| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Kinesiology, University of North Carolina-Charlotte, Charlotte, North Carolina; 2Malaria Research Institute, Johns Hopkins University, Baltimore, Maryland; 3Laboratory of Respiratory Biology, National Institute of Environmental Health Sciences, Durham; and 4Department of Biology, University of North Carolina-Charlotte, Charlotte, North Carolina
Submitted 22 November 2006 ; accepted in final form 2 April 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
candidate genes; aerobic capacity; genotyping
51% in humans (3) to 58–90% in mice (23, 26) to 39% in rats (18). However, the specific genes that determine an individual's maximal exercise endurance have not been identified. Relatively few investigations of the quantitative trait loci (QTLs) associated with maximal exercise endurance have been conducted. Two separate human studies by Bouchard and colleagues (4, 33) have identified possible genetic linkages with sedentary aerobic capacity on chromosomes 4, 8, 11, and 14 in a white population and on chromosomes 1 and 7 in a black population. Ways et al. (36) identified a significant QTL linked with aerobic running capacity on rat chromosome 16 and suggestive linkage on chromosome 3. Given the health implications of understanding the genetic regulation of a complex trait such as maximal exercise endurance, it is important that additional research be conducted in species that will allow the use of more powerful genetic tools. We established a foundation for such studies with our previously described strain distribution pattern of maximal exercise endurance in 10 different mouse inbred strains (26). Given earlier studies that have shown a genetic contribution to the regulation of exercise endurance, we hypothesized that several QTLs associated with maximal exercise endurance would be identified in inbred mice using progenitor strains earlier identified by their differential response to treadmill exercise.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
All procedures in this protocol were reviewed and approved by the University of North Carolina-Charlotte Institutional Animal Care and Use Committee and conformed to the humane animal care policies of the U.S. Department of Agriculture.
Model Used
We earlier identified Balb/cJ and DBA/2J mice as having high exercise endurance (HEE) and low exercise endurance (LEE), respectively (26). Ten female Balb/cJ mice were bred with nine male DBA/2J mice to produce BaD2 F1 mice. We chose the Balb/cJ mice as the maternal progenitor strain because earlier studies had suggested that exercise endurance was a maternally linked characteristic (11, 24). Additionally, the number of progenitor mice used was based on the statistical power (power = 0.99) of an earlier comparison of five mice per strain (26). F1 mice were weaned at 21–28 days and crossed between 6 and 8 wk of age when the mice had reached 20–25 g in weight. The F2 mice (n = 99; 60 female, 39 male BaD2 F2 mice) were weaned at 21–28 days and phenotyped for exercise endurance between 45 and 68 days of age. All mice were housed individually without access to running wheels (i.e., in the sedentary condition) in the same vivarium room that was maintained at 65–75°F and 20–40% humidity with a 12:12-h light-dark cycle. Mice were fed standard rodent chow (Harland Teklad 8604 Rodent Diet, Madison, WI; protein 24.5%, fat 4.4%, fiber 3.7%, nitrogen-free extract 48.6%) and provided water ad libitum.
Estimation of Exercise Endurance
The exercise endurance of the F2 mice was estimated using techniques described previously (26). In brief, after a 1-wk vivarium acclimation, the mice were oriented twice to running in a small cell suspended above a human treadmill. A shock grid was mounted at the rear of the cell (1.5 mA, 150 V) to provide stimulus for the animals to run (35). During testing, each mouse was placed in the treadmill cell and allowed 3 min of rest. The treadmill was then started at 22 m/min and increased in speed every 4 min thereafter to 28.4, 36.6, and 42.4 m/min, respectively. If the mouse completed 4 min at 42.4 m/min, the speed was held constant at 42.4 m/min, but the grade was increased 2% every 4 min thereafter. The grade of the treadmill was increased instead of further increasing the speed because it appeared in pilot testing that some mice could not biomechanically maintain running speeds above 45 m/min. The treadmill test was stopped when the mouse could not get off the shock grid within 3 s, and exercise endurance was estimated by total time that the mouse ran. This indirect measure of exercise endurance has been used as a phenotype in other rodent genetic studies (1, 6, 18, 26, 36), and running speed has been positively correlated with oxygen consumption in mice (13).
