Oculodentodigital dysplasia (ODDD) is an autosomal dominant disorder characterized by pleiotropic developmental anomalies of the limbs, teeth, face and eyes that was shown recently to be caused by mutations in the gap junction protein alpha 1 gene (GJA1), encoding connexin 43 (Cx43). In the course of performing an Nethyl-N-nitrosourea mutagenesis screen, we identified a dominant mouse mutation that exhibits many classic symptoms of ODDD, including syndactyly, enamel hypoplasia, craniofacial anomalies and cardiac dysfunction. Positional cloning revealed that these mice carry a point mutation in Gja1 leading to the substitution of a highly conserved amino acid (G60S) in Cx43. In vivo and in vitro studies revealed that the mutant Cx43 protein acts in a dominant-negative fashion to disrupt gap junction assembly and function. In addition to the classic features of ODDD, these mutant mice also showed decreased bone mass and mechanical strength, as well as altered hematopoietic stem cell and progenitor populations. Thus, these mice represent an experimental model with which to explore the clinical manifestations of ODDD and to evaluate potential intervention strategies.

The ovarian follicle in mammals is a functional syncytium, with the oocyte being coupled with the surrounding cumulus granulosa cells, and the cumulus cells being coupled with each other and with the mural granulosa cells, via gap junctions. The gap junctions coupling granulosa cells in mature follicles contain several different connexins (gap junction channel proteins), including connexins 32, 43, and 45. Connexin43 immunoreactivity can be detected from the onset of folliculogenesis just after birth and persists through ovulation. In order to assess the importance of connexin43 gap junctions for postnatal folliculogenesis, we grafted ovaries from late gestation mouse fetuses or newborn pups lacking connexin43 (Gja1(-)/Gja1(-)) into the kidney capsules of adult females and allowed them to develop for up to 3 weeks (this was necessitated by the neonatal lethality caused by the mutation). By the end of the graft period, tertiary (antral) follicles had developed in grafted normal (wild-type or heterozygote) ovaries. Most follicles in Gja1(-)/Gja1(-) ovaries, however, failed to become multilaminar, with the severity of the effect depending on strain background. Dye transfer experiments indicated that intercellular coupling between granulosa cells is reduced, but not abolished, in the absence of connexin43, consistent with the presence of additional connexins. These results suggest that coupling between granulosa cells mediated specifically by connexin43 channels is required for continued follicular growth. Measurements of oocyte diameters revealed that oocyte growth in mutant follicles is retarded, but not arrested, despite the arrest of folliculogenesis. The mutant follicles are morphologically abnormal: the zona pellucida is poorly developed, the cytoplasm of both granulosa cells and oocytes is vacuolated, and cortical granules are absent from the oocytes. Correspondingly, the mutant oocytes obtained from 3-week grafts failed to undergo meiotic maturation and could not be fertilized, although half of the wild-type oocytes from 3-week grafted ovaries could be fertilized. We conclude that connexin43-containing gap junction channels are required for expansion of the granulosa cell population during the early stages of follicular development and that failure of the granulosa cell layers to develop properly has severe consequences for the oocyte.

Studies of mice with targeted disruptions of specific connexin genes have revealed that at least two connexins, connexin37 (Cx37) and connexin43 (Cx43), play essential roles in ovarian follicle development. To explore the respective roles of these two connexins in gap-junctional communication between the developing murine oocyte and its surrounding cumulus granulosa cells, we used confocal immunofluorescence microscopy and oocyte preloading functional assays. Immunofluorescence microscopy located Cx37 within gap-junction plaques between granulosa cells and the oocyte, and Cx43 between surrounding granulosa cells. Preloading assays combining denuded oocytes and cultured granulosa cells expressing or lacking Cx37 or Cx43 revealed that Cx37 must be present in both cell types for the establishment of heterocellular gap-junctional coupling. Furthermore, immunofluorescence microscopy of cultured granulosa cells after incubation with denuded oocytes showed that the oocyte induces the formation of gap junctions containing Cx37 at the surface of granulosa cells. Continuous Cx37 expression in granulosa cells was confirmed using RT-PCR. Together, these results indicate that the growing murine oocyte is functionally coupled with granulosa cells by homotypic gap junctions composed of Cx37, and that the formation and/or stabilization of Cx37 junctions is selectively induced at the oocyte-granulosa interface by cell contact.

