Variability in relationships between semen quality and estimates of in vivo and in vitro fertility in boars

Animal Reproduction Science 81 (2004) 97–113

Variability in relationships between semen quality and estimates of in vivo and in vitro fertility in boars J.M. Popwell a , W.L. Flowers b,∗ a

b

San Francisco Center for Reproductive Medicine, 390 Laurel St., Suite 200, San Francisco, CA 94118, USA Department of Animal Science, 220-B Polk Hall, North Carolina State University, Raleigh, NC 97695-7621, USA Accepted 13 August 2003

Abstract The present experiment was designed to characterize relationships between common semen quality and fertility estimates for three boars known to differ in farrowing rate, number of pigs born alive, and monospermic penetration rate. The approach chosen to accomplish this was to monitor semen quality from these boars and use their semen alternately for either artificial insemination or in vitro fertilization for 40 weeks. This strategy relied on the variability in semen quality parameters that normally occurs in an individual boar over time. When comparisons were made among boars, farrowing rates, numbers of pigs born alive, and monospermic penetration rates were significantly different, but progressive motility, normal head and tail morphology, and acrosome morphology were not. However, when comparisons were made among ejaculates within individual boars, there were significant effects of semen quality on both in vivo and in vitro fertility. For boar 3495, the proportion of spermatozoa exhibiting progressive motility and distribution of spermatozoa in a percoll gradient had a positive linear effect on number born alive and monospermic penetration rate, respectively. For boar 2901, quadratic equations best described changes in litter size as a function of progressive motility and normal acrosomes. In addition, monospermic penetration rate increased linearly as normal acrosomes and the proportion of spermatozoa recovered from a percoll gradient increased. For boar 4291, the relationship between progressive motility and number born alive and between normal acrosomes and number of pigs born alive were also quadratic. However, a significant linear relationship was present only between normal acrosomes and monospermic penetration rate. These results demonstrate that simply relying on the means of common semen quality estimates from some boars has limited value in terms of being used as a prospective indicator of their in vivo or in vitro fertility. In contrast, characterization of relationships between semen quality and fertility estimates is useful for estimating differences in the fertility of ejaculates from individual ∗ Corresponding author. Tel.: +1-919-515-4003; fax: +1-919-515-4463. E-mail address: william [email protected] (W.L. Flowers).

0378-4320/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.anireprosci.2003.08.007

98

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

boars. However, both quantitative and qualitative differences in these relationships among boars are present and a given semen quality estimate that is a good predictor of in vivo or in vitro fertilization for one boar, may not be applicable for others. © 2003 Elsevier B.V. All rights reserved. Keywords: Boar; Semen quality; Fertilization

1. Introduction Boars are recognized as a significant source of variation with regards to the success of both in vivo (Flowers, 1997; Xu et al., 1998) and in vitro (Xu et al., 1996; Long et al., 1999) fertilization in swine. As a result, numerous studies have investigated relationships between semen quality estimates such as motility, morphology, and viability, and fertility estimates such as farrowing rates, numbers of pigs born alive, and in vitro fertilization rates in order to develop procedures for selecting boars prospectively for use in both systems. Unfortunately, conclusions drawn from these studies are equivocal. In some cases, characteristics such as normal acrosomes (Xu et al., 1998), normal head and tail morphology (Gadea and Matas, 2000), and progressive forward motility (Ivanova and Mollova, 1993; Flowers, 1997) had a positive relationship with boar fertility, while in others they did not (Xu et al., 1996; Flowers, 1997). While several factors could be responsible for this apparent dichotomy, it is interesting to note that a common feature of these studies was to compare means from boars that were obtained by averaging values from several ejaculates. With this type of analysis, the ability to examine relationships between semen quality and fertility estimates within a boar is lost. Observations from field studies demonstrate clearly that most estimates of semen quality and fertility vary significantly for boars over time (Flowers, 1997, 1998). Consequently, it is also possible that relationships between these two groups of measurements are not the same for all boars. In other words, a given semen quality estimate may be a good predictor of in vivo or in vitro fertilization for one boar, but not another. Consequently, the objective of this experiment was to examine variability in the relationships between common semen quality and fertility estimates among boars. 2. Materials and methods 2.1. Animals Three crossbred boars (Landrace×Large White×Duroc×Hampshire) that were 2.5±0.2 years of age and 215 ± 5 kg were used in the study. Historical data indicated that these three boars differed significantly in their in vivo (Table 4) and in vitro estimates of fertility. Boars were housed in individual pens (3.6 m × 2.7 m) in an environmentally controlled building. They were given ad libitum access to water and were fed a corn–soybean meal diet (14% crude protein) that exceeded all nutritional requirements for adult boars (NRC, 1998). Seven hundred twenty crossbred sows (Landrace × Large White × Yorkshire) with a mean parity

