| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Ecology and Evolution, University of Lausanne, CH-1015 Lausanne, Switzerland
ABSTRACT
The aim of the present study was to establish and compare the durations of the seminiferous epithelium cycles of the common shrew Sorex araneus, which is characterized by a high metabolic rate and multiple paternity, and the greater white-toothed shrew Crocidura russula, which is characterized by a low metabolic rate and a monogamous mating system. Twelve S. araneus males and fifteen C. russula males were injected intraperitoneally with 5-bromodeoxyuridine, and the testes were collected. For cycle length determinations, we applied the classical method of estimation and linear regression as a new method. With regard to variance, and even with a relatively small sample size, the new method seems to be more precise. In addition, the regression method allows the inference of information for every animal tested, enabling comparisons of different factors with cycle lengths. Our results show that not only increased testis size leads to increased sperm production, but it also reduces the duration of spermatogenesis. The calculated cycle lengths were 8.35 days for S. araneus and 12.12 days for C. russula. The data obtained in the present study provide the basis for future investigations into the effects of metabolic rate and mating systems on the speed of spermatogenesis.
male reproductive tract, spermatogenesis, testis
Spermatogenesis is the process whereby the spermatogonial stem cells on the basal membrane of the seminiferous tubules divide and differentiate, giving rise to testicular spermatozoa at the luminal surface. During this process, the germ cells are arranged into cellular associations and are divided into named stages. These stages proceed in succession over time in any given region of the tubule, and this process is defined as the seminiferous epithelium cycle [1]. The types of germ cell involved in this process are similar in all mammals [2, 3]. Differences between species exist with regard to the number of germ cells, the duration of spermatogenesis, and the topographical arrangement of the spermatogenetic stages [4].
The duration of the cycle has been studied in many species. It seems to be rather constant within species but varies between species. Within rodents, the cycle length may be as short as 6.7 days, as in the bank vole Clethrionomys glareolus [5], and up to 17 days, as in the Chinese hamster Cricetulus griseus [6]. In primates, the cycle length is 9.4 days in the cynomolgus monkey Macaca fascicularis [7] and 16 days in humans [8]. Within marsupials, the cycle duration is rather long, from 15.7 days in Trichosurus vulpecula [9] to 17.3 days in Didelphis albiventris [10]. There is no obvious factor to explain the differences between species.
Body size and the related energy metabolism appear to be important factors in determining the physiological rates [11–13]. However, a possible influence on spermatogenesis has never been tested. According to McNab [14], a higher metabolic rate may increase cell cycle speed and tissue synthesis. Therefore, we conclude that metabolic rate is a potential factor in spermatogenesis. In the context of energy metabolic rate, shrews represent an interesting model. The two recognized shrew subfamilies, Soricinae (red-toothed shrews) and Crocidurinae (white-toothed shrews), show important differences. The Soricinae (e.g., Sorex araneus) are known for their extremely high metabolic rates, which are 200–300% of the value expected for their body sizes [15–18]. In contrast, the Crocidurinae generally have metabolic rates just above the expected value (about 120%), with some desert species going as low as 75% of the expected value [19]. Therefore, we studied the duration of the spermatogenesis cycle in shrews.
To the best of our knowledge, this type of study has not been reported previously, with the exception of histological descriptions of the stages that characterize the cycle of spermatogenesis. According to Plöen et al. [20] and Garagna et al. [21], in Soricinae, the seminiferous tubules show a segmental arrangement, with usually just one stage found per tubule cross-section. This topographical arrangement is present in the majority of mammalian species. The situation is similar in the Crocidurinae, as shown for the Japanese Crocidura watasei [22] and the Asian Suncus murinus [23]. In these studies, the authors subdivided the cycle of the seminiferous epithelium according to the morphological characteristics of spermatids, with particular emphasis on their nuclei and acrosomic system [24]. Based on this method, the number of stages of the cycle varies between species, with 10 stages in S. araneus [21], 12 stages in C. watasei [22], and 13 stages in S. murinus [23]. However, the duration of each stage has yet to be determined.