Genotyping
After phenotyping, the mice were anesthetized using 2–4% inhaled isoflurane. When the animal showed no response to the paw pinch test, the heart was exposed and removed (i.e., euthanasia by exsanguination). The kidneys were then harvested, flash frozen in liquid nitrogen, and stored at –80°C for later genotyping. Initially, we selectively genotyped the 12 mice with the highest exercise endurance (9 females, 3 males) and the 12 mice with the lowest exercise endurance (8 females, 4 males). It has been noted that genotyping the upper 10% and lower 10% of the phenotypic distribution will capture approximately 80–85% of the genetic variance (8, 30). Our selectively genotyped subsample represented 24% of the total sample and the upper and lower 12% of the phenotypic strain distribution pattern.
DNA extraction from the kidneys and genotyping of the selective subsample were performed as previously described (16). The primers for the polymorphic simple sequence repeats markers (i.e., microsatellites) were selected from the Mouse Genome Database (MGD) (2) and were purchased from Research Genetics (Huntsville, AL). One hundred eight microsatellite markers, with an average 12.6-cM (SD 8.9) distance between them, provided coverage of the mouse genome with >95% confidence (34). One primer of each pair was 32P end-labeled. After amplification, the PCR products were resolved on 6% acrylamide gels. QTL locations within the selectively genotyped subsample were then estimated using statistical procedures. All animals (n = 99) in the F2 cohort were then genotyped around the identified QTL using informative microsatellites spaced as close together as possible to increase map resolution (i.e., fine mapping).
Statistics
Prior to the QTL analysis, a one-way ANOVA was used to test whether sex had a significant effect on exercise endurance in the F2 mice. Simple regression also was used to test whether body weight and exercise endurance were correlated to determine if exercise endurance should be adjusted for the effects of body weight. Exercise endurance for mice in the F2 population and in the two inbred populations (DBA/2J and Balb/cJ) was compared using a one-way ANOVA with Scheffé's post hoc analysis. Pre-QTL analyses were completed using JMP 5.1 (SAS, Cary, NC), 
values were set a priori at P < 0.05, and results are presented as means (SD).
Methods described elsewhere (20) were used to search for potential QTLs for maximum exercise endurance in the selective subsample of F2 progeny and in the subsequent full sample set, and the results were confirmed with MapManager QTX (28). In brief, the phenotypic data were first checked for normality. If the data were not normally distributed, the data were then log-transformed, resulting in a normal distribution of the exercise endurance scores. These scores were subsequently analyzed for linkage using the regression approach to interval mapping (15). For each chromosome, the regression analyses generated F values with their associated probabilities. These probabilities were converted to logs of the odds favoring linkage (i.e., LOD scores) by logarithmic transformation [LOD = log10(1/probability)]. If the highest LOD calculated for a given chromosome exceeded a specific threshold value, a QTL was considered to be present at the position associated with that LOD score. Threshold values were obtained from permutation tests involving 10,000 separate runs in which the aerobic capacity values for each individual mouse were randomly permuted and then run through the interval mapping analysis. These permutation runs provided 5% threshold values for each chromosome as well as a 5% experiment-wise threshold value applicable across all chromosomes (37). LOD scores exceeding the chromosomewise threshold values are suggestive QTLs, whereas LOD scores exceeding the experimentwise level provide significant evidence for QTLs (19).
Confidence intervals for each QTL were determined using the method of Darvasi and Soller (9). If a QTL was found on any chromosome, a test for the presence of two QTLs was conducted (20). The QTL analysis was repeated using program-determined linkage distances, as well as linkage distances from the MGD (2) and the Whitehead Institution database (10) using several different models to determine the dominance properties of identified QTLs. The models used were a "free" model where the additive and dominant components are allowed to have any value, an additive model where the dominant component is set to zero, a dominant model where a single coefficient is used for both the additive and dominant components, and a recessive model where the additive and dominant components are the same coefficient, but with opposite signs. Additionally, the data were checked for sex or breeding cross-direction differences in the phenotype to control for possible inflation of the LOD scores on the nonautosomal X chromosome (5).
