Background
Multiple sclerosis (MS) is a chronic inflammatory demyelinating disease in which myelin sheaths wrapping the axons in the brain and spinal cord are damaged [
1]. The symptoms of the majority of patients with MS usually begin with alternating periods of neurological disability and recovery that can last for many years. Slowly expanding demyelination accompanied by axonal damage and neuronal degeneration is well-accepted pathological hallmarks of neurological deficits in MS, and mechanisms resulting in remyelination of axons are thought to be important for restoration of neuronal function following relapses [
2].
Myelin loss affects the integrity of neuronal networks and their synaptic plasticity [
3]. Abnormalities in the spontaneous firing patterns of neurons have been reported in both in vitro and in vivo models of peripheral demyelinated axons [
4‐
6]. In the central nervous system (CNS), myelin loss was demonstrated to cause highly heterogeneous alterations in nodes of Ranvier, thereby affecting the excitability of neocortical pyramidal neurons [
7]. Scarce data exist on the effect of remyelination on the neuronal network activities. In addition to neuronal damage, demyelination induced abnormal patterns of cortical synaptic plasticity in a group of MS patients [
8] and in different animal models of demyelination [
9]. Enhancement of synaptic plasticity in the motor cortex due to the alteration of the functional properties of surviving neuronal circuits has been suggested as a crucial factor for recovery from relapse-associated neuronal damage in MS patients [
10].
The auditory thalamocortical pathway is the only neural substrate that sends precise frequency information to the auditory cortex [
11]. Impaired thalamocortical auditory connectivity underlies the functional abnormalities in several neurological disorders, such as epilepsy, autism, and schizophrenia [
12‐
14]. Increasing evidence suggests extensive involvement of the thalamocortical pathway in patients with MS [
15]. Any disruption in this topographically organized circuit and distortions in the temporal processing of frequency information may trigger the symptoms observed in MS [
16,
17]. Indeed, the myelin sheath breakdown along the auditory pathway was accompanied by abnormalities of axonal conduction and auditory deficits in MS patients [
18]. Abnormalities in auditory evoked potentials and the cognitive P300 wave indicated dysfunction of different regions of the central auditory pathway and pointed to a predictive value of different evoked potentials [
19,
20]. Several toxins have been used to produce demyelination in animals, including cuprizone. Feeding of cuprizone, a copper-chelating mitochondrial toxin, induces oligodendrocyte injury and demyelination in different brain regions, preferentially in the corpus callosum. Cuprizone-induced demyelination is usually followed by complete remyelination within a few weeks. Thus, cuprizone treatment serves a useful animal model for investigation of the processes involved in demyelination and consequent remyelination [
21‐
23]. To study how demyelination may affect neuronal function and synaptic transmission of the thalamocortical system, we used the cuprizone-mediated demyelination/remyelination model to investigate single cortical neuron properties as well as synaptic plasticity under demyelinated and remyelinated conditions in the auditory cortex in vitro.
Discussion
Our results revealed the impact of myelin integrity on firing properties and synaptic transmission of neurons in the auditory thalamocortical system. Myelin loss and restoration affected cellular activities in thalamorecipient layer 4 as well as cortical synaptic transmission in the third layer of the auditory cortex. In addition, the early and late phases of remyelination differentially changed neuronal excitability and synaptic plasticity of cuprizone-treated mice.
Myelination of the thalamic projections innervating the auditory cortex begins around 1 year of age and progresses until the fourth year [
31]. The main thalamocortical input from the ventral part of the medial geniculate body of the thalamus is directed in a tonotopic manner to layer 4 of the A1 [
32]. The information from cells in layer 4 of the auditory cortex project to the pyramidal neurons of layer 3, and from there, the inputs are distributed to the other cortical layers of both the ipsilateral and the contralateral auditory cortex through the corpus callosum. The corpus callosum, the largest fiber tract in the brain, is affected in most cases of MS patients as well as in cuprizone-treated mice [
33,
34]. The isthmus of the corpus callosum contains fibers from the motor, the somatosensory, and the primary auditory cortices [
34]. Our results revealed that demyelination altered thalamocortical inputs in primary auditory cortex, which might produce difficulty with sensory perception in MS [
35]. Using diffusion tensor imaging fiber tracking, a tenfold higher density of lesions in thalamocortical projections compared to other brain white matter regions was observed [
36].