The gap junction proteins connexin37 and connexin43 are required for ovarian folliculogenesis in the mouse. To define their respective roles in oogenesis, chimeric ovaries containing either null mutant oocytes and wild-type granulosa cells or the reverse combination were grafted to the renal capsules of immunodeficient female mice. After three weeks, the oocytes were tested for meiotic competence and fertilizability in vitro. Ovaries composed of connexin43-deficient oocytes and wild-type granulosa cells produced antral follicles enclosing oocytes that could develop to at least the two-cell stage, demonstrating that oocytes need not express connexin43 to reach maturity. Conversely, both follicle development and oocyte maturation were impaired in ovaries containing either wild-type oocytes and connexin43-deficient granulosa cells or connexin37-deficient oocytes and wild-type granulosa cells. Thus absence of connexin43 from granulosa cells or connexin37 from oocytes is sufficient to compromise both oocyte and follicle development. Wild-type oocytes paired with connexin37-deficient granulosa cells generated antral follicles containing oocytes that developed to at least the two-cell stage. Therefore, connexin37 absence from granulosa cells need not impair fertility in mice. Dye transfer experiments revealed persistent oocyte-granulosa cell coupling in those follicles, indicating functional compensation by another connexin. The results indicate that mouse oocytes do not need to express connexin43 in order to develop into meiotically competent, fertilizable gametes, but must express connexin37 for communication with granulosa cells, a requirement for oogenesis.

Intercellular communication is required for ovarian folliculogenesis. This is apparent in mice lacking connexin43 (Cx43, a gap junction protein strongly expressed in granulosa cells), or growth/differentiation factor-9 (GDF9, an oocyte-specific growth factor that stimulates granulosa cell proliferation and differentiation), or in mice expressing a mutant form of Kit ligand (KITL, a paracrine factor that, in the ovary, is secreted by granulosa cells to stimulate oocyte growth). In all of these mutant lines, follicle growth is impaired suggesting a possible interaction between paracrine signaling and gap junctional communication. To assess this possibility, we analyzed gene expression in mutant ovaries. Despite the lack of gap junctional coupling between granulosa cells of Cx43 null mutant ovaries, expression of the genes encoding KITL and its receptor, KIT, is maintained. Furthermore, GDF9 expression is maintained. In GDF9 null mutant ovaries, there is no apparent change in Cx43 expression and, correspondingly, the granulosa cells remain coupled. There is also no increase in granulosa cell apoptosis in ovaries lacking Cx43 or GDF9. Staining for proliferating cell nuclear antigen (PCNA) revealed that the granulosa cells of Cx43 null mutant ovaries have a reduced frequency of DNA synthesis. Using both radiolabeled thymidine incorporation and PCNA staining in vitro, we showed that recombinant GDF9 could restore the proliferation of coupling-deficient granulosa cells to the level of control cells. These results indicate that impaired folliculogenesis in mice lacking Cx43 is due at least in part to reduced responsiveness of granulosa cells to oocyte-derived GDF9, indicating an interaction between these two modes of intercellular communication.

Ovarian granulosa cells are coupled via gap junctions containing connexin43 (Cx43 or α-1 connexin). In the absence of Cx43, granulosa cells stop growing in an early preantral stage. However, the fact that granulosa cells of mature follicles express multiple connexins complicated interpretation of this finding. The present experiments were designed to clarify the role of Cx43 vs. these other connexins in the earliest stages of folliculogenesis. Dye injection experiments revealed that granulosa cells from Cx43 knockout follicles are not coupled, and this was confirmed by ionic current injections. Furthermore, electron microscopy revealed that gap junctions are extremely rare in mutant granulosa cells. In contrast, mutant granulosa cells were able to form gap junctions with wild-type granulosa cells in a dye preloading assay. It was concluded that mutant granulosa cells contain a population of connexons, composed of an unidentified connexin, that do not normally contribute to gap junctions. Therefore, although Cx43 is not the only gap junction protein present in granulosa cells of early preantral follicles, it is the only one that makes a significant contribution to intercellular coupling.

Genetic ablation of connexin37 (Cx37) or connexin43 (Cx43), the two gap junction proteins expressed by mouse ovarian granulosa cells, has been shown to result in impaired follicle development. We used patch-clamp techniques to evaluate quantitatively the contribution of these connexins to gap junctional intercellular communication (GJIC) among granulosa cells. The coupling conductance derived from a voltage step-induced capacitive current transient was used as a measure of GJIC in cultured granulosa cells. Using this method, we determined that the conductance of wild-type (84.1 +/- 28.6 nS; n = 6) and Cx37-deficient granulosa cells (83.7 +/- 6.4 nS; n = 11) does not differ significantly (P = 0.35), suggesting a limited contribution, if any, of Cx37 to granulosa cell coupling. In contrast, the conductance between granulosa cells of Cx43-deficient mice (2.6 +/- 0.8 nS; n = 5) was not significantly different from that of single, isolated wild-type granulosa cells (2.5 +/- 0.7 nS, n = 5; P = 0.83), indicating that Cx43-deficient granulosa cells were not electrically coupled. A direct measurement of transjunctional conductance between isolated granulosa cell pairs using a dual patch-clamp technique confirmed this conclusion. Interestingly, a partial rescue of folliculogenesis was observed when the Cx43-null mutation in C57BL/6 mice was crossed into the CD1 strain, and capacitive current measurement demonstrated that this rescue was not due to reestablishment of GJIC. These results demonstrate that folliculogenesis is impaired in the absence of GJIC between granulosa cells, but they also indicate that the severity is dependent on genetic background, a phenomenon that cannot be attributed to the expression of additional connexins.