C. Chevrier, IRBI, Faculté des Sciences, Parc Grandmont, 37200 Tours, France. Tel. +33 0247 367014; fax: +33 0247 367094; e-mail: firstname.lastname@example.org
Abstract The quality of a sperm population can be characterized physiologically and its fecundity predicted by its viable : non-viable sperm ratio. To improve the knowledge of reproductive strategies in two ectoparasitoid hymenopteran species, Eupelmus orientalis Crawford (Hymenoptera: Eupelmidae) and Dinarmus basalis Rondani (Hymenoptera: Pteromalidae), the assessment of sperm viability using the dual fluorescence staining procedure SYBR-14 : propidium iodide was developed. The aim of the study was to provide a comparative test in vitro applicable to both sexes to study the evolution of sperm quality at various stages of the reproductive processes. The reliability of propidium iodide to detect non-viable sperm (stained in red) was confirmed in both species on the basis of two stress tests (ethanol and Triton X-100) but our study also revealed that propidium iodide concentrations must be adequately adjusted for each single species. This experiment also demonstrated the physiological heterogeneity of sperm populations in E. orientalis and D. basalis males and females. In both species, 40% of the sperm in the seminal vesicles was found to be non-viable. By contrast with E. orientalis, the populations of non-viable sperm estimated from the seminal vesicles of D. basalis were found to be strongly different from those observed in the spermatheca. From the present results, the population of viable sperm detected in the spermatheca of females from both species proved a reliable predictor of fertilization achieved in ovipositing females.
In most species of insects, reproduction involves the long-term storage of sperm in a differentiated storage organ of the female, the spermatheca (review in Thornhill & Alcock, 1983), from which sperm are released progressively to fertilize ovae. In some species, sperm storage in the spermatheca may last up to weeks, months or even years (see Neubaum & Wolfner, 1999). Sperm quantity and quality are constraints acting on female fecundity, and quantitative studies have been carried out to understand the effect of sperm management on reproductive success (Smith, 1984). However, the quality of sperm populations has rarely been assessed at the various stages of the reproductive process, although such information is necessary for a better understanding of reproductive strategies (Tsubaki & Yamagishi, 1991; Yamagishi et al., 1992).
Among insects, Hymenoptera are valuable models for characterizing the quality of sperm. Through arrhenotoquous parthenogenesis, only female offspring is issued from fertilization (see Godfray, 1994, for a review). As a consequence, the number of functional sperm available in the spermatheca is reflected in the number of daughters produced by a given reproductive female. However, among species studied previously in that respect, the number of sperm estimated in the female spermatheca was vastly in excess of the number of daughters produced (Waage, 1984; Buys, 1990; Collins, 2000), indicating that the non-utilization of all sperm present in the spermatheca is, a priori, independent of their initial quality.
Both hymenopteran species chosen for the present study, Eupelmus orientalis Crawford (Hymenoptera: Eupelmidae) and Dinarmus basalis Rondani (Hymenoptera: Pteromalidae), are solitary ectoparasitoids of the West African bruchid larvae and pupae. In these species, the population of sperm involved at each stage of the reproductive sequence after a single mating is reduced (Bressac & Chevrier, 1998; Chevrier & Bressac, 2002). Moreover, the number of sperm stored in the female tract is quite similar to the number of fertilized eggs, a strong indication that a large majority, if not all, sperm stored in the female are viable and functional. Such biological models can therefore be considered as valuable tools to assess sperm quality in vivo indirectly.
