| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
a Section of Neurobiology, Physiology and Behavior, University of California at Davis, Davis, California 95616
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A screen for differences in gonadotropin-induced ovulation rate among 16 inbred strains of mice revealed a large proportion of strains with extreme differences in this trait [9]. Four out of 16 strains had low hormone-induced ovulation rate (HIOR) responses and only ovulated 9–12 eggs, while 3 of the 16 strains had HIOR responses that were 5- to 6-fold higher. For example, the HIOR of A/J mice was 9 ± 1 eggs while that of C57BL/6J mice was 54 ± 2 eggs (P < 0.001) [9]. Additional studies on the number of normal and atretic ovarian follicles found before and after gonadotropin treatment showed that the 6-fold higher HIOR of B6 over that of A/J is largely due to a greater induction of follicle maturation and a decreased incidence of atresia in B6 [11]. The higher response to gonadotropins of the B6 strain was also associated with higher ovarian aromatase activity and estrogen production [12, 13].
Most differences in reproductive traits like HIOR were thought to be quantitative, i.e., controlled by environmental effects and by many genes, each with small effects. Recent studies revealed that in addition to quantitative genetic control of reproduction, variant alleles at a small number of loci can also have dramatic effects on reproductive performance. For example, it is now recognized that a single gene, the Booroola fecundity gene (FecB), has a major effect on the high natural and hormone-induced ovulation rate in sheep developed by selection for large litter size [14–16]. The Inverdale Fecundity gene (FecXI) increases natural ovulation rate in the heterozygous state [15]. Thus, two naturally occurring genes with major effects on ovulation rate are present in sheep. Although these loci have been mapped to sheep chromosomal regions, little is known about the genes that control the more common quantitative genetic differences in ovarian responsiveness to gonadotropins. Study of the genetics of HIOR using crosses of inbred mice indicates that a small number of loci control major strain differences in this trait [9]. This led to the development of a backcross model suitable for mapping these genes, i.e., quantitative trait loci (QTL) controlling HIOR [17].
Evidence that large quantitative differences in reproductive performance are due to a small number of genes with major effects is extremely important. For example, identification of molecular markers for desirable or for deleterious reproductive quantitative trait loci (QTL) would allow a rapid increase in the frequency of favorable alleles in livestock populations using conventional animal breeding techniques. In several mammalian species including humans, such molecular genetic markers could be used to aid in the diagnosis of an individual's hormone response genotype. This approach would facilitate hormonal treatments for divergent hormone response genotypes to more accurately produce the desired level of reproduction, i.e., fertility or infertility.
On the house mouse (Mus musculus) genetic background, 6-fold strain differences in HIOR segregate in F2 crosses as though this trait is controlled by approximately three to four QTL with major effects [9]. This shows that genetic differences in ovarian responsiveness to gonadotropins exist within a species. While thousands of microsatellite and other molecular genetic markers are available in mice, the level of polymorphisms is approximately 2-fold greater between than within Mus subspecies. This is especially true for chromosomal regions where closely related Mus musculus strains seem to have inherited the same chromosomal segment from a common ancestor. Whether or not even greater differences in HIOR exist between subspecies of Mus is unknown. To investigate this possibility, widely divergent subspecies of Mus were used in the present study to provide even more potential genetic variation in HIOR and to insure a large number of polymorphic markers for mapping any observed differences in reproduction [18]. Specifically, the magnitude of genetic differences in HIOR was determined between and among Mus musculus and Mus spretus strains or lines of mice.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hormone-induced ovulation rate (HIOR) was measured on inbred A/J and B6 parentals, B6AF1 females, and 474 (B6xA)xA backcross females. Briefly, females were injected s.c. with 5 IU equine chorionic gonadotropin (eCG) at 4 wk of age, and then with 5 IU human chorionic gonadotropin (hCG) 2 days later. Eighteen to 22 h following the s.c. hCG injection, mice were killed by cervical dislocation, their oviducts removed, and the contents of the oviducts examined with the aid of a dissecting microscope. The number of ova in the cumulus oophorus clump was scored as "eggs in cumulus" and is a measure of HIOR. The number of eggs out of the cumulus clump was also scored. The number of eggs in and out of cumulus were added to calculate the total number of eggs ovulated. Previous studies have shown that ova or eggs in cumulus are from follicles matured by eCG and then ovulated by hCG [9, 11]. Eggs out of cumulus are derived from follicles that mature under the influence of endogenous gonadotropins, ovulate in response to the LH-like activity of eCG, and remain in the oviduct following the dissolution of the cumulus mass [9].
