The present manuscript by Decourt et al., aims at investigating the role of AgRP neurons in puberty onset, using a transgenic mouse model expressing excitatory (hM3Dq) and inhibitory (hM4Di) DREADDs. Authors induced either chronic or acute chemogenetic activation or inhibition of AgRP neurons in pups at PND 26 and followed their puberty, food intake or body weight. They found that AgRP neurons inhibition in male mice reverses the delayed puberty induced by undernutrition.
While the rationale of the study is original and innovative, there are several concerns with the study design and lack of consistency, interpretations and the conclusions drawn from the data.

Comments:

1- With regards to the CNO exposure period/time, showing showing FI after the chronic manipulation of AgRP neurons would have been informative given the protocol used in the experiments. Thus, the acute measurement is less relevant for the interpretation of the study. In any case, the very large difference in FI in the control groups (Fig 1b vs 1c) compromises the interpretation of the data. Are the mice used for FI measurements the same as the ones used for body weight (BW) monitoring?

2- There is no rationale for the start day of CNO treatment(PND26). While the exact time of puberty onset is unknown, at PND26 the activation of the HPG axis has likely started already. The authors could have started the CNO treatment in water as early as PND21 or before that during lactation through chronic injections to the pups. This may explain the lack of effect in the experimental models used since the effect of AgRP on puberty has been already documented before.

3- It is mentioned in the methods section that the pre-weighted pellet was measured at frequent intervals. What are these time intervals? The methods section should include detailed information of the procedures performed.

4- The graph in figure 1 (a), shows that all mice have a BW over 20g, while these pups were first monitored at PND age 26, which is an unusually high body weight for their age, before any manipulation of AgRP neurons. Authors should address/explain this.

5- Graphs 1(a) and 1(c) show that AgRP neurons inhibition have no effect on BW after chronic CNO exposure while it decreases the FI significantly following acute CNO injection. This discrepancy needs to be addressed.

6- The GTT and ITT studies have been performed in mice under acute CNO injections. What is the reason behind choosing to do this experiment in acute, but not chronic, manipulation of AgRP neurons?
7- With regards to GTT AND ITT data, authors mention in the discussion that the study by Steculorum et al., 2016 reported that “DREADD-mediated stimulation of AgRP neurons had no effect on glucose tolerance but decreased insulin sensitivity in mice”. These claims are incorrect. Steculorum et al reported that the chemogenetic activation of AgRP neurons acutely impaired glucose tolerance and insulin tolerance while it’s the optogenetic activation of AgRP neurons that induced insulin resistance, without affecting glucose. Please correct this statement in the discussion (page 18, line 400). How do authors explain the discrepancy between their data and those reported by Steculorum et al., 2016?

8- The pictures of Immunohistochemistry/RNAscope (figure 2) show AgRP labeling and co-expression with cFos in the anterior portion of the arcuate nucleus (ARC), while AgRP are widely expressed throughout the 3 portions of the ARC (anterior, median, posterior). Did the authors check AgRP expression and co-expression with cFos throughout the ARC or was the analysis only inclusive of the ventrolateral ARC, as shown in the pictures?

9- Page 9, line 208: authors claim that “in experiment 1, the age of preputial separation was significantly advanced in CNO treated AgRP-hM4Di mice”. This statement is incorrect, in experiment 1 (figure 3) puberty onset was not affected neither in males not females following AgRP manipulation.

10- There is inconsistency in the choice of using one or both sexes in the experiments. A large array of experiments was performed in males only, before females are introduced in the undernutrition experiments.

11- There are many overstatements throughout the discussion section. Example: In page 16, line 371, the authors claim that the present study aims to evaluate the reproductive role of AgRP neurons, while this study has only investigated puberty and is missing additional important component of reproductive axis (estrous cycles, fertility test, LH pulses, ovarian histology, etc).

12- In experiment 2, AgRP inhibition advances puberty onset only in males and only in the under-fed group. Why do the authors think AgRP neurons inhibition didn’t advance puberty onset in the normally fed group? Both literature and the data from this study suggest the effect of AgRP on the reproductive axis to be central (exerted directly on GnRH or Kiss1 neurons), and not peripheral following BW lost. In this case, the normally fed mice should also show signs of an earlier puberty onset compared to their control littermates. Have all mice from each of these experiments been validated for a proper hM3Dq and hM4Di expression?