Loss of MeCP2 increases GABA uptake by astrocytes to suppress tonic inhibition of CA1 pyramidal neurons

Abstract

Rett syndrome (RTT) is a neurodevelopmental disorder characterized a spectrum of phenotypes affecting neuronal and glial populations. Recent work by Dong et al. (Dong Q, Kim J, Nguyen L, Bu Q, Chang Q. J Neurosci 40: 6250–6261, 2020) suggests that augmented GABA uptake by astrocytes diminishes tonic inhibition in the hippocampus and contributes to increased seizure propensity in RTT. Here, I will review evidence supporting this possibility and critically evaluate how increased expression of a GABA transporter might contribute to this mechanism.

Rett syndrome (RTT) is a neurodevelopmental disorder caused by loss-of-function mutations in the gene encoding methyl-CpG-binding protein 2 (Mecp2), a transcription factor that can activate or repress transcription of a wide variety of genes [see the following reviews for details regarding how loss of Mecp2 impacts gene expression and function in neurons (Ref. 1) and glial cells (Ref. 2)]. Symptoms of RTT include autistic-like behavior, seizures, and severe respiratory dysfunction (3). MeCP2 is highly expressed in the brain, particularly neurons but also other cell types including astrocytes (4), and so multiple cell types as well as non-cell-autonomous effects contribute to features of RTT. In particular, inhibitory GABAergic signaling is preferentially suppressed in RTT (5). Consistent with this, conditional deletion of Mecp2 from forebrain GABAergic interneurons compromised inhibitory synaptic transmission and recapitulated several aspects of RTT (5). These results identify loss of GABAergic inhibitory signaling as an essential component of the pathogenesis of RTT. Furthermore, astrocytes have emerged as important regulators of excitatory and inhibitory synaptic transmission, in part, by rapidly removing transmitters from the interstitial space (6). Importantly, astrocyte specific re-expression of Mecp2 in global Mecp2 deficient mice improved several RTT-like features including anxiety, respiratory abnormalities, and mortality (7). Together these results beg the question, does Mecp2 deficiency disrupt astrocyte regulation of inhibitory signaling and contribute to the etiology of RTT?

Inhibitory signaling involves both phasic and tonic components. Phasic GABA inhibition is mediated at the synapse by GABA receptors with a low affinity for GABA and rapidly desensitize (8). Once released, GABA can diffuse from the synapse where it acts on nondesensitizing high-affinity extra-synaptic GABA receptors, which in the hippocampus contain α5-subunit-containing GABA receptors (8). The level of extracellular GABA is regulated by four members of the solute carrier 6 (SLC6) transporter family (GAT1-4) (9) that transport GABA typically in conjunction with Na⁺ and Cl⁻ at a stoichiometry of 2 or 3 Na⁺: 1 Cl⁻: 1 GABA (9). Of these, GAT1 and GAT3 are the most thoroughly characterized and evidence shows that pharmacological disruption of either or both transporters increased extracellular GABA levels and preferentially augmented desensitization resistant tonic inhibition (6). Although understudied, it is conceivable that increasing GAT function may lower extracellular GABA levels to the point of limiting tonic inhibition and increasing neural excitability as observed in RTT.

To address this, Dong et al. (10) tested the possibility that MeCP2 deficiency altered cell type specific GAT expression in the hippocampus and tonic inhibition of CA1 pyramidal neurons. They showed in vitro that amplitude of the tonic inhibitory current—measured as the change in holding current induced by bath application of the GABA_A receptor blocker bicuculline—was diminished in symptomatic Mecp2 null mice compared with controls. Amplitude of the tonic inhibitory current was similar between Mecp2 null and control mice at an earlier developmental time point before manifestation of RTT-like symptoms (10), suggesting that loss of tonic inhibition coincides with development of the RTT phenotype. This observation was also confirmed using an astrocyte specific inducible Mecp2 knockout model (10).

This deficit in tonic inhibition does not appear to involve loss of postsynaptic receptor function since CA1 neurons from both genotypes showed similar current responses to GABA receptor agonists as well as an inverse agonist specific to α5 subunit containing receptors that are thought to mediate tonic inhibition in this region (8). Furthermore, phasic GABAergic drive to CA1 neurons was similar between genotypes (10). However, the decay time constant of evoked inhibitory postsynaptic current, which is inversely related to the amount of GABA available to activate receptors, was enhanced in Mecp2 null compared to control (10), suggesting that GABA uptake was enhanced in the RTT model, Consistent with this, the addition of GABA to the perfusion medium normalized amplitude of tonic inhibition measured in hippocampal neurons from Mecp2 null and control mice (10).

