Poor Prediction Value of Sperm Head Morphometry for Fertility and Litter Size in Rabbit

This study was conducted to investigate the predictive capacity of fertility and litter size of sperm head morphometric measurements when the ejaculates fulfilled the minimum requirements commonly used in artificial insemination (AI). Semen samples from 11 rabbits (77 ejaculates) were evaluated for sperm motility, abnormal spermatozoa and sperm head morphometry using computer automated sperm analysis system. Morphometric dimensions for length, width, area and perimeter were analysed. Only ejaculates with more than 70% of motility rate and <15% of abnormal sperm were used for AI. A total of 1031 individual AI were performed in commercial rabbitries. Our results showed significant differences among animals for all sperm head measurements. The mean values for fertility and litter size obtained were 68.4 ± 0.01% and 9.3 ± 0.1% respectively. To assess the predictive value of morphometric dimensions in fertility, a logistic regression analysis was applied. Moreover, multiple linear regression analyses were used to examine the relationship between litter size and sperm head morphometric parameters. Logistic regression analysis rendered a significant model between fertility and area and perimeter, explaining the 0.65% variation. Multiple linear regression analysis rendered a significant model between litter size and width, area and perimeter that explained the 1.3% variation. By conclusion, the sperm head morphometric parameters assay showed low potential to predict fertility and litter size when the ejaculates fulfilled the minimum requirements commonly used in AI (motility and abnormal spermatozoa) in rabbit. As the advent of artificial insemination (AI), researchers have sought laboratory assays that would accurately predict the fertilising potential of a semen sample using a rapid and inexpensive procedure (Graham and Mocé 2005; Mocé and Graham 2008). However, the evaluation of the ejaculate is currently of utmost importance to determine its degree of normality before being processed for AI (Rodríguez-Martínez 2007). Poor semen quality is a good indicator of fertility failures, but good semen quality is no guarantee of suitable fertility (Colenbrander et al. 2003). Our ability to use in vitro assessments of semen quality to predict the fertility potential of a semen sample seldom explains more than 50–60% of the variation among males (Dejarnette 2005). Semen evaluation usually includes recordings of appearance, volume and sperm concentration, motility and morphology. These evaluation practices are enough to rule out ejaculates and to define the number of sperm per AI dose, but insufficient to ensure high fertility. Conventional subjective assays used are unreliable in predictive fertility (Parrish et al. 1998). Graham et al. (1980) reported that correlations between fertility and assays for motility, morphology and viability ranged from 0.06 to 0.86. Despite this, some studies have shown that semen samples possessing greater percentages of morphologically abnormal sperm exhibit reduced fertility (Saacke and White 1972; Barth and Oko 1989; Dejarnette 2005; Lavara et al. 2005). In rabbits, Lavara et al. (2005) reported that, from a predictive point of view regarding fertility, spermatozoa motility and morphological parameters can account for up to 45% of the variation. On the other hand, Quintero-Moreno et al. (2007) reported that the combination of sperm motility, sperm concentration and viability could explain the 16% variability in litter size. Advances in computer technologies have facilitated the automation of many established laboratory assays and development of new laboratory assays to evaluate the quality of semen samples (Mocé and Graham 2008). Sperm motility and morphology are increasingly evaluated by computer assisted semen analysis systems (CASA). Assisted sperm morphometry analysis (ASMA) has been applied in a number of livestock species (Davis et al. 1993; Ball and Mohammed 1995; Gravance and Davis 1995; Gravance et al. 1995, 1996a,b, 1997a,b, 1998a,b, 1999, 2008; Sancho et al. 1998; Buendía et al. 2002; Rijsselaere et al. 2004; §Marco-Jiménez et al. 2005, 2006; Martínez et al. 2006; Hidalgo et al. 2007). The use of sperm head morphometric parameters has been considered a good indicator of semen quality in bull (Phillips et al. 2004) and it is recommended as part of the spermiogram for domestic animals (Rodríguez-Martínez 2007). To date, there are not many studies describing the fertility predictive potential of ASMA test (Gravance et al. 1996a, 2008; Schmidt 1997; Hirai et al. 2001; Marco-Jiménez et al. 2005; Hidalgo et al. 2007). However, to