In recent decades, the presence or absence of sperm competition has been shown to be an important factor with influence on sperm production [25]. When sperm competition occurs, males are often selected on the basis of having an increased quantity of sperm. This is evidenced by a significant correlation between the strength of competition and testis size [26–29]. Larger testis volume is linked to higher total sperm production. In this context, shrews are again an interesting model. In multiple copulators, such as S. araneus, sperm competition seems to occur, and a litter may have up to five fathers [30–32]. In contrast, within the Crocidurinae, which is the most extensively studied species, Crocidura russula seems to be rather monogamous [33] and multiple paternity has never been demonstrated [34]. Besançon [35] has reported rather small testes for C. russula compared with the large testes of S. araneus. If increased sperm production is linked to an increase in testis size, sperm competition may also be expected to increase sperm production per unit time and testis volume. As shown previously, both an increase in metabolic rate (for any reason) and sperm competition may lead to a higher rate of sperm production. The present study examines the duration of the sperm production cycle in two species with very different life history strategies [36, 37], which should establish a standard for further comparisons.
In the present study, we determined the duration of the cycle of spermatogenesis in both the common shrew S. araneus and the greater white-toothed shrew C. russula. The dynamics of sperm production were determined by tracking 5-bromodeoxyuridine (BrdU) in the DNA of S-phase germ cells, since BrdU was incorporated into the nuclei of cells that were duplicating their DNAs in preparation for mitosis or meiosis.
S. araneus is widely distributed in the northern Palaearctic region. From April to June, we captured twelve adult S. araneus (chromosomal race of Valais, now considered to be S. antinorii [38]), weighing 10.9 ± 0.7 g, in the regions of Trient and Grand St-Bernard in the Alps at an altitude of 1400–2200 m.
C. russula, which originated in Africa [39], is widespread in south-western Europe. Fifteen adult males weighing 14.2 ± 1.3 g were captured during the breading season (February to August) in the region of Lausanne, Switzerland (altitude 300 m). The shrews were kept in a special building with a roof that protected them from direct rain but with walls of wire mesh that exposed the animals to natural fluctuations in temperature and humidity according to the weather. The shrews were caged individually, on natural soil. They had ad libitum access to water and food, which was composed of minced meat with vitamin supplements and mealworms.
BrdU (BD Biosciences, Pharmingen) was administered to each male as a single i.p. injection at a dose of 50 mg/kg. S. araneus males were treated in groups of three animals and killed by halothane overdose at 3 h, 8 days 3 h, 12 days 3 h or 16 days 3 h after injection of BrdU. The C. russula males were killed according to the following schedule: four males at 3 h, three at 8 days 3 h, three at 12 days 3 h, three at 16 days 3 h, and two at 24 days 3 h after the administration of BrdU. The testes were removed, weighed, fixed in a 4% solution of paraformaldehyde for 3 days, dehydrated in a graded series of ethanol, and embedded in Paraplast. Histological sections (3-µm thickness) of the testes were dewaxed and rehydrated. BrdU was localized using the Zymed BrdU Staining Kit (Invitrogen). Endogenous peroxidase was inactivated with 3% (v/v) H2O2 in methanol for 10 min at room temperature. To expose BrdU for immunohistochemical localization, DNA was denatured in 0.17% (v/v) trypsin solution for 10 min at 37°C in a humid chamber. Incubation with a biotinylated anti-BrdU antibody was performed for 1 h at room temperature. Sections were then incubated with streptavidin-HRP for 10 min at room temperature. DAB substrate solution was added to cover the tissue sections and they were incubated for 5 min or until color development. PAS-hematoxylin staining was performed for the identification of spermatogenic stages. Incubation with 1% periodic acid (for 10 min at room temperature) and staining with Schiff reagent (for 30 min at room temperature) was followed by incubation with Mayer hematoxylin solution (5 min at room temperature). Histological preparations were observed under an Axiophot microscope (Zeiss, Germany).
Stage Frequencies of the Seminiferous Epithelium Cycle
The ten stages (I–X) of the seminiferous epithelium cycle were determined following the criteria applied previously to S. araneus [20, 21]. In C. russula, we identified the 12 stages of the seminiferous epithelium according to the classification proposed by Adachi et al. [22] for Crocidura watasei. For direct comparisons with S. araneus, we pooled stages VIII-IX and X-XI of Crocidura, and renamed them stages VIII and IX respectively; thus, now stage XII corresponds to stage X in S. araneus (see Table 2).
|
The relative duration of each stage of the cycle in both species was determined for testes sections that were counterstained with PAS-hematoxylin. We counted only round cross-sections of the seminiferous tubules.
