In this study, we have shown that BMPR2 has approximately 25% lower expression in lymphocytes cultured from female humans compared with males (Figure ), and approximately 15% lower expression in whole lung from ovariectomized female mice as compared to males (Figure ). In both of these cases, the actual estrogen exposure was the same for male and female groups, but the level of estrogen receptor was increased in females (Figure ). This led us to believe that estrogen effect on BMPR2 expression was through canonical estrogen receptors. There was an evolutionarily conserved estrogen receptor binding site in the BMPR2 promoter; we confirmed its functionality by gel-shift assay and its strength by its ability to out-compete the canonical estrogen receptor target vitellogenin (Figure ). We found that direct administration of estrogen suppressed BMPR2 signal by 15% in cultured lymphocytes and in pulmonary microvascular smooth endothelial cells, but that in the presence of a proliferative signal this suppression could increase to 80% (Figure ). We then used a cell line missing the estrogen receptor, and showed that we could add it back to get a six-fold reduction in BMPR2 expression in a dose-responsive manner (Figure ). However, the reduction in BMPR2 gene expression at baseline was not reflected in a reduction in Smad-dependent effects including ID1; non-canonical BMPR2 signaling was not assessed (Figure ). In total, these data indicate that under normal circumstances, increased estrogen or increased estrogen receptor decreases BMPR2 expression by approximately 15%, but under extreme circumstances (for example, proliferation or complete lack of estrogen receptor) the total potential effect is a five- to six-fold change. If this reduction is relevant to PAH pathogenesis, the mechanism may involve downstream signaling by BMPR2 that is not along the canonical BMPR2 signaling cascade.
These results are relevant to the disease we study, pulmonary arterial hypertension, both because of the high female prevalence, but also because we have shown that among those with a BMPR2 mutation, it is women who metabolize estrogen into the relatively more estrogenically active 16α-hydroxyestrone compound that develop disease [15
]. In contrast, women who metabolize estrogen into 2-hydroxyestradiol or 2-hydroxyestrone are less likely to develop PAH [19
]. Several of the enzymes responsible for estrogen metabolism, including CYP1B1, have been studied and are expressed at the mRNA and protein levels in lung tissue; not surprisingly, there is heterogeneity of expression among individuals [34
]. It has been demonstrated that human bronchial epithelial cells metabolize estrogen in vitro
to hydroxyestrogens such as 2-hydroxyestrone and 16α-hydroxyestrone; however, while biologically plausible, studies in pulmonary microvascular endothelial cells are lacking [35
Since in most studies estrogen is protective of the pulmonary vasculature, the increased prevalence among women might be best explained by the simple hypothesis that gender differences occur through suppression of BMPR2. On a broader level, these results, while novel, might be expected. Whether or not the changes in BMPR2 expression we detected are specific to estrogens or also seen with other hormones, such as thyroid hormone and glucocorticoids requires further investigation. Thyroid abnormalities are well documented among PAH patients, and there is experimental data to suggest that glucocorticoids potentiate cellular proliferation among the vascular smooth muscle cells derived from idiopathic PAH patients [36
]. Insulin resistance and perhaps the metabolic syndrome appear more prevalent among PAH patients, as well [37
While this study evaluates BMPR2 expression at the molecular level, the PAH phenotype and the evolution of clinical disease among PAH patients is quite complicated. Data now exist to suggest that while PAH is more prevalent among females, incidence rates may not fully explain the gender discrepancy. For example, females have better right ventricular contractility compared to males at the time of diagnosis, which may contribute to a potential survival advantage [38
]. Females with PAH also may respond better to certain therapies, as was recently shown to be the case in a pooled analysis of subjects treated with endothelin receptor antagonists [40
]. Ultimately, there are two critical issues for human subjects at risk: (1) do you become a patient (develop PAH); and, (2) how long will you survive with PAH. Determining the diverse array of pieces that contribute to the 'gender puzzle' in PAH should ultimately contribute to better therapy for all PAH patients, and perhaps disease prevention. Estrogen-associated modification of BMPR2 expression may be one piece of that puzzle, although much work remains to be done.
This study has several limitations. For example, the human lymphocyte samples used were immortalized, so that they were equal in hormone exposure; this allowed for milieu uniformity but may not accurately replicate the actual circulating hormone exposure in human subjects. Also, we were unable to distinguish between ERα and ERβ effect, although we understand that in many cases the effect of the two is indistinguishable. Also, since our previous work implicated specific products of estrogen metabolism as a risk factor in PAH, comprehensive studies matching specific estrogen metabolites to specific receptors in the context of suppression of BMPR2 expression are warranted PAH [19
Furthermore, while we detected a reduction in BMPR2 gene expression, the measurable functional consequences on BMP signaling in the unstimulated state did not suggest reduced Smad-dependent (canonical BMP signaling) effects. However, Smad-independent effects were not measured. We suspect that under normal circumstances the influence of estrogens upon BMPR2 is not relevant, and only becomes meaningful in the context of injury repair, when BMPR2 appears to be in disequilibrium with the canonical TGFβ signaling pathway [42
]. As with BMPR2, estradiol decreases expression of TGFβ1 and the type 1 TGFβ receptor in the lung vasculature [44
]. Further investigations are needed, but our findings may reflect the complicated balance of BMPR2, the TGFβ pathway and associated events that result in PAH among susceptible subjects (such as those with low BMPR2 at baseline), or the possibility that Smad-independent effects mediate downstream ramifications of BMPR2 gene expression changes. In the context of cellular injury, a delicate balance of cellular differentiation and renewal characterize the normal healing process, and it is possible that enhanced estrogen exposures disrupt this balance. The fact that estrogen exposure alters BMPR2 expression and downstream signaling, while clearly complex, is novel and may be relevant to PAH pathogenesis. Further studies will be required to separate these related findings.