our best knowledge, only two reports in the literature study the fertility with ASMA systems when sperm is suitable for use in AI (Hirai et al. 2001 and Gravance et al. 2008, for fresh and frozen sperm respectively). The aim of this study was to determine the predictive value of sperm head morphometric parameters (length, width, area and perimeter) related to fertility and litter size when the ejaculates fulfilled the minimum requirements commonly used in AI in rabbit. The males used (n = 11) were from a high growth line of rabbits, selected for growth rate from weaning to slaughter (28–63 days). Selection methodologies were described by Estany et al. (1992). Males were housed at the experimental farm of the Animal Science Department, Polytechnic University of Valencia (Valencia, Spain). Crossbred female rabbits (n = 1031) from two selected lines on the basis of litter size at weaning were used in the experiment (Estany et al. 1989). All rabbits to be inseminated were multiparous lactating females. Females were housed on two commercial farms located in the provinces of Castellón (Spain). Animals were housed under photoperiod of 16L : 8D, in individual cages, fed with a commercial diet and provided water ad libitum. Each week, two ejaculates per male were collected on a single day using an artificial vagina, with a minimum of 30 min between ejaculate collections. All collections occurred within a 10-week period in late spring. Semen was collected early in the morning. Only ejaculates exhibiting a white colour were used, and if gel was present, it was removed. Ejaculates for each male were mixed and diluted (dilution 1 : 5) with a Tris–citric acid–glucose extender (Viudes-de-Castro et al. 1999), split into separate samples. Percentage of motile spermatozoa was assessed using light microscopy at a 100× magnification and image recorded (four sequences by sample were analysed). One sample per male was fixed (dilution 1 : 10) in a solution of glutaraldehyde 2% (Electron Microscopy Science, Washington) in Dulbecco’s phosphate buffered saline (Pursel and Johnson 1974) and the sperm concentration, sperm abnormalities and sperm head morphometry analysis (ASMA) were determined. Sperm concentration was determined using a Thoma-Zeiss cell counting chamber (Marienfield, Germany). Sperm abnormalities were determined using a phase contrast microscope at a magnification of 400×, evaluating a minimum of 100 sperm cells. Only ejaculates with more than 70% of motility rate and <15% of abnormal spermatozoa were used for AI. For ASMA test, one slide per male (mixed ejaculates collected on a single day) using a 10 μl drop of the diluted-fixed sample was placed onto a slide and covered with a cover slip. The equipment was a Vimas system (IMAGESP, Barcelona, Spain) and an Eclipse E400 Nikon microscope with 40× objective (Plan Fluor, Ph2 DLL) with phase contrast optics. A video camera (Sony CCD-IRIS SSC-M370CE; Sony Corporation, Tokyo, Japan) was mounted on the microscope to capture images and transmit them to the video digitiser board (Meteor II; Matrox Electronic Systems Ltd., QC, Canada) located in a Pentium processor. The array size of the video frame grabber was 512 bit × 512 bit × 8 bit providing digitized images of 262 144 pixels and 256 grey levels. Resolution of images was 0.08 μm per pixel in the horizontal and vertical axes. The morphometric dimensions of length (L), width (W), area (A) and perimeter (P) of a minimum of 100 normal sperm heads were randomly chosen manually and analysed from each slide. Only receptive multiparous lactating does (red colour of vulvar lips) were inseminated, using a standard curved plastic pipette (Imporvet, S.A., Barcelona, Spain). Each female was inseminated with 0.5 ml (about 6 million sperms per doe) and it was performed within 2–4 h of semen collection (Viudes de Castro and Vicente 1997). At the time of insemination, each female was injected intramuscularly with 1 μg of buserelin acetate (Hoechst Marion Roussel, S.A., Madrid, Spain) to induce ovulation. Fertility rate (calculated as the ratio: number of females that gave birth/number of inseminations) and litter size were recorded. Morphometric differences in individual spermatozoa per male were studied by performing an anova analysis. Fertility and litter size between males were also compared performing an anova analysis. The multiple comparison method used in anova was Tukey’s significant difference (HSD) procedure. Values were considered to be statistically significant when p < 0.05. The intra-analysis coefficient of variation from each analysis was calculated by the ASMA instrument and recorded. Estimates of the intra-analysis variation of analyses