Stages of short duration should appear rarely, while stages of long duration should appear frequently. Therefore, the stage frequencies (expressed as percentages) correspond to the relative durations of the stages of the cycle. We counted 200–400 tubular cross-sections per testis. In all, 7000 tubules from 12 S. araneus shrews and 6000 tubules from 15 C. russula shrews were scored. Stage frequencies were calculated for all animals using the equation:
|
| 76 |
The estimation of staining frequency was based on the percentage of tubules of a given stage that contained the most-advanced germ cells that were labeled with BrdU, as follows:
|
| 76 |
Duration of the Seminiferous Epithelium Cycle
Two methods were used to estimate the cycle duration. The first method was that of Rosiepen et al. [40, 41]. Following this formula, the duration of the seminiferous epithelium cycle was calculated based on the stage frequencies and BrdU staining frequencies for two time-points. We used the time-point of 3 h after BrdU injection as a reference group. Between this time-point and the other time-points, we calculated the duration of one cycle of spermatogenesis. For example, in S. araneus, the duration was calculated as follows: at 3 h after BrdU injection, (reference group), the most-advanced labeled cells were at stage VI, while at 8 days 3 h after BrdU injection in animal #2, the most-advanced labeled cells were also in stage VI.
The time interval 
T was calculated using the equation:
|
| 76 |
|
| 76 |
For example, in S. araneus, the most-advanced labeled cells at 3 h after BrdU injection were in stage VI. This corresponds to the sum of the frequencies of stages I to V plus the (staining frequency x stage frequency of stage VI). This is the relative time in percentage at 3 h. At 8 days 3 h (195 h real time), the most-advanced labeled cells were in the same stage. We added stages VII–V plus the (staining frequency x stage frequency of stage VI) at 8 days 3 h. In the same way, we calculated the relative time (%) for the other time-points of 12 days 3 h and 16 days 3 h after BrdU injection.
The duration of one cycle (100% relative time) was inferred from the regression function and is given as the value that was used for the comparisons. The standard deviation was also inferred from the linear regression.
This project was performed under authorization number 1707.1 VD. The number of samples was kept as low as possible, since all species of shrew are protected animals in Switzerland. However, the loss of some males from a local population has no detrimental consequences due to the rather high population sizes of these two species. In the Lausanne population on the campus of the University, we have trapped and individually marked over 500 individuals of C. russula in 4 yr [34, 42].
All the data are presented as the mean ± SEM. Analysis of regression was performed using software R (R Development Core Team, 2006). The significance level was considered to be P < 0.05. To test the differences between Sorex and Crocidura, we applied the t-test after testing for normality.
During the period from the administration of BrdU to killing, we examined the general health of the treated animals. BrdU administration was well-tolerated and no negative effects were observed in either species. Measurements of body mass and testicular size are given in Table 1. The absolute testes mass of S. araneus (0.17 ± 0.02 g) was significantly higher than that of C. russula (0.040 ± 0.004 g) (Welsch modified t-test for non-equal variance; t = –20.8759, df = 11.481, P < 0.0001), and the relative testes size expressed as a percentage of body weight was also significantly higher in S. araneus than in C. russula (Welsch modified t-test for non-equal variance; t = –20.5506, df = 11.85, P < 0.0001).
|
In S. araneus, 3 h after BrdU injection, the label was localized in the nuclei of preleptotene spermatocytes at stage VI. This stage was used as a reference, and it contained the most-advanced labeled cells at 3 h (Fig. 1a). Stage VI began after spermiation and was characterized by the presence of round spermatids in the luminal compartment of the tubule. Residual bodies were found in the seminiferous epithelium. Two spermatocyte generations were present. The preleptotene spermatocytes were visible in the vicinity of the basement membrane, which was marked by BrdU. Pachytene spermatocytes were situated in the middle region of the seminiferous epithelium. At 8 days 3 h, the most-advanced cells that contained BrdU were pachytene spermatocytes in the same stage (Fig. 1b). At 12 days 3 h, the most-advanced BrdU-labeled cells were newly formed round spermatids at stage I (Fig. 1c). This stage began after the second meiotic division of the secondary spermatocytes and was characterized by the presence of two generations of spermatids: round spermatids in the middle region of the seminiferous epithelium and elongated spermatids in the luminal compartment. Type A spermatogonia overlying the basal membrane and small pachytene spermatocytes were also observed. At 16 days 3 h in two animals, the most-advanced BrdU-labeled cells were round spermatids at stage V (Fig. 1d) and at stage VI in one animal (Fig. 1e). Stage V was characterized by the presence of elongated spermatids that were ready to be released into the seminiferous lumen and round spermatids with an acrosome that covered half of the nucleus.