Estimating the Number of Loci Controlling Differences in Reproduction
The number of loci controlling the observed differences in induced ovulation rate was estimated using the procedures of Sewall Wright [9, 19, 20], where the number (k) of QTLs segregating in a backcross between two strains with a phenotypic difference of D is estimated by the formula: k = D2/(16 x genetic variance). The genetic variance is estimated as the difference in the variance of the backcross progeny (which includes components of genetic and environmental variance) from the variance within the inbred parental strains and/or F1 (which includes only environmental variance).
Data Analysis
Data were subjected to analysis of variance using SuperANOVA [21] on a Macintosh (Cupertino, CA) IICi computer. When significant heterogeneity of variances between groups was detected, a logarithmic or square-root transformation of the data was performed prior to one-way ANOVA. Differences between strain and cross means were compared using a Tukey-Kramer Multiple range test, which keeps the experiment-wise type I error rate at an alpha of P < 0.01.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
As shown in Figure 1, strains and subspecies also differed in the number of eggs out of cumulus (P < 0.001). The number of eggs matured by endogenous gonadotropins and ovulated by eCG averaged 7.3 ± 1.3 in the Mus musculus strain B6; 0.7 ± 0.4 in the Mus musculus strain DBA/2J; and 0.0–0.3 eggs in Mus spretus lines and crosses, a difference of up to 10-fold between Mus musculus strains and 40-fold between subspecies. Thus, genotypes differ dramatically in the number of follicles that are matured by endogenous gonadotropins and can be ovulated by the initial dose of eCG.
Strains and lines showed highly significant differences in the total number of eggs ovulated, i.e., the sum of the number of eggs in and out of cumulus (P < 0.0001). The mean ± SEM total number of eggs ovulated by each strain was as follows: A/J (11.9 ± 1.4); DBA/2J (16.5 ± 3.2); B6 (54.6 ± 3.1); Mus spretus/Ros (5.0 ± 1.1); Mus spretus/Ros x SPRET/Ei F1 (32.0 ± 5.5); (Mus spretus/Ros x SPRET/Ei) x SPRET/Ei backcross N2 (16.8 ± 1.8); and (Mus spretus/Ros x SPRET/Ei) x SPRET/Ei backcross N3 (12.8 ± 2.4).
The low level of reproduction in SPRET/Ei inbreds did not allow measurement of HIOR in this strain. Nonetheless, the much higher HIOR of the Mus spretus/Ros x SPRET/Ei F1 over that of the Mus spretus/Ros parent and of each successive (Mus spretus/Ros x SPRET/Ei) x SPRET/Ei backcross suggests a large amount of heterosis in HIOR in this cross within Mus spretus. This finding demonstrates substantial genetic variation in HIOR within Mus spretus.
Strains and lines also showed significant differences in body weight, with Mus spretus lines and crosses found to be significantly smaller than Mus musculus strains (P < 0.05). The mean ± SEM body weight in grams for each strain or cross was as follows: A/J (15.3 ± 0.4); DBA/2J (15.7 ± 0.4); B6 (17.2 ± 0.3); Mus spretus/Ros (11.1 ± 0.2); Mus spretus x SPRET/Ei F1 (11.5 ± 0.2); (Mus spretus x SPRET/Ei) x SPRET/Ei backcross N2 (10.8 ± 0.2); and (Mus spretus x SPRET/Ei) x SPRET/Ei backcross N3 (12.5 ± 0.5). However, the effect of body weight was not significant for any of the ovarian response traits examined.