are expressed as coefficient of variations to illustrate the relative variability among the various measurements. To assess the predictive value of morphometric dimensions of L, W, A and P on fertility, a logistic regression analysis was used. The variable fertility was dichotomous and the factors were the means of sperm head parameters (L, W, A and P) calculated for each ejaculate. Moreover, multiple linear regression analysis was used to examine the relationship between litter size and sperm head parameters (L, W, A and P). Statistical analyses were performed using a commercially available statistics package (Statgraphics Plus, Version 5.1, STSC Inc., Rockville, MD, USA). Morphometric parameters of 7753 sperms were analysed. The measurements of sperm heads from all bucks are summarized in Table 1. Significant differences were observer for L, W, A and P among bucks. Length ranged between 8.01 μm vs 8.70 μm, width ranged between 4.36 μm vs 4.68 μm, area ranged between 29.02 μm2 vs 32.90 μm2 and perimeter ranged between 21.46 μm vs 22.90 μm. Within-animal coefficient of variation ranged from 3.98% (P) to 8.85% (A). Comparison of fertility between males indicated a significant difference, ranged between 52.1 ± 0.5% to 81.4 ± 0.4% for male E and J respectively (Fig. 1). No differences for litter size were observed between males (Fig. 2). The values ranged between 8.1 ± 0.5 to 10.2 ± 0.4 for male K and H respectively. Fertility mean obtained for each male. Data represent the means ± SE. abcdBars with unlike superscripts differ (p < 0.05). Numbers inside bars indicate the artificial inseminations per male Litter size obtained for each male. Data represent the means ± SE. Numbers inside bars indicate the numbers of artificial inseminations per male The mean ± standard error value obtained for fertility and litter size in this study were 68.4 ± 0.01% and 9.3 ± 0.1% respectively. Logistic regression analyses reported one model with low percentage of concordance between sperm head morphometric parameters and fertility. The model explained a deviance of 0.65% (p = 0.0149), directly related with A (p = 0.0498) and P (p = 0.011). The multiple linear regression analysis was derived by backward elimination of variables for sperm head morphometric parameters applied to a litter size generated model, which could explain the 1.3% variation in litter size (p = 0.0391). The model included the variables W (p = 0.058), A (p = 0.043) and P (p = 0.068). As in the previous findings of Gravance and Davis (1995), a highly significant animal effect was observed for all measurements evaluated. However, the mean values for all parameters observed in this study were partly in agreement with those reported by Harding et al. (1979) for electron microscopy (L, 8.06–8.42 μm; W, 4.59–4.86 μm), Cummins and Woodall (1985) for wet mounts (L, 8.51 μm; W, 4.98 μm), and Beatty and Napier (1960) for stained sperm (L, 8.08–8.14 μm; W, 4.69–4.70 μm) but higher than those by Gravance and Davis (1995) for stained sperm (L, 7.27–7.38 μm; W, 3.88–3.91 μm; A, 21.70–22.10 μm2; P, 19.00–19.20 μm). These differences may be a result of the differences in specimen preparation methods. Studies in others species have reported similar differences for wet vs dry preparations (Katz et al. 1986). In this study, samples were evaluated using a microscope with phase contrast optics for wet mount (Marco-Jiménez et al. 2006). To correlate sperm head parameters with fertility, it is first necessary to reduce the high variability of the AI processes by means of the standardization of several factors, such as semen collection, dilution, number of sperm per insemination dose, reproductive status and genetic origin of females inseminated, or the methods used for synchronization and insemination of the females (Lavara et al. 2005). Our objective was to determine how much new information on fertility was obtained by sperm head morphometric parameters when previously evaluated samples fulfilled the minimum requirements for AI. Assessment of sperm motility and normal sperm morphology is an essential component of the spermiogram (Dejarnette 2005). With this semen quality, the fertility and litter size values obtained in this study coincide with the previous reports using multiparous lactating does (Lavara et al. 2005; Viudes-de-Castro et al. 2005). Previously, Lavara et al. (2005), using a regression model including motility (determined with CASA system) and morphological parameters, obtained a model capable of explaining 45% of the variation in fertility. However, sperm head morphometry, when motility and abnormalities were previously fixed, accounted for 0.65% of the variation in fertility. The comparison