Prenatal androgen treatment impairs the suprachiasmatic nucleus arginine-vasopressin to kisspeptin neuron circuit in female mice

Abstract

Polycystic ovary syndrome (PCOS) is associated with elevated androgen and luteinizing hormone (LH) secretion and with oligo/anovulation. Evidence indicates that elevated androgens impair sex steroid hormone feedback regulation of pulsatile LH secretion. Hyperandrogenemia in PCOS may also disrupt the preovulatory LH surge. The mechanisms through which this might occur, however, are not fully understood. Kisspeptin (KISS1) neurons of the rostral periventricular area of the third ventricle (RP3V) convey hormonal cues to gonadotropin-releasing hormone (GnRH) neurons. In rodents, the preovulatory surge is triggered by these hormonal cues and coincident timing signals from the central circadian clock in the suprachiasmatic nucleus (SCN). Timing signals are relayed to GnRH neurons, in part, via projections from SCN arginine-vasopressin (AVP) neurons to RP3V^KISS1 neurons. Because rodent SCN cells express androgen receptors (AR), we hypothesized that these circuits are impaired by elevated androgens in a mouse model of PCOS. In prenatally androgen-treated (PNA) female mice, SCN Ar expression was significantly increased compared to that found in prenatally vehicle-treated mice. A similar trend was seen in the number of Avp-positive SCN cells expressing Ar. In the RP3V, the number of kisspeptin neurons was preserved. Anterograde tract-tracing, however, revealed reduced SCN^AVP neuron projections to the RP3V and a significantly lower proportion of RP3V^KISS1 neurons with close appositions from SCN^AVP fibers. Functional assessments showed, on the other hand, that RP3V^KISS1 neuron responses to AVP were maintained in PNA mice. These findings indicate that PNA changes some of the neural circuits that regulate the preovulatory surge. These impairments might contribute to ovulatory dysfunction in PNA mice modeling PCOS.

Keywords: GnRH; LH surge; PCOS; androgen receptor; circadian; electrophysiology; tract-tracing.

Figure 1

PNA increases Ar expression in the SCN of female mice. (A) (Left) Example confocal images of sections from VEH and PNA mice, showing expression of Avp (green) and Ar (red) mRNA in the SCN. (Right) DAPI counterstain merged with Avp and Ar staining revealed Ar-positive cells containing (filled arrowheads) or not (empty arrowheads) Avp mRNA. (B) Summary graph of the numbers of SCN cells expressing transcripts for Avp (Avp ⁺), Ar (Ar ⁺), Ar and Avp (Ar ⁺ and Avp ⁺ ), or Ar only (Ar ⁺ and Avp ^-) in VEH and PNA mice. (C) Summary of the proportions of Avp-containing cells also expressing Ar [(Ar ⁺ + Avp ⁺)/Avp ⁺] and of Ar-containing cells also expressing Avp [(Ar ⁺ + Avp ⁺)/Ar ⁺] in the SCN of VEH and PNA mice. (D) Summary graph of the numbers of Ar particles in SCN cells with and without Avp expression. *p < 0.05 unpaired t-tests.

Figure 2

PNA decreases innervation of the RP3V by SCN^AVP neuron projection fibers. (A) Example maximum projection images showing mCherry-ir neurons (red) in the SCN of VEH and PNA Avp-cre mice injected with AAV-DIO-mCherry. White dashed lines delineate the SCN, the third ventricle (3V) and optic chiasm (OX). (B) Maximum projection images showing mCherry-ir SCN^AVP neuron fibers in the RP3V of the VEH and PNA mice shown in (A). White dashed lines delineate the RP3V, 3V and OX. (C) Bar graphs summarizing SCN transfection (left) and RP3V innervation by mCherry-expressing SCN^AVP neuron projection fibers (right) in VEH and PNA Avp-cre mice with successful SCN transfection. (D) Linear regression analyses of mCherry-ir fiber density in the RP3V in relation to SCN transfection in VEH and PNA mice. *p < 0.05 unpaired t-test; ****p < 0.0001 ANCOVA.

Figure 3

PNA reduces SCN^AVP neuron fiber innervation of RP3V^KISS1 neurons. (A) Example maximum projection images showing mCherry-ir SCN^AVP neuron fibers (red) around kisspeptin-ir neurons (green) in the RP3V of VEH and PNA Avp-cre mice injected in the SCN with AAV-DIO-mCherry. Example appositions between mCherry-ir SCN^AVP neuron fibers and kisspeptin-ir somata are indicated with filled arrowheads. i) and ii) show single confocal planes and orthogonal views for two kisspeptin-ir neurons with appositions. Empty arrowheads point to kisspeptin-ir somata that do not receive such appositions. (B) Bar graph summarizing the average number of kisspeptin-ir neurons per section in the RP3V of VEH and PNA mice. (C) Mean proportion of kisspeptin-ir neuron somata with at least one close apposition by an mCherry-ir fiber. (D, E) Average numbers of appositions per kisspeptin-ir neuron (D) and per innervated kisspeptin-ir neuron (E). *p < 0.05 and **p < 0.01 unpaired t-tests.

Figure 4

SCN^AVP→RP3V^KISS1 neuron circuit function in diestrous VEH and PNA mice. (A) Experimental set-up. Cre-dependent AAV-DIO-ChR2-mCherry vectors were injected in the SCN of Avp-cre:Kiss1-hrGFP female mice. GFP-expressing neurons were targeted for cell-attached recordings in brain slices containing the RP3V and SCN^AVP neuron projection fibers stimulated with trains of blue LED light pulses. The recording protocol, with timing of light stimulation and periods considered for analysis [immediate (Imm.) and delayed (Del.)], is illustrated in the lower panel. (B) Example traces showing spontaneous action potential firing of RP3V^KISS1 neurons from VEH and PNA mice. Blue boxes indicate when the light stimulation occurred. (C) Summary graph of normalized RP3V^KISS1 neuron firing rate during (Imm.) and within 5 minutes after (Del.) optogenetic stimulation in VEH and PNA mice. (D) Proportions of RP3V^KISS1 neurons displaying responses during and after optogenetic stimulation in VEH and PNA mice. Numbers in bars are cell numbers. Scale bars = 50 pA/60 s.

Figure 5

RP3V^KISS1 neuron responses to exogenous AVP are maintained in PNA mice. (A) Example traces illustrating the effect of exogenous AVP (green bars) on RP3V^KISS1 neuron action potential firing in slices from VEH (top) and PNA (bottom) mice. (B) Bar graph summarizing AVP-induced changes in RP3V^KISS1 neuron firing rate in VEH and PNA mice. (C) Proportions of RP3V^KISS1 neurons responding to exogenous AVP in VEH and PNA mice. Numbers in bars are cell numbers. Scale bars = 50 pA/60 s.