Sex-specific QTL effects were tested by first assigning a code for each sex (males = 1, females = 2). A whole genome scan was then run as before but with multiple regression of exercise endurance on the interaction of the genotypic values with sex. Threshold values again were estimated by permutation tests, and chromosomes with significant LOD values were assumed to contain a QTL whose effects differed in the two sexes. For those chromosomes, separate QTL runs were made for males and for females, and new chromosomewise and experimentwise threshold LOD values were obtained for significance testing.
In all QTL analyses, if the P value resulting from the comparison of the LOD score with the calculated specific threshold value was less than 0.05, but greater than 0.01, the QTL was considered "suggestive." If the P value was less than 0.01 but greater than 0.001, the QTL was classified as "significant," and if less than 0.001, considered "highly significant." These analyses and classification of P values follow the standardized nomenclature from Lander and Kruglyak (19).
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Body weight for the F2 female mice [16.33 g (SD 3.36)] was less (P < 0.001) than that for the F2 male mice [21.0 g (SD 3.10)]. Body weight of progenitor DBA/2J [n = 9 males; 15.44 g (SD 1.74)] and Balb/cJ mice [n = 10 females; 16.5 g (SD 1.72)] was significantly less (P < 0.05) than the F2 male mice but did not differ (P > 0.05) from that of the F2 female mice. Despite the difference in body weight between sexes in the F2 cohort, no significant difference was found between sexes in exercise endurance [males = 8.02 min (SD 5.86), females = 9.33 min (SD 6.90); P = 0.33], but the statistical power of this comparison was relatively low (power = 0.16). Neither body weight (r2 = 0.0007) or age when phenotyped (r2 = 0.063) correlated with exercise endurance.
Mean exercise durations for DBA/2J and Balb/cJ mice (Fig. 1) were similar to those noted previously (26), where Balb/cJ mice ran significantly longer (P = 0.005) than DBA/2J mice. Exercise duration for the BaD2 F2 [8.82 min (SD 6.51)] animals was not significantly different from the LEE DBA/2J mice [4.82 min (SD 3.07), P = 0.08] but was significantly less than the HEE Balb/cJ mice [14.06 min (SD 8.15), P = 0.02, power = 0.81]. The frequency distributions of exercise duration in the F2 mice were shifted toward the DBA/2J parental average (Fig. 2) and exhibited significant skewness [1.67 (SD 0.24)] and kurtosis [3.101 (SD 0.49)]. The application of the Shapiro-Wilk (W = 0.847, P < 0.0001) and the Komogorov-Smirnov (D = 0.147, P < 0.01) tests for normality indicated that the distribution was not normal; thus the exercise endurance data were log-transformed for the QTL mapping analysis.
|
|
Selective mapping. The 12 highest endurance mice [endurance = 22.8 min (SD 6.3)] and the 12 lowest endurance mice [endurance = 1.77 min (SD 0.7)] were selectively genotyped. The body weights of the 12 highest and 12 lowest endurance mice were not different [HEE = 19.5 g (SD 0.89) vs. LEE = 17.8 g (SD 0.89), P = 0.18]. Because there were no sex differences in phenotype and because the maternal progenitors all derived from Balb/cJ (i.e., no breeding cross-direction differences in phenotype), the LOD scores from the X-chromosome were not adjusted (5). Results of the selective QTL mapping for exercise endurance in the 12 highest and 12 lowest endurance mice revealed two potential QTLs, one on chromosome 8 in females only and one on chromosome X in all mice. The significant QTL on chromosome X (designated Exercise Endurance on chromosome X; EEX) was located at 59.6 cM, 6 cM beyond the DXMit121 marker, and achieved a LOD score of 2.45. A suggestive (LOD score = 2.21) QTL in females (designated Exercise Endurance on chromosome 8 in females; EE8f) was found on chromosome 8 at 42.1 cM, 6 cM beyond the D8Mit359 marker.
