We genotyped 204 married Hutterite men for the Met470Val and intron 8 polyT and (TG)m repeat polymorphisms; allele and haplotype frequencies are shown in and . The CBAVD-associated 5T, TG12, and TG13 alleles are either absent (5T, TG13) or rare (TG12) in the Hutterites. On the other hand, the Val 470 allele occurs at high frequency (frequency 0.29). The Val allele resides on two haplotypes in the Hutterites, one common (TG11-7T-Val470) and one rare (TG12-7T-Val470).
Frequencies of the CFTR polymorphisms.
Frequencies of the CFTR haplotypes in the Hutterites.
To assess the effects of these polymorphisms on male fertility, we defined a measure of “birth rate” as the number of births per year of marriage in men with at least two children (see Materials and Methods
). Associations between the CFTR
alleles and haplotypes and birth rate were examined in Hutterite men using a regression-based test designed for large complex pedigrees, and which corrects for the relatedness between all pairs of men in this study 
. The results of the association studies are summarized in .
Results of association tests with birth rate in Hutterite men.
Homozygosity for the Met470 allele was associated with significantly lower birth rates in Hutterite men (P
0.0096; ), and accounted for 4.59% of the residual variance (after adjusted for covariates, see Materials and Methods
) in birth rate between males. The association remained significant when Val470 homozygotes (N
14) and heterozygotes (N
89) were combined (Model 2), consistent with a recessive effect of the Met470 allele on increased birth rates (P
0.0029; ). Both models remain significant after adjusting for multiple comparisons by a conservative Bonferroni correction (Model 1 Pc
0.038, Model 2 Pc
0.012). There was no significant association between alleles at the (TG)m locus and birth rate. A marginal association was observed between the 7T allele and increased birth rates (7T vs. 9T, P
0.060), but we attribute this to linkage disequilibrium (LD) with the Val470 allele (D'
CFTR Met470Val genotypes and birth rate in Hutterites men.
Nine men in this sample carried the M1101K mutation (none were ΔF508 carriers). In all cases, the M1101K mutation was on the Met470 background. Therefore, to remove the potential confounding effects of M1101K, we repeated our analyses after excluding these nine men. The association with Met470Val remained equally significant (Model 1 P
0.0059, Model 2 P
0.0020), suggesting that the observed fertility effects associated with Met470Val are not due to this pathogenic CFTR
On the other hand, the association with the Met470Val locus in Hutterite men is quite robust. shows the cumulative distribution of the number of years from marriage to each birth by genotype. On average, Met/Met men achieve each birth in more time than men with one or two copies of the Val allele. The difference between the means of the genotype groups increases with increasing birth number, reflecting a cumulative, positive effect of the Val allele (relative to Met/Met) on male fertility. The average effect of homozygosity for the Met470 allele compared to carrying one or two copies of the Val470 is a decrease of 0.049 births per year of marriage (). This corresponds to 0.56 fewer births over the course of an average reproductive period (11.5 [±5.0] years in this cohort). For example, Met470 homozygous men who are married 11.5 years or longer have a median of 7 children compared to 8 children in Val470 carrier men (Wilcoxon P
0.0002; ). Finally, the time required to achieve 6 births (the overall mean and median family size in our sample) is significantly longer for Met470 homozygotes (). The median time to having a sixth child is 11.9 years (upper, lower quartiles: 10.2, 14.0) among Met/Met men and 10.18 years (upper, lower quartiles: 8.7, 11.9) among Met/Val+Val/Val men (Log-rank P
Met470Val genotypes and fertility in Hutterite men.
