| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gamete Biology; |
Faculty of Medical and Health Sciences, The University of Auckland, Auckland 1142, New Zealand
ABSTRACT
Although the sex of the offspring in mammals is commonly viewed as a matter of chance (depending on whether an X or a Y chromosome-bearing spermatozoon reaches the ovum first), evolutionary biologists have shown that offspring sex ratios are often significantly related to maternal dominance, a characteristic that has been shown to be linked to testosterone in female mammals, including humans. Hence, we hypothesized that variations in female testosterone might be related to reproductive mechanisms associated with sex determination, with higher levels of follicular testosterone being associated with a greater likelihood of conceiving a male. To investigate this hypothesis we collected follicular fluid and cumulus-oocyte complexes from bovine antral follicles. Individual matched samples of follicular fluid were assayed for testosterone, whereas the oocytes were matured, fertilized, and cultured in vitro. The resultant embryos were sexed by PCR. The level of testosterone in the follicular fluid was then compared with sex of the embryo (n = 171). Results showed that follicular testosterone levels were significantly higher for subsequently male embryos (Mann-Whitney U = 2823; P [one-tailed] = 0.016). When we excluded embryos from follicles in which the estradiol-to-testosterone ratio was more than 1 (leaving a sample size of 135), the same result held (Mann-Whitney U = 1667; P [one-tailed] = 0.009). Thus, bovine ova that developed in follicular fluid with high concentrations of testosterone in vivo were significantly more likely to be fertilized by Y chromosome-bearing spermatozoa.
embryo, fertilization, follicle, oocyte development, steroid hormones, testosterone
For many years, theorists in sex allocation have viewed the absence of evidence for a satisfactory system of adaptive control of the sex ratio in mammals as a serious problem, and many have suggested this problem may only be solved when more is known about a proximate mechanism for sex determination [1–4]. At the same time, there has been growing evidence to support the hypothesis that the mammalian female could have some control over the sex of the offspring she conceives. Evidence for a maternal input to sex determination or, more accurately, sex predetermination, comes primarily from animal behaviorists and evolutionary biologists who have documented atypical sex ratios according to maternal dominance and maternal condition in both wild and captive populations [5–9]. In humans, dominant women have been shown to conceive more sons [10].
Researchers have described a number of ways in which there could be a maternal influence on the sex of the offspring, both before and after conception [11]. For example, preconception effects include the timing of insemination within the estrus or menstrual cycles [12], the length of the follicular phase [13], and the potential for maternal control of differential access through the zona pellucida [14]. The possibility of a maternal influence on the sex ratio immediately after conception has also been described [15–17].
While many of these suggested maternal influences may be relevant to offspring sex ratio, we chose to investigate the role of maternal testosterone. The main reason was the weight of evidence from evolutionary biology and psychology [18, 19]. Briefly, this is based on the finding that good maternal condition and/or maternal dominance, especially maternal dominance measured around the time of conception [20], is associated at a statistically significant level with more male offspring. It seems likely that good condition in animals is subsumed by dominance, and since dominance has been shown to be a behavioral characteristic underpinned by testosterone in mammals [21], including humans [22], we hypothesized that maternal testosterone could provide the link between the behavioral characteristics known to be associated with atypical offspring sex ratios and a proximate mechanism for the determination of the sex of the offspring. Since higher maternal dominance has been shown to be related to male-biased offspring sex ratios, we hypothesized that higher female testosterone would be associated with the conception of male offspring.
Follicular steroids are known to vary widely between individuals [23] and to fluctuate over time within the estrous or menstrual cycles [24]. In human follicular fluid, testosterone occurs at a higher concentration than it does in female blood by a factor of 10 000 to 30 000 [25]. Although testosterone in the antral follicle is thought to act primarily as a precursor for the production of estrogen [23], we hypothesized that high testosterone in vivo would be associated with more male embryos resulting from in vitro fertilization.