The assessment of sperm quality by procedures in vitro was first developed in vertebrates submitted to artificial insemination procedures. In such species, determining sperm quality has led to an impressive series of tests, including the evaluation of morphology, functional status (motility, etc.) and cell integrity (acrosome, membrane, midpiece, etc.). Preliminary (unpublished) observations in E. orientalis and D. basalis indicated that morphological abnormalities of sperm are not visible, at least at the head level. Moreover, in these insects, sperm remain immobile in vitro, which prevents the assessment of sperm viability by analysis of sperm displacements. Other tests, including those currently in use to assess sperm membrane integrity by differential staining directly detectable by microscopy, may therefore be of special interest (Williams et al., 1998). Thus, in honey bees, Apis mellifera, several supra-vital dual staining techniques have already been developed to study sperm viability on the basis of the dual combination of propidium iodide either with Hoechst 33342 (Locke et al., 1990; Peng et al., 1990) or with SYBR-14 (Collins & Donoghue, 1999; Collins, 2000). In the present study, the latter test was chosen to assess sperm viability in E. orientalis and D. basalis, as it has proved to be a valuable tool in assessing sperm viability and, in several cases, in predicting fertilizing potential (Garner et al., 1986; Garner et al., 1994; Donoghue et al., 1995; Garner & Johnson, 1995; Chalah & Brillard, 1998; Chalah et al., 1999). The SYBR-14 : propidium iodide test contains a membrane-permeant nucleic acid stain, SYBR-14, which reveals live sperm cells in bright green and propidium iodide, a conventional dead-cell stain which stains nuclei bright red in cells with damaged plasma membranes. The promising results obtained with this test in insect sperm (Collins & Donoghue, 1999; Collins, 2000) led to the present attempt to validate this test to assess sperm viability in E. orientalis and D. basalis.
Materials and methods
Eupelmus orientalis and D. basalis were obtained from colonies established in the laboratory, but that originated from West Africa. All experiments were carried out in a climate-controlled chamber in the presence of Callosobruchus maculatus hosts (from West Africa), as previously described by Doury & Rojas-Rousse (1994) and Gauthier et al. (1996). In all experiments, 24-h-old virgin males and 2-h-old virgin females were used in both species (Chevrier & Bressac, 1997; Bressac & Chevrier, 1998). Mated females were used 24 h after a single controlled mating with a 24-h-old virgin male.
Females and males were dissected in a drop of Beadle saline (128.3 mm NaCl, 4.7 mm KCl, 23 mm CaCl2), as described previously by Bressac & Chevrier (1998). The paired seminal vesicles from each male were first transferred into 20 µL of saline solution, then opened with microforceps and finally rinsed with a controlled volume of either 100 µL (E. orientalis) or 30 µL (D. basalis) of saline. The resulting suspensions were shaken gently to disperse sperm and then homogenized. Two 3 µL drops of the final suspension were deposited on clean slides. In females, the spermathecae were opened in one drop (3 µL) of saline on a clean slide and their content dispersed with circular movements of the forceps.
Estimate of sperm viability by SYBR-14 : propidium iodide test
Staining procedure. Sperm samples collected from males and females were treated with SYBR-14 : propidium iodide (Fertilight kit; Molecular Probes, Eugene, OR, U.S.A.) as follows: 0.5 µL of SYBR-14 solution were added to sperm samples on the microscope slide. The preparation was gently stirred and incubated at room temperature (20 ± 2 °C) for 4 min and then 0.25 µL of propidium iodide solution was added. Observations under fluorescence microscopy were performed 1 min after incubation (20 ± 2 °C, blue excitation filter at λ = 490 nm; 200 × magnification). The percentage of red, red + green, or green subpopulations of sperm were determined in each drop. The total population of sperm present under each coverslip was also estimated by addition of the three subpopulations of stained sperm. When the two drops derived from one male sample showed over 20% difference in sperm staining distribution, the sample was discarded.
Quantitative assessment of sperm populations. The total population of sperm present in each suspension was estimated by DAPI (4 -6-diamidino-2-phenylindole) staining, as previously described (Bressac & Chevrier, 1998), to compare this procedure with estimates obtained after dual staining (see above).
Assessment of the concentrations of propidium iodide and SYBR-14. In order to determine optimal propidium iodide concentration revealing non-viable sperm, sperm samples from E. orientalis and D. basalis males were prepared as described above with, for each preparation, a series of dilutions adjusted at 10.6, 21.3, 42.6, 106, 420 or 666 µg/mL of propidium iodide (final volume: 3 µL each). SYBR-14 concentration was held constant in all preparations (8000 nm).