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

99

of 3.2 ± 0.7 were bred with semen collected from the three boars during the experimental period. Sows were maintained in individual crates (1.6 m×2.8 m) during breeding, gestation and farrowing. Sows were provided ad libitum access to water and fed corn–soybean meal diets formulated to meet or exceed their specific nutritional needs (NRC, 1998) during breeding, gestation, and lactation. 2.2. Experimental design The initial 30 ml of the sperm-rich portion of ejaculates from each boar was collected once per week for 40 weeks by the gloved-hand technique (Almond et al., 1998). The collection period began in October and ended in June. During the odd-numbered weeks (1, 3, . . . , 39), the following semen quality estimates were evaluated for each ejaculate: the percentage of progressively motile spermatozoa; the percentage of spermatozoa with normal head and tail morphology; the percentage of spermatozoa with normal acrosome morphology; and the percentage of spermatozoa with acrosin activity. In addition, 1 × 109 spermatozoa were placed on a percoll gradient and their distribution within different fractions of the gradient was determined. The remainder of the spermatozoa in each ejaculate were used at four different sperm/oocyte ratios, 2 × 10, 2 × 103 , 2 × 104 , and 2 × 107 , to estimate in vitro monospermic and polyspermic penetration rates of porcine oocytes. During the even-numbered weeks (2, 4, . . . , 40), the percentage of spermatozoa exhibiting progressive forward motility, normal head and tail morphology, normal acrosome morphology, and normal acrosin activity, as well as the distribution of spermatozoa in a percoll gradient were determined. After being evaluated, each ejaculate was extended in Beltsville Thawing Solution (Johnson and Pursel, 1975) and used to inseminate sows. The insemination dose was standardized to 3 × 109 total spermatozoa in 80 ml of extended semen and used within three days of collection. Sows (at least 10 per ejaculate per week) were bred once each day of estrus beginning on the first day of detected estrus. One experienced technician administered all the artificial matings during the experimental period. Farrowing rate and number of pigs born alive were recorded. All procedures involving sows and boars were approved by the North Carolina State University Institutional Animal Care and Use Committee. 2.3. Semen quality estimates Ejaculates were observed immediately after collection for the percentage of spermatozoa exhibiting progressive forward motility, percentage of spermatozoa with normal head and tail morphology, and percentage of spermatozoa with normal acrosome morphology. Progressive forward motility was evaluated by diluting semen 100-fold; placing diluted semen on a warmed microscope stage (35 ◦ C); and videotaping 10 different microscopic fields at 40× magnification with a Sony CCD color video camera attached to an Olympus Van-Ox S microscope (Opelco, Washington, DC). The proportion of spermatozoa exhibiting progressive forward motility in each field was determined by observing the video recording of each field in slow motion. For morphological evaluations, prepared slides were stained with Trypan Blue and observed at 40× magnification. All spermatozoa in 10 different microscopic fields were evaluated according to the morphological criteria established by Lunstra

100

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

and Echternkamp (1982). Acrosome morphology was evaluated only on spermatozoa with normal head and tail morphology. All spermatozoa in 20 microscopic fields were evaluated using the classifications of Pursel et al. (1972) with phase contrast microscopy at 100× magnification. Acrosin activity of spermatozoa and the distribution of spermatozoa in a percoll gradient were also evaluated in each ejaculate after collection. Acrosin activity was assessed according to slightly modified procedures of Penn et al. (1972). A 10 ␮l sample of each ejaculate was placed on a microscope slide coated with a mixture of 3.5% gelatin (Fisher Scientific, Sewanee, GA) and 0.03% detergent (Tween 80, Fisher Scientific, Sewanee, GA) and incubated for 75 min at 37 ◦ C. Slides were stained with Toludine Blue for 15 s and air dried. The presence of a circle of digested gelatin surrounding the head of spermatozoa (digestion halo) was considered to be indicative of normal acrosin activity. Two hundred spermatozoa from each ejaculate were evaluated to determine the percentage of cells with acrosin activity. Distribution of spermatozoa within percoll gradients was performed using a slightly modified version of the system described by Moohan and Lindsay (1995). An isotonic 90% percoll solution was prepared by combining 1 ml of a 10× stock solution of M199 (Sigma 2520 with 26.18 mM sodium bicarbonate and 0.01% penicillin-streptomycin) and 9 ml of percoll (Sigma, St. Louis, MO). From the 90% solution, 70 and 50% solutions were prepared by the addition of the appropriate amount of M199. Percoll gradients were prepared by the sequential addition of 1 ml of the 90, 70, and 50% solutions into a 15 ml conical tube. A total of 1 × 109 spermatozoa was added on top of the 50% layer. After the addition of spermatozoa, gradients were subjected to centrifugation at 1500 rpm and 30 ◦ C for 30 min. Each layer was then removed and placed into a separate tube and washed with 1 ml of M199 to remove excess percoll. The number of spermatozoa in each fraction was determined with a hemocytometer (Almond et al., 1998). The proportion of spermatozoa exhibiting progressive forward motility was also evaluated in each fraction of the percoll gradient. 2.4. In vitro penetration estimates Chemicals and media were purchased from Sigma (St. Louis, MO) unless otherwise stated. All media contained the following supplements: 26.18 mM sodium bicarbonate; 8.25 mM calcium lactate; 0.90 mM pyruvate; 3.05 mM glucose; 0.04% BSA, and 0.01% penicillin-streptomycin. Ovaries were obtained from a local abattoir, washed twice in phosphate buffered saline, and placed in M199 (Sigma 2520) at 33 ◦ C while being transported to the laboratory (60 miles). Upon arrival at the laboratory, all follicles 3–7 mm in diameter were aspirated with an 18-gauge needle attached to a 10 ml syringe. The fluid that was collected was pooled and placed in 50 ml of M199 (Sigma 2520) that was maintained at 37 ◦ C in 5% CO2 in air. Oocytes were removed and washed five times with M199 (Sigma 2520) and placed into 3 ml culture dishes (Becton Dickson, Franklin Lakes, NJ). They were then washed an additional 10 times with M199 (Sigma 3769) supplemented with 0.68 mM glutamine. Maturation conditions for oocytes were similar to those described by Funahashi and Day (1993). Oocytes were cultured in M199 (Sigma 3769) supplemented with 0.68 mM glutamine, 100 ␮l LH, 100 ␮l FSH, 100 ␮l ITS (insulin: 5 ␮g/ml, tranferrin: 5 ␮g/ml, and