The present study built on an earlier finding that showed the potential for a significant relationship between the level of testosterone in the mother's follicular fluid and the subsequent sex of the embryo [26]. In that study we took ova mainly from primary follicles. In these follicles we found no relationship between level of testosterone in the follicular fluid and the subsequent sex of the embryo. But in a small sample of antral follicles there was such a relationship—ova developing in vivo in follicular fluid that contained high levels of testosterone being more likely to be fertilized by Y-bearing spermatozoa in vitro.
We hypothesized, therefore, that if there is an effect of follicular testosterone on subsequent sex of offspring, it may be occurring at an earlier stage of development than first thought, hence our decision in this study to concentrate on antral follicles. It has been suggested [27] that the ratio of estrogen to testosterone in the follicular fluid may be a more accurate indicator of oocyte maturity than follicle size in both cows [28] and humans [29]. Since the particular timing within the antral phase may be important, we used both gross follicle size and the ratio of estrogen to testosterone in the follicular fluid as indicators of follicular development.
On the basis of our earlier results [26], we undertook a prospective power analysis to calculate the sample size needed to achieve a power of 0.8 at a significance level of 0.05 with an expected mean difference of 100 nM follicular testosterone and a standard deviation of 200 nM testosterone. The analysis indicated a total sample size of 126 was required.
Ovaries were recovered from heifers immediately after slaughter at a registered slaughterhouse (Auckland Meat Processors, Auckland, New Zealand). They were placed in a large thermos for transport to the laboratory. The diameter of each follicle was measured and recorded. Follicles measuring between 4 and 10 mm in diameter were punctured with a 19x half-needle attached to a sterile 3-cc syringe, and the follicular contents, fluid, granulosa cells, and the cumulus-oocyte complex (COC) were aspirated and then expelled into a sterile 35-mm Petrie dish. The dish was searched to recover the COC, which was graded and placed into HEPES-buffered 199 medium (Earls salts, supplemented with 15 mM HEPES, and 10 mM bicarbonate; Gibco, Auckland, New Zealand).
Follicular fluid (15 µl) was transferred to a sterile 0.6-ml Eppendorf tube, which was labeled with the follicle identity. The remaining follicular fluid was transferred to a sterile 1.6-ml Eppendorf tube, labeled with matching ID, and stored on ice until assayed.
All oocytes were matured, fertilized, and cultured using the single-embryo method of culture [30], allowing each embryo to be traced back to its follicle of origin. Maturation was as follows: follicular fluid (10%) from the follicle the oocyte was recovered from replaced fetal calf serum, and 10 µl/ml FSH and LH was added. Oocytes were cultured at 38.5°C in a mixture of 5% CO2 in air for 22 h before in vitro fertilization was performed using semen from a single bull at a final sperm concentration of 106/ml. Of the 526 oocytes harvested and matured for this study, 191 cleaved, giving a cleavage/fertilization rate of 36.3%. Presumptive zygotes were cultured for either 2 or 3 days in synthetic oviduct fluid [31] until reaching the six- to eight-cell stage. Each embryo was graded and washed in PBS before transferring in 5 µl PBS into a labeled PCR tube and frozen at –80°C until all runs were completed.
Follicular Fluid and Steroid Assay
Each follicular fluid sample, recovered as described above, was centrifuged to remove the cellular debris before addition to maturation medium. The larger aliquot was stored at –20°C until being assayed for testosterone. Testosterone was measured by an in-house ELISA [32] with the samples diluted 1:100 in the assay buffer. The detection limit of the assay is 0.3 nM. Estradiol was measured using a commercial BioQuant (San Diego, CA) ELISA according to the manufacturer's instructions. The estradiol assay had a sensitivity of 10 pg/ml. Samples were initially assayed at a dilution of 1:10, but for samples with values above the highest standard or below the lowest standard the dilution was adjusted appropriately and the sample reassayed.
