Infertility in Females with Cystic Fibrosis Is Multifactorial: Evidence from Mouse Models

Abstract

Infertility is commonly associated with cystic fibrosis (CF). Although infertility in men with CF has been thoroughly investigated, the infertility observed in women with CF has not been well studied. To investigate female infertility associated with CF, we used two independently derived mouse models of CF. Both of these models displayed decreased fertility characterized by a reduction in litter number and litter size. Our findings suggest that much of the reduced fertility in these mice originates from decreased fertilization due to inadequate sperm transport within the female reproductive tract. However, our data indicate that additional reproductive phenotypes in the CF female mice also contribute to the reduced fertility including small ovarian and uterine size, aberrant estrous cycles, and decreased oocyte ovulation rates. These data, along with previous work demonstrating that the gene mutated in CF, the cystic fibrosis transmembrane conductance regulator (CFTR), is normally expressed in tissues vital to reproduction, raises the possibility that CFTR may have a direct effect on fertility. If so, CFTR may also play an important role in normal female fertility within the general population.

CYSTIC FIBROSIS (CF) is the most common life-shortening genetic disease in Caucasians and is caused by mutations in the gene encoding a cAMP-regulated chloride channel, the CF transmembrane conductance regulator (CFTR) (1,2). CF is a systemic illness that affects various organ systems including the pulmonary, endocrine, epithelial, gastrointestinal, pancreatic, immune, and reproductive systems (1,3). As clinical advances over the last 2 decades have resulted in increasing longevity, reproductive function in both men and women with CF has become more important. Infertility in men with CF has been extensively studied and found to be due to malformation or atrophy of the vas deferens leading to congenital bilateral absence of the vas deferens (4). Reduced fertility has also been observed in women with CF (5,6), but the underlying causes for this infertility remain unclear. The prominent hypothesis for the decreased fertility in CF females is viscous mucus in the cervix that may create a barrier to sperm passage (6,7,8). Whereas abnormal cervical mucus has been observed in CF women (6,7), no functional tests have been completed to observe sperm passage in the CF reproductive tract or determine whether abnormal mucus fully explains the decreased fertility. Recent studies suggest a possible role for CFTR in the normal functioning of multiple female reproductive tissues (9,10,11) as well as the hypothalamic-pituitary-gonadal (HPG) axis (12,13,14,15,16). Abnormalities in these systems could also contribute to the impaired fertility in women with CF.

CFTR is expressed throughout the female reproductive tract in the cervix, ovary, oviduct, and uterus (9,10,11). Whereas the function of CFTR in these tissues remains unclear, much of the available research has focused on the control of fluid transport and its contribution to fertility. Murine Cftr expression in the uterus is regulated by ovarian hormones with increasing expression in response to estrogen and decreasing expression in response to progesterone, a pattern that correlates with cyclic changes in uterine fluid (11,17,18,19). Further involvement of CFTR in the regulation of fluid throughout the reproductive system is supported by its suggested involvement in two conditions in the general population: hydrosalpinx and ovarian hyperstimulation syndrome (20,21,22). Both of these conditions are detrimental to fertility due to a high influx of fluid within the reproductive tract, and higher expression of CFTR in the affected reproductive tissues has been observed in both disorders (21,22). Finally, the viscous mucus that is hypothesized to be the origin of female infertility in CF patients may also be attributed to inadequate fluid regulation in the cervix. Without properly functioning CFTR, a decrease in fluid across the cervical epithelium could cause mucus accumulation and lead to decreased sperm transport, analogous to models of gastrointestinal and airway mucus accumulation in CF.

Female CF patients display endocrine abnormalities that suggest that impaired HPG axis function could also play a role in the observed infertility. Delayed puberty is common in young women with CF, with an average age of menarche of 14.9 vs. 13.0 yr in controls (23). Systemic illness and compromised nutritional status clearly play a role in delayed puberty (24), but it has been noted that CF patients can experience delayed puberty, even with good clinical and nutritional status (23). Hormone changes have been observed in CF female adolescents who displayed reduced estradiol and FSH levels (25) and CF female adults who displayed increased testosterone and reduced estradiol and progesterone levels, compared with age-matched controls (26). CF has also been associated with menstrual irregularities including amenorrhea, irregular cycles, and anovulation observed in women (26,27). Interestingly, abnormalities in the reproductive endocrine axis have been viewed as an indirect consequence of CF and have been largely ignored as possible contributors to the observed female infertility (27).

