http://www.breast-cancer-research.com/content/16/4/R79/figure/F5 (panel B)
We did not have decent negative controls (as this was just an addition to the main story), so I am not sure about the specificity. As ERbeta is quite ubiquitous, I think finding the negative control could be a challenge, but crucial if you are really concerned about the specificity.
What I've learnt while participating in this exactly topic duscussion during 37th Annual Breast Cancer Symposium in San Antonio (December 9-13, 2014) was that the mouse mammary tumor cell lines, D2A1 and MM51 both significantly express ER-alpha and ER-betha. Specifically D2A1 cells are dependent on estrogen for growth and have increased expression of aromatase. Additionally, D2A1 cells are very aggressive and highly metastatic - thus representing a good model of breast cancer progression. MM51 cells are Her2/neu positive and are an adequate model to study the crosstalk between growth factors ans ER signaling.