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

101

sodium selenite: 5 ng/ml), 1 ␮g of estradiol: 17␤, and 1 ml of porcine follicular fluid (v/v) at 37 ◦ C in 5% CO2 in air. After 22 h, oocytes were transferred to M199 (Sigma 3769) with glutamine and cultured for an additional 20 h. After maturation, oocytes were transferred to M199 (Sigma 3769) with glutamine and 1.02 mM caffeine in preparation for fertilization. After collection, spermatozoa were separated from seminal fluids by centrifugation at 37 ◦ C and 500 × g. The supernatant was decanted and spermatozoa were washed two times with BTS followed by two more washes with M199 (Sigma 2520) supplemented with 10% fetal calf serum (v/v) (Hyclone, Logan, UT). After washings, spermatozoa were resuspended in 9 ml of M199 with fetal calf serum and incubated for 3 h at 37 ◦ C in 5% CO2 in air. Following capacitation, concentrations of 1 × 103 , 1 × 105 , 1 × 106 , and 1 × 109 spermatozoa/ml were prepared and added to 3 ml culture dishes containing 50 oocytes in 2 ml of M199 (Sigma 3769) with glutamine and 1.02 mM caffeine. This resulted in sperm/oocyte ratios of 2 × 10, 2 × 103 , 2 × 104 and 2 × 107 , respectively. All dishes were placed in 5% CO2 in air in an incubator at 39 ◦ C for 5 h. There were two replicates of 50 oocytes per replicate for each ratio for each ejaculate. After incubation with spermatozoa, oocytes were placed on slides and fixed for 48 h in 25% acetic alcohol at room temperature. Oocytes were then stained with 1% aceto-orecin and examined with phase contrast microscopy at 40× magnification. All oocytes were evaluated for maturational status and the presence of spermatozoa within their cytoplasm (Yoshida, 1987). Oocytes in metaphase I, metaphase II and diakinesis were considered to be mature. Those with an intact germinal vesicle or in the initial stages of germinal vesicle breakdown were considered to be immature. Oocytes were classified as being penetrated when at least one sperm head or male pronucleus was present in the vitellus. Oocytes with more than one sperm head or male pronucleus were considered to be polyspermic. Degenerating oocytes were not included in the analysis. 2.5. Statistical analyses Two separate analyses were used. The first group of analyses was performed to examine differences among boars in semen quality parameters, penetration rates, farrowing rate and number of pigs born alive. In general, data were analyzed with analysis of variance procedures using general linear model estimations (SAS, 1990) and Student–Newman–Keul’s multiple range test was used to determine differences among means of independent variables when significant effects were observed (Snedecor and Cochran, 1998). For semen quality parameters and penetration rates, ejaculate was considered to be the experimental unit and data were normalized with an arcsine transformation prior to analyses (Snedecor and Cochran, 1998). The statistical model for percentage of spermatozoa exhibiting progressive forward motility, normal head and tail morphology, normal acrosome morphology, and normal acrosin activity included the main effect of boar. Distribution of spermatozoa within percoll gradients was analyzed with a model that consisted of boar, fraction (50, 70, or 90%) and the boar by fraction interaction. A boar by fraction interaction was present (P < 0.05). Therefore, additional analyses were conducted to evaluate differences among boars within each fraction. Finally, a model that included boar, sperm/oocyte ratio and their interaction was used to analyze monospermic and polyspermic penetration rates.

102

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

Differences among boars in the number of pigs born alive and farrowing rate were analyzed using analysis of variance procedures for repeated measures for continuous (Snedecor and Cochran, 1998) and categorical data (Koch et al., 1977), respectively. Sows, which were bred with semen from each ejaculate (at least 10 per ejaculate per boar), were considered to be the experimental unit. In addition, sows were divided into three categories based on their parity at the time of breeding: parities 1 and 2; parities 3–5; and parities 6 and greater. The statistical model consisted of boar, ejaculate (week of study), parity group and appropriate interactions. Ejaculate was considered to be the repeated variable and tested accordingly in the analysis (Gill and Hafs, 1971). The previous number of pigs born alive was used as a covariate in the model. Student–Newman–Keul’s multiple range test was used to determine differences among means of independent variables when significant effects for number of pigs born alive were observed (Snedecor and Cochran, 1998). The second group of analyses used polynomial regression techniques to examine relationships between an individual semen quality estimate of an ejaculate and estimates of in vivo and in vitro fertility within each boar (Snedecor and Cochran, 1998). For these analyses, the percentage of spermatozoa exhibiting progressive forward motility, normal head and tail morphology, normal acrosome morphology, normal acrosin activity and the proportion of spermatozoa in the 90% fraction of the percoll gradient were treated as the independent variables. The number of pigs born alive, farrowing rate, and monospermic penetration rate were treated as the dependent variables. For this analysis, data from sperm/oocyte ratios of 2 × 103 were used because they produced the highest monospermic and lowest polyspermic penetration in vitro for each boar. 3. Results Semen quality estimates for motile and morphologically normal spermatozoa are summarized in Table 1. There were no differences among boars in the percentage of spermatozoa exhibiting progressive forward motility (P > 0.32); normal head and tail morphology (P > 0.45); or normal acrosome morphology (P > 0.47). However, boar 4291 produced Table 1 Differences among boars for semen quality estimates (mean ± S.E.) Semen quality estimate