The next experiment examined the HIOR (i.e., the number of eggs ovulated and found in cumulus), the number of eggs out of cumulus, the total number of eggs ovulated, and the body weight of highly inbred A/J and B6 parental strains and B6AF1 and (B6xA)xA backcrosses (Table 1). Data were analyzed both with and without correction for the effect of body weight. Since body weight did not significantly affect any of the traits examined, nor did it reverse the ranking of strains or crosses, data are presented without correction for the effect of body weight. Strains and crosses differed significantly in HIOR (P < 0.0001) (Fig. 2), log HIOR (P < 0.0001), square-root HIOR (P < 0.0001), eggs out of cumulus (P < 0.001), total eggs ovulated (P < 0.0001), and body weight (P < 0.001). Relative to A/J mice, B6 females showed a 4.8-fold higher HIOR (P < 0.0001), 17-fold higher number of eggs out of cumulus (P < 0.001), 4.8-fold higher number of fragmented eggs out of cumulus (P < 0.0001), 7.8-fold higher total number of eggs out of cumulus (P < 0.0001), and 5.1-fold higher total number of eggs ovulated (P<0.0001).
|
|
Relative to the B6-A/J mid-parental average, the B6AF1 showed 38.9% heterosis for the number of eggs in cumulus (P < 0.01), 210% heterosis for the number of eggs out of cumulus (P < 0.01), and 61.6% heterosis for the total number of eggs ovulated (P < 0.01). Relative to a 1/4 B6 + 3/4 A/J mid-parental average, the (B6xA)xA backcross showed 22.6% heterosis for the number of eggs in cumulus (P < 0.01), 147% heterosis for the number of eggs out of cumulus (P < 0.01), and 37.5% heterosis for the total number of eggs ovulated (P < 0.01). Square-root transformation of the data did not alter the finding of positive heterosis in F1s and backcrosses for each of these induced ovulation rate traits.
Figure 3 depicts frequency distributions for square-root transformed HIOR of highly inbred A/J and B6 parental strains as well as B6AF1 and (B6xA)xA backcrosses. The number of loci controlling the observed differences in induced ovulation rate was estimated using the procedures of Sewall Wright [9, 19, 20]. In essence, this procedure estimates the number of loci controlling strain differences in HIOR by considering the differences in parental strain means relative to how the trait segregates in the backcross relative to the parentals and F1. These estimates assume that 1) each QTL acts additively, 2) the effects of each QTL are similar in magnitude, 3) the parental phenotypic data are normally distributed and show homogeneity of variances, 4) the QTLs are unlinked, and 5) the parental strains are true extremes, i.e., all of the B6 alleles increase this trait while all of the A/J alleles decrease this trait. It should be emphasized that these estimates of the number of loci with major effects on HIOR do not preclude the existence of additional genes with small effects.
|
The differences in HIOR segregated as though they were controlled by the action of approximately 3.6 loci (for raw data) and 3.0 loci with major effects (for square-root transformed data). These estimates are in agreement with previous estimates of approximately 3–4 loci [9], but additional genes with small effects may be operative. The differences in total number of eggs ovulated between B6 and A/J mice segregated as though they were controlled by the action of approximately 2.5 loci (for square-root transformed data) to 3.5 loci (for raw data). The B6AF1 showed significant heterosis in the number of ova out of cumulus (P < 0.01). Therefore, the strain difference in this trait was estimated from the B6AF1 and A/J means. Using this approach with square-root transformed data, the differences in the number of eggs out of cumulus segregated as though they were controlled by the action of approximately 1–1.3 loci with major effects. Given the magnitude of differences between parental strains or F1s, and the approximate number of QTL controlling the difference in each trait, each B6 QTL should increase HIOR in this backcross population by approximately 4–5 eggs, the number of eggs out of cumulus by approximately 3–9 eggs, and the total number of eggs ovulated by approximately 5–14 eggs.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the present study, major differences in HIOR were observed between strains A/J and C57BL/6J (B6), and between Mus spretus lines and crosses. These dramatic differences provide convincing evidence that HIOR is highly regulated by genetic controls or perhaps by genotype x environment interactions. Fivefold genetic differences in this trait were observed among Mus musculus strains of mice, and 6-fold differences in HIOR were found in Mus spretus lines and crosses. While the HIOR of some Mus musculus strains, e.g., A/J and DBA/2J, was similar to that of some Mus spretus lines and crosses, B6 showed an HIOR that was 10-fold higher than that of the lowest Mus spretus population. These genetic differences in HIOR are particularly striking because A/J, B6, and several other strains with widely divergent HIOR were found in a screen of strains that had not been developed by selection for differences in reproduction per se. Indeed, SJL/J, C57BL/10J, and B6 showed 4- to 6-fold higher mean HIOR than strains A/J, AKR/J, CBA/J, and Sec/1ReJ [9].
