Impaired astrocytic gap junction coupling and potassium buffering in a mouse model of tuberous sclerosis complex

Abstract

Abnormalities in astrocytes occur in the brains of patients with Tuberous Sclerosis Complex (TSC) and may contribute to the pathogenesis of neurological dysfunction in this disease. Here, we report that knock-out mice with Tsc1 gene inactivation in glia (Tsc1(GFAP)CKO mice) exhibit decreased expression of the astrocytic connexin protein, Cx43, and an associated impairment in gap junction coupling between astrocytes. Correspondingly, hippocampal slices from Tsc1(GFAP)CKO mice have increased extracellular potassium concentration in response to stimulation. This impaired potassium buffering can be attributed to abnormal gap junction coupling, as a gap junction inhibitor elicits an additional increase in potassium concentration in control, but not Tsc1(GFAP)CKO slices. Furthermore, treatment with a mammalian target of rapamycin inhibitor reverses the deficient Cx43 expression and impaired potassium buffering. These findings suggest that Tsc1 inactivation in astrocytes causes defects in astrocytic gap junction coupling and potassium clearance, which may contribute to epilepsy in Tsc1(GFAP)CKO mice.

Fig. 1

Astrocytic expression of the gap junction protein Cx43 is downregulated in Tsc1^GFAPCKO mice. (A, B) Representative western blots from neocortex and hippocampal extracts of brains from 4–5 week old Tsc1^GFAPCKO mice and littermate controls showed a decrease in Cx43 protein expression. Neuronal specific gap junction protein Cx36 is unchanged. β-actin is included as an internal control for protein loading. (C) After normalization to β-actin levels, quantitative summary of all experiments shows that Cx43 expression was decreased by ~40% in both neocortex and hippocampus in Tsc1^GFAPCKO mice compared with controls. The Cx43/β-actin ratio was normalized to the control group. (D) Although there appeared to be a slight increase in Cx36 expression in neocortex and hippocampus of Tsc1^GFAPCKO mice, there was no significant difference between the two groups. Con – control mice, KO - Tsc1^GFAPCKO mice. **p < 0.01, ***p < 0.001, by t-test (n = 9 mice/group).

Fig. 2

The decrease in Cx43 expression in Tsc1^GFAPCKO mice is mTOR-dependent. (A,B) Representative western blots show Cx43 expression in Tsc1^GFAPCKO mice and control mice administered 3mg/kg rapamycin or vehicle for 3 weeks starting at P14. While vehicle-treated Tsc1^GFAPCKO mice have significantly decreased Cx43 levels in both neocortex and hippocampus compared with control mice, rapamycin prevented this decrease in Cx43 expression in Tsc1^GFAPCKO mice. (C,D) Quantitative summary of all experiments demonstrates that rapamycin treatment of Tsc1^GFAPCKO mice increased Cx43 levels back to control levels. The Cx43/β-actin ratio was normalized to the vehicle-treated control group. Con_Ve – vehicle-treated control mice, Con_Ra – rapamycin-treated control mice, KO_Ve – vehicle-treated Tsc1^GFAPCKO mice, KO_Ra – rapamycin-treated Tsc1^GFAPCKO mice. ***p < 0.001, by ANOVA (n = 8 mice/group).

