Aminoacylation reaction refers to the formation of an ester bond between an amino acid and specific transfer RNA (tRNA), resulting in aminoacylated tRNA. It is involved in biological protein synthesis, mainly catalyzed by aminoacylated tRNA synthetase (aaRS) family enzymes (Wei et al., 2019). Based on enzyme catalytic domains and tRNA recognition sites, aaRS can be divided into classes I and II. Class II mainly performs biological functions in the form of homodimers (Banik and Nandi, 2012; Chaliotis et al., 2017). Asparaginyl tRNA synthetase (NARS) is a subtype of class II aaRSs, similar to most aaRSs. It includes the cytoplasmic type NARS1 and the mitochondrial type NARS2 (Vinogradova et al., 2021), both of which are associated with human neurodegenerative diseases. However, cytoplasmic and mitochondrial differences in protein localization and catalytic function can cause other diseases as well. Among them, heterozygous, biallelic NARS1 mutations can lead to autosomal recessive neurodegenerative disorders, with microscopic, language, and gain abnormalities (OMIM#619091), and autosomal dominant neuropathological disorders, with microscopic, language, epilepsy, and gait abnormalities (OMIM#619092). The main clinical characteristics are growth retardation, seizures, peripheral neuropathy, and ataxia (Manole et al., 2020).

NARS2 mutations can result in combined oxidative phosphorylation deficiency 24 (COXPD24, OMIM#616239). The main clinical characteristics are infantile seizures (which may progress to intractable epileptic retardation), overall growth retardation, mental retardation, hypotonia, and hearing impairment. Some cases may be accompanied by kidney disease or liver abnormalities (Simon et al., 2015; Sofou et al., 2015; Mizuguchi et al., 2017; Seaver et al., 2018; Štěrbová et al., 2021; Zhang et al., 2022). Compared to NARS1 mutations, NARS2 mutations are associated with earlier onset and worse condition of related nervous system diseases. To date, fewer than 30 NARS2 variants have been reported worldwide; therefore, the disease and the variation spectrum should be further explored (Štěrbová et al., 2021).

Herein, we report two infants with premature mitochondrial encephalopathy with epilepsia partialis continua (EPC). Genetic testing revealed novel compound heterozygous NARS2 mutations, including a rare intron mutation at the donor splicing site. The minigene experiment verified that the mutation led to abnormal splicing. The results of this study improved our understanding of the phenotype and the variation spectrum of COXPD24 and indicated that compound heterozygous NARS2 mutations are involved in the pathogenesis of nervous system disorders.

We reported two patients from two unrelated families with clinical characteristics of early-onset mitochondrial encephalopathy accompanied by EPC. The phenotypes of the infants were similar and included intractable EPC, hypotonia of the limbs, and general psychomotor retardation in early life. Brain MRI showed abnormal signals in the cortex and the hippocampus. Seizures in patient #1 was characterized by right limb twitching in the initial stage and left limb twitching 2 weeks later, not consistent with any epileptic type in previously reported cases. Three novel NARS2 mutations (p.Leu62Pro, p.Phe170Cys, and c.251 + 2T > G) were identified with genetic testing, including one non-coding sequence mutation. Our structural analysis of the two missense mutations showed that the decreased local structural hydrophobicity affected the structural stability. MD results suggested the consequent decrease in the dimer binding free energy of the NARS2 protein, possibly affecting the protein function. The minigene experiment showed that c.251 + 2T > G affected mRNA splicing and causing frameshift mutations, which may lead to reduced protein expression or degradation. Taken together, clinical examination and genetic analysis of two patients revealed a diagnosis of COXPD24, the role of NARS2 non-coding sequence mutations in the disease pathogenesis, and novel phenotype and mutation spectrum of COXPD24.