Mouse models of CF can help determine which factors contribute to female infertility associated with CF. Although not the focus of previous research, mouse studies provide anecdotal evidence suggesting that CF female mice are subfertile (28,29,30,31,32). To test the hypothesis that CF female mice are subfertile and that abnormalities other than excess cervical mucus contribute to the decreased fertility, we investigated two independently created mouse models of CF. Mice from both models displayed decreased fertility evidenced by fewer litters and fewer pups per litter than non-CF mice. Our findings support the hypothesis that sperm transport is impaired in the female CF reproductive tract leading to decreased oocyte fertilization, but we also observed decreased size of reproductive organs, abnormal estrous cycling, and aberrant ovulation patterns indicating that additional abnormalities likely contribute to the decreased fertility observed in women with CF.

Materials and Methods

Mouse strains

The CF mouse models (Cftr^−/−) used in these studies contained either the Cftr^tm1Unc mutation (referred to as the S489X mutation in this manuscript), which generates a stop codon in the coding region of exon 10 of Cftr (32) (Jackson Laboratory, Bar Harbor, ME; stock no. 002196), or the ΔF508 mutation, which generates a 3-bp deletion of Cftr in exon 10 and results in the loss of a phenylalanine residue (33) (a gift from Dr. Kirk Thomas, University of Utah, Salt Lake City, UT). Both mutations have been bred to be congenic on the C57BL/6J genetic background.

Animal housing

All animals used in this study were cared for according to a Case Western Reserve University approved protocol and Institutional Animal Care and Use Committee guidelines. Animals were housed in standard polysulfone microisolator cages in ventilated units with corncob bedding. Mice were given ad libitum access to irradiated chow (Harlan Teklad 7960; Harland Teklad Global Diets, Madison, WI) and either sterile water or half-concentration Colyte solution (Schwarz Pharma, Madison, WI) to prevent intestinal obstruction in Cftr^−/− animals. All animals were maintained on a 12-h light, 12-h dark schedule at a mean ambient temperature of 22 C.

To assess fertility quantitatively, mating pairs of wild-type males crossed with 10- to 12-wk-old wild-type or Cftr^−/− females were continuously mated for 5 months and monitored daily for litters born and number of pups.

Evaluation of estrous cycle

Estrous cycling was assessed using 14- to 16-wk-old females housed in cages of three to four females per cage. Vaginal smears were collected daily between 0900 and 1100 h for 20 consecutive days. The stage of estrous cycle was determined as follows: proestrus (epithelial cells), estrus (cornified epithelial cells), metestrus (cornified epithelial cells and leukocytes), and diestrus (majority leukocytes) (34). Estrus frequency and cycle length were compared among groups.

Tissue collection, histology, and hormone measurements

Female mice aged either 6–8 or 14–16 wk were weighed, killed, and reproductive organs collected. Tissues for strain comparisons were taken at the same stage of the estrous cycle: uterine and cervical sections were taken from all mice during estrus and reproductive organ weights and ovarian sections were taken from all mice during diestrus. Ovaries, uteri, and cervix were isolated and fixed in 10% formalin. Uteri and cervix were sectioned and hematoxylin and eosin stained and inspected for the presence of physical blockage or mucus. Ovaries were sectioned every 8 μm, hematoxylin and eosin stained, and placed on glass slides. To estimate the number of corpora lutea (CL), sections were used to count each CL, and the CL was followed through consecutive sections to avoid double counting (35). To ensure appropriate CL stage comparison between wild-type and mutant ovaries, CL counts were calculated from mice at diestrus that had undergone estrus in the previous 5 d.

To measure circulating hormone levels, blood was collected from 14- to 16-wk-old wild-type and Cftr mutant females late in the afternoon during proestrus. FSH and LH levels were measured by the National Hormone and Peptide Program (Dr. A. F. Parlow, Torrance, CA) using mouse LH and FSH RIA kits.

