GABAergic interneurons, Decreased leads to Synaptogenesis, Decreased

Early in the development of the neocortex, GABAergic interneurons play a role in the formation of spontaneous synchronized activity, which has a fundamental role in the activation of glutamatergic synapses, the synchronization of synaptogenesis and the establishment of long–range cortico-cortical connections (Voigt et al., 2001; 2005). Increasing evidence suggests that GABAergic signaling is the main regulator of this early neuronal activity, as it is established before the glutamatergic one in the neocortex (Wang and Kriegstein, 2009; Owens et al., 1999; Tyzio et al., 1999; Hennou et al., 2002; Ben-Ari, 2006). Despite the fact that GABA is the main inhibitory neurotransmitter in the adult CNS, it exerts depolarizing actions in the immature brain (Ben-Ari et al., 2007), caused by the low levels of Cl^- concentration in the post-synaptic cells (Rivera et al., 1999; Ehrlich et al., 1999). K-Cl co-transporter 2 (KCC2) is the main Cl^- efflux mechanism with a developmentally-regulated expression profile in the brain and it is therefore thought to be the regulator of GABA signalling during early neuronal development. The effects of KCC2 on the levels of [Cl-]I in immature neurons and the subsequent effects on the shift of the GABA signaling has been extensively studied during the last decades:

• Existing data indicate that KCC2 expressed by GABA neurons is sufficient to shift from the depolarizing and excitatory period of GABA during cortical neuron development (Lee et al., 2005; Chudotvorova et al., 2005) and to effectively decrease the [Cl-]I in immature rat neurons (Chudotvorova et al., 2005).

• Transcriptional repression of KCC2 in rat cortical neurons delayed the GABA switch corresponding to significant changes of [Cl-]I in the same neurons (Yeo et al., 2009).

Several studies focused on the effects of GABA signaling on synaptogenesis and they all had convergent results leading to a strong biological plausibility of this KER.

• Too early shift of GABA-induced excitation-to-inhibition not only affects synaptic integration, but it also results in deficient circuitry development (Wang and Kriegstein, 2008). This has been demonstrated in rodents and mammals cortical neurons in culture.

• Premature GABA switch has also morphological effects in cortical neurons, as it has been shown to drive in fewer and shorter dendrites with defective effects in synaptic formation (Cancedda et al., 2007).

• In the dentate gyrus of the adult hippocampus, newborn granule cells are tonically activated by ambient GABA before being sequentially innervated by GABA- and glutamate-mediated synaptic inputs. GABA initially exerts an excitatory action on newborn neurons owing to their high cytoplasmic Cl^- ion content (Ge et al., 2006).

• An early hyperpolarizing shift in Cl⁻ reversal potential, by premature expression of KCC2, has been shown to increase the ratio of inhibitory-to-excitatory inputs both in Xenopus tectal neurons and rat cortical neurons in vitro (Chudotvorova et al., 2005; Akerman and Cline, 2006).

The mechanistic details of this relationship are not entirely known, but the most possible mechanism entails a functional relationship between GABA and NMDA receptor activation (Wang and Kriegstein, 2008; Cserép et al., 2012). Cortical neurons begin to express functional NMDA receptors when they migrate to the cortical plate, but these initial glutamatergic synapses are “silent” because of the Mg²⁺ block of NMDA receptors at the resting membrane potential (LoTurco et al., 1991; Akerman and Cline, 2006). GABAergic depolarization can facilitate relief of this voltage-dependent Mg²⁺ block and allow Ca²⁺ entry to initiate intracellular signalling cascades (Leinekugel et al., 1997). This mechanism suggests that the initial depolarizing GABAergic transmission is required for the formation of the glutamatergic synapses and is therefore responsible for the regulation of the balance between excitation and inhibition in the developing cortex (Wang and Kriegstein, 2009). 

The correlation between the GABA function and synaptogenesis has been mainly studied through the developmental modifications of intracellular Cl^- gradient and the subsequent GABA switch. In all available cases, this is performed by disturbing KCC2 or NKCC1 expression with genetic or mechanical manipulations of the neuronal models.

The temporal concordance of GABA shift and synaptogenesis is extensively reviewed by Ben-Ari et al., 2007 and 2012. It is widely accepted that the first spontaneous synaptic activity in the cortex is driven by the GABA-mediated depolarization and it is necessary for the subsequent synapse formation in the brain. Furthermore, the absence of T3 in cultures of cortical GABAergic interneurons can delay the typical developmental KCC2 up-regulation and subsequently the GABA shift, with a profound decrease in the number of synapses (Westerholz et al., 2010; 2013).

