Gap Junctions As An Adaptive Response

Published Date: 02 Nov 2017

Newcastle University

Biomedical Sciences

Discuss the potential role of gap junctions in the pathophysiology of temporal lobe epilepsy

BMS2013

Name: Hannah Langlands

Student Number: 110191164

Due Date: 15/04/2013

Introduction

Temporal lobe epilepsy (TLE) is characterised by reoccurring spontaneous seizures which occur as a result of abnormal neuronal activity localised in brain structures such as the amygdala and hippocampus. Indeed, it is lesions of the hippocampus which have most frequently been associated with the onset of temporal lobe seizures (Engel, J., 2001). Such seizures arise as a result of a change in balance between inhibitory and excitatory transmission which ultimately favours an over excitation of a network of neurones. Initially, this over excitation was thought to be generated by a single considerable EPSP in synaptic transmission (Johnston and Brown, 1981). However, recent research has proved this inaccurate and has led to indicate an involvement of electrical transmission via gap junctions. Several hypotheses have suggested different roles of gap junctions as a fundamental neuronal mechanism which underlies temporal lobe epilepsy; these involve their ability to augment repeated synaptic signals to result in hypersynchrony of neurones as well as their ability to generate exceptionally high frequency oscillations.

It has been known that hippocampal tissue slices of an epileptic brain exert high frequency patterns of up to several hundred hertz during a seizure (Traub and Wong, 1982). In order to communicate this type of frequency, neuronal interactions must have a rapid and discrete impact on the neighbouring cell. On the basis of these properties, it becomes difficult to explain epilepsy via single excitatory synaptic interactions because of two reasons: firstly, synaptic EPSP transmission works on a scale of several ms which ultimately is too slow and secondly, synaptic EPSP transmission is also open to modification by additionally incoming presynaptic signals. This means, post EPSP effects are prone to being lost and thus unable to transmit the concluding high frequency pattern. Both features therefore suggest it highly improbable that sole synaptic transmission could result in such high frequencies. It then becomes plausible to suggest simultaneous, reoccurring chemical firing by many neurones which have been augmented by gap junctions is needed in order to generate such high frequency patterns (Traub, Cunnigham and Whittington, 2011).

Hypersynchrony

Gap junctions are primarily made up of two functional units known as connexins. Each connexon comprises of 6 transmembrane connexins (Cx) which surround a central pore and mediate the rapid, bidirectional, passive movement of ions, secondary messengers and other small molecules. Gap junctions in the CNS are often located on interneurons as well as astrocytes. The final function of these gap junctions is ultimately determined by a major connexin; interneuronal gap junctions express Cx36 and astrocytic gap junctions express Cx43 (Meldrum and Rogawski, 2007). These connexins regulate the extent to which gap junctions connect two cells together and their overall ability to mediate the formation of a syncytium in a neuronal network. On the dendrites of interneurons, these connexins regulate the synchrony of gamma oscillations which ultimately controls the firing of action potentials. It follows that changes in expression of these connexins may therefore alter this function and consequently may play a vital role in the generation of neuronal hypersynchony (Traub et al, 2001a). Indeed, it is this hypersynchrony of gamma oscillations which is thought to result in the augmentation of over excitation in neurones observed in epileptic seizures.

Anticonvulsant effects of specific drugs which target gap junctions provide evidence to illustrate an involvement of electrical transmission in seizure generation and therefore shows TLE not to be solely generated by a single chemical synapses. Medina-Ceja et al (2008) indicate how application of carbenoxone, a gap junction antagonist, almost entirely reduces the amplitude and frequency of epileptic discharges induced by 4-AP; overall, demonstrating the role of gap junctions in neuronal generation and propagation of seizure activity.