Exogenous hormone treatment and in vivo and in vitro fertilization assessment

Superovulation was achieved by injecting 14- to 16-wk-old wild-type and Cftr^−/− females with pregnant mare serum gonadotropin (ip, 5 IU/mouse) and 48 h later injecting with human chorionic gonadotropin (hCG; ip, 5 IU/mouse) (36). Oviducts of plugged females were flushed with saline 48 h after hCG injection and embryo stages were recorded (36). For in vitro fertilization assessment, superovulated females were mated with wild-type males after hCG injection; oocytes were retrieved from the oviducts 12–13 h after hCG injection, placed in a 500-μl drop of human tubal fluid (HTF) medium (Irvine Scientific, Santa Ana, CA), covered with mineral oil and placed in an incubator at 37 C in an atmosphere of 5% CO₂ in air. Sperm was collected from the vas deferens of a wild-type male mouse, placed in 1 ml HTF at 37 C in an atmosphere of 5% CO₂ in air for 20 min. Sperm counts were completed and 15 × 10⁶ sperm/ml was added to the previously collected oocytes. The oocytes were washed 6 h later in three drops of HTF and placed a clean drop of HTF covered with mineral oil and placed in the incubator. Fertilization rate was assessed the next morning by counting the number of two-cell embryos.

Oviductal sperm number and capacitation assessment

Wild-type and Cftr^−/− females were mated to wild-type males. Oviducts from females with vaginal plugs were flushed with M2 medium (Sigma, St. Louis, MO), and sperm number was determined. Sperm were assessed for the presence of acrosomes by staining with Coomassie brilliant blue as described by Feng et al. (37). Briefly, sperm were air dried on glass slides and fixed with 5% paraformaldehyde for 15 min. The slides were washed with PBS and stained for 5 min in 0.25% Coomassie brilliant blue in 10% glacial acetic acid and 25% methanol and rinsed in water. The presence or absence of the acrosome was assessed in 200 sperm per animal. If fewer than 200 sperm were observed, every sperm was analyzed.

Data analysis

Data are displayed as means ± sd. Because it is not known whether all variables assessed followed a gaussian distribution, to be conservative nonparametric Mann-Whitney U tests were used for comparisons of litter number, litter size, estrous cycle number and length, reproductive weights, and CL numbers. A t test was used for comparison of hormone measurements. A χ² test was used for comparisons of mice without estrus cycles and mice without litters. Statistical significance was attributed to P < 0.05.

Results

Fertility in Cftr mutant females

To determine the effect of the absence of Cftr on female fertility, two independently derived mouse models of CF (S489X and ΔF508) and wild-type females were mated to wild-type males for 5 months and examined daily for subsequent birth of litters and number of pups. Both CF models displayed decreased fertility with fewer numbers of litters and smaller litter sizes, compared with wild-type females (Table 1). Both CF models also displayed an increase in females unable to give birth (29%; P < 0.05) over the observed time period (Table 1).

Table 1.

Fertility of wild-type and Cftr mutant females

Mouse type	Mice, n	Females with no litters, n (%)	Average litter per month	Average no. of pups/litter
Wild type	18	0 (0)	0.92 ± 0.13	6.56 ± 2.36
ΔF508	14	5 (35.7)	0.32 ± 0.26a	3.81 ± 1.43a
S489X	10	2 (20)	0.36 ± 0.36a	3.55 ± 1.92a
Both mutations	24	7 (29.2)	0.34 ± 0.30a	3.70 ± 1.61a

^aP < 0.001 vs. wild type. 

Because there were no appreciable differences in fertility or other reproductive phenotypes between the two CF models (Tables 1–7), they will be collectively referred to as Cftr^−/− in the following experiments.

Table 2.

Uterine and ovarian size in wild-type and Cftr mutant females

Mouse type	Mice, n	Average ovarian weight (mg)	Average uterine weight (mg)	Average mouse weight (g)
6–8 wk
Wild type	7	3.1 ± 0.6	61.4 ± 9.8	19.0 ± 1.45
ΔF508	6	1.4 ± 0.2a	27.4 ± 6.4a	16.4 ± 1.59b
S489X	6	1.8 ± 0.6a	26.6 ± 14.3a	15.9 ± 1.46a
Both mutations	12	1.6 ± 0.5a	27.0 ± 10.6a	16.1 ± 1.48a
14–16 wk
Wild type	11	4.2 ± 1.1	74.8 ± 11.8	21.6 ± 1.92
ΔF508	6	2.6 ± 0.6a	45.8 ± 20.5b	18.6 ± 0.88a
S489X	8	2.8 ± 0.6a	50.6 ± 17.1b	19.4 ± 1.33a
Both mutations	14	2.7 ± 0.6a	48.2 ± 18.2b	19.0 ± 1.11a

^aP < 0.005 vs. wild type. 