A large amount of genetic variation in ovarian responsiveness to gonadotropins within and between related species is also demonstrated by the present study. Previous investigations using variance estimation procedures [19] showed that the 6-fold difference in HIOR between A/J and B6 segregated in a B6xAF2 cross as though it was controlled by the action of approximately 3–4 loci with major effects [9]. Our work using (B6xA)xA backcrosses confirms this observation. While the factor analysis shows that a small number of loci are responsible for major strain differences in HIOR, the existence of additional loci with small effects on these differences cannot be excluded. Furthermore, if any of the statistical assumptions made in estimating the number of loci controlling this trait are violated, the actual number of loci could be substantially greater or fewer.
While most of the genetic differences in the number of eggs in cumulus following gonadotropin treatment in B6xA crosses appear to be additive, nonadditive gene action was also evident as shown by the 26–39% heterosis in HIOR in the B6AF1. In contrast, a previous study found little heterosis for the number of eggs in cumulus in a B6AF1 [9]. Several other crosses also showed little or no heterosis for HIOR, although some crosses showed positive and others showed negative heterosis for HIOR [9]. Nevertheless, the present data suggest that much of the genetic variation in the number of eggs out of cumulus following gonadotropin treatment is largely due to nonadditive effects. The observed deviations from midparental means suggest that one or more of the genes controlling the number of eggs matured by endogenous gonadotropins, ovulated by eCG, and then found out of cumulus exerts significant dominance or epistasis. However, these dominance deviations may also be scale effects [31].
It is clear that HIOR and ovarian responsiveness to gonadotropins are highly genotype dependent. As a result, conclusions obtained from studies with one strain of laboratory animal may not be fully applicable to other genotypes or strains that differ dramatically in hormone response genotype. This problem has been largely overlooked because most studies on the physiological, hormonal, and biochemical control of ovarian function in rodents were conducted in a relatively small number of lines of "outbred" laboratory animals such as Sprague Dawley or Long-Evans rats, or CD-1 mice, with most studies only involving one strain or line. Early work showed that many of these outbred populations had a large amount of genetic variation in ovarian response to gonadotropins [32–34]. Subsequently, for economic and production reasons, most of these commercially available outbred strains of laboratory animals were developed by intense, long-term selection for large litter size and vigor, with random mating of selected individuals in each generation. This represents selection for large litter size with avoidance of inbreeding rather than true random selection and random mating. In fact, the majority of outbred laboratory lines of rodents have been selected almost solely for large litter size and vigor for about 100 consecutive generations. Such intense, long-term selection for large litter size, especially in large populations, would dramatically increase the frequency of genes associated with elevated prolificacy and its components of natural ovulation rate, embryonic, and fetal survival [2, 5, 8]. Animals from such highly selected populations will tend to be much more uniform in their response to a given reproductive hormonal stimulus or pharmacological treatment, which makes it much easier to detect significant differences between endocrine treatments. However, these commercially supplied strains of laboratory rodents are likely to reveal very little, if any, of the large amount of genetic variance in the reproductive responses found in the original populations [32–34]. Unfortunately, control populations not intensely selected for large litter size are generally unavailable for many of the commonly used outbred lines of laboratory animals. This lack of an unselected control population for most commercially available outbred laboratory animal stocks is in sharp contrast to genetic selection experiments where both selected and randomly selected lines, or divergently selected lines, are maintained continuously [2, 28, 35]. While some commercial suppliers have recently started selecting for more intermediate litter size and growth rates in their outbred stocks, such selection is highly unlikely to restore the genetic variance in reproductive responses found in the original populations.