- Westerholz et al., 2013 This study showed that in rat T3-deficient cultures of cortical GABAergic PV⁺ interneurons, the number of synaptic boutons (presynaptic terminals containing the presynaptic marker synaptophysin) was reduced, an effect that was abolished after exogenous BDNF application.

- López-Espíndola et al., 2014: Human brain sections from MCT8-deficient subjects (30^th gestational week male fetus and an 11-year-old boy) were studied in comparison with relevant healthy control brain tissues. The MCT8-deficient fetal cerebral cortex showed 50% reduction of TH (i.e., T4, T3, and rT3), while T3 and T4 levels were normal in the liver. This TH deficiency in the brain produced an expected increase in type 2 deiodinase and decrease in type 3 deiodinase mRNA expression. Also, MCT8-deficient fetus showed a delay in cortical and cerebellar development and myelination, loss of parvalbumin (marker of GABAergic interneurons) expression, abnormal calbindin-D28k content, impaired axonal maturation (impaired synaptogenesis), and diminished biochemical differentiation of Purkinje cells. The 11-year-old boy displayed altered cerebellar structure, deficient myelination, deficient synaptophysin (presynaptic marker of synapse) and parvalbumin expression and abnormal calbindin-D28k expression.

Possible indirect KERs (not created due to limited evidence)

- Fisher et al., (2013), recently published a quantitative biologically-based dose-response model (BBDR) for iodine deficiency in the rat. In particular, HPT axis adaptations to dietary iodide intake in euthyroid (4.1-39 µg iodide/day) and iodide-deficient (0.31 and 1.2 µg iodide/day) conditions were evaluated. In rat pups that were iodide deficient during gestation and lactation, decreases in serum T4 levels were associated with declines in TH levels in the fetal brain. A 15% reduction in cortical T4 in the fetal brain was sufficient to induce permanent reductions in synaptic function in adults, confirming that even modest developmental TH disruption can cause synaptic dysfunctions also in hippocampal region (critical for learning and memory) of the adult brain (Gilbert et al., 2013). These data support indirect KER between decrease TH level in the brain and decreased synaptogenesis.

In regards to toxicological studies, bisphenol A (BPA), an environmental toxicant known to inhibit NIS-mediated iodide uptake (Wu Y et al., 2016), has also been found to delay and decrease both KCC2 expression and the developmental Cl^- shift (Yeo et al., 2013):

- Yeo et al., 2013 In primary rat cortical neurons and primary human cortical neurons (obtained from human fetal brain tissue specimens), 100 nM BPA caused decrease of KCC2 mRNA expression (≥ 25% decrease in rat cells at 4-5 DIV, ~ 70% decrease in human cells at 10 DIV) and attenuated [Cl⁻]_i shift in migrating cortical inhibitory precursor neurons. These data support indirect KER between NIS inhibition and GABAergic interneuron alteration.

These findings also concur with studies in other brain areas, such as the auditory brainstem and the hippocampus (Friauf et al., 2008; Hadjab-Lallemend et al., 2010), supporting correlation between KCC2 expression and GABA presence in the brain, and their implication in synaptogenesis.

In vivo evidence for the role of GABA in synaptogenesis is controversial. Ji et al., 1999 have shown that in GAD65/67-deficient mice, in which the production of GABA was reduced to less than 5%, the development of brain morphology until birth was normal. These mice die at birth and therefore synaptogenesis and circuit development could not be controlled, however no morphological defects were detected in the neocortex, cerebellum and hippocampus of these animals by the time of their death. The authors of this study suggested that GABA may not be crucial for development. However, functional changes were not assessed in this study. One hypothesis is that glutamate, glycine and taurine could compensate for the lack of GABA (LoTurco et al., 1995; Flint et al., 1998).

In KCC2 knock out mice, apart from lung atelectasis, no other obvious histological changes in the brain were observed in neonatal mice (Hubner et al., 2001). Moreover, these mice died at birth, before the GABA switch takes place, and neuronal electrical activity or synaptogenesis were not evaluated.

Additionally, after premature expression of KCC2 transporter an increase of the excitatory synapses was observed, but the glutamatergic synapses were not affected (Chudotvorova et al., 2005), as in the case of NKCC1 knock out mice (Wang and Kriegstein, 2008). These contradictory results reveal the complexity of the developmental brain and suggest that many different mechanisms are involved in the regulation of the temporal profile of the two main neuronal co-transporters, namely the KNCC1 and KCC2. However, in all cases the importance of Cl^- homeostasis in the developmental cortex and its correlation with the proper synapse formation is demonstrated.