Initially, carbenoxone was thought to target GABAergic interneuronal gap junctions, as hypersynchrony of these interneurons could result in a spread of excitatory potentials which could therefore lead to the onset of an epileptic seizure. Naturally, Cx36 became a target of anti-epileptic drugs (AEDs). However, studies have revealed Cx36 knockouts to promote hyperexctiability of neurones and result in the generation of epileptic discharge; this proved Cx36 as an unreliable target of carbenoxolone and other developing AEDs. Subsequently, the antiepileptic affects of carbenoxolone was attributed to astrocyte gap junctions (Vos et al, 2010). Samoilova et al (2008) showed blocking of Cx43, in a rat hippocampus, by a synthetic mimetic peptide reduced coupling of gap junctions and attenuated spontaneous recurrent epileptic activity; suggesting the generation of epileptic seizures was dependent on astrocyte coupling by gap junctions. Furthermore, Takahashi et al (2010) not only showed an increase in Cx43 expression and thus an increase in coupled astrocytes in the early stages of epileptogenisis but also reduced overall excitation on Cx43 knockouts. Both studies illustrate an involvement of astrocytic gap junctions which are thought to facilitate the synchrony of neuronal excitation via the propagation of calcium waves and the promotion of glutamate release resulting in excess neuronal excitability and consequently seizure activity (Beaumont and Maccaferri, 2011).

Very Fast Oscillations

In a different way, gap junction coupling is also thought to contribute to the pathogenesis of TLE by the generation of very fast oscillations (VFO). Advances in electroencephalogram (EEG) recordings, used as a diagnostic tool in TLE, have revealed high frequency oscillation patterns to be strongly associated with the early stages of a TLE seizure (Roopun et al, 2010). On recordings, cellular behaviour preceding and at the initial stages of seizures, has been shown to reach frequencies between 80-500 Hz. This, in turn, could act as a significant functional indicator of the underlying mechanisms of epileptic discharge generation (Traub et al, 2001b). Although similar VFO expression (200 Hz ripples) has been observed in the cortex of non epileptic hippocampus tissue (Chrobakk and Buzsáki, 1996), this has been at lower amplitude. Hence, it is a more dominant VFO expression which is thought to underlie an epileptic seizure. This prevailing VFO has been associated with a dependence on, an ectopic location of, gap junctions which connect the axons of pyramidal hippocampal neurones (Roopun et al, 2010).

Originally, computing simulations predicated the vital involvement of axonal gap junctions (Traub and Bibbig, 2000). This model demonstrated how a network of pyramidal neurones could produce ripple oscillations, 80-200 Hz, dependent on gap junctions. This is illustrated in a block diagram in Figure 1A: When connected by gap junctions, pyramidal axons are able to work as a signal generator which aims to excite interneurons. After incoming excitation, via gap junctions, has overcome pyramidal hyperpolarisation, interneurons can become excited via AMPA receptors. This allows interneurons to fire at higher frequencies and thus produces VFO. Further in vivo studies have confirmed this fundamental role of gap junctions in the generation of VFO preceding epileptic activity and have also shown it to be independent of chemical synaptic transmission. Roopun et al (2010) established a reduction in VFO activity, in the hippocampus, once gap junction conduction had been inhibited by carbenoxolone. Furthermore, the study not only showed little correlation between synaptic inputs and VFO amplitude but on elimination of all synaptic transmission spontaneous VFO activity proved to continued; showing no involvement of chemical transmission in the generation of VFO and therefore also no involvement in the initiation of seizures.

Fig. 9.

Figure 1: Block diagram to illustrate two gap junction dependent networks believed to generate high-frequency oscillations (>100Hz) and gamma oscillations. (Traub and Bibbig, 2000)

It is also to be noted that both in vitro and in vivo studies have determined several conditions of gap junctions needed in order for the generation of VFO. Firstly, the number of gap junctions between axons must be adequate and each interneuron must be coupled to more than one axon (Traub et al, 2001b). However, this number must be neither too low (under 1.5 axon per interneuron) nor too high (more than 3 axon per interneuron) in order to generate sufficient excitation of the interneuron and consequent VFO activity (Traub et al, 2002). Furthermore, the conductance of these gap junctions much also be high enough in order to efficiently transfer action potentials between pyramidal cell axons. Finally, an initial spontaneous action potential, (Traub et al, 2001b., Traub et al, 2002) which has not been modulated, must occur; if repressed by an IPSP which is too large, gamma frequency oscillations, instead of VFO, are produced as a result and no seizure will arise (Traub and Bibbig, 2000).

