In the present study, we identified a patient with ASD, which is possibly caused by two novel compound-heterozygous SCN10A mutations. Through electrophysiological examination, we found that both mutations lead to different degrees of loss-of-function in Nav1.8. In addition, behavioral experiments showed that Nav1.8 knockout mice present behavioral abnormalities that are in line with a typical ASD phenotype.
The most apparent symptoms of ASD comprise deficits in social interaction or repetitive, stereotypical behavior. Accordingly, these symptoms can be found in the diagnostic criteria of this group of neurodevelopmental disorders[40]. But additionally, a significant fraction of ASD patients also show abnormal reactions to environmental stimuli such as somatosensory input[41]. In fact, hyper- or hyporeactivity to different stimuli can act as predictor for diagnosis and/or severity of ASD[42, 43]. The social impairment that is most prominent in many ASD cases may also be the result of this altered perception of the environment. It has been shown that sensory abnormalities in early life relate to the development of social symptoms in children diagnosed with ASD[44].
The molecular mechanisms underlying these somatosensory abnormalities are associated with distinct peripheral somatosensory neurons, such as low-threshold mechanoreceptors (LTMRs) which are responsible for the perception of innocuous tactile stimuli[45, 46]. Apart from the mainly studied myelinated Aβ- and Aδ-LTMRs, there are also unmyelinated LTMRs (C-LTMRs, reviewed in [47]), that are suggested to play a role both in neuropathic pain[48] and affective or social touch[49], showing another connection between sensory perception and social interaction.
There is also growing evidence for an involvement of nociceptive neurons in ASD pathophysiology, as ASD patients often show abnormal reactions to potentially painful stimuli[50–52]. This involvement of nociceptors is well in line with the observations in our patient, which was reported to be less sensitive to painful stimuli. This fits our hypothesis of a role of Nav1.8 in the pathogenesis of ASD in the patient, because Nav1.8 is expressed in nociceptors and is important for their function.
Another link between ASD and Nav1.8 can be found in the example of SHANK3, which is a scaffolding protein that has been found to be responsible for a subset of monogenetic ASD cases[53, 54]. Mutations in SHANK3 are also linked to another neurodevelopmental disorder, the Phelan-McDermid syndrome[55]. This syndrome also shows prominent changes in somatosensory responses, namely hyperreactivity to light touch and at the same time hyporeactivity to painful stimuli[56, 57]. On a molecular basis, apart from different brain regions, SHANK3 can be found at presynaptic terminals of somatosensory neurons in the dorsal horn of the spinal cord[58, 59]. Han et al.[58] observed a marked reduction of nociceptive behavior in mice with SHANK3 haploinsufficiency. They found reduced pain responses in models of inflammatory as well as neuropathic pain especially concerning heat hyperalgesia. This effect could be narrowed down to interaction of SHANK3 with TRPV1 in neurons positive for Nav1.8.
VGSCs such as Nav1.8 are responsible for the fast upstroke of the action potential and are thus an important regulator of neuronal excitability. Neuronal excitability is commonly implicated in pathophysiological theories of ASD. Especially the theory of excitatory/inhibitory imbalance[9] offers a way of explaining ASD symptoms. Alterations of VGSCs could create such an imbalance and therefore cause autistic traits. Accordingly, VGSC isoforms expressed preferentially in the brain, such as Nav1.1 and Nav1.2[60–62], but also Nav1.7[63], which can be found in nociceptors, were already linked to ASD. Nav1.8, which is preferentially expressed in the peripheral nervous system, has so far not been associated with ASD. Nav1.8 was shown to be involved in diseases of the peripheral nervous system, such as chronic pain disorders[64–67], and arrhythmias[68]. The channel’s expression pattern and its associated diseases suggest a mainly peripheral effect Nav1.8 and its mutations. Our investigated mutations represent a loss-of-function of the channel. Interestingly, loss-of-function in the isoform Nav1.7, which is also mainly expressed in the peripheral nervous system, cause the disorder congenital insensitivity to pain (CIP)[69]. This disease does not show any signs of autistic symptoms in affected patients. Therefore, one could hypothesize that loss of function in Nav1.8 may affect peripheral neurons in a different way than loss of function in Nav1.7.
Another explanation would be an additional involvement of Nav1.8 also in the central nervous system. In fact, the channel can be found in the CNS[70], offering the possibility of a broader involvement in neurological diseases. Additionally, recent research from Kambouris et al. revealed mutations in SCN10A that can be linked to epilepsy[10].
The cardinal symptoms of our patient include mental retardation, reduced verbal skills and hyperactivity. Through next-generation sequencing including 6110 possibly related genes, we found two heterozygous SCN10A mutations (one missense mutation and one terminator mutation). Both of the two mutations were not reported before. We verified the screening results and investigated his parents using Sanger sequencing, which indicated that it is in accordance with the separation of Mendelian law. Both parents carry one of the mutations, and do not show signs of ASD. Thus, the genetic data indicated that the combined occurrence of the screened SCN10A mutation may account for the patient’s symptoms. However, SCN10A was mostly correlated to pain phenotypes before, questioning the mutations to be responsible for ASD. Thus, we further carried out whole genome sequencing and a whole exome sequencing, but failed to detect another potential mutation site except for the SCN10A mutations.
To investigate possible functional changes introduced by the two mutations, electrophysiological experiments were performed. The results of our examinations reveal that the p.R512X mutation leads to a complete functional knockout of the channel. Therefore, all Nav1.8 function in the patient’s neurons arises from channels carrying the p.I1511M mutation. Our detailed patch-clamp experiment regarding this mutation show a slight loss-of-function regarding a speeded entry of fast inactivation as well as an enhanced slow inactivation. We conclude that these changes in the channel’s biophysical properties may lead to impaired function of Nav1.8-expressing neurons and to an altered neuronal development. This alteration may have led to the deficits that the patient developed and which resulted in the diagnosis of ASD. Adding to our in vitro data from patch-clamp experiments, we investigated Nav1.8 knockout mice to check for ASD-like behavior. And in fact, we saw several changes in behavior compared to wild type mice that demonstrate autistic traits in the knockout mice.
However, there are some limitations in this study. First, we only verified the functions of p.I1511M and p.R512X mutations in vitro. The functions of these mutations in vivo need to be further verified. The mice with these mutations can be constructed to verify the mutation functions in the future investigations. Second, since we have only found one ASD patient which may relate to SCN10A mutations, the role of Nav1.8 in ASD needs to be confirmed in more cases. Third, the cellular mechanism underlying the effect of Nav1.8 on ASD needed to be explored in future studies.
In summary, this study expands the potential function of SCN10A in autism according to the evidence at molecular, in vivo and human level. Our work presents Nav1.8 as a protein potentially involved in ASD pathophysiology and adds to the knowledge about the genetic basis of the disorder.