To determine optimal SYBR-14 stain concentration, sperm samples from E. orientalis and D. basalis males were prepared as above and a 3-µL drop was taken. Staining was performed by increasing SYBR-14 concentrations from the maximal concentration of 100 nm suggested in the Product Information of Molecular Probes (commercial stock at 1 mm in DMSO). Preliminary observations with SYBR-14 : propidium iodide used as recommended in the protocol (100 nm) indicated that a large proportion of sperm present remained entirely unstained. In order to determine the appropriate concentrations of SYBR-14 for each species, trials were conducted in the following range: 100, 500, 1000, 2000, 4000, 8000 and 16 000 nm. In this experiment, propidium iodide concentration was held constant within each species (106 µg/mL in E. orientalis and 21.3 µg/mL in D. basalis). The final concentrations of fluorochromes in each species were chosen on the basis of the lowest tested concentrations providing easily readable and reliable fluorescence.
Validation of fluorescence procedures. To verify the non-viability of propidium iodide-fluorescent stained sperm (red), a series of male sperm suspensions (in Beadle saline) was separated into a duplicated series and then treated with either ethanol (1/1 v/v) or Triton X-100 (1/99 v/v). Cells treated by either of the compounds lose their plasma membrane integrity (Crissman & Tobey, 1974; Taylor, 1980) and become readily permeable to propidium iodide. All preparations were incubated for 2 min at 20 ± 2 °C and then counterstained with SYBR-14 : propidium iodide as described in the staining procedure section.
Analysis of variance (anova) was used to investigate the response of the proportions of sperm to different concentrations of propidium iodide when the data conformed to the assumptions of anova. Tukey–Kramer multiple comparisons tests for pairwise comparisons were performed on the mean of those factors with significant F-values. Otherwise, the Kruskal–Wallis test and Dunn's multiple comparisons tests were used when the differences among the standard deviations were significant using Bartlett's Test. The interrelationships between the variables were evaluated using Pearson correlation coefficients. Slopes of the regressions were compared by t-tests to assess possible significant differences between the treatments. The likelihood ratio test (G-test, derived from a χ2 test) was performed to compare the sperm proportions of the stained subpopulations from an equilibrated three-thirds distribution and to compare sperm proportions between seminal vesicles and spermatheca.
Assessment of the concentration of propidium iodide
In E. orientalis, no significant differences were observed between samples treated with concentrations of propidium iodide from 10.6 to 106 µg/mL (Table 1), but a significant increase in the percentage of red sperm (propidium iodide-stained) was observed at the highest concentrations tested (420 and 666 µg/mL). In D. basalis, a significant increase in the percentage of red sperm (propidium iodide-stained) was observed for concentrations of propidium iodide ranging from 10.6 to 106 µg/mL (Table 1). No significantly greater percentages of propidium iodide-stained sperm were observed for concentrations higher than 106 µg/mL (Table 1). Therefore, working concentrations of propidium iodide for subsequent experiments were chosen as 106 µg/mL in E. orientalis and 21.5 µg/mL in D. basalis.
Table 1. The effect of varying propidium iodide concentrations on the non-viability percentage (mean ± SE) in sperm samples of E. orientalis and D. basalis males. Different letters indicate significant differences between data (Kruskal–Wallis test and Dunn's multiple comparisons test in E. orientalis, KW = 51, P < 0.05; anova and Tukey–Kramer multiple comparison test in D. basalis, F5,114 = 249, P < 0.05).
% of non-viable sperm
% of non-viable sperm
Assessment of the concentration of SYBR-14
No significant differences were observed in the proportion of propidium iodide-stained sperm in concentrations of SYBR-14 ranging from 500 to 8000 nm in E. orientalis and 4000–8000 nm in D. basalis (Table 2). However, in both species, at the lowest or highest concentrations of SYBR-14, the proportion of propidium iodide-positive sperm was higher than at any of the other concentrations tested (Table 2).
Table 2. The effect of varying SYBR-14 concentrations on the nonviability percentage (mean ± SE) in sperm samples of E. orientalis and D. basalis males. Different letters indicate significant differences between data (Kruskal–Wallis test and Dunn's multiple comparisons test in E. orientalis, KW = 40, P < 0.05; anova and Tukey–Kramer multiple comparison test in D. basalis, F5,54 = 114, P < 0.05).
