Ventromedial hypothalamic glucose sensing and glucose homeostasis vary throughout the estrous cycle

Abstract

Objective: 17β-Estradiol (17βE) regulates glucose homeostasis in part by centrally mediated mechanisms. In female rodents, the influence of the ovarian cycle on hypoglycemia counterregulation and glucose tolerance is unclear. We found previously that in prepubertal females, 17βE modulates glucose sensing in nonadapting glucose-inhibited (GI) and adapting GI (AdGI) neurons within the ventrolateral portion of the ventromedial nucleus (VL-VMN). Nonadapting GI neurons persistently decrease their activity as glucose increases while AdGI neurons transiently respond to a glucose increase. To begin to understand if endogenous fluctuations in estrogen levels across the estrous cycle impact hypothalamic glucose sensing and glucose homeostasis, we assessed whether hypoglycemia counterregulation and glucose tolerance differed across the phases of the estrous cycle. We hypothesized that the response to insulin-induced hypoglycemia (IIH) and/or glucose tolerance would vary throughout the estrous cycle according to changes in 17βE availability. Moreover, that these changes would correlate with estrous-dependent changes in the glucose sensitivity of VL-VMN glucose-sensing neurons (GSNs).

Methods: These hypotheses were tested in female mice by measuring the response to IIH, glucose tolerance and the glucose sensitivity of VL-VMN GSNs during each phase of the estrous cycle. Furthermore, a physiological brain concentration of 17βE seen during proestrus was acutely applied to brain slices isolated on the day of diestrous and the response to low glucose in VL-VMN GSNs was assayed.

Results: The response to IIH was strongest during diestrous. The response of nonadapting GI and AdGI neurons to a glucose decrease from 2.5 to 0.5mM also peaked during diestrous; an effect which was blunted by the addition of 17βE. In contrast, the glucose sensitivity of the subpopulation of GSNs which are excited by glucose (GE) was not affected by estrous phase or exogenous 17βE application.

Conclusion: These data suggest that physiological fluctuations in circulating 17βE levels across the estrous cycle lead to changes in hypothalamic glucose sensing and the response to IIH.

Keywords: Estrogen; Glucose tolerance; Glucose-excited neurons; Glucose-inhibited neurons; Hypoglycemia counterregulation.

Figure 1. Glucose tolerance and the response to insulin-induced hypoglycemia vary with estrous phase

(A, B) Blood glucose levels in response to injection of D-glucose (A, 1 g/kg IP) or insulin (B, 2 U/kg, SC) in cycling females. Data sets marked with an asterisk (*) are significantly different according to repeated measures two-way ANOVA followed by Bonferoni post-hoc tests. (C, D) Quantification of AUC for glycemia. Baseline was set at 90mg/dL and 40mg/dL for GTT and IIH curves, respectively. Numbers within columns indicate the n number for each group. Columns which share the same symbols are not significantly different from each other via standard one-way ANOVA followed by Tukey post-hoc tests. AUC: area under curve, IP: intraperitoneal, SC: subcutaneous, n.s: not significantly different.

Figure 2. Glucose sensing in VL-VMN nonadapting GI and adapting GI neurons varies throughout the estrous cycle

Glucose sensing was evaluated in hypothalamic slices taken at each phase of the estrous cycle. (A) Representative whole cell current-clamp recordings of a nonadapting GI (left) and adapting GI (AdGI, right) neuron. Glucose changes are schematically displayed above each recording; dashed grey line represents resting Vm. (B, C) Quantification of %ΔVm and %ΔIR in response to decreasing glucose from 2.5 mM to 0.1mM (2.5G→0.1G) or 0.5mM (2.5G→0.5G) in nonadapting GI (B) and AdGI neurons (C). For nonadapting GI, n=7, 6, 5, 8 at 0.1mM glucose and n=5, 5, 5, 7 at 0.5mM glucose. For AdGI, n=23, 12, 10, 10 at 0.1mM and n=6, 7, 9, 7 at 0.5mM respectively. n# indicated in this figure legend is for diestrous, proestrous, estrous and metestrous, respectively. (D, E) Quantification of %ΔVm and %ΔIR in response to decreasing glucose from 2.5 mM to 0.1mM (2.5G→0.1G) or 0.5 mM (2.5G→0.5G) in nonadapting GI (D; prepubertal n=18 and diestrous n=8) and AdGI neurons (E; prepubertal n=37 and diestrous n=23) from prepubertal and diestrous females. *p<0.05, **<p<0.01 via unpaired student t-tests. Columns which share the same symbols are not significantly different from each other via standard one-way ANOVA and Tukey posthoc tests. G: mM glucose, IR: input resistance, n.s: not statistically different, Vm: membrane potential. ¹For AdGI neurons only, Bartlett’s test for unequal variances determined the variances between phases was significantly different (p=0.03) for %ΔIR in response to 0.1mM glucose.

Figure 3. Glucose sensing in VL-VMN GE neurons does not vary throughout the estrous cycle

Glucose sensing was evaluated in hypothalamic slices taken at each phase of the estrous cycle. (A) Representative whole cell current-clamp recording of a GE neuron. Glucose changes are schematically displayed above each recording; dashed grey line represents resting Vm. (B) Representative voltage responses to a hyperpolarizing pulse in response to 0.1mM glucose at each phase of the estrous cycle. Vm was normalized to 2.5mM glucose to emphasize changes in IR (C) Quantification of %ΔVm and %ΔIR in response to decreasing glucose from 2.5 mM to 0.1mM (2.5G→0.1G) (left; n=21, 9, 10, 7) or 0.5mM (2.5G→0.5G) (right; n=6, 4, 5, 5). n# indicated in this figure legend is for diestrous, proestrous, estrous and metestrous, respectively. (D) Quantification of %ΔVm and %ΔIR in response to decreasing glucose from 2.5 mM to 0.1mM (2.5G→0.1G) in GE neurons from prepubertal (n=17) and diestrous (n=7) females. n.s: not statistically different via standard one-way ANOVA and Tukey posthoc test. G: mM glucose, IR: input resistance, Vm: membrane potential.

Figure 4. 17βE blunts the response of VL-VMN nonadapting GI and AdGI neurons to low glucose

In this experiment, 17βE was added to hypothalamic slices taken during diestrus only. (A) Representative voltage responses to a hyperpolarizing pulse in response to 0.1mM glucose in the presence and absence of 17βE (4nM, diestrous only). Vm was normalized to 2.5mM glucose to emphasize changes in IR. (B, C) Quantification of %ΔVm and %ΔIR in response to decreasing glucose from 2.5 mM to 0.1mM (2.5G→0.1G) in nonadaptive GI (B left; n=4), AdGI (B right; n=6) and GE (C; n=6) neurons during diestrous in the presence and absence of 17βE (4nM). *p<0.05, +p<0.01 via paired students t-test. G: mM glucose, IR: input resistance, Vm: membrane potential.