The method of choice worldwide for artificial insemination (AI) in pigs involves the use of liquid-preserved boar spermatozoa stored at 15–18°C (Johnson et al., 2000). When Beltsville thawing solution (BTS), a so-called short-term extender, is used, AI usually takes place within 72 h after semen collection. During this time, standard semen parameters, i.e., motility and plasma membrane integrity, usually give no indication of a substantial reduction in semen quality; both these features are considered to be minimum prerequisites for satisfactory fertility outcomes after insemination (1). However, results from an insemination trial under experimental conditions have suggested that, despite relatively high values for standard semen parameters, storage in vitro reduces the ability of sperm to survive and fertilize in vivo (2); the study even demonstrated that semen storage for as little as 24–48 h in BTS extender leads to a significant reduction in the fertilization rate of ova in vivo. It is, therefore, clear that better laboratory methods for detecting changes in sperm fertilizing capability are required.
In the context of semen preservation and quality assessment, special emphasis is regularly laid on the phrase that the plasma membrane should retain its “functional integrity.” This expression implies that not only should macroscopic and microstructural integrity be maintained but also that cells should remain receptive to external stimuli and respond in an appropriate manner (3, 4). A particular sperm function that is crucial for fertilization and requires external stimuli is the ability to undergo the capacitation process. In vivo, capacitation takes place in the female genital tract. It is considered to be a coordinated series of events leading to a gradual destabilization of the spermatozoon that enables it to undergo the acrosome reaction in response to contact with the zona pellucida and then to fuse with the egg beneath; in sperm that have not succeeded in fertilizing, the destabilization is believed to lead eventually to cell death (5).
A key molecule that triggers and maintains the capacitation process in vitro (and that probably plays a similar role in vivo) is the bicarbonate ion. Several studies have indicated that many molecular changes during capacitation are highly dependent on the presence of extracellular bicarbonate (reviewed by 6). Although monitoring of most of the changes require sophisticated experimentation, detection of early changes in membrane stability is readily performed via monitoring of intracellular calcium levels. Under stable conditions and even in the face of external levels of some 2 mM calcium, cytoplasmic levels within the sperm cell are less than 40 nanomolar (cf. Ref. 7,8); however, as the plasma membrane destabilizes, the cell loses its ability to maintain this low level. The consequent rise in intracellular calcium can be detected via fluorescent calcium probes such as Fluo-3.
Such rises can be observed in single cells using fluorescence microscopy, but more useful population information can be obtained using flow cytometry to monitor very large numbers of cells at frequent intervals. It has been shown that individual spermatozoa respond to bicarbonate (destabilize) under in vitro capacitating conditions at different rates (9–11). By means of flow cytometry, the rate and degree of response within the sperm population can be readily followed. Indeed, an earlier study by our group (12) used Fluo-3 in combination with flow cytometry to detect the effect of different dilution and storage temperatures on sperm response to bicarbonate.
The aim of this study was to investigate the effect of long-term storage of extended boar semen, comparing motility and viability parameters with sperm responses to incubation under capacitation conditions as monitored by flow cytometry of calcium permeability. Special emphasis was put on the detection of possible changes in specific responses to bicarbonate because there have been indications that spermatozoa may lose their responsiveness to bicarbonate during storage (13, 14). Our hypothesis was that this loss of response could prove to be a sensitive measure of storage-associated loss of semen quality (i.e., loss of fertilizing capacity).
MATERIALS AND METHODS
General Experimental Design
Diluted semen samples were evaluated after 12, 24, 72, 120, and 168 h of storage at 17°C. At each time point, basic semen quality was characterized by sperm motility and the percentage of cells with intact plasma and acrosomal membrane. Subsequently, subsamples were incubated under in vitro capacitating conditions and at intervals assessed for changes in viability (plasma membrane integrity) and intracellular calcium levels using a flow cytometer with Fluo-3 as the calcium reporter probe. The Kd of Fluo-3 for binding Ca2+ at 37°C is reported to be 864 nM (15), that is, at least 20-fold higher than the resting level of intracellular calcium in sperm, but some 2000-fold lower than the extracellular calcium level (2 mM) in the capacitating medium; thus, capacitating sperm with increased permeability to calcium would be readily detectable. Control subsamples were incubated under non-capacitating conditions and similarly assessed. Emphasis was placed on specific effects of bicarbonate on sperm subpopulations over incubation time.
All chemicals used were of analytical grade. Unless otherwise stated they were purchased from Merck (Darmstadt, Germany) and Roth (Karlsruhe, Germany). Propidium iodide (PI) was obtained from Sigma-Aldrich (Steinheim, Germany), whereas Fluo-3/AM and fluorescein-isothiocyanate-conjugated peanut agglutinin (FITC-PNA) were both purchased from Axxora (Lörrach, Germany).