We next attempted to replicate this association in another population that is also characterized by high natural fertility rates and large families, the Old Order Amish of Lancaster County, Pennsylvania 
. Three hundred fifteen Amish men, for whom reproductive histories were available, were genotyped for the Met470Val polymorphism. In this Amish population, the derived Val allele is the major allele, with a frequency of 0.65. As a result, only 37 men were homozygous for the Met allele. Consistent with results in the Hutterites, Met/Met men had lower birth rates (0.46±0.13 births/year) than Met/Val (0.50±0.14 births/year) or Val/Val (0.49±0.17 births/years) men (). This difference, however, was not statistically significant (P
0.22), most likely due to the small number of Met/Met homozygous men, and the corresponding lack of power. In addition, (TG)m and polyT genotypes were not available in the Amish population. Therefore, we can not rule out possible interactions with the haplotype background or independent effects of these repeat polymorphisms, especially if their allele frequencies are notably different from the Hutterites, as in the case of Met470Val polymorphism.
If the fertility effect associated with Met470Val genotypes in the Hutterites is generalizable, then the fitness advantage associated with the Val470 allele would be expected to leave a signature of positive selection on the pattern of variation at this locus 
. Therefore, we examined Met470Val genotype data from the International HapMap Project 
) and the Human Genome Diversity Project (HGDP) 
). The derived Val allele is very rare in sub-Saharan Africa (allele frequency ranges from 0 in Yorubans to 0.10 in Sans), whereas it occurs at high frequencies in non-African populations, and is even the more common allele in some European and Asian populations (reaching frequencies as high as 0.93 in Tuscans and 0.80 in Mongolians; ), as has been noted previously 
and as we observed in the Amish. The differences in the allele frequency distributions are also reflected in HapMap samples, where the Fst between the European (CEU) and Yoruban (YRI) populations is 0.43 (compared to genome-wide average of 0.11). Moreover, extended haplotype homozygosity (EHH) in the CEU population is apparent on the Val background compared to the Met background (). The integrated haplotype score (iHS), a measure of EHH 
, is −1.93 (genome-wide average is 0). Compared to genome-wide distributions in HapMap Phase 2 data, an Fst of 0.43 falls in the upper 3.3% (CEU vs. YRI) and an iHS of −1.93 falls in the lower 2% (CEU) of SNPs (). The fraction of SNPs in these data with an Fst ≥0.43 and an iHS ≤−1.93 is 0.003 (). The combined observations of a high frequency derived Val allele outside of Africa, a high Fst value, and a long EHH on haplotypes carrying the Val allele are suggestive of positive selection, and is consistent with the advantageous fertility effects of the Val allele relative to the Met allele, as observed in this study.
Geographic distribution of the Met470Val polymorphism in HGDP samples.
Population genetic parameters of the CFTR Met470Val locus.
Pompei et al. previously reported a lack of genetic variation in the CFTR
gene in carriers of the Val470 allele in healthy Europeans sampled from six different geographical areas in Europe 
, and speculated that the Val470 allele was under positive selection by conferring an advantage in the presence of pathogenic diseases. While we can not rule out that the Val470 allele confers resistance to pathogens, our study provides support for an alternate hypothesis: the Val470 allele rose to high frequencies outside of Africa due to a fertility advantage in carrier men. The fact that this allele is either absent or very rare in African populations further suggests either that the allele arose after early humans left African or that there is additional (negative) selection on the Val470 allele in certain (African) environments.