^bP < 0.05 vs. wild type. 

Table 3.

Characterization of estrous cycle in wild-type and Cftr mutant females

Mouse type	Mice, n	Mice with no estrus, n (%)	Total no. of estrus	Avg no. of cycles/mousea	Cycle lengtha
Wild type	14	0 (0)	58	4.14 ± 0.86	5.11 ± 1.58
ΔF508	18	4 (22.2)	29	2.07 ± 1.20b	11.79 ± 5.64b
S489X	12	5 (41.7)	16	2.29 ± 1.44b	11.19 ± 6.29c
Both mutations	30	9 (30.0)	45	2.14 ± 1.28b	11.59 ± 5.71b

^aCalculated using only females with at least one estrous cycle. 

^bP < 0.0001 vs. wild type. 

^cP < 0.005 vs. wild type. 

Table 4.

Wild-type and Cftr mutant females’ response to exogenous hormones

Mouse type	Mice, na	Average ovarian weight (mg)	Average uterine weight (mg)	Average mouse weight (g)	Oocytes ovulated, n
Wild type	10	7.3 ± 2.3	110 ± 10	21.5 ± 3.08	24.3 ± 9.62
ΔF508	9	7.5 ± 1.9	110 ± 30	19.8 ± 2.10b	24.2 ± 7.87
S489X	6	6.0 ± 1.7	92 ± 14	19.4 ± 1.85b	18.6 ± 5.64
Both mutations	15	6.9 ± 1.8	100 ± 25	19.6 ± 1.98b	22.2 ± 7.47

^aOvulated oocytes were present in all injected females. 

^bP < 0.05 vs. wild type. 

Table 5.

Fertilization of oocytes 48 h after hCG in wild-type and mutant females

Mouse type (no. of mice)	Total cells, n	Cells at stage, %
		Oocyte	One cell	Two cell	Three cell	Four cell
Wild type (4)	71	8.5	1.4	76.0	5.6	8.5
ΔF508 (4)	63	98.4	0	1.6	0	0
S489X (4)	65	100	0	0	0	0
Both mutations	128	99.2	0	0.8	0	0

Table 6.

In vitro fertilization of oocytes from wild-type and Cftr mutant females

Mouse type (no. of mice)	Fertility rate (%)a
Wild type (4)	49 of 72 (68.1)
ΔF508 (4)	52 of 74 (70.3)
S489X (4)	46 of 68 (67.6)
Both mutations (8)	98 of 142 (69.0)

^aFertility rate is expressed as the number two-cell embryos over the total number of oocytes (percent). 

Table 7.

Number of oviductal sperm and capacitation in wild-type and Cftr mutant females

Mouse type (no. of mice)	Total no. of sperm (average per oviduct)	Capacitation (n = 200 per mouse) (%)
Wild type (3)	1536 ± 291	100
ΔF508 (3)	142 ± 105a	100
S489X (3)	132 ± 133a	100
Both mutations (6)	137 ± 114a	100

^aP < 0.001 vs. wild type. 

Gross morphology of reproductive organs from Cftr mutant females

Because initial observations of Cftr^−/− female reproductive organs suggested that ovaries and uteri were reduced in size (Fig. 1), weights of uteri and ovaries of wild-type and Cftr^−/− animals aged 6–8 wk were compared. On average, the ovarian weight from Cftr^−/− females was 50% reduced, compared with controls, and uterine weight was reduced on average 56%, compared with controls (Table 2). Although CF mice are known to have growth deficiencies, the relative weight reduction in the reproductive organs was surprising, considering the total animal weight was reduced by only 15%. When the reproductive organ weights were corrected for weight of each mouse, the differences were still significant (ovaries, P < 0.005; uteri P < 0.005, data not shown), suggesting that differences in whole-body growth could not explain the differences in the weight of reproductive organs.

Figure 1.