For the experiments three ejaculates were used from each of 14 mature, healthy boars (Pietrain, German Large White, and crossbred animals) housed either at the Unit for Reproductive Medicine of Clinics or at a nearby AI station. Semen was collected using the “gloved hand”-technique and filtered through gauze to remove gel particles. Immediately after collection, semen samples were transferred to the laboratory and diluted with prewarmed (35°C) BTS (Minitube, Tiefenbach, Germany) to a final concentration of 20 × 106 spermatozoa per ml. After dilution semen samples were cooled at room temperature for 1.5 h and subsequently stored at 17°C. Samples from the native semen were evaluated for motility and morphology of the spermatozoa as described by Petrunkina et al. (16) to ensure that only normospermic ejaculates were used for the experiments (i.e., ejaculates with more than 70% motile and less than 25% morphologically abnormal sperm).
Assessment of Sperm Motility
An aliquot of 4 ml diluted semen was incubated for 15 min at 38°C in a water bath and motility assessed with the CASA-system SpermVision® (Minitüb, Tiefenbach, Germany) using four-chamber slides (Leja, Nieuw Vennep, The Netherlands) with a chamber depth of 20 μm. The CASA-system was equipped with a 20-fold objective, a camera adapter (U-PMTVC tv-0.75, Olympus, Hamburg, Germany), and a camera with a resolution of 800 × 600 pixels (AccuPixel TM6760 CL, JAI A/S, Glostrup, Denmark). The system was operated by SpermVision® software (Version 3.5, Minitüb, Germany). For each sample, 10 successive fields in the central axis of a chamber were recorded at a rate of 30 pictures per 0.5 s for each field. The parameters assessed were the percentage of motile sperm (total motility), the percentage of progressively motile sperm (progressive motility), and for the group of progressively motile sperm also the average straight-line velocity (VSL), curved-line velocity (VCL), average path velocity (VAP), average amplitude of lateral head-displacement (ALH), and beat cross frequency (BCF). A spermatozoon was considered to be motile when its average head orientation change was higher than 2.5°, and considered to be progressively motile when the distance moved from A to B in a straight line exceeded 4.5 μm.
Assessment of the Integrity of the Plasma and Acrosomal Membrane
An aliquot of the diluted semen sample was stained with PI (final concentration 2.5 μg/ml) and FITC-PNA (final concentration 3.0 μg/ml) for 5 min at room temperature. A subsample of 5 μl was transferred to 895 μl of a saline medium buffered with 4-(2-hydroxyethyl)-1-piperazine-ehtanesulfonic acid (HEPES) (HBS: 137 mM NaCl, 20 mM HEPES, 10 mM Glucose, 2.5 mM KOH, pH 7.4 at 20°C, 300 ± 5 mOsmol/kg) and 10,000 events analyzed with a DAKO “Galaxy” flow cytometer (DAKO, Hamburg, Germany) controlled by “FloMax®” software (version 2.4, Partec, Münster, Germany). The cells were excited at a wavelength of 488 nm by an argon ion laser (20 mW). Fluorescence signals for FITC-PNA were detected using a 537.5/22.5-nm bandpass filter and for PI using a 630 nm long pass filter. Forward and side scatters were plotted on linear scales, fluorescence data were plotted on logarithmic scales. Traditionally, the sperm population is identified by characteristic forward scatter )FSC) and side scatter (SSC) distribution patterns (so-called L-shape—see Fig. 1A). Voltages and gains were set to record L-shaped light scatter signal and to detect a negative control for FITC-PNA and PI in first logarithmic decades, respectively. However, identification exclusively by means of light scatter results in misestimation of sperm counts in diluted semen samples (17); therefore, a correction of data for non-DNA containing particles in the samples was performed as proposed by Petrunkina et al. (17, 18). The percentage of sperm with intact plasma membrane (PI-negative, PIneg) and intact acrosomal membrane (FITC-PNA-negative, FITC-PNAneg) was determined. The overlap of emission spectra between PI and FITC-PNA was compensated after acquisition using “FloMax®” software.