Our data revealed a decreased excitability of layer 4 neurons from A1 in cuprizone-treated mice, characterized by more negative RMPs, longer evoked spiking intervals, and smaller EPSPs. Axonal demyelination increases stability of the membrane potential and leads to impaired neuronal function in demyelinating diseases by increasing the input capacitance and changing the surface exposure of ionic channels, especially K
+ channels [
37]. In the present study, layer 4 neurons revealed a significant hyperpolarization of the RMP after cuprizone treatment. Intracellular recording from A1 neurons has shown that depolarization and hyperpolarization of the membrane potential underlie excitatory-inhibitory response properties to thalamic inputs [
38]. Alterations of depolarizing shifts have been also reported in the membrane potential of peripheral nerves when recordings were performed near the site of demyelination [
5]. In contrast to layer 4 neurons in our study, demyelination of the axon of layer 5 neurons in the auditory cortex causes spontaneous intrinsic and network excitability of pyramidal neurons of cuprizone-treated mice [
7]. In addition to the well-known effect of layer 4 neurons to drive feed-forward and recurrent inhibition in other cortical layers [
39,
40], it has been shown that enhanced layer 4 activity directly suppressed layer 5 neurons by activating deep, fast-spiking inhibitory neurons in awake, behaving mice [
41]. A pronounced reduction in layer 4 activity by hyperpolarization of neurons enhanced layer 5 firing and broadened the representation of horizontal space across the population of layer 5 regular spiking neurons [
41]. Hyperpolarization and decreased excitability of layer 4 neurons observed in cuprizone-treated mice in our study caused disinhibition and thereby increased excitability in other cortical layers. Disturbed spatiotemporal connections between different areas of cortical networks have been suggested to underlie cognitive impairment, commonly observed in MS patients [
41,
42]. Damage to heterogeneous profiles of myelination of different neocortical layers may disorganize different arrays of inter-layer communication and prevent the emergence of complex cellular behaviors [
43].
In general, it has been shown that the spatiotemporal patterns of de- and remyelination vary across the brain after 6 weeks of cuprizone administration and subsequent remyelination. This suggests varying susceptibility to injury and/or ability to repair in the brain in the cuprizone mouse model [
44]. Therefore, partially diversified results may be expected. Following cessation of a 6-week cuprizone diet, demyelinated lesions demonstrate about 50 % recovery after 1 week and an almost complete remyelination was observed within 4 weeks [
45]. Our data indicate a partial recovery of neuronal excitability and synaptic plasticity in the early and late phases of remyelination. In keeping with our results, it has been shown that while immunohistochemical staining points to extensive remyelination, neurotransmission along previously demyelinated neuronal tissue has not been completely recovered to normal [
46]. Following 6 weeks of cuprizone ingestion, the recovery of action potentials was incomplete in an ex vivo slice preparations model even after feeding with normal chow for another 6 weeks [
47]. In a similar way, cuprizone feeding increased the response latency between the left and right sensorimotor cortices and only partial recovery of axonal conduction was observed after remyelination [
48]. Early periods of remyelination, prior to new myelin formation, are associated with a redistribution and reaggregation of sodium channels close to the lesion site, a process that may render neurons more susceptible to injury [
49]. It has been shown that production of de novo synapses with recruited oligodendrocyte progenitor cells by demyelinated bioelectrically active axons and release of glutamate are crucial for remyelination and recovery of lost function [
50]. Different subtypes of sodium channel that are expressed during remyelination, as well as the effect of inflammatory mediators on its function, may be responsible for alterations of neuronal properties in the early and late phases of myelin restoration [
51].
In our study, EPSPs evoked by stimulation of the thalamocortical projections in the control group were mainly monosynaptic in its early part and polysynaptic in its late part. The late components appeared with a smooth time course since the velocity of transmission in the pathways involved was not very different. Cuprizone-induced demyelination flattened the EPSPs as expected for demyelination. The late EPSPs became multiphasic during the remyelination process after both 7 and 25 days after withdrawal of cuprizone. It can be assumed that this is due to the alteration of the synaptic network which followed alterations in axonal conduction velocities on the basis of partial and/or patchy remyelination, thus leading to prolonged multiphasic EPSPs [
52].
Our data indicated that demyelination reduced the amplitude of EPSPs and decreased synaptic plasticity at cortical synapses in the auditory cortex. Remyelination after 7 days restored the synaptic strength in these synapses. However, further remyelination was associated with impairment of LTP. In vivo intracellular recordings of the primary auditory cortex have shown that modulation of synaptic plasticity in this region is a prominent feature of synaptic responses to auditory stimuli [
53]. These observations are in keeping with our results that LTP in the hippocampal CA1 area was decreased and LTP-related spatial memory was impaired in animal models of MS [
54,
55]. Conversely, some other studies found no differences in hippocampal LTP between demyelinated and normal neuronal network [
56]. Clinical studies indicate that the degree of neuroinflammation is associated with changes of LTP in MS patients [
10]. It has been suggested that synaptic plasticity may be enhanced or impaired during experimental and clinical demyelination-remeylination, depending on the degree of neuroinflammation and the time point at which it is investigated [
57].
Abbreviations
AHP, afterhyperpolarizations; APs, action potentials; EPSPs, excitatory postsynaptic potentials; fEPSPs, field excitatory postsynaptic potentials; IPSPs, inhibitory postsynaptic potentials; LTP, long-term potentiation; MS, multiple sclerosis; PFA, paraformaldehyde; PLP, myelin proteolipid protein; RMP, resting membrane potential; THP, the threshold potential of action potentials