Fig. 3

Hippocampal astrocytic gap junction coupling is impaired in Tsc1^GFAPCKO mice. (A, B) Representative hippocampal sections showing the extent of biocytin-stained astrocytes from control mice and Tsc1^GFAPCKO mice following injection of a single astrocyte in stratum radiatum. Scale bars, 50 μm. (C, D) Representative hippocampal sections showing density of GFAP-positive astrocytes in control and Tsc1^GFAPCKO mice. Scale bars, 20 μm. (E) Typical “passive” current profile of recorded astrocytes. A range of depolarizing and hyperpolarizing voltage steps between −160 to +70 mV (holding potential −70mV) activated large time- and voltage- independent membrane currents. Scale bars, 500 pA, 10 ms. (F)Summarized data from all experiments showed reduced numbers of biocytin stained astrocytes in hippocampal slices from Tsc1^GFAPCKO mice mice. ***p < 0.001, by t-test (n = 20 injections, Tsc1^GFAPCKO mice; n=16 injections, control mice). (G) After injection of a single astrocyte in the stratum radiatum along the superficial surface of the hippocampal slice, 40 μm thick sections obtained from the original, injected 400 μm slice were processed and counted individually for biocytin-positive cells. The number of biocytin-positive cells decreased in sections more distant from the injected astrocyte (in Section 1 or 2) in both Tsc1^GFAPCKO and control mice and was lower in Tsc1^GFAPCKO mice in all sections. Dark triangle, control mice; open triangle, Tsc1^GFAPCKO mice.(H) When comparing the distribution of biocytin-positive cells in different strata within the hippocampus after injection of a single astrocyte in the stratum radiatum, control mice exhibited an expected decrease in coupled cells in the more distal stratum lacunosum-moleculare compared to the stratum radiatum. In contrast, Tsc1^GFAPCKO mice had equal numbers of coupled cells in these two strata. Thus, compared to control mice, the number of coupled cells in Tsc1^GFAPCKO mice was only decreased in the stratum radiatum, not the stratum lacunosum-moleculare. Con – control mice, KO – Tsc1^GFAPCKO mice, s.r – stratum radiatum, s.l.m – stratum lacunosum-moleculare. ***p < 0.001 by ANOVA.

Fig. 4

Potassium buffering is impaired in Tsc1^GFAPCKO mice. (A)Representative traces generated from a reference electrode for field potential (f.p.) recording (showing an antidromic population spike immediately following the stimulation artifact) and potassium-selective electrode in response to a 20Hz, 200 pulse train (starting at the arrow) at 100% stimulation intensity in hippocampal slices from control mice and Tsc1^GFAPCKO mice. Scale bars: for field potential 5mV, 25ms; for [K⁺]₀ 2mM, 5s. (B) Summary of maximal absolute rises in [K⁺]₀ (delta [K⁺]₀) evoked at 50 and 100% stimulation intensity. With both levels of stimulation, the peak increase in [K⁺]₀ was significantly great in Tsc1^GFAPCKO mice compared to controls. (C) Summary of maximal evoked [K⁺]₀ elicited by stimuli at 50 and 100% stimulation intensity, normalized to amplitudes of respective population spikes. With both levels of stimulation, the normalized peak increase in [K⁺]₀ was significantly great in Tsc1^GFAPCKO mice compared to controls. (D) Absolute rises in stimulation-induced [K⁺]₀ changes in the hippocampal slices before and after perfusion with the gap junction inhibitor, carbenoxolone. Carbenoxolone treatment induced a significant, additional increase in peak [K⁺]₀ in control mice, but had no effect in Tsc1^GFAPCKO mice. Black bars, before carbenoxolone; grey bars, after carbenoxolone. *p < 0.05, **p < 0.01, ***p < 0.001, by ANOVA (n = 6 mice/group).

Fig. 5

The impairment in potassium buffering in Tsc1^GFAPCKO mice is mTOR-dependent. (A,B) Summary of maximal rises in [K⁺]₀ evoked at 50 and 100% stimulation intensity in control mice, and untreated or rapamycin-treated Tsc1^GFAPCKO mice. nAs previously, with both levels of stimulation, the absolute rise in [K⁺]₀ (delta [K⁺]₀) (A) and the normalized peak increase in [K+]₀ (B) were significantly greater in Tsc1^GFAPCKO mice compared to controls. Rapamycin treatment of Tsc1^GFAPCKO mice decreased [K⁺]₀ levels back to control levels. *p < 0.05, **p < 0.01, ***p < 0.001, by ANOVA (n = 6 mice/group).