This is a relatively straight forward study by Watanabe et al that presents the neuroanatomical description of AR in neurons of the arcuate nucleus known to be involved in the control of GnRH release at gestational day 15.5. The authors chose this age as it is included in the window of treatment with DHT that generates the PNA model of PCOS in mice. In addition, the authors assessed the distribution of AR in other areas of the brain at this age as well. The collebaling in the ARC, done through HIS (RNAscope) revealed the interesting finding that despite the documented (by some of the coauthors) that GABAergic inputs onto GnRH neurons are affected in PNA mice, only a small fraction of GABAergic neurons (~10%) expressed AR at this time. Neither do other neurons of the ARC like POMC or TH neurons, although TH neurons despite being predominantly GABAergic are described to express AR in ~22% of them. Perhaps not a high level of coexpression but it doesn’t rule out a biological role and in strike contract with the overall GAD-AR coexpression (~10% total). This indicates that, perhaps, a general description of the overall GABAergic population (even if it is divided in rostral, middle and caudal) might not be enough to identify potential subpopulations of GABAergic neurons involved in the changes induced by PNA. Strikingly, Kiss1 neurons vastly express AR at this age. This is in line with numerous studies showing increased Kiss1 expression in adult PNA treated mice and raises the question of whether he DHT treatment is acting directly on these neurons. This could be readily measured trough different markers of cell activation at this time. This concept is critical for their GABAergic model of PNA modifications leading to PCOS because a plausible pathway would be that increased Kiss1 activity (even at this early age) could lead to increased activation of the HPG axis and higher levels of sex steroids that could be activating estrogen receptor in the neurons of interest. While addressing this pathway is probably beyond the intent of the authors, as this study would require PNA mice, comparing the distribution and co-localization levels of AR with those of ERa would be informative in this context.

The authors have addressed my comments.