Fine mapping. All 99 F2 animals were genotyped for fine mapping. Thirteen microsatellite markers surrounding the two QTLs were used for fine mapping and were spaced an average of 2.08 cM (SD 1.51) apart. Significant segregation distortion was present in the markers surrounding the QTL for chromosome X; thus QTL analysis methods that allow for segregation distortion were used (27). From this analysis, our previous selective-genotype QTLs were confirmed. Significant linkage with exercise endurance for all mice was found at 57.9 cM (DXMIT31) on chromosome X (EEX; Table 1). Further suggestive linkages were found for D8MIT359 (36.1 cM; EE8F) in the female mice (using both additive and recessive models) and in the region of the chromosome X from 53.6 to 56.8 cM (DXMIT121, DXMIT5, DXMIT236) in all mice using all statistical models.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Investigators have used both human and rat models in previous attempts to identify the chromosomal regions linked with maximal exercise endurance. Bouchard et al. (4) phenotyped 481 Caucasian subjects using a volitional fatigue cycle ergometer protocol. The autosomal chromosomes were then genotyped using 289 microsatellite markers with an average spacing of 11 cM. The investigators found four significant QTLs linked with maximal exercise endurance (4q12, 61.658 cM; 14q21.3, 49.255 cM; 8q24.12, 128.157 cM; and 11p15.1, 21.174 cM) and an additional 21 regions that were suggestive of QTL linkage. Rico-Sanz and colleagues (33) extended this work and using similar methods genotyped 862 subjects of white and black ethnicity. They reported "promising linkage" of sedentary 
O2 max on chromosome 11 (11p15.1) in the white subjects and four suggestive QTLs for O2 max for the black subjects on chromosomes 1 and 7 (1p31, 7q32, 7q36, 7q36) but no significantly linked QTLs. Ways and colleagues (36) identified three QTLs linked with exercise endurance in an F2 rat model (n = 224) derived from Copenhagen (COP) and DA inbred progenitors. On chromosome 16, they found significant linkage of exercise endurance with D16Rat17, suggestive linkage with D16Rat55, a marker located 30 cM distal to D16Rat17, and suggestive linkage with D3Rat56 on chromosome 3.
Maximal exercise endurance is a complex trait regulated by many genetic and environmental factors. While this study has identified two QTLs associated with exercise endurance, including one that suggests a sex influence, it is probable that no studies have identified all of the QTLs that play a role in the regulation of maximal exercise endurance. However, one of the strengths of genetic/QTL studies is the conserved gene synteny among species that allows the comparison of identified genetic loci in multiple models. Thus the exercise endurance QTLs and nearby markers can be compared directly to similarly identified QTLs and markers derived from other species (e.g., humans or rats). For example, the significant D16Rat17 exercise endurance QTL identified by Ways et al. (36) is homologous to our mouse EE8F QTL. While none of the mouse exercise endurance QTLs identified in the present study were homologous with the significant human QTLs reported by Bouchard's group, EE8F is homologous with two "potentially useful" linkages that Bouchard et al. identified on human chromosome 13 (D13S787; 13q12.11) and chromosome 22 (D22S274, 22q13.2) (4). At this time, no comparisons between our EEX QTL and the results of either Bouchard et al. (4) or Ways et al. (36) are possible since neither investigation reported results from chromosome X.