The embryos (six to eight cells) were sexed by PCR using primers specific for the S4 repeat sequences on the X and Y chromosomes (forward primer 5'-CAAGTGCTGCAGAGGATGTGGAG-3'; reverse primer 5'-GAGTGAGATTTCTGGATCATATGGCTACT-3') [33]. Amplicons were stained with ethidium bromide and visualized under ultraviolet light following electrophoresis on either a 3% agarose or 10% polyacrylamide gels. We successfully determined the sex of 173 of the 191 embryos in this study, the sex ratio of the resultant sample being 0.59.
As found in the earlier study [26], distribution of testosterone in the follicular fluid was markedly skewed to the right. The distributions were not significantly different from exponential (Kolmogorov-Smirnov Z: 0.971 for female embryos and 1.135 for male embryos). Two samples with extremely high levels of testosterone were identified as outliers by Method E6 of Bartlett and Lewis [34] for identifying outliers in an exponential distribution (n = 173 and test statistic T = 0.9056, P = 0.00). These two samples were excluded, leaving complete results for 171 embryos. Follicular testosterone levels were compared to the sex of embryos using the nonparametric Mann-Whitney U test. We also calculated the Killeen [35] replication statistic for the probability that equipotent experiments would support the effects we obtained.
Testosterone levels in each matched sample of follicular fluid were compared with the subsequent sex, male or female, of the embryo (n = 171). The analysis showed that follicular testosterone levels were significantly higher for subsequently male embryos than for subsequently female embryos (U = 2823; P [one-tailed] = 0.016). The probability of replicating this effect with experiments equal in power to ours is 0.935 [35]. The median testosterone level for subsequently male embryos was 122.60 nM, and for subsequently female embryos it was 90.75 nM. Figure 1 shows the distributions of follicular testosterone for embryos (n = 171) of each sex.
|
Of the embryos originating from follicles that had more than 300 nM testosterone (n = 20), only three were subsequently fertilized by X-bearing spermatozoa, whereas in embryos originating from follicles that had more than 300 nM estradiol, there were similar numbers of male and female embryos produced (n = 16; male = 9, female = 7).
When we restricted our sample to those embryos that had matured in follicular fluid in which the oestradiol to testosterone ratio was less than one (n = 135) the same result held (Mann-Whitney U = 1667, P [one-tailed] = 0.009), with a probability of replication of 0.954 for an equipotent experiment.
Our results show that oocytes exposed to high levels of testosterone in the follicular fluid are more likely to produce male embryos after in vitro fertilization. The results also suggest there is not an absolute testosterone level at which oocytes become committed to fertilization by a Y chromosome-bearing spermatozoon, but rather that testosterone, by some as-yet unknown mechanism, can influence the susceptibility of an oocyte to fertilization by an X or Y chromosome-bearing sperm.
It is unclear from our work how the high levels of testosterone in individual follicles were generated. Were they the consequence of increased testosterone production or of decreased aromatization of testosterone to estradiol, or a combination of both? The mechanism leading to raised testosterone may be important, since earlier studies [23, 24] suggested that high levels of follicular testosterone at the midantral stage of development might indicate forthcoming atresia of the oocyte. Yet, despite studying only follicles of relatively constant size (4–10 mm in diameter), which we used as a surrogate for follicular developmental stage, we found a wide range of follicular testosterone levels (11–977 nM), and all of these were associated with subsequently fertilizable oocytes. It may be that the cause of high follicular testosterone rather than the high testosterone itself determines whether an oocyte will become atretic or go on to be ovulated with an increased likelihood of being fertilized by a Y chromosome-bearing spermatozoon. This suggestion is supported by the finding that estrogen prevents follicular atresia while androgens antagonize this effect of estrogen rather than directly inducing atresia [36]. Thus, failure to aromatize testosterone to estrogen might lead to atresia, whereas increased testosterone production, provided the follicle can aromatize sufficient testosterone to estrogen to prevent atresia, might lead to predominantly male offspring.