In fact, the tendency for most researchers to use only a single highly selected outbred stock for their studies may have led to an underestimation of the amount of genetic variation in reproductive traits. Given the relatively large amount of genetic variance observed in ovarian response traits, comparison of several divergent genotypes, i.e., strains or lines that have not been highly selected in the same direction for the same trait, should be made before drawing conclusions about the hormonal or biochemical control of a reproductive trait.
During the quantification of genetic differences in HIOR, we developed animal models suitable for mapping genes that control major strain differences in HIOR and other reproductive traits. It was recognized at the outset of this process that identifying molecular genetic polymorphisms for mapping studies would be considerably easier in an inter-specific than an intra-specific cross. The Mus spretus line derived from Roswell Park had the lowest HIOR of any strain or line examined. Although the present study did not examine HIOR in interspecific crosses of Mus, the major 10-fold difference in HIOR of Mus spretus/Ros and the Mus musculus strain (B6) suggests that a cross between these strains may be suitable for mapping genetic differences in ovarian response to gonadotropins. While the HIOR of Mus spretus/Ros was about half that of the lowest highly inbred Mus musculus strain (A/J), this spretus population was not highly inbred. We have not yet characterized the physiological and biochemical control of ovarian function in the Mus spretus/Ros population. In contrast, several reproductive traits have been extensively studied in A/J and B6 strain mice including follicle populations, the incidence of follicular atresia, the induction of LH receptors, aromatase activity, ovarian steroidogenesis, and other parameters [9–11]. These well-characterized genetic differences in HIOR between A/J and B6 mice, and the indication that many of the component controls of follicular function have been conserved through mammalian speciation, led to adoption of a (B6xA)xA backcross model for mapping genes controlling major differences in HIOR. The results of this mapping study are presented in a companion paper [17].
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Jimmy Spearow, Section of NPB, Rm 196 Briggs Hall, University of California at Davis, One Shields Ave., Davis, CA 95616. FAX: 530 752 5582; jlspearow{at}ucdavis.edu
Accepted: June 3, 1999.
Received: January 5, 1995.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
location of publisher: Boston: Butterworths; 1985: 115–125.This article has been cited by other articles:
|
|
 |
|
  R. A. Davey and H. E. MacLean Current and future approaches using genetically modified mice in endocrine research Am J Physiol Endocrinol Metab, September 1, 2006; 291(3): E429 - E438. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Liljander, M.-A. Sallstrom, S. Andersson, P. Wernhoff, A. Andersson, R. Holmdahl, and R. Mattsson Identification of Genetic Regions of Importance for Reproductive Performance in Female Mice Genetics, June 1, 2006; 173(2): 901 - 909. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wang, W.-g. Ma, L. Tejada, H. Zhang, J. D. Morrow, S. K. Das, and S. K. Dey Rescue of Female Infertility from the Loss of Cyclooxygenase-2 by Compensatory Up-regulation of Cyclooxygenase-1 Is a Function of Genetic Makeup J. Biol. Chem., March 12, 2004; 279(11): 10649 - 10658. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Totsuka, Y. Nagao, T. Horii, H. Yonekawa, H. Imai, H. Hatta, Y. Izaike, T. Tokunaga, and Y. Atomi Physical performance and soleus muscle fiber composition in wild-derived and laboratory inbred mouse strains J Appl Physiol, August 1, 2003; 95(2): 720 - 727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T. Hentges and M. J. Low Ovarian Dependence for Pituitary Tumorigenesis in D2 Dopamine Receptor-Deficient Mice Endocrinology, December 1, 2002; 143(12): 4536 - 4543. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Spearow, P. A. Nutson, W. S. Mailliard, M. Porter, and M. Barkley Mapping Genes That Control Hormone-Induced Ovulation Rate in Mice Biol Reprod, October 1, 1999; 61(4): 857 - 872. [Abstract] [Full Text]
|
|
|
|
 |
|
 J. L. Spearow, P. Doemeny, R. Sera, R. Leffler, and M. Barkley Genetic Variation in Susceptibility to Endocrine Disruption by Estrogen in Mice Science, August 20, 1999; 285(5431): 1259 - 1261. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