Preparation of Sperm and Media for Kinetic Studies
For each day of analysis, aliquots of diluted semen were centrifuged through a discontinuous gradient of 35 and 70% iso-osmotic Percoll®-saline (GE Healthcare, Munich, Germany) essentially as described by Harrison et al. (9). Briefly, 4 ml of diluted semen were layered over a two-step gradient of 4 ml of 35% Percoll-saline on 2 ml of 70% Percoll-saline. Tubes were centrifuged at 300g for 10 min followed by 15 min at 750g. After centrifugation, the supernatant was aspirated to leave a residual of 0.8 ml of 70% Percoll-saline, in which the sperm pellet was resuspended. Next, 4 μl of a Fluo-3/AM stock solution (1 mM in dimethylsulfoxide) were added to the sperm suspensions and incubated in the dark at ambient temperature. After 10 min, the samples were diluted sixfold using HBS supplemented with polyvinylpyrrolidone and polyvinyl alcohol (each 0.5 mg/ml) and incubation continued for another 20 min in the dark. The diluted sperm suspension was then layered on 4.8 ml of a sucrose washing medium (232 mM sucrose, 20 mM HEPES, 10 mM Glucose, 2.5 mM KOH, pH 7.4 (20°C), 300 ± 5 mOsmol/kg) and centrifuged for 10 min at 300g followed by 10 min at 700g to remove free dye. Subsequently, the supernatant layers were aspirated off to leave 0.5 ml of sucrose medium, in which the sperm pellet was resuspended. The suspension of loaded sperm was kept in the dark at ambient temperature and used within 1 h of preparation. Aliquots of prepared spermatozoa were incubated at a concentration of 1 1.5 × 106 cells per ml in three different variants of Tyrode's medium (9). The complete medium consisted of 96 mM NaCl, 20 mM HEPES, 5 mM glucose, 3.1 mM KCl, 0.4 mM MgSO4, 0.3 mM KH2PO4, 100 μg/ml gentamycin sulfate (SERVA, Heidelberg, Germany), 20 μg/ml phenol red, 1.0 mM sodium pyruvate, 21.7 mM sodium lactate, 3 mg/ml bovine serum albumin (Cohn's fraction V, fatty acid free), 15 mM NaHCO3 and 2 mM CaCl2. The complete medium is referred to as TyrBicCa. The control media both lacked bicarbonate and contained either 2 mM CaCl2 (TyrCa) or 1 mM Na2-EGTA (disodium ethylene glycol tetracetate) instead of calcium (TyrControl). All media were adjusted to a pH of 7.4 at 38°C and an osmolality of 300 ± 5 mOsmol/kg. The osmolality of the control media was adapted by modifying the NaCl content. Before use, all media were passed through a 0.22-μm filter (PES membrane, Roth, Karlsruhe, Germany) to reduce “noise” in the flow cytometric analyses, and PI (final concentration 2.5 μg/ml) was added. For equilibration, TyrBicCa was kept in an incubator under 5% CO2 and 100% humidity, whereas TyrCa and TyrControl were kept sealed in a heating cabinet.
Flow Cytometric Assessments for Kinetic Studies
Samples were analyzed after 3, 20, 40, 60, 90, 120, 150, and 180 min of incubation on the DAKO “Galaxy” flow cytometer as described earlier. Each time, a total of 10,000 events were assessed at a flow rate of 400–800 events per second. The sperm population was identified by characteristic forward and side scatter distribution patterns (“L-shape”). Only non-agglutinated events were considered for further analysis. Figure 1 illustrates the gating strategy and process of analysis. Among cells gated on the L-shaped FSC/SSC signal (Region R1, Fig. 1A), signals for PI distinguished between cells with defective plasma membranes (PI-positive, PIpos: Region R2, Figs. 1B and 1Ca) and intact plasma membranes (PI-negative, PIneg: Regions R3 and R4), whereas Fluo-3 further subdivided the PIneg sperm population into cells with a low Fluo-3 fluorescence signal (live low-Ca2+ cells, Fluo-3neg Region R3, Figs. 1B and 1Cb) and those with a higher Fluo-3 fluorescence signal (live high-Ca2+ cells, Fluo-3pos, Region R4, Figs. 1B and 1Cc). For kinetic analysis, quadrants based on the above strategy were defined from data collected after 3 min incubation to distinguish three sperm subpopulations. These quadrants were kept constant for all following incubation time points and were used to evaluate changes occurring during incubation (cf. Fig. 3). The overlap of emission spectra from PI and Fluo-3 was compensated after acquisition using “FloMax®” software.
Data were analyzed using “FloMax,” Excel®, and the Statistical Analysis System software (SAS®, version 9.1; SAS Inst., Cary, NC). Selected list mode data files were exported from FloMax and processed with FlowJo™ (Treestar) software for graphical illustrations. For description of kinetic changes in sperm subpopulations during in vitro capacitation, linear and nonlinear modeling was performed using PROC REG and PROC NLIN. Calculations for both procedures were limited to a maximum of 200 iterations. The “Gauss-Newton” method was used for nonlinear regressions. When more than one model could be derived for a sperm subpopulation, the model achieving the highest goodness of fit was chosen, unless otherwise stated.
Selected parameters from the kinetic studies were tested for normality and compared between media and times of storage (PROC UNIVARIATE). Data for motility and membrane integrity were also tested for normality and comparisons made between storage time points. Influence of storage length on all semen parameters was tested by one-way ANOVA (PROC GLM).
Estimation of variance components for the factors “boar” and “ejaculate” on the specific responsiveness to bicarbonate of a sample were performed using the method of minimum variance quadratic unbiased estimation (MIVQUE0) from the VARCOMP procedure.
Unless otherwise stated, data are presented as means ± standard deviations. Differences were considered to be significant when their probability of occurring by chance was less than 5% (P < 0.05).