Percentage of spermatozoa exhibiting progressive forward motility Percentage of spermatozoa with normal head and tail morphology Percentage of spermatozoa with normal acrosome morphology Percentage of spermatozoa with acrosin activity

N

Boar 2901

3495

4291

40

72.3 ± 7.3

77.1 ± 8.3

76.2 ± 7.1

40

74.5 ± 2.1

76.3 ± 2.0

72.2 ± 2.8

40

85.1 ± 6.1

82.1 ± 5.9

78.4 ± 4.1

40

84.3 ± 3.1a

86.3 ± 4.3a

74.7 ± 4.0b

N: number of ejaculates evaluated. Means with different superscripts (a and b) within the same row differ (P < 0.05).

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

103

Table 2 Spermatozoa in collections from different fractions of percoll gradients (mean ± S.E.) Percoll gradient fraction (%)

50 70 90

N

40 40 40

Boar 2901

3495

4291

22.3 ± 4.3 40.5 ± 4.1a 37.2 ± 5.1a

23.1 ± 6.3 37.3 ± 4.0a,b 39.6 ± 5.9a

24.2 ± 5.1 21.2 ± 4.5b 54.6 ± 4.0b

N: number of ejaculates evaluated. Means with different superscripts (a and b) within the same row differ (P < 0.05).

ejaculates with fewer (P < 0.05) spermatozoa with normal acrosin activity compared with boars 2901 and 3495. There was a significant interaction (P < 0.05) between boar and percoll gradient fraction. Consequently, differences among boars within each fraction are shown in Table 2. Ejaculates from boar 4291 had a greater and lower proportion of sperm cells in the 90 and 70% fractions, respectively (P < 0.05) compared with boar 2901. Ejaculates from boar 4291 also had a greater proportion of sperm cells in the 90% fraction than boar 3495. No differences (P > 0.50) were present among boars in the proportion of spermatozoa per ejaculate in the 50% fraction. The percentage of spermatozoa exhibiting progressive forward motility was highest (P < 0.05) in the 90% fraction (88.7 ± 1.3%), intermediate in the 70% fraction (63.5 ± 2.4%), and lowest (P < 0.05) in the 50% fraction (50.5 ± 2.3%). There was no boar by percoll gradient fraction interaction (P > 0.63) for progressive forward motility. The effects of boar and sperm/oocyte ratio on in vitro penetration rates are shown in Table 3. There were no differences among boars for polyspermic (P > 0.45) penetration rate. However, boar 4291 had higher (P < 0.05) monospermic penetration rates than boar 2901. More spermatozoa (P < 0.05) penetrated oocytes at ratios of 2 × 104 and 2 × 107 compared with ratios of 2×10 and 2×103 . Monospermic penetration was greater (P < 0.05) with 2 × 103 than with 2 × 104 or 2 × 107 sperm/oocyte. Similarly, 2 × 10 sperm/oocyte produced better (P < 0.05) monospermic penetration than a ratio of 2 × 107 . Polyspermic Table 3 Effect of boar and sperm/oocyte ratio on monospermic and polyspermic penetration rate of porcine oocytes matured in vitro (mean ± S.E.) Independent variable

N

Monospermic penetration rate (%)

Polyspermic penetration rate (%)

Boar 2901 3495 4291

8000 8000 8000

12.1 ± 2.6a 15.3 ± 3.1a,b 20.1 ± 3.0b

51.3 ± 7.8 49.4 ± 8.2 41.3 ± 6.9

Sperm/oocyte ratio 2 × 10 2 × 103 2 × 104 2 × 107

6000 6000 6000 6000

17.3 ± 3.1d,e 21.2 ± 2.9e 12.3 ± 2.8d,e 8.5 ± 3.2f

10.1 ± 5.1d 18.3 ± 6.3d 72.4 ± 12.3e 88.5 ± 14.7e

N: number of ejaculates evaluated. Means with different superscripts within a column for boar (a–c) and sperm/oocyte ratio (e–g) differ (P < 0.05).