|
In C. russula, the most-advanced BrdU-labeled cells at 3 h, 12 days 3 h, and 24 days 3 h after BrdU injection were in stage VI (Fig. 2). In this stage, we observed cell associations similar to those seen for S. araneus. One generation of round spermatids and residual bodies were recognized in the seminiferous epithelium and two generations of spermatocytes (preleptotene and pachytene) were present in this stage. At 3 h, the most-advanced labeled cells were the preleptotene spermatocytes (Fig. 2a). At 12 days 3 h, BrdU was detected in pachytene spermatocytes (Fig. 2d), and at 24 days 3 h, the label was detected in round spermatids (Fig. 2f). At 8 days 3 h, the most-advanced labeled cells were pachytene spermatocytes at stage III in two animals (Fig. 2b) and at stage IV in one animal (Fig. 2c). In stage III, the acrosomal granule began to flatten on the surface of the spermatid nucleus. Maturing spermatids were also present at this stage. Stage IV was characterized by the beginning of the formation of the head caps on the nuclei of the spermatids. At 16 days 3 h after BrdU injection, the most-advanced labeled cells were newly formed round spermatids in stage I (Fig. 2e). This stage was characterized by the presence of two generations of spermatids: maturing spermatids and round spermatids that contained BrdU.
|
Stage Frequency, Staining Frequency, and Duration of the Seminiferous Epithelium Cycle
The stage frequencies are shown in Table 2. The last stage is always the shortest, and stages I, V, and VI are the longest. The staining frequencies for Sorex and Crocidura are given in Table 3 and Table 4.
|
|
According to the method of Rosiepen et al. [40, 41] (equation 3), we estimated the cycle duration in S. araneus as 8.47 ± 0.26 days (Table 1). The observed variation was from 8.11 days to 8.79 days. In C. russula, the mean cycle duration was 12.42 ± 0.60 days (Table 1), varying from 11.72 days to 13.40 days.
With the second method, we estimated the duration of the seminiferous epithelium cycle by linear regression. For example, in S. araneus #1 at 3 h after BrdU injection, the most-advanced labeled cells were at a position that corresponded to 74.96% of the relative duration of the cycle, i.e., the sum of the stage frequencies (I+V) + (staining frequency x stage frequency of stage VI). At 8 days 3 h (195 h) after BrdU injection, the most-advanced labeled cells in another animal (#2) were at stage VI, corresponding to 171.76% of the relative duration. The label passed through almost 100% of one cycle of spermatogenesis. At 12 days 3 h (291 h), the most-advanced labeled cells in animal #10 were at 212.01% from the starting point. We also analyzed the position of the most-advanced labeled cells at 16 days post-BrdU injection. We applied the same method to the determination of the cycle duration in C. russula.
The cycle durations obtained by linear regression were 8.35 ± 0.13 days (r2 = 0.998) for S. araneus and 12.12 ± 0.19 days (r2 = 0.997) for C. russula (Fig. 3). As a consequence, the duration of spermatogenesis based on 4.5 cycles is approximately 37.6 days for S. araneus and 54.7 days for C. russula.
|
The present study is the first to estimate the cycle length of spermatogenesis in shrews and to apply linear regression as a new method for cycle length determination. Comparing the new linear regression method with the traditional method, the former calculation gave, as expected, a mean that was similar to that provided by the latter. However, with regard to variance, and even with a relatively small sample size (12 and 15 individuals), the new method seems to be more precise. Indeed, our results show that the regression method is significantly more accurate for both Sorex (F = 0.243, df1 11, df2 8, P = 0.017) and Crocidura (F = 0.097, df1 14, df2 10, P < 0.001). Moreover, while the classical method deals with pairs, making each value dependent upon at least two animals, the regression method allows the inference of information for each animal tested. Thus, it is possible to compare factors, such as basal metabolic rates or testosterone levels, with the residuals of the regression, thereby generating a score for spermatogenesis speed with regard to the mean speed for each animal tested. This new method also allows the detection of any abnormalities in spermatogenesis speed, as individuals that diverge from the regression line are depicted graphically. For all of these reasons, we conclude that determination of the spermatogenesis cycle using this new method is statistically more precise and allows more possibilities than the traditional method.