of these results are difficult, because, to our best knowledge, morphometric sperm head parameters have received more attention in recent years, but few current studies have examined the relations between fresh sperm head morphometry and fertility in livestock species. These studies could be grouped according to their initial approach: (i) comparing fertile and infertile males (Gravance et al. 1996b; Sailer et al. 1996; Casey et al. 1997) and (ii) grouped above and below the median value of fertility (Hirai et al. 2001) or according to the sperm head size (Marco-Jiménez et al. 2005). When fertile and infertile males were compared, the differences in sperm quality were so high that no laboratory assays could clearly explain the differences in fertility. In the second group of studies, Hirai et al. (2001) established a cut-off value for fertility (more and <86%) using normal boars housed in an AI centre (70% motile spermatozoa and maximum of 20% of cytoplasmatic droplets) and significant differences for ASMA parameters were obtained between groups. In our study, (statistical analysis not shown) using similar criteria and establishing a cut-off value for fertility: more than 75% (male J and H) and <55% (male E and K) (Fig. 2), the rabbit with lower fertility had a significantly more elongated sperm head shape (8.43 ± 0.01 μm and 4.49 ± 0.01 μm vs 8.37 ± 0.01 μm and 4.66 ± 0.01 μm, for length and width and lower and less fertility respectively), results similar to those of Hirai et al. (2001). In accordance with these results, Marco-Jiménez et al. (2005) in rabbit, established cut-off values reporting that males with lower head size had lower fertility (45.0% vs 77.9%). These findings could support the hypothesis that the use of cut-off values when results are previously known could remove validity from the laboratory assays to predict fertility. Similarly, Hirai et al. (2001) found that boars with a litter size of more than 10 piglets per litter had significantly more elongated spermatozoa than boars with <10 piglets per litter. Our study reported that sperm head morphometric parameters explained a variation in the litter size of 1.3%. Previously, Quintero-Moreno et al. (2007), using motility parameters (determined by CASA system), sperm concentration and viability, explained the 16% variability in litter size. The minor prediction potential shown between laboratory assays and litter size could be explained because litter size has been found to be dependent on a large number of factors affecting the female, including a physical state (Dziuk 1994; Quintero-Moreno et al. 2007). Considerable effort is being invested in identifying markers for functional sperm capacity that can more accurately predict fertility (Colenbrander et al. 2003). The aim of most current researchers is instead to identify a combination of tests that together analyse the most important sperm function parameters, as combining sperm function tests allows more accurate prediction of fertility (Wilhelm et al. 1996). Assessment of sperm morphology, as a result of its relationship to fertility of the ejaculate/male, is considered a major component of the spermiogram (Phillips et al. 2004). In this line, the ASMA method was developed to reduce the technical variation in the sperm morphology assay (Gravance and Davis 1995). However, the ASMA assay does not increase the predictive potential of fertility and litter size in rabbit if the sperm morphology has been evaluated previously using a classical procedure, probably as a consequence of significant correlation among spermatozoa attributes (Dejarnette 2005). Our results are similar to those of Gravance et al. (2008) for frozen bull sperm, who reported that sperm morphometry may be of minimal value in predicting post-thaw fertility. In conclusion, the sperm head morphometric parameters showed low potential to predict fertility and litter size when the ejaculates fulfilled the minimum requirements commonly used in AI (motility and abnormal spermatozoa) in rabbit. Nevertheless, in the future it could be interesting to study the predictive value of sperm head morphometric parameters without selection of the ejaculates. This work was supported by Grants from the Ministry of Education and Science of Spain AGL 2004-01722 (GAN). English text version was revised by N. Macowan English language service. The design of the study has been performed by Vincent and Marco-Jiménez. Viudes-de-Castro, MP and Lavara R have collaborated actively in the development of the study (seminal evaluation and artificial insemination). Finally Balasch S has advised and involved in the data analysis

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.