% of non-viable sperm
% of non-viable sperm
Finally, the species-specific recommended final stain concentrations in the sperm drops are 106 µg/mL propidium iodide – 8000 nm SYBR-14 in E. orientalis, and 21.5 µg/mL propidium iodide – 8000 nm SYBR-14 in D. basalis.
Validation of propidium iodide to detect sperm with damaged membrane
Samples of sperm from E. orientalis or D. basalis treated with ethanol or Triton X-100 showed 100% red-stained sperm.
In both species, there was a highly significant correlation (F1,28 = 263 in E. orientalis, F1,28 = 149 in D. basalis, P < 0.001) between the numerical change of sperm counted after SYBR-14 : propidium iodide staining and the number of sperm counted by DAPI staining in a same sperm suspension. The slopes of the regression linear lines differed from 1 (t-test, t = 9, d.f. = 56 in E. orientalis, t = 16, d.f. = 56 in D. basalis, P < 0.001) and the fitted equations were, respectively, y = 0.74x + 4.57 with r = 0.95 in E. orientalis and y = 0.58x + 8.97 with r = 0.92 in D. basalis. Thus, SYBR-14 : propidium iodide staining results in an underestimate compared with DAPI staining.
Evaluation of sperm viability by SYBR-14 : propidium iodide
In both seminal vesicles and spermatheca of E. orientalis and D. basalis, three subpopulations of sperm were identified: red, red + green, and green (Fig. 1).
The mean number of sperm estimated in males after SYBR-14 : propidium iodide staining was 7593.3 ± 636.0 (mean ± SE, n = 10) in E. orientalis and 2680.0 ± 259.5 (n = 10) in D. basalis. The percentages of the three subpopulations are shown in Fig. 1. In both species, the proportions of different stained-sperm from seminal vesicles were statistically different from an equilibrated three-thirds distribution (G-test, G= 7.4 in E. orientalis and G = 9.1 in D. basalis, d.f. = 2, P < 0.05).
The mean number of sperm counted in females after SYBR-14 : propidium iodide staining was 127.6 ± 11.7 (mean ± SE, n= 30) in E. orientalis and 78.6 ± 5.1 (n = 30) in D. basalis. The percentages of the three subpopulations are shown in Fig. 1. In E. orientalis, proportions of different stained-sperm in the spermatheca were not statistically different from an equilibrated three-thirds distribution (G-test, G = 1.9, d.f. = 2, P > 0.05). In D. basalis females, proportions of different stained sperm from the spermatheca were statistically different from an equilibrated three-quarters distribution (G-test, G = 39.5, d.f. = 2, P < 0.05). Moreover, the proportions of propidium iodide-stained sperm in E. orientalis were similar (whether significant G-test, G = 10.4, d.f. = 2, P < 0.05) in males and females but strongly different in D. basalis (G = 91.6 d.f. = 2, P < 0.05; Fig. 1).
The present results indicate that SYBR-14 : propidium iodide can be considered as an effective method to differentiate living from dead sperm in seminal vesicles and spermatheca from E. orientalis and D. basalis. Similar observations had already been reported in honey bee sperm by Collins & Donoghue (1999). The present work also reveals that the SYBR-14 : propidium iodide method must be validated for each individual species.
In E. orientalis and D. basalis all sperm samples collected in the seminal vesicles and treated with either ethanol (cell drying; Crissman & Tobey, 1974) or the non-ionic detergent Triton X-100 (cell membrane permeant; Taylor, 1980) and counterstained with SYBR-14 : propidium iodide stained readily in contrasted red, confirming that propidium iodide is an efficient indicator of insect sperm with damaged membranes (Collins & Donoghue, 1999).