Motility and Integrity of Plasma and Acrosomal Membranes in Stored Semen Samples
Diluted semen samples at 12 h of storage were characterized on average by 83.1% PI-negative (PIneg) and FITC-PNA-negative (FITC-PNAneg) sperm, 89.0% total motility, and 86.4% progressive motility (Table 1). These parameters were influenced by storage length (P < 0.05), decreasing slightly to 81.7% PIneg/FITC-PNAneg sperm, 86.5% total motility, and 83.5% progressive motility at 72 h (P < 0.05). Further storage until 168 h led to a further ongoing reduction of motile and membrane-intact spermatozoa. One semen sample at 120 h and an additional 14 samples at 168 h were not assessable for kinetic studies, because of an insufficient yield of sperm after centrifugation through Percoll®. Thirteen of these 15 samples had a progressive motility of less than 30% at the time points mentioned. The samples that were useable for kinetic studies after 168 h storage contained on average 71.0% PIneg/FITC-PNAneg sperm and exhibited 74.4% total motility and 69.9% progressive motility before density gradient centrifugation. The other motility descriptors (VAP, VCL, VSL, ALH, and BCF) for the semen samples used in the kinetic studies did not change with storage time compared to values at 12 h of storage (P > 0.05; Table 1).
Table 1. Conventional assessment parameters of stored semen samples (n = 14 boars, m = 42 ejaculates) used to establish kinetic profiles for calcium influx
The percentage of sperm with intact plasma and acrosomal membranes (PI and FITC-PNA-negative) was determined by flow cytometry. Motility parameters were assessed by a CASA-system. All data are means and standard deviations. Different letters within a row indicate significant differences (P <0.05).
n = 14 boars; m = 41 ejaculates
n = 11 boars; m = 27 ejaculates
PI- & FITC-PNA-neg. (%)
83.1 ± 4.2a
82.1 ± 4.6a,b
81.7 ± 4.8b
79.6 ± 6.1c
71.0 ± 20.9d
Total motility (%)
89.0 ± 3.2a
87.6 ± 2.9b
86.5 ± 2.9c
83.4 ± 7.0d
4.4 ± 10.4e
Progressive motility (%)
86.4 ± 3.6a
84.0 ± 3.3b
83.5 ± 2.9b
79.0 ± 7.0c
69.9 ± 11.1d
71.5 ± 5.4a
69.0 ± 9.5a
72.0 ± 8.4a
72.7 ± 6.5a
70.5 ± 9.3a
99.2 ± 8.6a
98.2 ± 12.6a
99.5 ± 11.4a
103.1 ± 12.9a
100.8 ± 16.9a
61.0 ± 5.6a
58.4 ± 9.0a
61.1 ± 8.4a
61.2 ± 7.2a
59.3 ± 7.7a
1.82 ± 0.20a,c
1.81 ± 0.21a
1.78 ± 0.24a
1.81 ± 0.31a
1.87 ± 0.39a
42.6 ± 1.9a
41.9 ± 2.2a
41.7 ± 3.0a
41.0 ± 3.0a
38.6 ± 3.5a
Kinetic Changes in Sperm Subpopulations During Incubation
At 12 h storage, incubation in the complete medium with bicarbonate and calcium (TyrBicCa) resulted in a continuous and major decline in the percentage of live spermatozoa with low intracellular calcium concentration (Fluo-3-negative and PI-negative: Fluo3neg/PIneg), as shown in Figure 2 and in “Region 3 Low Ca2+ live” in Figures 3A and 3B. At the same time, the percentage of Fluo-3-positive/PI-negative (plasma-membrane-intact sperm with high intracellular calcium concentration: Fluo-3pos/PIneg) increased (Fig. 2 and “Region 4 High Ca2+ live” in Figs. 3A and 3B). The percentage of PI-positive (plasma-membrane damaged: PIpos) spermatozoa also increased (“Region 2 PI-positive” in Figs. 3A and 3B). As expected, the majority of cells died in the media where they were incubated with bicarbonate and calcium ions, both known triggers of destabilization and cell death. In the control medium (TyrControl), changes occurred only to a minor extent.
Table 2. Model equations for kinetic changes in sperm subpopulations of samples incubated for 180 minutes in three variants of a Tyrode's medium after different storage length at 17°C
Equations are given for models that resulted in the highest goodness of fit. Exceptions were made for Fluo-3-positive/PI-negative spermatozoa incubated in Tyrode's medium only with calcium (TyrCa) after 72 and 168 hours of storage, respectively, to improve comparability of successive times of storage. Models with the best goodness of fit are 2.14t0.3511 (R2 = 0.98) at 72 h and 1.38 ln(t) + 7.54 (R2 = 0.93) at 168 h, respectively. All equations have P < 0.05. In Tyrode control all linear or logarithmic models for Fluo-3/PI negative population have missed significance levels; power function models were significant but at a very low goodness-of-fit.
TyrBicCa = Tyrode's medium with 15 mM bicarbonate and 2 mM calcium.
TyrCa = Tyrode's medium without bicarbonate, with 2 mM calcium
TyrControl = Tyrode's medium without bicarbonate and calcium.