104

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

Table 4 Boar effect on farrowing rate and number of pigs born alive (mean ± S.E.) Boar

N

Farrowing rate (%)

Number of pigs born alive

Before experiment 2901 3495 4291

531 417 521

83.9 ± 3.4a 88.7 ± 3.9a 70.2 ± 4.1b

9.6 ± 0.2a 11.4 ± 0.2b 9.4 ± 0.2a

During experiment 2901 3495 4291

219 221 224

84.6 ± 4.8a 86.2 ± 4.3a 72.4 ± 4.5b

9.8 ± 0.4a 11.2 ± 0.3b 9.5 ± 0.4a

N: number of sows bred to each boar. Means within the same column with different superscripts (a and b) differ (P < 0.05).

penetration was higher (P < 0.05) with sperm/oocyte ratios of 2×107 and 2×104 compared with ratios of 2 × 103 and 2 × 10. Differences among boars in farrowing rate and number of pigs born alive are shown in Table 4. Boars 3495 and 2901 exhibited similar (P > 0.05) farrowing rates which were higher (P < 0.05) than that achieved by boar 4291. In contrast, number of pigs born alive was greater (P < 0.05) for sows bred by boar 3495 compared with those bred to boars 2901 or 4291. Results of the polynomial regression analyses are summarized in Table 5. Significant effects for all three boars were observed only between the percentage of spermatozoa Table 5 Summary of polynomial regression analyses of relationships between semen quality estimates and different measures of in vitro and in vivo fertilization Semen quality estimates

In vitro and in vivo estimates of fertility Number born alive

Farrowing rate

Monospermic penetration rate

2901 Progressive motility Normal morphology Normal acrosomes Normal acrosin activity Spermatozoa in 90% fraction

Quadratic (P < 0.05) No effect (P > 0.23) Quadratic (P < 0.01) No effect (P > 0.45) No effect (P > 0.32)

No effect (P > 0.15) No effect (P > 0.32) No effect (P > 0.17) No effect (P > 0.26) No effect (P > 0.43)

No effect (P > 0.41) No effect (P > 0.32) Linear (P < 0.01) No effect (P > 0.34) Linear (P < 0.05)

3495 Progressive motility Normal morphology Normal acrosomes Acrosin activity Spermatozoa in 90% fraction

Linear (P < 0.01) No effect (P > 0.21) No effect (P > 0.87) No effect (P > 0.43) No effect (P > 0.47)

No effect (P > 0.11) No effect (P > 0.32) No effect (P > 0.31) No effect (P > 0.22) No effect (P > 0.42)

No effect (P > 0.98) No effect (P > 0.24) No effect (P > 0.89) No effect (P > 0.32) Linear (P < 0.01)

4291 Progressive motility Normal morphology Normal acrosomes Acrosin activity Spermatozoa in 90% fraction

Quadratic (P < 0.05) No effect (P > 0.19) Quadratic (P < 0.01) No effect (P > 0.26) No effect (P > 0.15)

No effect (P > 0.17) No effect (P > 0.31) No effect (P > 0.21) No effect (P > 0.34) No effect (P > 0.20)

No effect (P > 0.68) No effect (P > 0.23) Linear (P <0.01) No effect (P > 0.29) No effect (P > 0.15)

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

14

105

2901

12 10

Number of Pigs Born Alive

8 6 0 14

3495

12 10 8 6 0 14

4291

12 10 8 6 0 0

60

70

80

90

100

Progressive Motility (%) Fig. 1. Scatter diagrams for the relationship between the proportion of spermatozoa exhibiting progressive motility and number of pigs born alive for boars 2901, 3495 and 4291. Significant quadratic relationships were present for boars 2901 (y = −0.003x2 + 0.495x − 10.2; r2 = 0.76) and 4291 (y = −0.005x2 + 0.884x − 27.8; r2 = 0.81). A significant linear relationship (y = 0.084x + 4.7; r 2 = 0.70) existed for boar 3495.

exhibiting progressive motility and number of pigs born alive (Fig. 1). The relationships between the percentage of spermatozoa exhibiting progressive motility and number of pigs born alive for boars 2901 and 4921 were significant (P < 0.05) at the quadratic level, while for boar 3495 it was linear (P < 0.01). There were no significant relationships (P > 0.41) between the proportion of spermatozoa exhibiting progressive motility and monospermic fertilization rate (Fig. 2). For boars 2901 and 3495, there was a linear effect (P < 0.05) of the proportion of spermatozoa in the 90% percoll gradient fraction on monospermic penetration rate (Fig. 3). However, for boar 4291, this relationship was not significant (P > 0.15).

106

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

25

2901

20

Monospermic Penetration Rate (%)

15 0 25

3495

20

15 0 25

4291

20

15 0 0

60

70

80

90

100

Progressive Motility (%) Fig. 2. Scatter diagrams for the relationship between progressive motility and monospermic penetration rate for boars 2901, 3495 and 4291. No significant effects were present.

