Vassalli, A., Matzuk, M.M., Gardner, H.A.R., Lee, K.-F., and Jaenisch, R. (1994). Activin/inhibin bB subunit gene disruption leads to defects in eyelid development and female reproduction. Genes and Development 8, 414-427.
Analyses of their function in vivo have been initiated recently through the inactivation of individual subunits and receptors by homologous recombination in mice. Inactivation of the a subunit results in the absence of inhibins and causes infertility due to the development of gonadal tumors at an early age. Independent inactivation of the bA and bB genes has demonstrated that the bA and bB subunits display non-overlapping activities. In bB-deficient mice the activities of activins B and AB as well as inhibin B are eliminated. bB-deficient pups are born with open eyelids but their overall development is normal. Fertility of the females is slightly reduced and a prolonged gestation time may be indicative of systemic endocrine defects. Most prominently, females exhibit a lactational defect and cannot support their litters. The inactivation of the bA gene eliminates activin A and AB as well as inhibin A and causes malformation of the secondary palates and an absence of teeth and whiskers. Since the newborn mice are incapable of suckling and die within 24 hours the effects on mammary gland development are not known. Mice deficient in both bA and bB exhibit a combination of the phenotypes seen in each of the mutants but have no additional defects.
The development of the mammary gland proceeds in distinct phases, and functional differentiation of secretory epithelial cells is a critical step in the reproductive cycle of mammals. Pronounced ductal growth occurs at the onset of puberty and extensive development in cycling virgins leads to the formation of a ductal tree which fills the entire mammary fat pad. Alveolar proliferation occurs during pregnancy and terminal differentiation of alveolar epithelial cells is completed at the end of gestation with the onset of milk secretion at parturition. In order to understand the roles of inhibins and activins in mammary gland function we studied mammogenesis in bB-deficient mice. In particular, we determined whether systemic or mammary derived activins and inhibins are required for mammary development.
Sections through mammary whole mounts from late pregnant (slide 6) and post-partum mice revealed an even more pronounced failure of ductal outgrowth and severe underdevelopment of lobulo-alveolar structures in bB-deficient mice. At late pregnancy (Fig. 3A) the alveoli were sparse and small, and they contained round cells with large nuclei. Very little secretion was found in the lumina (Fig. 3A). The persisting end buds in late pregnant and post-partum bB-deficient mice had a disorganized appearance. The cap cell layer was missing, mitoses were frequently seen in the body cells and the central lumen was often absent (Fig. 3C,D). Concomitant with the underdevelopment of the alveoli, the stroma contained many adipocytes (Fig. 3A-D) while they were mostly replaced by alveolar epithelial cells in the wild type glands (Fig. 3E,F). After parturition, the alveoli of bB-deficient mice were still small, the secretory cells remained rounded and the small lumen contained a dense secretion (Fig. 3B). In contrast, alveoli of post-partum control mice displayed the typical flattened appearance of secretory cells and exhibited extended lumina (Fig. 3F).
Transplantation of the entire mammary fat pad containing the mammary anlage from a wild type mouse into a bB-deficient host addressed the question whether a systemic, endocrine effect of bB was responsible for the mammary underdevelopment. The transplanted glands were harvested at late pregnancy, and normal development was observed (Figure 11A). The contralateral endogenous bB-deficient gland had the typical appearance of a mutant gland (Figure 11B). In the reverse experiment, the transplantation of bB -/- fat pads into wild type mice, we succeeded only in one case to get limited development of the mammary epithelium (data not shown). This demonstrates that systemic bB is not necessary for mammary development, and it identifies the stroma and/or the epithelium as the sites for the defect in the bB-deficient mice.
The second set of transplantation experiments was aimed to identify whether the growth stimulatory effect of bB was of paracrine nature and mediated by the stroma, as compared to an autocrine or intracrine effect in the epithelial compartment. A small piece of mammary tissue from bB deficient females was transplanted into cleared fat pads from virgin wild type mice. In this situation the mutant mammary epithelium penetrates the host fat pad and becomes associated with wild type stroma. These experiments allowed us to analyze development of bB-deficient mammary epithelium in a wild type fat pad in a hormonal environment unaffected by a possible perturbance of pituitary and ovarian hormone levels caused by the absence of the bB subunit. While one fat pad hosted the mutant epithelium the other carried a wild type transplant. These mice were mated two months after transplantation and mammary tissues were analyzed after the mice had given birth. Both, the wild type (Figure 10) and the mutant (Figure 10) transplants grew out to a similar extent. The ducts fully penetrated the fat pad and extensive lobulo-alveolar units had formed. This demonstrates that bB produced by the stroma can rescue the development of mutant mammary epithelium. Clearly, mammary epithelial cells do not require an autocrine, cell autonomous function of the bB subunit.
Stromal dependence of mammary development
Mammary gland development is dependent on epithelial-stromal interactions. From its initiation as a small epithelial bud in the embryo, reciprocal and sequential tissue interactions are required for mammary morphogenesis. The expression patterns of many growth factors in mammary development suggest their involvement in these processes. LEF 1, keratinocyte growth factor and neuregulin have been identified as factors which are produced in the mesenchyme and thus may mediate short range signals to influence mammary epithelial growth. All of these molecules are also required for the development of other organs and their inactivation does not show the same exclusive effect on mammary glands as does the bB subunit.
The transplantation studies localize the defects of ductal growth and alveolar differentiation in bB-deficient mice to the mammary stroma. There is no requirement for systemic bB since wild type fat pads can support full epithelial development in bB-deficient hosts. Moreover, epithelium derived bB is not required as bB-deficient epithelia develop to the same extent in wild type fat pads as control epithelia. A paracrine action of activins and inhibins is further supported by the observation that the organs affected in mice with a bA mutation, namely development of the secondary palates, teeth and whiskers, all depend on epithelial mesenchymal interactions. Furthermore, branching morphogenesis of embryonic salivary gland, kidney and pancreas in organ culture can be reversibly disrupted by activin A while no effect was seen with inhibin A. In testicular development, inhibin appears to have paracrine function while activin has an autocrine effect on Sertoli cells. It is still unclear whether the effective growth factor which is missing in the bB-deficient gland is activin B, AB, inhibin B or all of these. Pure preparations of these factors will be required to dissect the mechanisms of their actions on the different cell populations of the mammary gland and help understand the unique function of these multipotent and widely expressed growth factors in the mammary gland.
Gertraud Robinson and Lothar Hennighausen
National Institutes of Health
Bethesda, MD 20892
Phone: 301-496-5004
Fax: 301-496-0839
e-mail: gertraur@bdg10.niddk.nih.gov
last update: June 1998