From Figure 2, it can be seen that the greatest changes in redistribution of subpopulations in TyrBicCa after 12 h storage occurred between 3 min and 40–60 min of incubation. After prolonged storage, the plateau phase for the Fluo-3pos/PIneg sperm population in TyrBicCa was reached after 40 min (120 h) and 20 min (168 h), respectively. Therefore, values at 60 min incubation were chosen as parameters for evaluating all effects of storage.
Influence of Storage Length on Kinetic Profiles
Kinetic changes in the Fluo-3neg/PIneg and Fluo-3pos/PIneg subpopulations during incubation in variants of Tyrode's medium were readily derived as functions of incubation time (t) at all time-points of storage (see Table 2). For all populations in all media at all points of storage, with the exception of the Fluo-3pos/PIneg population of sperm cells in TyrControl, the regression curves had very high goodness-of-fit and high significance levels. The Fluo-3pos/PIneg population were essentially absent during incubation in TyrControl because the presence of EGTA in this medium resulted in a free calcium concentration of less than 2.4 nM—see Ref. 9; thus, spermatozoa whose calcium homeostasis was compromised would be likely to lose intracellular calcium in this medium in contrast to the situation in TyrCa—a relevant experiment is illustrated in Harrison et al. (9).
In the bicarbonate-containing medium (TyrBicCa) changes of the Fluo-3neg/PIneg population followed logarithmic model curves with a negative slope. With increasing storage length, the absolute value of this slope decreased (verified by comparing the first derivatives), thereby indicating a less pronounced reaction of the sperm cells to in vitro capacitating conditions; the most prominent reduction in rate of response could be observed between 12 and 72 h of storage, where factors changed from −17.6 to −13.6. In parallel with the lessening of slope, the y-axis intercept decreased. This development was due to two processes. On the one hand, the initial distribution of spermatozoa among subpopulations changed with ongoing storage. The values of Fluo-3neg/PIneg sperm after 3 min incubation in TyrBicCa decreased (P < 0.05) from 87.9% at 12h to 83.0% at 72 h and 75.2% at 168 h (see Table 3). This initial loss in sperm stability due to storage length was also reflected in an increase (P < 0.05) of Fluo-3pos/PIneg spermatozoa in TyrBicCa at 3 min from 3.3% after 12 h to 6.1% after 72 h and finally to 10.1% after 168 h (see Table 3). A similar but lesser trend could be detected in the PIpos sperm population. On the other hand, the percentage of the Fluo-3neg/PIneg sperm population remaining after 60 min incubation in TyrBicCa increased with ongoing storage time. Compared to the situation at 12 h (29.3%), values were higher at 72 h (38.1%) and 168 h of storage (38.9%; P < 0.05 each). This trend toward a reduced destabilization rate could also be observed in the regression curve equations derived for the Fluo-3pos/PIneg sperm population in TyrBicCa (Table 2). Table 3 shows that while values for this population at 3 min incubation time increased steadily with storage length, values at 60 min incubation decreased significantly from 21.6% at 12 h to 15.7% at 72 h of storage (P < 0.05). Further storage until 120 h and 168 h resulted in values after 60 min incubation equal to those after 72 h (P > 0.05). One may note that storage was not accompanied by a higher amount of PIpos spermatozoa in TyrBicCa after 60 min incubation.
Table 3. Sperm subpopulations after 3 min and 60 min incubation in three variants of Tyrode's medium after different semen storage times
Washed spermatozoa were incubated in Tyrode&'s based media containing either 15 mM bicarbonate and 2 mM calcium (TyrBiCa), or 2 mM calcium without bicarbonate (TyrCa), or no bicarbonate and calcium (TyrControl). Results are given for plasma-membrane-intact spermatozoa (propidium iodide (PI)-negative) with either low or high intracellular calcium concentration (Fluo-3-negative and Fluo-3-positive, respectively) and for plasma-membrane-damaged spermatozoa (PI-positive). Values with different superscripts within a row differ significantly (P < 0.05).
n = 14 boars; m = 41 ejaculates.
n = 11 boars; m = 27 ejaculates.
&#142;§Indicates equal amounts of a sperm subpopulation in different media after a given storage time (P >0.05).
During incubation under non-capacitating conditions in the medium without bicarbonate but including calcium (TyrCa), the kinetic changes for Fluo-3neg/PIneg and Fluo-3pos/PIneg spermatozoa fitted best to linear regression equations (negative and positive, respectively) during the first 24 h of storage. An increasing instability of semen during storage during 72 h was evidenced by an approximately doubled slope for regressions of the Fluo-3neg/PIneg sperm population changes during incubation, accompanied by a decreasing y-axis intercept (Table 2). Pronounced instability became obvious when data for this population after 120 and 168 h of storage were fitted best to a negative logarithmic equation. The instability was mirrored by positive increases in the slopes of the regression equations derived for changes in the Fluo-3pos/PIneg population (linear and power equations, see Table 2).