Morphology and histology of ovaries, uteri, and cervix of 7-wk-old wild-type and Cftr^−/− females. A, Gross morphology of reproductive tracts of wild-type (left) and Cftr^−/− females (center and right). Bar, 500 mm. Note the overall size reduction in Cftr^−/− female tracts as well as the thinner uteri. B and C, Histology of ovaries from wild-type (B) and Cftr^−/− females (C). Bar, 500 μm. Wild-type ovaries displayed CL (marked by asterisk), whereas the markedly smaller Cftr^−/− ovaries rarely contained CLs, even though other follicle stages were present. D–F, Histology of cervices from wild-type (D) and Cftr^−/− females (E and F). Of the 15 Cftr^−/− females examined, the Cftr^−/− mouse in (F) was the only one with obvious physical blockage in the cervix (marked by arrowhead). Bar, 250 μm.

Cftr^−/− mice are known to have a later onset of puberty (38), making it likely that attainment of adult sexual maturation is also delayed. Thus, a comparison of reproductive organ weights was repeated at age 14–16 wk to ensure a direct comparison between sexually mature females. Both ovarian and uterine weight of CF mice remained significantly reduced by 36%, compared with those of wild-type animals, whereas total animal weight reduction was only 12% (Table 2). After reproductive organ weights were corrected for the weight of each animal, the differences were still significant (ovaries, P < 0.05; uteri P < 0.05, data not shown) suggesting that differences in whole-body growth could not explain the differences in the weight of reproductive organs, even in sexually mature female mice.

Ovulation in Cftr mutant females

The reduction in number of pups per litter in Cftr^−/− females and the reduced size of the ovaries may indicate a problem with oocyte maturation and/or ovulation. Ovarian histology from Cftr mutants displayed normal follicle development with appropriate follicle stages present (Fig. 1). We investigated the number of ovulatory follicles that ovulated oocytes by observing number of CL in ovaries of both mutant and wild-type females (Fig. 1). The average number of CLs in wild-type females was 9.3 ± 1.4 per ovary, whereas mutants displayed an average of only 1.5 ± 2.0 CLs per ovary (P < 0.001), indicating that when ovulation occurs in mutants, fewer oocytes are ovulated, compared with control females.

Estrous cycle in Cftr mutants

Given the reduced number of litters and ovulation abnormalities, we investigated the effect of Cftr mutation on estrous cyclicity. The stage of estrous cycle was examined in 14- to 16-wk-old animals for a period of 20 d. Interestingly, 30% of Cftr^−/− females never entered estrus but were constantly in diestrus, whereas none of the control mice showed a similar phenotype (P < 0.05). Of the females that displayed at least one estrous cycle during the observation period, mutant females displayed half as many cycles as wild-type females, resulting in a doubling of average cycle length (Table 3).

Hormone levels

FSH and LH circulating levels were measured in wild-type and Cftr mutant females in 14- to 16-wk-old animals at proestrus to investigate possible differences. FSH levels were slightly elevated in both Cftr mutant strains, compared with wild-type females (Fig. 2), although values were still within the normal range reported for mice (e.g. Refs. 39 and 40). There were no detectable differences in LH levels between wild-type and Cftr mutant females.

Figure 2.

FSH and LH levels in wild-type and Cftr mutant female mice. FSH (A) and LH (B) levels in wild-type (open bars), ΔF508 (black bars), and S489X (striped bars) female mice. *, Significantly different from wild-type mice (P < 0.05).

Ovulation induction in Cftr mutant mice

We hypothesized that superovulation of CF females with the commonly used exogenous hormones pregnant mare serum gonadotropin and hCG could correct many of the observed defects. Indeed, after exogenous hormone injections of both wild-type and Cftr^−/− females, there was no significant difference in the numbers of oocytes ovulated (Table 4). The weights of both ovaries and uteri after superovulation were also not significantly different between wild-type and mutant females (Table 4). Because normal ovulation occurred in hormone-injected mutant females, we examined whether Cftr^−/− females could produce normal number of pups per litter after superovulation. Despite normal ovulation, only one of 10 mutant females that displayed vaginal plugs gave birth, but the lone female did give birth to 20 pups.