Distribution of sperm cells in percoll gradients did not influence (P > 0.15) number of pigs born alive (Fig. 4). The proportion of spermatozoa with normal acrosomes exhibited a quadratic relationship (P < 0.01) with number of pigs born alive (Fig. 5) and a linear relationship (P < 0.01) with monospermic penetration rate (Fig. 6) in boars 2901 and 4291, but had no effect (P > 0.85) on these parameters for boar 3495. 4. Discussion The most common way that semen quality data are used to determine whether boars should be used for breeding is to compare means that are generated by averaging values

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113 25

107

2901

20

Monospermic Penetration Rate (%)

15 0 25

3495

20

15 0 25

4291

20

15 0 0

35

40

45

50

55

60

Spermatozoa in 90% Percoll Fraction (%) Fig. 3. Scatter diagrams for the relationship between the proportion of spermatozoa recovered from the 90% fraction of a percoll gradient and monospermic penetration rate for boars 2901, 3495 and 4291. Significant linear relationships were present for boars 2901 (y = 0.474x + 1.5; r 2 = 0.71) and 3495 (y = 0.555x − 2.0; r 2 = 0.74), not for boar 4291.

across ejaculates (Woelders, 1991; Flowers, 1997). In the present study, use of this approach, prospectively, would have led to the conclusion that the three boars were equal in terms of their in vivo and in vitro fertility, because the means for the majority of the semen quality estimates were not different. The only exception to this was for boar 4291 who had the best and worst values for distribution of spermatozoa in percoll gradients and proportion of spermatozoa with acrosin activity, respectively. Percoll gradients separate populations of spermatozoa, in part, based on the velocity of their forward motility. The basic premise is that spermatozoa with increased velocity of movement will be able to enter the layers with high concentrations of percoll (Grant et al., 1994), which, in turn, is correlated positively with their ability to penetrate oocytes (Fraser, 1984). Acrosin is an enzyme that is released during the acrosome reaction and also plays an important role in oocyte penetration during

108

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

14

2901

12 10

Number of Pigs Born Alive

8 6 0 14

3495

12 10 8 6 0 14

4291

12 10 8 6 0 0

35

40

45

50

55

60

Spermatozoa in 90% Percoll Fraction (%) Fig. 4. Scatter diagrams for the relationship between the proportion of spermatozoa recovered from the 90% fraction of a percoll gradient and number of pigs born alive for boars 2901, 3495 and 4291. No significant effects were present.

fertilization (Polakoski and Parrish, 1977). Therefore, from a physiological perspective, it is not unreasonable to predict that the advantage in the velocity of motion for spermatozoa might be offset by the disadvantage in acrosin, creating a situation where his fertility was similar to the other two. In contrast, boar 3495 produced larger litters than boars 2901 and 4291, while more sows bred to boars 3495 and 2901 farrowed compared with those bred to boar 4291. Based on these data, a logical ranking for the boars in terms of in vivo fertility is 3495, 2901 and 4291. Monospermic penetration rates over a wide range of sperm/oocyte ratios were greater for boar 4291 than boar 2901, with comparable values for boar 3495 being intermediate.

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

14

109

2901

12 10

Number of Pigs Born Alive

8 6 0 14

3495

12 10 8 6 0 14

4291

12 10 8 6 0 0

60

70

80

90

100

Spermatozoa with Normal Acrosomes (%) Fig. 5. Scatter diagram for the relationship between the proportion of spermatozoa with normal acrosomes and number born alive for boars 2901, 3495 and 4291. Significant quadratic relationships were present for boars 2901 (y = −0.004x2 + 0.83x − 28.8; r 2 = 0.79) and 4291 (y = −0.008x2 + 1.313x − 43.6; r 2 = 0.84), but not for boar 3495.

Consequently, in the present study, means based on common semen quality estimates were of limited value for predicting the in vivo or in vitro fertility for the three boars examined. As mentioned earlier, this finding is consistent with some studies (Xu et al., 1996; Flowers, 1997), but contradicts others (Ivanova and Mollova, 1993; Flowers, 1997; Xu et al., 1998; Gadea and Matas, 2000). When means from semen quality data are used to evaluate boar fertility, one of the inherent assumptions is that relationships between these two groups of measurements are equivalent for most boars. In other words, increases in a given semen quality parameter,

110

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

25

2901

20

Monospermic Penetration Rate (%)

15 0 25

3495

20

15 0 25

4291

20

15 0 0

60

70

80

90

100

Spermatozoa with Normal Acrosomes (%) Fig. 6. Scatter diagrams for the relationship between the proportion of spermatozoa with normal acrosomes and monospermic penetration rate for boars 2901, 3495 and 4291. Significant linear relationships were present for boars 2901 (y = 0.202x + 2.37; r 2 = 0.70) and 4291 (y = 0.183x + 5.0; r 2 = 0.66), but not for boar 3495.

such as progressive motility or normal acrosomes, would be expected to produce reasonably consistent increases in farrowing rates, number of pigs born alive, and monospermic penetration rates, regardless of boars being studied. In contrast, results from the polynomial regression analyses indicate that both quantitative and qualitative differences among boars exist for these relationships. For boar 3495, there was a linear relationship and increases in progressive motility consistently resulted in an increase in litter size. For boars 2901 and 4291, the same relationships were best described by quadratic functions and once progressive motility reached about 75 and 80%, respectively, no additional improvements in number of pigs born alive were observed. Similarly, relationships between the proportion

