kcal/mol, respectively. We propose that these EA values are characteristic of the plasma membrane relatively unperturbed by cytoskeletal interactions. In untreated sperm, decrease of Lp with decreasing temperature and presence of cryoprotectant and the cytoskeletal interactions all act to hamper the sperm cells' ability to respond to osmotic stress encountered during freezing and thawing, such that these cells are especially sensitive to cryodamage.

Determination of water permeability coefficient for human spermatozoa and its activation energy.

Four experiments were conducted to determine the permeability coefficient of human sperm to water (Lp) and its activation energy (Ea). Critical tonicity (tonicity at which 50% of the cells swell and lyse) was determined by equilibrating sperm to 22 degrees C (experiments 1a and 1b), 30, 22, 8, or 0 degrees C (experiment 2a), and 0, -1, -3, -5, or -7 degrees C (experiment 2b) and then exposing them to various hypotonic media (215-3 mOsm). For Lp determination, sperm were equilibrated to 30, 22, 8, or 0 degrees C (experiment 3a), 8, 0, or -3 degrees C (experiment 3b), and -1, -3, -5, or -7 degrees C (experiment 3c), and then were exposed for increasing times to hypotonic (40 mOsm) media. Activation energies were calculated from the results of the latter experiments (experiment 4). Results indicate a temperature-dependent (p < 0.05) critical tonicity, with sperm exhibiting an increased membrane fragility at 8, 0, and -7 degrees C, relative to 30, 22, -1, -3, or -5 degrees C (67.5 +/- 2.4, [mean +/- SEM], 62.7 +/- 2.3, and 61.9 +/- 3.7 mOsm vs. 57.4 +/- 3.4, 57 +/- 1.2, 54.8 +/- 3.4, 60.1 +/- 5.3, and 59.8 +/- 5.2 mOsm, respectively). Human sperm have an Lp of 2.40 +/- 0.20 microns/min/atm at 22 degrees C and an Ea of 3.92 +/- 0.59 kcal/mol between 30 and -7 degrees C. The Ea for cells incubated at temperatures above 0 degrees C (3.92 kcal/mol) show an apparent discontinuity (p < 0.004) in water permeability in supercooled conditions (7.48 kcal/mol). These data suggest that 1) human sperm have a high Lp and low Ea, relative to other cell types, above 0 degrees C; and 2) this high Lp and its low Ea change significantly below 0 degrees C.

Osmotic effects on feline spermatozoa from normospermic versus teratospermic donors.

Effects of osmolality stresses on the sperm of normospermic (>60% normal sperm/ejaculate) versus teratospermic (<40% normal sperm) domestic cats and the normospermic leopard cat and the teratospermic clouded leopard were studied. Spermatozoa were exposed to various anisotonic solutions in a single step or returned to near isotonic conditions in a single step after exposure to anisotonic solutions. The percentage of sperm motility was measured subjectively, and dual fluorescent stains were used to assess membrane integrity by flow cytometry. The percentage of sperm motility declined (P < 0.05) in domestic cat sperm exposed to osmolalities <200 and >450 mOsm. Spermatozoa from all felines underwent marked (P < 0.05) membrane disruption following a hypotonic stress, but sperm from teratospermic donors experienced greater (P < 0.05) membrane disruption in response to decreased osmolality. While feline spermatozoa appeared to be highly resistant to hypertonic (600, 1200, and 2400 mOsm) conditions, with >85% of the cells maintaining intact membranes, severe membrane disruption occurred when cells were returned to isotonicity in a single step. There was no difference (P > 0.05) between a 1- and 5-min exposure to various anisotonic solutions. Similarly, sperm from normospermic and teratospermic domestic cats responded identically after exposure to ionic or nonionic solute. Results demonstrate that: (1) spermatozoa from teratospermic males are more vulnerable to a hypotonic stress than sperm from normospermic counterparts; (2) in response to small deviations in osmolality, feline sperm experience a more rapid decline in motility than membrane integrity; and (3) an abrupt return to isotonicity after a hypertonic stress causes extensive sperm membrane damage regardless of ejaculate quality.

Glycerol permeability of human spermatozoa and its activation energy.

Glycerol has commonly been employed as a cryoprotectant in cryopreservation of human spermatozoa. However, the addition of glycerol into the sperm before freezing and the removal of glycerol from the sperm after freezing and thawing result in anisotonic environments to the cells, which can cause cell injury. To define optimal procedures for the addition/removal of glycerol and to minimize the cell injury, one needs to know the kinetics of glycerol permeation across the sperm plasma membrane at different temperatures. For this, one has to determine the permeability coefficient of glycerol (Pg) and its activation energy (Ea). Values of Pg at different temperatures and at different glycerol concentrations were determined by measuring the time required for 50% spermolysis in hyperosmotic glycerol solutions which were hypotonic with respect to electrolytes. Value of the Ea was determined assuming an Arrhenius type temperature dependence of Pg. A dual fluorescent staining technique (propidium iodide and 6-carboxyfluoroscein diacetate) and flow cytometry were used to measure the spermolysis. The values of Pg in 0.5, 1.0, 1.5, and 2.0 M glycerol at 22 degrees C are 1.62, 1.88, 1.68, and 1.54 x 10(-3) cm/min, respectively. The values of Pg in 1 M glycerol at 0, 8, 22, and 30 degrees C are 0.33, 0.54, 1.88, and 2.60 x 10(-3) cm/min, respectively. The value of Ea is 11.76 kcal/mol.