In fact, given the large fertility effects observed in the Hutterites, it is surprising that the Val470 allele has not gone to fixation in non-African populations. However, there might be several reasons why this has not occurred. First, the combined data on fertility effects of the Val470 allele indicate that this allele can be associated with both increased and decreased fertility, depending on genetic background. In the presence of the 5T allele at the intron 8 polyT locus, Val470 increases the risk of CBAVD and male infertility 
. In the absence of the 5T allele (as in the Hutterites), the Val470 allele is associated with increased male fertility relative to Met470. Although the mechanism of this interaction is obscure, it provides one example of counteracting variation that could increase the time to fixation of the Val470 allele. Second, as mentioned above, the Val allele could also be deleterious in certain environments, such as in the presence of specific pathogens or the 5T allele, as a result of its pleiotropic effects in other organ systems. Third, the fertility advantage we observed is restricted to males; we found no such association in Hutterite women (data not shown). This would further slow the spread of the allele as there would be no selection advantage in half of all Val carriers. Lastly, this study was conducted in a population living under optimal conditions for reproductive success, including excellent nutrition and abundant food, access to modern health care, and negligible maternal mortality. Thus, estimates of fitness effects based on Hutterite fertility rates are likely inflated compared to the effects in human populations throughout most of evolutionary history, when competing selective pressures were likely more prevalent. Taken together, the lack of fixation of the Val470 alleles in populations outside of African may not be inconsistent with the fertility effects observed in the Hutterites, but rather suggestive of antagonistic effects of other genetic variations or environment factors that tempered these effects during most of human evolution.
To our knowledge, this is the first report demonstrating that a common variation in the CFTR
gene influences reproductive fitness in fertile, healthy men. Nearly all previous studies on CFTR
mutations and reproduction in males have focused on patients with infertility. Increased prevalences of CFTR
mutations in men with reduced sperm quality, with azoospermia without CBAVD, and with isolated CBAVD have been reported 
, suggesting the involvement of CFTR in sperm production and development 
. Moreover, heterozygous Cftr+/−
mice have reduced sperm fertilizing capacity and lower overall fertility 
. Although little is understood about the physiological role of CFTR protein in the normal male reproductive system 
, it is known that the reproductive tissues are more sensitive to changes in CFTR function 
. It is, therefore, possible that subtle differences in CFTR conductive properties between the Met and Val alleles may result in changes in the fluid environment of male reproductive tract, which would eventually lead to differences in sperm transport activity, morphology or quality 
, and could account for the observed fertility differences reported here. On the other hand, it is possible that the fertility effects of the Met470Val polymorphism described in this study are unique to the Hutterites and would not be replicated in other populations with measures of natural fertility and large family sizes. However, combined with the evolutionary signatures at this locus, the consistent (if not significant) results in the Amish, and the plausible biological mechanism, we believe that our data provide support for at least one specific variant in the CFTR
gene influencing natural variation in fertility in healthy men.
Lastly, there has been a long-standing debate as to whether disease-causing CF mutations, such as ΔF508, confer a fertility advantage to healthy carriers (for example see Danks et al. 
). Unfortunately, the results we report here do not provide insight into this question. The most common CF causing mutations in Europeans (i.e. ΔF508, G542X, N1303K, W1282X) and the most common mutation in the Hutterites, M1101K 
, all reside on haplotypes carrying the ancestral, Met470 allele in exon 10 
, the 9T allele at the polyT locus, and (by inference) the TG10 or TG11 alleles at the (TG)m locus in intron 8 
. Therefore, any positive fertility effects of the Val470 allele would not be expected to affect the frequencies of the common CF disease-causing mutations in European populations.
In conclusion, the combined observations of high levels of variation in the CFTR gene, decreased fertility among CF patients and some CF carriers, and our observation of lower fertility associated with homozygosity for the Met470 allele in healthy men suggest that there are multiple independent, and possibly competing, evolutionary forces acting on the CFTR locus. The modifying effects of the haplotype background (i.e., 5T) on specific variants further imply important epistatic interactions between variants in the CFTR gene. Lastly, the high frequency of Val470 outside of Africa raises the possibility of interaction between CFTR alleles and changing environmental conditions. Thus, understanding the complex evolutionary history of the CFTR gene may require detailed studies of variation in worldwide samples of patients with CF and CF-related disorders, as well as healthy individuals. Regardless, this gene continues to provide surprises and represents outstanding examples of epistasis, in which the same allele (e.g., Val470) can have beneficial or deleterious effects depending on genetic background, and of a locus influenced by both positive (due to fertility advantage) and negative (due to CF and CF-related phenotypes) selection.