In vivo and in vitro fertilization in superovulated Cftr mutant females

Our findings that exogenous hormone injections in Cftr^−/− females led to a normal number of ovulated oocytes but still produced rare litters suggests that the oocytes may not have been fertilized or failed to develop once fertilized. To distinguish between these two possibilities, Cftr^−/− and wild-type females were superovulated, mated to wild-type males, and embryos then flushed from the female reproductive tract. Whereas the majority of embryos from wild-type females were at the two- to four-cell stages, the mutant females contained greater than 98% unfertilized oocytes (Table 5). Retrieval of later stage embryos displayed similar results in that the majority of embryos from wild-type animals were at the 16- to 32-cell stage, whereas oocytes and only a few embryos at this stage were observed in Cftr^−/− females (data not shown). These findings as well as the fact that delayed birth was not observed in Cftr^−/− mothers suggest that the absence of embryos in our study was not due to delayed fertilization. To assess the capacity of Cftr^−/− oocytes to be fertilized outside the female reproductive tract, in vitro fertilization was completed for both wild-type and mutant females. No significant differences in in vitro fertilization rates were observed (Table 6). These findings indicate that reduced fertilization within the female reproductive tract of Cftr^−/− females, rather than failure to develop, is a major reason for infertility in these females.

Sperm transport and capacitation in Cftr mutant females

Physical obstruction to sperm passage from cervical mucus is hypothesized to be the main reason for reduced fertility in women with CF. Given the poor fertilization rate in CF mice, we next examined the cervices and uteri of Cftr^−/− and wild-type mice for excess mucus accumulation and/or other physical signs of obstruction. Obvious cervical mucus was observed in one of 15 mutant females, whereas no wild-type females displayed a similar phenotype (Fig. 1). No other obvious obstructions were found in any uteri examined (seven per mouse strain) from either mutant or wild-type females. These findings suggest that physical obstructions within the reproductive tract cannot be the sole explanation for the decreased fertilization rate observed in Cftr^−/− females.

Although obvious mucus accumulation cannot explain the unfertilized oocytes in Cftr^−/− females, studies have shown that fluid control within the female reproductive tract may be important in sperm movement (41,42,43). Thus, we examined the ability of sperm to transverse the mutant reproductive tract by flushing the oviducts of mutant and wild-type animals the morning of a positive vaginal plug and counting the sperm present. We observed a significantly reduced number of sperm on average in Cftr^−/− female oviducts, approximately 10% that of wild-type females, indicating that abnormalities in the mutant female reproductive tract may decrease sperm motility (Table 7).

Finally, a previous study suggested that impaired bicarbonate transport in the mutant reproductive tract could impede the final steps of sperm maturation, known as capacitation, necessary for fertilization (44). To test this possibility, we examined the oviductal sperm from both mutant and wild-type females for the removal of the acrosome, a vital step in capacitation. All sperm examined had undergone an acrosome reaction (Table 7), suggesting a reduction in sperm transport to the oviduct was the cause for unfertilized oocytes rather than the inability of the female reproductive tract to transport bicarbonate necessary for capacitation.

Discussion

Whereas CF has long been associated with female infertility, the underlying causes have not been fully examined and remain unclear. The study of mouse models of CF can provide important insights into the human condition and allow for experiments not feasible in humans. Past studies of CF mouse models included only cursory studies of female fertility but indicate that the females have decreased fertility (28,29,30,31,32). To our knowledge, this is the first report to quantify CF female mouse fertility and to examine in depth the mechanism(s) of the observed infertility. We used two independent mouse models of CF to decrease the likelihood that our findings were model dependent. Both models have fewer litters and fewer pups per litter over a 5-month period than wild-type females, demonstrating that the females could reproduce but did display decreased fertility. To understand the basis for this impaired female fertility, we thoroughly examined multiple aspects of female reproduction in both CF mouse models.

Our studies indicate that a major cause of decreased fertility in mice with CF is impaired sperm transport within the mutant female reproductive system. In contrast to hypotheses that suggest a cervical mucus barrier to sperm transport, this phenomenon, or any other appreciable physical blockage along the female reproductive tract, was observed in only one mouse and seems to play only a minor role in mice. Although it is possible that this discrepancy is a technical artifact, it is interesting to note that excess cervical mucus should have little effect on female fertility in the rodent because rodent sperm is ejaculated into the uterine cavity, past the cervical canal (45,46,47). Thus, cervical secretions are not a barrier to sperm transport. Our data suggest that instead of a physical barrier impeding sperm transport, the decreased fertilization observed in mutant female mice is more likely due to inadequate fluid control in the reproductive tract, resulting in decreased sperm number in the oviduct. Uterine fluid is known to be important in sperm transport and CFTR has been suggested to play an important role in fluid control in the female reproductive tract (18,21,22). Women with CF have achieved reproductive success with intrauterine insemination, which deposits sperm past the cervix into the uterus, suggesting cervical mucus is a barrier to fertilization in humans. However, repeated unsuccessful attempts with this type of insemination have also been reported (48,49), suggesting that further abnormalities, such as inadequate fluid control throughout the rest of the reproductive tract, could also contribute to the infertility in humans.