Gap Junctions as an Adaptive Response

So far it has been illustrated that gap junctions play a notable role in the pathophysiology of TLE in two discrete ways: through the generation of hypersynchrony and the generation of VFO. Ultimately, both ways share the same result of an excess excitable neuronal gain which is the cause of seizure. Evidence has shown this to indicate an increase in gap junction activity is established in an epileptic brain, in comparison to a healthy non epileptic brain. Indeed, an increase in gap junction expression has been observed in hippocampal brain slices (Samoilava et al, 2003). However, in a contrast to VFO formation, this increase in gap junctions has been found to be expressed post-seizure as well as pre-seizure. With plausible evidence to suggest the VFO hypothesis correct, a paradox is formed; why would a neuronal network further increase the number of gap junctions if gap junctions are the initial cause seizures to begin with?

A recent model, by Volman et al, which aims to explain this paradox, hypothesises that gap junctions determine the stability of a neuronal network which ultimately establishes the likelihood of seizure. It is known that mature healthy individuals exhibit weak gap junction coupling which can be easily regulated to prevent excess neuronal activity. However, individuals suffering from epilepsy possess a more dynamic and adaptive response to incoming neuronal activity, with the ability to increase the number of gap junctions present. This increase successfully suppresses and stabilises ‘weak and transient’ excitatory neuronal activity. However, upon ‘persistent and strong’ neuronal inputs, a critical tipping point is reached by which this increase leads to an escalation of neuronal excitation by initial means of VFO and following hypersynchrony and therefore, a consequent seizure; Figure 2. Thus, gap junctions could serve beneficial in prevention of further escalations or indeed derogatory as they encourage augmentation of neuronal firing on continual stimulation. Overall, this computing model, based on interneurons, demonstrates a dynamic role of gap junction expression in epileptic patients which is ultimately determines the strength and duration of a seizure. Although based on interneurons, the model is suitable to be applied to any excitable system which can undergo gap junction coupling, for example glial cells. This further supports the previously suggested possible involvement of astrocytic gap juncitons as an alternative to interneurons.

Figure 2: Schematic presentation showing how correlative increase between neuronal activity and gap junctions can lead to seizure escalation. (Volman et al, 2011)

Figure 7 Schematic presentation of the effect that topological gap junction connectivity can have on the regulation of activity in networks that are prone to seizing.

Conclusion

To conclude, there is evidence to strongly suggest gap junctions are factors which promote epileptic seizure activity. However, the precise underlying mechanism is still to be confirmed. It is apparent that gap junctions ultimately generate augmentation of excitatory activity which is known to result in seizures. Studies have shown gap junctions function to form a neuronal syncytium in order to generate hypersychrony of neurones which in turn could contribute to the propagation of excitation and result in a seizure. Further investigation illustrated this not to be mediated by Cx36 but strongly associated with Cx45 found on astrocytic gap junctions. However, additional studies are needed to determine exactly which connexins are involved and to what extent they are modified. Furthermore, studies have also reliably shown gap junctions to mediate VFO activity which has been shown to precede and initiate seizures. Despite evidence to establish that chemical transmission is unable to solely generate TLE seizure, it seems that the epileptic effects of gap junctions are indeed dependent on an initial incoming chemical signal. Ultimately, it is the properties of this incoming signal which will determine if a total gain of excitation by gap junctions will occur and thus a resultant seizure. One proposal which associates all previous hypotheses is that individuals suffering from epilepsy have an already established predisposition to an increase in gap junction numbers, which in conjunction with a powerful and reoccurring chemical signal will result in both VFO and hypersynchrony of neurones to result in seizure activity. Therefore, overall, evidence would suggest gap junctions play a vital role in the cause of temporal lobe epilepsy and should be considered as a target for developing anti-epileptic drugs.

Word Count: 1942