Regression equations for changes in the Fluo-3neg/PIneg sperm population during incubation in the absence of bicarbonate and calcium (TyrControl) showed a steady increase in instability with storage length (Table 2), although this instability was less apparent than that seen during incubation in TyrCa.
Changes in Responsiveness to Incubation Conditions
The change in the amount of a sperm subpopulation during incubation in a medium within a given time is referred to as the responsiveness of a sperm sample to the incubation conditions (11). For this work, the responsiveness was calculated as the difference (Δ) between the data values at the beginning (3 min) and after 60 min of incubation. Values for the Fluo-3neg/PIneg sperm population were calculated as Δ = 3 min − 60 min, whereas for the Fluo-3pos/PIneg and PIpos populations, it was calculated as Δ = 60 min − 3 min. The responsiveness values are illustrated in Figure 4.
The responsiveness of the Fluo-3neg/PIneg population in TyrBicCa dropped gradually from 58.6% at 12 h to 36.4% at 168 h (P < 0.05; Fig. 4A). In contrast, the responsiveness in TyrCa calculated on the basis of this population increased from 6.8% at 12 h to 24.0% at 168 h (P < 0.05; Fig. 4B). Even though the responsiveness calculated on the basis of the Fluo-3neg/PIneg sperm population was always smaller in TyrCa than in TyrBicCa (P < 0.05), this difference continuously diminished, thus by extrapolation so did the response that could be attributed to a specific action of bicarbonate: the latter specific responsiveness, that is, the difference between the response in TyrBicCa and TyrCa was reduced from 51.8% at 12 h to 28.3% at 72 h and 12.4% at 168 h (P < 0.05 each).
The responsiveness of the Fluo-3pos/PIneg sperm population during incubation in TyrBicCa declined as a result of storage relatively more rapidly than did the decline in responsiveness of the Fluo-3neg/PIneg population, falling by approximately half between 12 h storage and 72 h (18.3–9.5%, respectively) and even further to 5.3% at 168 h (P < 0.05; Fig. 4A). Conversely, the responsiveness of the Fluo-3pos/PIneg population in TyrCa increased from 1.2% (12 h) to 5.1% at 72 h (P < 0.05, Fig. 4B) and remained at this level until 168 h of storage (Fig. 3B).
A comparison between the values for responsiveness in TyrCa and TyrControl based on all three sperm populations (Fluo-3neg/PIneg, Fluo-3pos/PIneg and PIpos) indicates that extracellular calcium is a source of increased destabilization of liquid-preserved spermatozoa during incubation in vitro (compare Figs. 4B with 4C). The destabilization due to calcium, calculated as the difference between the responsiveness in TyrCa and TyrControl in terms of loss of Fluo-3neg/PIneg sperm, increased within the first 72 h of storage from 2.2% (12 h) to 7.4% (72 h; P < 0.05) but during further storage remained constant (P > 0.05). Nevertheless, lengthy storage during liquid preservation at 17°C resulted in a significant increase in instability during incubation, even in the bicarbonate-free and calcium-free medium Tyrcontrol: loss of Fluo-3neg/PIneg in TyrControl increased from 4.7% (12 h) to 16.4% (168 h; P < 0.05) (Fig. 4C).
Differences between Boars in their Specific Responsiveness to Bicarbonate
The quantitative responses of Fluo-3neg/PIneg spermatozoa to capacitating and noncapacitating media varied greatly between boars and between ejaculates. After 12, 24, and 72 h of storage the individual boar effect accounted for 69%, 68%, and 72% of the variance observed for the specific response to bicarbonate within 60 min of incubation (see Table 4). With longer storage length the proportion of variance due to boar decreased, the factors “boar” and “ejaculate” then accounting in equal parts for the observed variance of this parameter. Examples showing the different evolution of specific responsiveness to bicarbonate in individual boars following 72 h storage at 17°C are given in Figures 5A–5C. Samples from some boars (e.g., boar no. 12; Fig. 5C) exhibited constant responsiveness during 72 h of storage without major signs of destabilization in either TyrBicCa or TyrCa. Conversely, the responsiveness for others in TyrBicCa showed a major decrease between 24 and 72 h storage [cf. boars no. 6 (Fig. 5A) and 7 (Fig. 5B)]. This change in specific responsiveness to bicarbonate was accompanied by a varying increase in instability in TyrCa. To what extent these results would be reflected in fertility measures remains to be shown; sufficient fertility data were not available for these boars.
Table 4. Components of variance for specific responsiveness to bicarbonate calculated on the basis of the Fluo-3-negative/PI-negative sperm population
The calculations were carried out using the method MIVQUE0 (minimum variance quadratic unbiased estimation) from the VARCOMP procedure within the Statistical Analysis System software. The specific responsiveness to bicarbonate, as described by the difference in responsiveness in the media TyrBiCa and TyrCa is dominated during the first 72 hours of storage by the factor ‘boar’ (n = 14 boars, 3 ejaculates per boar)
n = 14 boars; m = 41 ejaculates.
n = 11 boars; m = 27 ejaculates.