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

111

of spermatozoa with normal acrosomes and number of pigs born alive were significant for boars 4291 and 2901, but not 3495. This phenomenon was also observed for in vitro fertility. Relationships between the proportion of spermatozoa with normal acrosomes and monospermic penetration rates were linear for boars 2901 and 4291, while a significant effect did not exist for boar 3495. However, a linear increase in monospermic penetration rate was observed as the proportion of spermatozoa in the 90% fraction of percoll gradients increased for boars 2901 and 3495, but not for boar 4291. A physiological explanation for these differences is not apparent at the present time. However, fertilization is a progressive process that involves a number of physiological and biochemical events including hyperactivation (increases in motility); capacitation; initiation of an acrosome reaction; penetration of the zona pelucida; binding to the ovum’s plasma membrane; and formation of a male pronuclei (Sirard et al., 1993). Failure of any of these steps hinders its success. Most semen quality measures, in general, and certainly the ones examined in the present study, typically estimate the occurrence of only one of these events in a population of spermatozoa. Therefore, it is possible that divergent patterns among boars reflect the fact that individuals differ in terms of which of the fertilization-related events is limiting. In the present study, for boar 3495, number of pigs born alive increased in a linear manner as the proportion of spermatozoa exhibiting progressive motility increased from 60 to 90%. Consequently, it is tempting to speculate that for boar 3495 the proportion of spermatozoa exhibiting progressive motility may have been a limiting factor for his in vivo fertility. In contrast, for boars 2901 and 4291, litter size reached plateaus between 75 and 80%. As a result, some other characteristic of spermatozoa, possibly one not evaluated in the present study, probably was preventing ejaculates with more than 80% progressive motility from producing increased litter sizes in these boars. Additional support for this speculation is provided by the observation that number of pigs born alive was consistently lower for 2901 and 4291 compared with 3495 at each level (i.e., 60, 70%, etc.) of progressive motility. In essence, their spermatozoa were inferior in some other trait that caused them to fertilize fewer ova despite similar levels of progressive motility. This same rationale could explain other observed differences for relationships between semen quality and fertility estimates among boars. The profile of plasma membrane proteins (Ash et al., 1994; Berger et al., 1996) and the chromatin structure (Evenson et al., 1994) are two properties of spermatozoa that are involved with binding of spermatozoa to oocytes and formation of male pronuclei, respectively, and have been correlated with fertility. Additional studies are required to determine whether differences in these characteristics among ejaculates can explain some of the variability in relationships between semen quality and fertility estimates in boars. It also appears that some measures of semen quality can be used to evaluate ejaculates for some boars with reasonable confidence in vitro, but not in vivo and vice versa. For each of the boars studied, there were significant, positive relationships between progressive motility and number of pigs born alive, but not between progressive motility and monospermic penetration rate. Conversely, for two of the three boars, the distribution of spermatozoa in percoll gradients influenced monospermic penetration rate in a positive manner, but had no effect on number born alive. These observations probably are related to differences in the environments to which spermatozoa are exposed during each process. During in vitro fertilization, spermatozoa are incubated for short periods of time (three hours in the present

112

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

study) and with compounds (caffeine) that enhance sperm cell velocity measurements before they are placed with ova. Consequently, motility characteristics observed at collection probably are very similar to those during fertilization in vitro. During in vivo fertilization, spermatozoa are transported to the oviduct and are stored in crypts formed by oviductal cells in the isthmus (Rodriguez-Martinez et al., 2001). While they are in the crypts, their metabolism and motility decreases significantly (Hunter et al., 1998). Once ovulation occurs, they are released and transported to the junction of the ampulla and isthmus where fertilization takes place. This creates a situation in which motility characteristics of spermatozoa observed at collection presumably are actually suppressed for extended periods of time, often 12–24 h, and then reactivated before fertilization occurs. Whether the initial estimates of motility at collection are correlated positively with those of fertilizing spermatozoa after their release from the oviductal crypts is not clear. However, if they are not, then it seems plausible that the distribution of spermatozoa within a percoll gradient could be a reasonable estimate of in vitro, but not in vivo fertility for some boars. The observation that there were no relationships between semen quality estimates and farrowing rate was surprising. Previous studies have reported improvements in farrowing rate as the proportion of spermatozoa exhibiting progressive motility or having normal acrosomes in an ejaculate increases (Flowers, 1997). However, recently, Braundmeier et al. (2002) were able to use a competitive sperm binding assay to evaluate fertility differences among boars in litter size, but not farrowing rate. Thus, there is, at least, one other report that indicates detecting differences in farrowing rate may be more difficult than finding those associated with litter size in boars. Alternatively, Xu et al. (1998) was not able to detect differences in numbers of pigs born alive among boars when 3 billion spermatozoa were inseminated. However, reducing the number of sperm cells to 2 billion produced sufficient results to rank boar fertility based on litter size. It is possible that a similar situation existed in the present study—an insemination dose of 3 billion spermatozoa was not sufficient to examine critically differences in farrowing rate.

5. Conclusions Means of common semen quality estimates involving motility, acrosome morphology, and head and tail morphology have limited value in terms of being used as a prospective indicator of in vivo and in vitro fertility for some boars. When this occurs, characterization of relationships between semen quality and fertility estimates appears to be quite useful for estimating differences among ejaculates for individual boars. However, both quantitative and qualitative differences among boars exist for these relationships and semen quality estimates used to evaluate in vivo fertility are not necessarily the best ones to use for optimization of in vitro results and vice versa.

References Almond G., Britt J., Flowers B., Glossop C., Levis D., Morrow M., See T., 1998. The Swine A.I. Book, second ed. Morgan Morrow Press, Raleigh, NC, USA.

