CRISPR knockdown of ESR1 in preoptic GABA neurons.

A,B. Photomicrographs showing distribution of mCherry (gRNA from AAV) and expression of GFP (Cas9) in VGAT neurons and nuclear-located ESR1 (blue) in PVpo of two mice receiving either gRNA-LacZ (A) or gRNA-2 (B). The mCherry signal is removed in the adjoining plates a. and b. so that the VGAT neurons (green) co-expressing ESR1 (blue or teal nuclei) are more easily identified. C. Individual data points and mean±SEM percentage of GFP/VGAT neurons expressing ESR1 within injected regions of the RP3V and MPN for the three gRNA groups. *** p<0.001 versus gRNA-LacZ (ANOVA, post-hoc Dunnett’s tests). D-G, representative examples of AAV injection sites (pink) in four mice; 16337 (bilateral gRNA-2), 14945 (unilateral RP3V, bilateral MPN gRNA-2), 13631 (bilateral gRNA-3), 13786 (bilateral gRNA-LacZ).

Deletion of ESR1 from preoptic GABA neurons and estrous cyclicity.

A-C. Individual paired data points and mean±SEM estrous cycle length before and after AAV gRNA injection of lacZ, gRNA-2, and gRNA-3 into the RP3V and MPN. D-E, Individual paired data points and mean±SEM estrous cycle length before and after AAV gRNA injection in mice with bilateral AAV injections in the RP3V and MPN analyzed separately. Four mice (3 x gRNA-2, 1 x gRNA-3) enter constant estrus and one gRNA2 mouse was in constant diestrus, scored as a cycle length of 0. No significant effects of gRNA injection were detected (p>0.05 Wilcoxon paired tests). F-H, Examples of estrous cycle patterns from three mice including one (G) that entered constant estrous following gRNA-2 injection. The individual animal number is given in each frame.

Deletion of ESR1 from preoptic GABA neurons does not alter pulsatile LH secretion.

A-D. Representative LH pulse profiles from female mice given AAV gRNA-lacZ, gRNA-2 and gRNA-3. The mouse identification number is given in brackets. E-G. Histograms show the individual data points and mean±SEM for parameters of pulsatile LH secretion in mice given gRNA-lacZ, gRNA-2 and gRNA-3 into the RP3V and MPN. No significant effects are detected (p>0.05, Kruskal-Wallis test). H-J. Correlations between the % VGAT neurons with ESR1 in the RP3V and parameters of pulsatile LH secretion. Individual mice are color-coded according to their gRNA treatment. No significant correlations were detected (Pearson r < 0.34 in all cases). K-M. Correlations between the % VGAT neurons with ESR1 in the MPN and parameters of pulsatile LH secretion. Individual mice are color-coded according to their gRNA treatment. No significant correlations were detected (Pearson r < 0.41 in all cases).

Effects of ESR1 deletion in preoptic GABA neurons on surge parameters.

A,B. Individual data points and mean±SEM values showing the percentage of GnRH neurons with cFos and single-point LH levels for mice killed at the time of the expected surge given gRNA-LacZ (black), gRNA-2 (blue) and gRNA-3 (green) injections centered on the RP3V and MPN. ** p=0.0023 (Krusakl-Wallis) compared with LacZ. C,D. Correlations between the % of RP3V VGAT neurons with ESR1 and cFos expression by GnRH neurons or LH secretion. Individual mice are color-coded according to their gRNA treatment. A significant correlation for cFos in GnRH neurons exists (p=0.008, Pearson r = 0.66) but not for LH (p=0.26, Pearson r = 0.31). E,F. Correlations between the % of MPN VGAT neurons with ESR1 and cFos expression by GnRH neurons or LH secretion. Individual mice are color-coded according to their gRNA treatment. No significant correlations were found.

Suppression of RP3V kisspeptin expression is associated with the loss of estrous cycles and the surge mechanism.

A,B. Dual-label immunohistochemistry for kisspeptin (brown) and ESR1 (black) shows the normal high level of co-expression (white arrowheads) in a representative gRNA-lacZ mouse (A) but near absence of kisspeptin immunoreactivity in a representative “bilateral loss” mouse (#16337) (B). 3V, third ventr iCc. lEest.rous cycle length before and after gRNA injection in gRNA “lacZ” mice, gRNA2/3 mice with “normal” kisspeptin expression, gRNA2/3 mice with a “unilateral” reduction in kisspeptin, and gRNA2/3 mice with a near complete “bilateral” loss of kisspeptin. * p<0.05 compared to pre-values. D,E. Single point LH levels and % of GnRH neurons with cFos at the time of the expected surge in gRNA “lacZ” mice, gRNA2/3 mice with “normal” kisspeptin expression, gRNA2/3 mice with a “unilateral” reduction in kisspeptin, and gRNA2/3 mice with a near complete “bilateral” loss of kisspeptin. * p<0.05 compared to lacZ. F-H. Parameters of pulsatile LH secretion in gRNA “lacZ” mice, gRNA2/3 mice with “normal” kisspeptin expression, gRNA2/3 mice with a “unilateral” reduction in kisspeptin, and gRNA2/3 mice with a near complete “bilateral” loss of kisspeptin. No significant differences were detected.

Kisspeptin-ESR1 co-expression in re-grouped gRNA mice. Table showing the numbers of kisspeptin neurons/section in the AVPV and PVpo and percentage expression with ESR1. “LacZ” refers to all mice given gRNA-LacZ, “Normal” refers to all gRNA-2/3 mice with normal kisspeptin expression (unilateral cell counts shown), “Unilateral” refers to gRNA-2/3 mice in which only one side of the brain had reduced kisspeptin cell numbers with the cell count on the affected side given, “Bilateral” refers to gRNA-2/3 mice with essentially no cytoplasmic kisspeptin expression bilaterally in the RP3V. Too few kisspeptin neurons existed to reliably determine co-expression with ESR1. ** p<0.01, *** p< 0.001 compared with lacZ group (Kruskal-Wallis with Dunn’s tests, exact p values given below).

VGAT-ESR1 in mice grouped on the basis of kisspeptin expression. Table showing the percentage of VGAT neurons in the RP3V and MPN expressing ESR1 in gRNA mice re-grouped on the basis of kisspeptin expression (see Table 1 for explanation of groups). * p<0.05, *** p< 0.001 compared with lacZ group (Kruskal-Wallis with Dunn’s tests, exact p values given below).