Spermatogenesis Cycle Length Comparisons
The aim of the present study was to determine the durations of the seminiferous epithelium cycles in shrews with different metabolic rates (Soricinae versus Crocidurinae) and different mating systems (S. araneus versus C. russula). In mammals, the duration of spermatogenesis is considered to be species-specific, although strain and breed differences are found among members of the same species [24]. However, the duration of spermatogenesis is not necessarily the same for closely related species [24, 43]. Spermatogenic cycle length is under the control of the germ cell genotype [44] and is not altered by Sertoli cells or by exposure to the gonadotropic hormone environment of a different species. Recently, it has been reported that the length of the spermatogenic cycle is conserved in testis tissues xenografted from pigs and sheep into recipient mice [45]. According to our working hypothesis, the shorter cycle length of spermatogenesis in S. araneus than in C. russula should be attributable to the higher metabolic rate or the promiscuous mating system in the former. We confirmed that the cycle length of S. araneus (8.35 ± 0.13 days) was 68.7% of the cycle length of C. russula (12.12 ± 0.19 days). Therefore, with spermatogenesis of 4.5 cycles, total spermatogenesis lasts 37.6 days in S. araneus and 54.7 days in C. russula.
This hypothesis of an influence of metabolic rate on cycle length is in agreement with the hypothesis of MacNab [14] with regard to the intensity of tissue synthesis and growth rate. A higher metabolic rate in mammals is correlated with a high body temperature. As in all shrews, the testes lie in a cremaster sac in the abdominal cavity [46], so body temperature may act directly on cycle length, as shown experimentally for mice [47]. A high metabolic rate is possible only in temperate or cold climates, otherwise overheating would occur. This situation is well-known in Sorex, which has a body temperature of about 39°C [48] and a holarctic distribution. In contrast, Crocidura has a lower metabolic rate [15, 17] and a lower body temperature of about 35.5°C [48], and originally had a rather tropical distribution. According to Vogel [15], the high metabolic rate probably evolved to ensure strong heating competence in a temperate climate under sufficient nutritional conditions. Tropical shrews developed a lower metabolic rate to avoid overheating [14], particularly desert shrews, such as C. viaria [19].
Until now, a correlation between the rate of sperm production and the metabolic rate has not been demonstrated, and our present results do not reveal the underlying mechanism. If this relationship holds true, there should also be a relationship between sperm production rate and body size, which shows a strong correlation with metabolic rate. More data are needed to understand how metabolic rate and body mass influence the cycle length of spermatogenesis.
According to the Parker paradigm of sperm competition [25], the mating system is of primary importance in sperm production. When the frequency of copulation is higher, the testes are larger (multiple mating systems), and when copulation is rare, the testes are small (single mating systems). Apparently, the selective pressure of multiple insemination and competition among spermatozoa in the female genital tract (sperm competition) has led to the evolution of larger testes in species with multiple mating systems [49]. Larger testes have a greater volume of seminiferous tissue, which is required to achieve higher levels of sperm production. There is a strong relationship between the mating system and relative testis size, with polyandrous species tending to have relatively larger testes [26, 50].
Our results show that both increased testis size and increased speed of spermatogenesis lead to increased sperm production. As a rough estimation, sperm production in S. araneus could be six-fold higher than in C. russula, due to the volume difference (a factor of 4.25) multiplied by the speed difference (a factor of 1.45). A similar finding in Australian mice has been reported by Peirce and Breed [51]. In that study, the promiscuous Pseudomys australis had a relative testis size of 3.21 and a cycle length of 11.2 days, whereas Notomys alexis had a relative testis size of only 0.16 and a cycle length of 14 days. This was interpreted by the authors as the result of a single-male mating system [51].
Finally, this study raises questions as to the plesiomorphic situation and the derived situation. Evolution can go in either direction, depending on the selective forces. Sperm production is certainly energy-consuming, and a lack of sperm competition may quickly lead to a reduction in sperm production. On the other hand, species that replace sperm competition with mate-guarding or even paternal investment in the survival of the litter, as is the case for C. russula [33], may simply invest their energy differently, while a trade-off may lead to equal fitness.
To our knowledge, basal metabolic rate and spermatogenic cycle length have not been shown to be correlated, and our present investigation was not designed to reveal such a relationship. Indeed, the speed of the cycle may depend on different physiological mechanisms. To separate the effect of metabolic rate from the effect of the mating system, future studies of sperm cycle length should include, on the one hand, several Soricidae that have low metabolic rates (e.g., Notiosorex [52]) or that lack multiple paternity, and on the other hand, some Crocidura with multiple paternity, if such types exist in nature.
ACKNOWLEDGMENTS
We thank R. V. Krstic, S. Kasas and C. Wedekind for critical reading of the manuscript and Anne-Marie Mehmeti for the provision of laboratory facilities.
FOOTNOTES
1Supported by the University of Lausanne, Switzerland. 
Correspondence: 2FAX: 41 21 692 4165; e-mail: Roumen.Parapanov{at}unil.ch
Received: 12 October 2006.
First decision: 13 November 2006.
Accepted: 29 January 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