The conventional parameters for assessing sperm quality, namely motility and plasma and acrosomal membrane integrity within the stored semen, showed only little change in sperm characteristics up to 120 h of storage. Notably, these parameters were especially insensitive to changes between 24 and 72 h of storage, confirming an earlier study by our group (19). Thus, on the basis of conventional parameters, the stored ejaculates appeared to maintain good quality during 120 h of storage, as they contained at least 80% motile spermatozoa and would normally be considered entirely suitable for AI. Commonly, a minimum of 60–70% motile sperm is considered to be necessary for AI in pigs to result in normal fertility outcomes (1, 20).
However, flow cytometric assessment of calcium levels in incubated sperm samples revealed major storage-related changes in the sperm population. By comparing the effects of incubation in three different parallel media, namely bicarbonate-plus-calcium-containing (TyrBicCa) with calcium-containing (TyrCa) and calcium-free (EGTA-containing TyrControl), we were able to distinguish effects due to bicarbonate from effects due to calcium (see Fig. 4). The most noticeable effect of storage was on the sperm population response to bicarbonate.
The original article by Harrison et al. (9) studying bicarbonate-mediated calcium influx in Fluo-3 loaded sperm described the effect of bicarbonate as causing a decrease in the population of live cells with low intracellular calcium with the concomitant appearance of a subpopulation of plasma-membrane-intact spermatozoa with high calcium together with an increase in the dead cell population. Our findings essentially confirmed these observations: incubation in the presence of bicarbonate caused a progressive loss of low-calcium live cells and an increase in both high-calcium-live and dead cells. However, an increasing period of storage progressively reduced the specific response of sperm to bicarbonate. This could be seen in two ways. First, there was an overall reduction in the loss of low-calcium live cells which was not accompanied by concomitant increases in high-calcium live cells and dead cells (Fig. 4A). Second, increasing storage resulted in an increased sensitivity to incubation per se, especially in the presence of calcium. This sensitivity was revealed as a loss of low-calcium live cells coupled with a small increase in high-calcium live cells but marked increase in dead cells (cf. Fig. 4B with 4C and see Table 3). Overall, therefore, the effect of storage may be interpreted as causing a loss of specific responsiveness to bicarbonate and an increase in overall instability. The increase in overall instability does not account, however, for the loss in responsiveness to bicarbonate. In fact, surprisingly, storage appears to render a significant subpopulation of intrinsically stable spermatozoa refractory to bicarbonate.
The extent of this subpopulation can be calculated as follows (see Table 3): In the original population stored for 12 h, there were on average 91.2% low-calcium live cells after 3 min incubation in the control medium; after 60 min in TyrBicCa, there were only 29.3%; of the 61.9% lost, 4% represented an intrinsic loss due to the control incubation and 3.5% further loss due to the presence of calcium (TyrCa 60–TyrControl 60), thus, the percentage of sperm specifically responsive to bicarbonate at 12 h was 54.4%. In the population stored for 120 h, there remained 85.1% of low-calcium live cells at 3 min; after 60 min in TyrBicCa, there were only 40.8%; of the 44.3% lost, 11.3% represented the loss due to control incubation and 12.2% further loss due to the presence of calcium, thus, the percentage of sperm specifically responsive to bicarbonate at 120 h was only 20.8%. On this basis, 33.6% of the low-calcium live population had become refractory. It may be noted that a similar reduction in responsiveness to bicarbonate of stored spermatozoa was described by Harrison et al. (13), and Guthrie and Welch (14), in terms of fewer plasma membrane-intact spermatozoa exhibiting bicarbonate-induced plasma-membrane phospholipid disorder as detected by merocyanine-540 staining.
Harrison et al. (9) presented data to indicate that the high-calcium live population represents an early stage of instability through which sperm individuals pass during the overall destabilizing capacitation process. Our results agree with that general concept. However, inspection of the responsiveness data in Figure 4B shows that although longer-term storage results in a marked increase in the loss of low-calcium live sperm and a concomitant increase in dead sperm during incubation in TyrCa, there is no parallel increase in the high-calcium live population. The situation for samples incubated in TyrBicCa is more complicated due to the increasing presence of refractory sperm, nevertheless, it can be seen by comparing Figure 4A with 4B that, as storage lengthens, the proportion of high-calcium live that accompanies bicarbonate-induced loss of low-calcium live actually decreases until it approaches that detected during incubation in TyrCa. We interpret our observations as indicating that the high-calcium live state in the presence of bicarbonate becomes increasingly short-lived as storage renders most of the spermatozoa more intrinsically unstable.
