IntroductionRetinoic acid signaling pathways are disabled in human breast cancer suggesting a controlling role in normal mammary growth that might be lost in tumorigenesis. We tested a single receptor isotype, RARα1 (retinoic acid receptor isotype alpha, isoform 1), for its role in mouse mammary gland morphogenesis and mouse mammary tumor virus (MMTV)-wingless-related MMTV integration site 1 (wnt1)-induced oncogenesis.MethodsThe role of RARα1 in mammary morphogenesis was tested in RARα1-knockout (KO) mice and in mammary tumorigenesis in bi-genic (RARα1/KO crossed with MMTV-wnt1) mice. We used whole mounts analysis, stem cells/progenitor quantification, mammary gland repopulation, quantitative polymerase chain reaction (Q-PCR), test of tumor-free survival, tumor fragments and cell transplantation.ResultsIn two genetic backgrounds (129/Bl-6 and FVB) the neo-natal RARα1/KO-mammary epithelial tree was two-fold larger and the pubertal tree had two-fold more branch points and five-fold more mature end buds, a phenotype that was predominantly epithelial cell autonomous. The stem/progenitor compartment of the RARα1/KO mammary, defined as CD24low/ALDHhigh activity was increased by a median 1.7-fold, but the mammary stem cell (MaSC)-containing compartment, (CD24low/CD29high), was larger (approximately 1.5-fold) in the wild type (wt)-glands, and the mammary repopulating ability of the wt-gland epithelium was approximately two-fold greater. In MMTV-wnt1 transgenic glands the progenitor (CD24low/ALDHhigh activity) content was 2.6-fold greater than in the wt and was further increased in the RARα1/KO-wnt1 glands. The tumor-free survival of RARα1/KO-wnt1 mice was significantly (P = 0.0002, Kaplan Meier) longer, the in vivo growth of RARα1/KO-wnt1 transplanted tumor fragments was significantly (P = 0.01) slower and RARα1/KO-wnt1 tumors cell suspension produced tumors after much longer latency.ConclusionsIn vitamin A-replete mice, RARα1 is required to maintain normal mammary morphogenesis, but paradoxically, also efficient tumorigenesis. While its loss increases the density of the mammary epithelial tree and the content of luminal mammary progenitors, it appears to reduce the size of the MaSC-containing compartment, the mammary repopulating activity, and to delay significantly the MMTV-wnt1-mammary tumorigenesis. Whether the delay in tumorigenesis is solely due to a reduction in wnt1 target cells or due to additional mechanisms remains to be determined. These results reveal the intricate nature of the retinoid signaling pathways in mammary development and carcinogenesis and suggest that a better understanding will be needed before retinoids can join the armament of effective anti-breast cancer therapies.

We generated a transgenic (Tg)-mouse model expressing a dominant negative-(DN)-RARα, (RARαG303E) under adipocytes-specific promoter to explore the paracrine role of adipocyte retinoic acid receptors (RARs) in mammary morphogenesis. Transgenic adipocytes had reduced level of RARα, β and γ, which coincided with a severely underdeveloped pubertal and mature ductal tree with profoundly decreased epithelial cell proliferation. Transplantation experiments of mammary epithelium and of whole mammary glands implicated a fat-pad dependent paracrine mechanism in the stunted phenotype of the epithelial-ductal tree. Co-cultures of primary adipocytes, or in vitro differentiated adipocyte cell line, with mammary epithelium showed that when activated, adipocyte RARs contribute to generation of secreted proliferative and pro-migratory factors. Gene expression microarrays revealed a large number of genes regulated by adipocyte-RARs. Among them, pleiotrophin (PTN) was identified as the paracrine effectors of epithelial cell migration. Its expression was found to be strongly inhibited by DN-RARα, an inhibition relieved by pharmacological doses of all-trans retinoic acid (atRA) in culture and in vivo. Moreover, adipocyte-PTHR, another atRA responsive gene, was found to be an up-stream regulator of PTN. Overall, these results support the existence of a novel paracrine loop controlled by adipocyte-RAR that regulates the mammary ductal tree morphogenesis.

Background: Retinoic acid suppresses cell growth and promotes cell differentiation, and pharmacological retinoic acid receptor (RAR) activation is anti-tumorigenic. This begs the question of whether chronic physiological RAR activation by endogenous retinoids is likewise antitumorigenic.

No abstract