The strength of any QTL investigation is determined by multiple factors; three of the most important are the variability of the phenotype, the effect size, and the population size. While the variance of the exercise endurance of the F2 population was larger than the variances observed in our earlier inbred strain screen study (26), we had anticipated a larger variance in the F2 animals vs. the inbred strains given the heterozygous genome in these animals. Otherwise, there do not appear to be explanations for the variation we observed given that the mice in this study were tested at the same time of year as our strain screen mice and were housed in similar circumstances and in the same location, and the testing was conducted in a similar fashion by the same technicians using the same equipment as was used in the earlier study (26). Li et al. (25) have noted that although an estimated 60% of the F2 mouse mapping experiments in the literature used fewer than 300 individuals, smaller cohorts of mice decrease the chances of finding any but the strongest QTLs. Although our sample size was relatively small, our observed effect size was large as there was a 12.9-fold difference in maximal exercise endurance between the two selectively genotyped groups. In addition, the homology of our QTLs with those reported by Ways et al. (36) and Bouchard and colleagues (4) further strengthens the validity of our QTLs. As noted in the consensus statement by the Complex Trait Consortium, "there is no single ‘gold standard’ for the identification of a QTL...rather, there should be a predominance of evidence that supports its identity" (29). Thus, given that our genotyping plan in these inbred mice theoretically captured 80–85% of the variability in the cohort (8, 30), we are confident that these results significantly add to the currently sparse literature investigating the underlying genetic regulating mechanisms of exercise endurance and provide a further foundation for the identification of candidate genes linked to exercise endurance.
Identification of candidate genes in the past has been speculative, consisting primarily of searching for known genes in the QTL region that are involved in the basic physiological regulating mechanism(s) of the phenotype of interest. Due to the multifactorial nature of exercise endurance control and because the number of genes located in each QTL region is large, candidate gene lists generated from any species are often quite lengthy (see discussion in Refs. 4, 32, 36). Similarly, a large number of candidate genes that could be involved in the regulation of maximal exercise endurance are located within 15 cM of the QTLs identified in this study; however, discussion of any of these "candidate genes" from the genes located within the identified QTL without further corroborating evidence such as microarray results would be speculative at best. Thus the future challenge is to narrow the identified QTLs to a more manageable size, thus enabling the pursuit of more precise identification of the location and identity of genes involved in the regulation of exercise endurance (17).
While not its primary function, the identification of QTLs also provides a means for investigating the validity of potential candidate genes influencing exercise endurance that have not been determined through genetic mapping studies. For example, there has been much investigation regarding the role of angiotensin-converting enzyme (ACE) genes (14), mitochondria synthesis genes (38), and the peroxisome proliferator-activated receptor (PPAR) genes (12) in regulating exercise endurance. These genes (ACE, PPAR, and mitochondrial synthesis) are not found in QTL regions thus far associated with exercise endurance in humans (4, 33), rats (36), or mice (present study). Since most investigators agree that the heritable regulation of maximal exercise endurance is likely polygenic, it is probable that genes within and external to the identified QTLs participate in this system. Therefore, the use of a variety of techniques that provide evidence for the involvement of specific genes as well as further work narrowing the involved chromosomal areas need to be completed before meaningful candidate gene investigations can commence.
In summary, we developed, phenotyped, and genotyped F2 mice derived from Balb/cJ (high exercise endurance) and DBA/2J (low exercise endurance) progenitors. Significant linkage of maximal exercise endurance for all mice was found on chromosome X (EEX; 57.9 cM) and a suggestive linkage was found in the female mice on chromosome 8 (EE8F; 36.1 cM). EE8F was homologous to exercise endurance QTLs found in the rat (D16Rat17) and human models (D13S787 and D22S274). These data provide a foundation for further study of possible candidate genes located within each QTL region that influence exercise endurance.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
O2 max in the sedentary state: the HERITAGE family study. Med Sci Sports Exerc 30: 252–258, 1998.O2 max, and response to endurance training. Med Sci Sports Exerc 23: 177–185, 1991.This article has been cited by other articles:
|
|
 |
|
  H. S. Yang, M. H. Vitaterna, A. D. Laposky, K. Shimomura, and F. W. Turek Genetic analysis of daily physical activity using a mouse chromosome substitution strain Physiol Genomics, September 1, 2009; 39(1): 47 - 55. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Copp, R. T. Davis, D. C. Poole, and T. I. Musch Reproducibility of endurance capacity and VO2peak in male Sprague-Dawley rats J Appl Physiol, April 1, 2009; 106(4): 1072 - 1078. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Lightfoot, M. J. Turner, D. Pomp, S. R. Kleeberger, and L. J. Leamy Quantitative trait loci for physical activity traits in mice Physiol Genomics, February 19, 2008; 32(3): 401 - 408. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