Astrocytes are primarily responsible for regulating extracellular GABA levels and since these cells express GAT1 and GAT3, Dong et al. (10) used a combination of biochemistry, astrocyte electrophysiology, and transporter specific pharmacology to identify the basis for limited extracellular GABA in Mecp2 deficient mice. They found GAT3 but not GAT1 protein expression was increased in the hippocampus of MeCP2 null mice compared with control. At the astrocyte level, they show that GAT3- but not GAT1-sensitive current was upregulated in slices from Mecp2 null mice compared with control. At the neural level, they show that diminished tonic inhibition coincided with increased neural excitability and seizure-like activity, whereas selective blockade of GAT3 rescued both tonic inhibitory tone and decreased neural excitability in Mecp2 null mice. They also showed that systemic application of a selective GAT3 blocker lowered seizure propensity and improved mortality in Mecp2 null mice. These results suggest that GAT3 is a novel MeCP2 target in hippocampal astrocytes.

Interestingly, although MeCP2 is ubiquitously expressed throughout the brain, Dong et al. (10) found that loss of Mecp2 minimally affected tonic inhibition or GAT3 expression in the cortex, thus suggesting that MeCP2 does not regulate astrocyte GAT3 expression in cortical astrocytes. This is consistent with evidence that astrocytes are genetically and functionally heterogenous (11) and so loss of Mecp2 can differentially effect astrocytes in a brain region-specific manner. Also, considering inhibitory signaling is diminished in RTT (5) and as synaptic transmission is the main GABA source for tonic inhibition (8), it is not likely increased GAT3 expression is an adaptive response, but rather, is a consequence of diminished MeCP2 repression of Slc6a11 (the gene encoding GAT3). This scenario is in-line with the well-established role of MeCP2 as a transcriptional repressor (3); however, evidence for direct interactions between MeCP2 and Slc6a5 is lacking.

Another important consideration is transporter stoichiometry, which determines the extent to which GAT activity can lower extracellular GABA levels. Considering extra synaptic GABA receptors have high GABA affinity [EC₅₀ of ∼0.2 µM (12)], it has been assumed that GATs are unable to lower extracellular GABA concentration to levels that limit receptor activation (13). However, GATs with the generally accepted stoichiometry of 2 Na⁺: 1 Cl⁻: 1 GABA expressed in astrocytes with a hyperpolarized resting membrane potential (−80 mV) that precludes reverse mode operation of GAT, are expected to lower extracellular GABA to ∼32 nM at their equilibrium potential (Fig. 1). GATs with a 3Na⁺: 1 Cl⁻: 1 GABA stoichiometry have a higher transmitter concentration capacity and are expected to lower extracellular GABA to <1 nM even at considerably more depolarized potentials (Fig. 1). These estimates are consistent with the work of others (9, 14) and suggest that even under baseline conditions GATs are not operating at their equilibrium potential. Consistent with this, extrasynaptic GABA is estimated to be around 160 nM (15), a level four times greater than at the equilibrium potential 2 Na⁺: 1 Cl⁻: 1 GABA stoichiometry transporters. Therefore, it is possible that under normal conditions in the hippocampus, GABA flux into the extrasynaptic space exceeds transporter uptake capacity. Increasing the number of transporters will not alter the transporter equilibrium potential, but rather will improve the rate at which transporter can reach equilibrium, and thus lower extracellular GABA to levels that limit receptor activation. This interpretation is consistent with evidence from Mecp2 deficient animals showing that increased GAT3 expression corresponded with diminished tonic inhibition. It should be noted that inhibitory transmission is also diminished in Mecp2 null mice (5), which is expected to decrease extracellular GABA levels and tonic inhibition independent of GABA uptake. However, this concern is partly mitigated by evidence showing that blocking GAT3 increased tonic inhibitory currents to levels similar to control (10), suggesting there is sufficient GABA release in Mecp2 deficient mice to mediate tonic inhibition.

Figure 1.