The causes of the decline in response to bicarbonate and the increasing inherent instability brought about by storage are not clear, and the mechanisms involved are unknown. It is obviously desirable to store spermatozoa at as low a temperature as possible to minimize deleterious processes. However, the large numbers of viable boar sperm required for successful AI render the use of frozen semen impractical for routine commercial usage. It was long ago established that boar spermatozoa are sensitive to cooling and low suprazero temperatures, thus, dilution with warmed extender and subsequent slow cooling of the semen to a temperature of ∼ 17°C was developed empirically as the optimal storage procedure. Nevertheless, preservation of boar spermatozoa at 17°C has to be viewed as a hypothermic process with uncertain effects on cell structures and molecular functions. First, before storage, the cooling procedure of the diluted semen, although slow (from 35°C to ∼ 20°C in 1.5 h), might bring about changes in the plasma membrane that lead to disorganization and instability (cf. 21,22). Although the high response to bicarbonate and low instability displayed by the 12 h stored samples argues against any damaging effect of this cooling process, long-term storage at the markedly nonphysiological temperature of 17°C could well have deleterious effects on cellular processes. Lowering of temperature is very likely to reduce intrinsic enzyme activity, thereby lowering energy metabolism and ion homeostasis, with uncertain results on cell “house-keeping.” Further focused research will be needed to investigate these possibilities.
Another factor often discussed in context with sperm ageing during liquid storage is the potential for damaging oxidation processes (23). Spermatozoa contain high amounts of polyunsaturated phospholipids and cholesterol which make them particularly vulnerable to lipid peroxidation (24, 25). However, lipid peroxidation during 5 days liquid storage of boar semen has been found to be low (26), as has the incidence of intracellular reactive oxygen species (ROS) in boar spermatozoa (27). Thus, little clear evidence so far exists to suggest that ROS are a principal source of sperm deterioration during liquid storage.
The deterioration in sperm function during storage that we have detected in this study has clear implications with respect to AI practice. The key sperm process with respect to fertilization is capacitation. To fertilize the egg successfully, a given individual spermatozoon needs to capacitate in a close time relationship to ovulation. In describing the possible influences on fertility and subfertility of different capacitation dynamics within the sperm population, Petrunkina et al. (11) proposed an optimal time window within which fertilization is most likely. Moreover, they demonstrated that subfertility may be related to either too slow or too fast a response to in vitro capacitating conditions. As our study demonstrates, storage influences the initial destabilization rate as well as the overall response in a semen sample. So, storage may skew capacitation dynamics, with the result that fewer spermatozoa capacitate at an optimal time relative to ovulation. Capacitation is believed to be suppressed or modulated by binding of spermatozoa to the epithelium of the oviductal isthmus, referred to as the sperm reservoir (28). It is not clear to what extent the oviductal epithelium may compensate for storage damage in spermatozoa by stabilizing the cells that reach the region of the sperm reservoir, or whether binding becomes even more difficult for such stored sperm. However, it is likely that the reduced accessory sperm numbers observed after insemination with aged semen samples (2) may be related to a more unstable plasma membrane which either impairs residence in the sperm reservoir (29) or makes the sperm unresponsive to bicarbonate. It may also be that spermatozoa with impaired stability and subtle differences in membrane structure or function are recognized by the oviductal epithelial cells, resulting in “negative” selection, that is, these spermatozoa are forced to further destabilize and disintegrate (30).
In the light of the previous paragraph, it was notable (see Fig. 5) that individual boars showed marked differences in their ejaculates' resistance to storage, as judged by the responsiveness to bicarbonate (incubation in TyrBicCa) and to calcium (incubation in TyrCa). Although there were within-animal ejaculate differences, boar differences accounted for more than two-thirds of the total variance for specific responsiveness to bicarbonate up to 72 h (see Table 4). Thus far, it has been essentially impossible to detect differences in the effects of storage on semen from boars deemed of “high quality” on the basis of conventional semen assessment parameters. Such differences emerge currently only after careful monitoring of fertility results after AI with semen of differing age, a lengthy and potentially costly process. The testing of storage effect on specific responsiveness to bicarbonate during incubation in vitro appears to offer a potentially sensitive and simple laboratory assessment procedure. The boar differences thereby revealed must now be related to accurate AI fertility data.
In conclusion, our findings show that testing the specific destabilization response of sperm to bicarbonate challenge offers a sensitive way of assessing sperm functional ability. Using a calcium probe in combination with flow cytometry and comparing responses in a bicarbonate-containing Tyrode's medium with those in a bicarbonate-free control is a relatively simple procedure which allowed us to detect marked changes in the sperm population that could not be revealed by standard measures of sperm motility and viability. Loss of calcium homeostasis (i.e., reduction of the low-calcium live population) appears the best parameter of membrane destabilization. Two opposing effects of storage length could be demonstrated: part of the sperm population became increasingly unstable during storage, whereas another part became refractory toward bicarbonate stimulation. Either way, these changes in sperm membrane stability and capacitation dynamics that take place during storage seem likely to compromise the chances for establishing a functional sperm reservoir in the oviduct and thereby the chances for fertilization. The methods and approach used in this study may offer a better means to optimize storage conditions and diluents in the future.