In addition to fluid control, bicarbonate transport by Cftr in the female reproductive tract, which may be involved in sperm capacitation, has been hypothesized to play a role in decreased fertilization (44). A recent study using an in vitro model of the CF uterine environment observed a significant decrease in sperm capacitation when Cftr was either absent or inhibited, compared with the control environment (44). To test this hypothesis in vivo, we examined capacitation of oviductal sperm from mutant and wild-type females and observed no significant differences between the groups. Whereas we did not test whether bicarbonate transport was impaired in the CF reproductive tract, we did observe that sperm capacitation was normal, suggesting that, at least in this instance, the in vitro model did not fully recapitulate the in vivo model.

Our studies also revealed that the abnormalities in CF reproductive function extend beyond physical factors to involve the endocrine system. Both CF models display delayed puberty (38), smaller reproductive organs, decreased oocyte ovulation numbers, and estrous cycle irregularities. These findings are similar to phenotypes observed in women with CF including amenorrhea, irregular cycles, anovulation, smaller uteri, and delayed puberty (25,26,27). We also observed a slight increase in FSH levels in Cftr^−/− female mice, compared with controls. Whereas these results contrast with human data showing decreased FSH levels in adolescents and normal levels in adult women with CF, it should be noted that the human studies used serum collected from either random cycling adolescents (25) or adults at a different cycle stage (26) than was used in our study. The meaning of the increased gonadotropin levels is unclear. The slightly increased FSH levels in Cftr mutant females could be the result of the decreased number of ovulatory follicles, leading to decreased estradiol and/or inhibin production and lack of feedback inhibition of FSH secretion. Increased FSH levels could also be a direct effect of Cftr on FSH regulation and/or a component of central insufficiency in animals that cannot fully compensate for the slowed/diminished ovarian development seen in the Cftr^−/− mice. Regardless, the slightly increased FSH is an interesting finding that provides further evidence of abnormal reproductive endocrine function in the CF mice and is worthy of further study.

Interestingly, when we exposed the CF female mice to exogenous hormones, we observed that the organ size and ovulation difficulties were corrected. This is an important finding because it suggests that the CF reproductive organs can respond to gonadotropins but that an impaired HPG axis may be a direct cause of reduced fertility in women with CF. This possibility deserves further investigation and is supported by recent studies suggesting that CFTR is functional in the central nervous system. CFTR has been identified in rodent brains and the human hypothalamus (12,13,14,15) and has been reported to regulate GnRH secretion in a hypothalamic neuronal cell line (16). Taken together these findings and our data provide early evidence that suggest that mutations in CFTR may contribute directly to the delay in sexual maturation and the infertility observed in CF women through a role in the modulation of the HPG axis. In future studies of Cftr^−/− female mice, it will be important to determine FSH and LH levels and their relationship to the levels of other reproductive hormones (e.g. estradiol, progesterone, testosterone) at various estrous stages across development.

Reproductive development and physiology are highly conserved among mammals, which allows for further understanding human reproduction and fertility through the study of model systems. The mouse has been a particularly good model system for expanding the understanding of genetic factors affecting reproduction (reviewed in Ref. 50). Our findings using the CF mouse models indicate that CF negatively impacts female fertility in mice as it does in humans. One important question that arises from our data is which of the observed phenotypes are a direct effect of absent Cftr function and which abnormalities are byproducts of systemic CF disease. In other words, does Cftr play a direct role in normal reproductive endocrine function and fertility or does its absence lead to a systemic illness that indirectly causes infertility? To answer this question, we are creating a conditional Cftr knockout mouse that will allow us to delete Cftr function in a specific tissue, cell type, or even time period. This resource will be a powerful complement to human studies and will allow us to study the role that Cftr plays in infertility without the confounding effects of a systemic knockout. We propose that by specifically deleting Cftr function in particular tissues within the neuroendocrine system and the reproductive tract, we will be able to distinguish between direct and indirect effects of Cftr’s absence. If Cftr has a primary role, the continued study of CF mouse models will not only help us better understand female infertility associated with CF but may uncover a role for Cftr in fertility in the general population.