Estimated extracellular GABA concentration at the equilibrium potential for GAT3. The predicted [GABA]o values were derived using the modified Nernst equation,

$$\left[\mathit{\text{GABA}}\right]\mathit{\text{o}}={\mathit{\text{e}}}^{\left(\frac{\mathit{\text{Vrev}}\times \mathit{\text{F}}({\mathit{\text{n}}}_{\mathit{\text{Na}}}{\mathit{\text{Z}}}_{\mathit{\text{Na}}}+{\mathit{\text{n}}}_{\mathit{\text{Cl}}}{\mathit{\text{Z}}}_{\mathit{\text{Cl}}})}{\mathit{\text{RT}}}\right)-\mathit{\text{ln}}\left\{{\left(\frac{[\mathit{\text{N}}{\mathit{\text{a}}}^{+}]\mathit{\text{o}}}{[\mathit{\text{N}}{\mathit{\text{a}}}^{+}]\mathit{\text{i}}}\right)}^{{\mathit{\text{n}}}_{\mathit{\text{Na}}}}{\left(\frac{\left[\mathit{\text{C}}{\mathit{\text{l}}}^{-}\right]\mathit{\text{o}}}{\left[\mathit{\text{C}}{\mathit{\text{l}}}^{-}\right]\mathit{\text{i}}}\right)}^{{\mathit{\text{n}}}_{\mathit{\text{Cl}}}}\right\}}\times [\mathit{\text{GABA}}]\mathit{\text{i}},$$

for both 2 Na⁺: 1 Cl⁻: 1 GABA and 3 Na⁺: 1 Cl⁻: 1 GABA transporter stoichiometry. The following values were used: T = 310 K, R = 8.314 J·K⁻¹·mol⁻¹, F = 96,485 C·mol⁻¹, nNa = 2 or 3, nCl = 1, nGABA zNa = +1, zCl = −1, [Na⁺]o = 145 mM, [Na⁺]i = 15 mM, [Cl⁻]o = 125 mM, [Cl⁻]i = 7 mM, and [GABA]i = 2 mM. At a potential of −80 mV (typical astrocyte resting membrane potential), the predicted [GABA]o using 3:1:1 and 2:1:1 stoichiometry was 0.32 nM and 61.25 nM, respectively.

Seizures are a prominent yet poorly understood feature of RTT. An important finding from Dong et al. (10) was that increased GAT activity and the resulting loss of tonic inhibition increased seizure-like activity in slices from Mecp2 null mice, whereas blocking GAT activity diminish seizure activity and prolong life in Mecp2 null mice. Furthermore, enhanced GAT activity is also associated with increased seizure propensity and epilepsy. For example, Schijns et al. (16) identified an Slc6a11 single nucleotide polymorphism in patients with temporal lobe epilepsy and febrile seizure+, suggesting that GAT-3 dysfunction and altered GABA clearance contributes to development of seizures and epilepsy. Furthermore, Zhang et al. (17) showed that the circadian clock transcriptional repressor Rev-erbα is upregulated in epileptic tissue and was found to increase seizure severity, in part, by promoting Slc6a11 expression and GABA uptake and limiting tonic inhibition. Therefore, understanding GAT function is broadly relevant to several disease states including RTT and epilepsy.

Another common feature of RTT that can negatively impact quality of life is disordered breathing, typified by forced or irregular breathing, hyperventilation, and frequent episodes of apnea (18). The basis of disordered breathing in RTT appears to involve disruption of central chemoreception, the mechanism by which the brain regulates breathing in response to changes in tissue CO₂, possibly due to disruption of inhibitory signaling. For example, previous work showed systemic application of a GABA uptake inhibitor improved breathing in Mecp2 deficient mice (19). Evidence also suggests that expression of extrasynaptic GABA receptors in a brainstem respiratory center called the locus coeruleus (LC) increased in Mecp2 null mice (20), possibly as a compensatory response to diminished tonic inhibition. However, it is not known whether enhanced GABA uptake by astrocytes is a contributing factor. On a related note, astrocyte dysregulation has been shown to contribute to loss of respiratory chemoreception at the level of the retrotrapezoid nucleus (RTN). For example, neurons (excitatory and inhibitory) and astrocytes in this region sense changes in CO₂/H⁺ to produce an integrated CO₂/H⁺-dependent drive to breathe. However, CO₂/H⁺-evoked Ca²⁺ responses were diminished in RTN astrocytes from Mecp2 deficient mice (21). It is not clear whether disruption of RTN astrocytes also impacts inhibitory tone in the RTN.

In summary, the work highlighted here identifies GAT3 in hippocampal astrocytes are novel targets of MeCP2 and potential therapeutic targets for treatment of neural hyperexcitability associated with RTT and epilepsy. Furthermore, my assessment of the extent to which GAT activity can lower extracellular GABA levels provides insight regulation of tonic inhibition.

GRANTS

This work was funded in part by National Institutes of Health (NIH)/F31NRSA Grant 1F31NS120467-01.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author.

AUTHOR CONTRIBUTIONS

B.M.M. prepared figures; drafted manuscript; edited and revised manuscript; approved final version of manuscript.