Since testosterone levels change with follicular development, it was also possible that the different levels of testosterone we observed in follicles were simply related to the developmental stage of the follicles rather than to the individual follicle's ability to produce or metabolize testosterone. Therefore, we employed the ratio of estrogen to testosterone, which has been suggested to be a more accurate measure of follicular development, to further explore the association between high follicular testosterone levels and fertilization by a Y chromosome-bearing spermatozoon. Using this measure to more accurately restrict our analysis to follicles at a similar stage of development resulted in a stronger association between follicular testosterone and sex of subsequent embryo than gross follicular size had done. This suggests that the range of testosterone levels we saw was not due to variation in the stage of follicular development, but represented differences in the intrinsic ability of viable follicles to produce and/or metabolize testosterone.
The full spectrum of the physiological roles of testosterone in follicular fluid may not yet have been elucidated. The most prominent role for this hormone in the follicle is to act as the precursor for estrogen synthesis via aromatization. Although testosterone is known to regulate the production of other molecules in the follicle via the action of androgen receptors, which in the bovine appear to be restricted to the granulosa cells [37], our results suggest an additional role for follicular testosterone in the mammalian ovarian follicle may be to modify the developing oocyte in such a way as to render it more susceptible to fertilization by a Y chromosome-bearing spermatozoon. Exactly what modification testosterone might induce in the developing oocyte is unclear, but the first interactions between an oocyte and a spermatozoon occur as the sperm binds to the oocyte's zona pellucida, and since two of the three zona pellucida proteins are produced during the antral stages of follicular development it is possible to speculate that testosterone levels may alter the nature or expression of these proteins in some subtle way. Alternatively, since spermatozoa must penetrate the cumulus cells surrounding the oocyte before fertilizing it, it is also possible that interactions between cumulus cells and spermatozoa are altered in some subtle way by high levels of follicular testosterone.
The relationship we have shown between level of follicular testosterone and subsequent sex of offspring suggests there might be a pathway whereby mammalian mothers could influence which sex offspring they conceive. As testosterone not only underpins dominance behaviors but also fluctuates over time, depending on environmental conditions, changes in the level of follicular testosterone could help account for both interindividual and intraindividual fluctuations in the sex ratio. Since chronic stress causes female (but not male) testosterone to rise [38, 39], conditions that cause chronic stress in females may lead to an increase in the number of male conceptions. Such conditions have already been associated with raised sex ratios in a number of species. For example, local resource competition has been shown to lead to raised sex ratios in animals in the wild [40, 41] and in humans following war [42, 43]. It may be that this sequence of events provides an adaptive counterbalance to male vulnerability, also known to rise under conditions of chronic stress [44–46]. Conversely, good conditions may give rise to somewhat lower birth sex ratios [47].
The suggestion that mammals may, after all, have a measure of adaptive control of the sex of their offspring would be consistent with findings in other taxa [48, 49]. If so, our results could help pave the way toward the resolution of a longstanding puzzle in sex allocation theory [50]. However, we wish to caution against the overinterpretation of the data presented here. We have made an interesting observation that suggests a mechanism whereby mammalian females could influence the sex of their offspring. Nevertheless, much more work will be required to confirm the existence of this mechanism, including independent confirmation by others and a demonstration of a similar mechanism in other mammals.
ACKNOWLEDGMENTS
We thank J. Lewis, Christchurch, for conducting the steroid assays. We are grateful to Auckland Meat Processors for supplying bovine ovaries and to Ambreed New Zealand for supplying sperm from a bull of proven fertility.
FOOTNOTES
1Supported by grants from The University of Auckland Research Committee and the Faculty of Medical and Health Sciences, The University of Auckland. 
Correspondence: 2V.J. Grant, Faculty of Medical and Health Sciences, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand. FAX: 64 9 3737 013; e-mail: vj.grant{at}auckland.ac.nz
Received: 11 October 2007.
First decision: 30 October 2007.
Accepted: 31 December 2007.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  P. Bermejo-Alvarez, D. Rizos, D. Rath, P. Lonergan, and A. Gutierrez-Adan Can Bovine In Vitro-Matured Oocytes Selectively Process X- or Y-Sorted Sperm Differentially? Biol Reprod, October 1, 2008; 79(4): 594 - 597. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