The final concentrations of SYBR-14 and propidium iodide recommended here are the lowest tested concentrations providing optimal (visible) fluorescence. The species-specific recommended final concentrations of fluorochromes are 106 µg/mL propidium iodide and 8000 nm SYBR-14 for E. orientalis and 21.5 µg/mL propidium iodide and 8000 nm SYBR-14 for D. basalis. The concentrations of propidium iodide recommended here are comparable to those used for honey bee (A. mellifera) associated with Hoechst 33342 (Locke et al., 1990), SYBR-14 or Calcein-AM (Collins & Donoghue, 1999). From the present observations, it appears that the concentration of 100 nm of SYBR14 recommended in the User's Guide to detect viable sperm in mammalian species, is inadequate for insect samples. A concentration of 8000 nm is an effective compromise, as at the highest molarity (16 000 nm) a majority of sperm were shown to be dead.
Almost all of the sperm samples from seminal vesicle stained and counted in this study had > 30% non-viable sperm. The experimental procedure used here cannot explain such a high proportion of non-viable sperm, as special care was taken to open and flush out the seminal vesicles gently. Such high proportions of non-viable sperm may, at least in part, be caused by the quite high initial dilution rate of sperm during flushing procedures, which in turn would also explain the differences observed between the present data and that of Collins & Donoghue (1999). Nevertheless, the aim of this study was not to determine the intrinsic viability of individual sperm, but to provide a comparative test in vitro between sexes based on the relative changes in sperm viability during the various processes of reproduction.
From a general standpoint, vital or supra-vital fluorescent probes used without fixation may not reveal fully the population of sperm present in a preparation (Williams et al., 1998). Moreover, dual fluorescent staining may not always differentiate between live and dead cells (Locke et al., 1990). However, in the present study, there is a highly positive correlation between sperm counts performed after SYBR-14 : propidium iodide dual procedure (no fixation) and those performed after DAPI procedure (with fixation). This indicates that the dual fluorescence test can be of practical interest for further qualitative and quantitative analyses in corresponding species. Moreover, this test also proved to be a valuable tool to estimate sperm concentrations indirectly by regression adjustment in E. orientalis and D. basalis.
A significant subpopulation of dual-stained sperm (red + green) is observed from samples collected both in the seminal vesicles and spermatheca from either species. The proportion of this transitional population is generally greater than that found by Collins & Donoghue (1999) in honeybee sperm. Hypothetically, such dual-stained sperm may have lost some functional capacities, or their membrane may be physically or chemically organized to facilitate this event in the species used here. Nevertheless, dual-stained sperm may not represent a population unfit for fertilization. In D. basalis after a single mating, Chevrier & Bressac (2002) observed on average 92 daughters produced by each female during its reproductive life. Taking into account the developmental mortality rate of D. basalis progeny (Damiens et al., 2001) and assuming that mortality is equivalent between males and females (Doury & Rojas-Rousse, 1994), a mean total of 104 daughters were produced initially by the females, corresponding to 104 sperm used in fertilization. According to the present results, using the regression line established in D. basalis after SYBR-14 : propidium iodide and DAPI staining, we can estimate that 107.3 ± 24.7 green and red + green sperm were stored in the spermatheca at the beginning of the female's life. The same correlation between SYBR-14 : propidium iodide stained sperm and daughter output (Bressac & Chevrier, 1998) exists in E. orientalis. In keeping with this sperm efficiency, those dual-stained sperm have to be considered as viable. Thus, the transitional population of dual-stained (red + green) sperm observed after the dissection of the spermatheca is an artefact due to the method.
In E. orientalis, the populations of SYBR-14 : propidium iodide-stained sperm from seminal vesicle and spermatheca presented a similar profile. By contrast, in D. basalis, a strong difference occurred in the populations of coloured sperm counted from seminal vesicles and spermatheca. A strong increase in the proportion of viable sperm occurred in the spermatheca. In E. orientalis, the quantity of sperm stored after one copulation exceeds the number of daughters produced (Bressac & Chevrier, 1998). By contrast, in D. basalis, after one mating the quantity of sperm stored in the spermatheca of females is insufficient to fertilize all eggs produced during their reproductive life (Chevrier & Bressac, 2002). The presence of a majority of viable spermatozoa may be due to this constraint. The physiological mechanism is unknown but females could assume a manipulation of the transferred sperm population during the spermathecal replenishment (see Neubaum & Wolfner, 1999) to increase the percentage of viable sperm.
The authors thank Christian Thibeaudeau for technical assistance. The English text was corrected by Sue Edrich.