J.M. Popwell, W.L. Flowers / Animal Reproduction Science 81 (2004) 97–113

113

Ash, K.L., Berger, T., Horner, C.M., Famula, T.R., 1994. Boar sperm plasma membrane protein profile: correlation with zona-free hamster ova assay. Theriogenology 42, 1217–1226. Braundmeier, A.G., Shanks, R.D., Miller, D.J., 2002. The relationship of porcine zona binding ability to fertility. Biol. Reprod. 66 (Suppl. 1), 274–275. Evenson, D.P., Thompson, L., Jost, L., 1994. Flow cytometric evaluation of boar semen by the sperm chromatin structure assay as related to cryopreservation and fertility. Theriogenology 41, 637–651. Flowers, W.L., 1997. Management of boars for efficient semen production. J. Reprod. Fert. Suppl. 52, 67–78. Flowers W.L., 1998. Management of reproduction. In: Wiseman, J., Varley, M.A., Chadwick, J.P. (Eds.), Progress in Pig Science. University of Nottingham Press, Nottingham, pp. 383–405. Fraser, L.R., 1984. Mechanisms controlling mammalian fertilization. Oxford Rev. Reprod. Biol. 6, 174–225. Funahashi, H., Day, B.N., 1993. Effects of the duration of exposure to hormone supplements on cytoplasmic maturation of pig oocytes in vitro. J. Reprod. Fertil. 98, 179–185. Gadea, J., Matas, C., 2000. Sperm factors related to in vitro penetration of porcine oocytes. Theriogenology 54, 1343–1357. Gill, J.L., Hafs, H.D., 1971. Analysis of repeated measurements of animals. J. Anim. Sci. 33, 331–336. Grant, S.A., Long, S.E., Parkinson, T.J., 1994. Fertilizability and structural properties of boar spermatozoa prepared by Percoll gradient centrifugation. J. Reprod. Fert. 100, 477–483. Hunter, R.H.F., Huang, W.T., Holtz, W., 1998. Regional influences of the Fallopian tubes on the rate of boar sperm capacitation in surgically inseminated gilts. J. Reprod. Fertil. 114, 17–23. Ivanova, M., Mollova, M., 1993. Zona-penetration in vitro test for evaluating boar sperm fertility. Theriogenology 40, 391–410. Koch, G.G., Landis, J.R., Freeman, J.L., Freeman, D.H., Lehron, R.G., 1977. A general methodology for the analysis of experiments with repeated measures of categorical data. Biometrics 33, 133–148. Long, C.R., Dobrinsky, J.R., Johnson, L.A., 1999. In vitro production of pig embryos: comparisons of culture media and boars. Theriogenology 51, 1375–1390. Lunstra, D.D., Echternkamp, S.E., 1982. Puberty in beef bulls: acrosome morphology and semen quality of bulls of different breeds. J. Anim. Sci. 55, 638–648. Moohan, J.M., Lindsay, K.S., 1995. Spermatozoa selected by a discontinuous percoll density gradient exhibit better motion characteristics, more hyperactivation, and longer survival than direct swim up. Fertil. Steril. 64, 160–165. National Research Council, 1998. Nutrient Requirements of Swine, tenth revised ed. National Academy of Sciences, Washington, DC. Penn, A., Gledhill, B.L., Darzynkiewicz, Z., 1972. Modification of the gelatin substrate procedure for demonstration of acrosomal proteolytic activity. J. Histochem. Cytochem. 207, 499–506. Polakoski, K.L., Parrish, R.F., 1977. Boar proacrosin. Purification and preliminary activation studies of proacrosin isolated from ejaculated boar sperm. J. Biol. Chem. 252, 1888–1894. Pursel, V.G., Johnson, L.A., Rampacek, G.B., 1972. Acrosome morphology of boar spermatozoa incubated before cold shock. J. Anim. Sci. 34, 278–283. Rodriguez-Martinez, H., Tienthai, P., Suzuki, K., Funahashi, H., Ekwall, H., Johannisson, A., 2001. Involvement of the oviduct in sperm capacitation and oocyte development in pigs. Reproduction Suppl. 58, 129–145. SAS, 1990. SAS/STAT User’s Guide. SAS Institute Inc., Cary, NC, USA. Sirard, M.A., Dubuc, A., Bolamba, D., Zheng, Y., Coenen, K., 1993. Follicle–oocyte–sperm interactions in vivo and in vitro in pigs. J. Reprod. Fert. Suppl. 48, 3–16. Snedecor G.W., Cochran W.G., 1998. Statistical Methods, eighth ed. Iowa State University Press, Ames, IA, USA. Woelders, H., 1991. Overview of in vitro methods for evaluation of semen quality. Reprod. Domest. Anim. Suppl. 1, 145–164. Xu, X., Seth, P.C., Harbison, D.S., Cheung, A.P., Foxcroft, G.R., 1996. Semen dilution for assessment of boar ejaculate quality in pig IVM and IVF systems. Theriogenology 46, 1325–1337. Xu, X., Pommier, S., Arbov, T., Hutchings, B., Sotto, W., Foxcroft, G.R., 1998. In vitro maturation and fertilization techniques for assessment of semen quality and boar fertility. J. Anim. Sci. 76, 3079–3089. Yoshida, M., 1987. In-vitro fertilization of pig oocytes matured in vivo. Jpn. J. Vet. Sci. 49, 711–718.