The disease phenotype spectrum resulting from biallelic NARS2 mutations is relatively broad. Most infants diagnosed with COXPD24 have sensorineural hearing loss in the early stage, followed by delayed neural development, intractable seizures, and hypotonia. MRI shows white matter lesions, mainly involving the frontal and parietal lobes and the deep white matter, and features of agenesis of the corpus callosum or diffuse brain atrophy (Simon et al., 2015; Sofou et al., 2015; Vanlander et al., 2015; Mizuguchi et al., 2017; Souza et al., 2017; Seaver et al., 2018; Han et al., 2019; Lee et al., 2020; Štěrbová et al., 2021; Tanaka et al., 2022; Vafaee-Shahi et al., 2022; Yagasaki et al., 2022; Yang et al., 2022; Zhang et al., 2022). The characteristics of Leigh syndrome can be observed in some patients’ brain MRI or autopsy results, including symmetric long T2 abnormal signals in the brain stem or the putamen or atrophy of the corpus callosum with lamellar necrosis and vacuolization (Simon et al., 2015; Tanaka et al., 2022). In addition, a few cases may be accompanied by renal dysfunction or elevated liver enzymes, possibly leading to a clinical misdiagnosis of Alpers syndrome (Seaver et al., 2018). The survival rate of patients with COXPD24 is low. The reported cases have shown a mortality rate of approximately 26%. The survivors may gradually lose their ability to take care of themselves and usually develop chronic progressive ataxia, severe psychomotor retardation or regression, loss of head control, and dysphagia (Souza et al., 2017; Tanaka et al., 2022). More rare phenotypes also include neonatal diabetes mellitus and metabolic acidosis (Yagasaki et al., 2022). Therefore, customized nursing for survivors of COXPD24 may help improve the quality of life. However, the complex and variable phenotypic disease characteristics pose some challenges to the accurate clinical diagnosis. As for the genetic etiology, including the three novel mutations identified in this study, 26 NARS2 mutations have been reported. However, according to the existing genotype–phenotype association analysis, whether the phenotype is non-progressive and mild or severe, infantile, and lethal cannot be determined. Further, no significant correlation seems to exist with the location or type of mutation (Sofou et al., 2015). However, this may also be because of the lack of sufficient genotype and phenotype data.

The NARS2 locus on chromosome 11q14.1 spans 14 exon regions and encodes a protein structure consisting of ABD (amino acids 38–122) and catalytical domain (amino acid 137–471), lacking the unique N-terminal extension structure of the NARS1 protein (Zhang et al., 2022). The literature review revealed 23 NARS2 mutations, most of which were missense mutations and mainly distributed in the catalytic domain region. In this study, three novel NARS2 mutations were identified, including the two missense mutations Leu62Pro and Phe170Cys located in the ABD region and the catalytic domain, respectively. The crystal structure of the human NARS2 protein has not been analyzed. Molecular modeling data showed that wild-type Leu62 and Phe170 were located in hydrophobic pockets and adjacent to the homodimer structure. The mutations decreased the local structural hydrophobicity. MD predicted that the mutations Leu62Pro and Phe170Cys decreased the dimer binding free energy of the NARS2 protein, possibly indicating that the mutation affected protein dimerization. The results of the hazard prediction algorithm for amino acid changes caused by these point mutations also indicated the existence of deleterious and damaging.

NARS2 non-coding sequence mutations have rarely been reported. Only one classical splice-site mutation has been reported in a pair of siblings with severe intellectual disability and status epilepticus (Mizuguchi et al., 2017). In this study, one patient with c.251 + 2T > G showed EPC at the age of 2 months, followed by progression to generalized seizures and death at the age of 11 months. This suggested that non-coding sequence mutations leading to mRNA splicing changes may aggravate the patient’s phenotype. c.251 + 2T > G is not included in the distribution frequency database of the normal population. The minigene experiment indicated that the mutation produced three mRNA products. Because of the small sequence length difference between these products, we differentiated them with cloning sequencing. The results showed that the three mRNA products led to truncated proteins. Thus, we strengthened the PS3 evidence of c.251 + 2T > G through experiments and further revealed that the non-coding sequence mutation played a role in the pathogenesis of COXPD24.

We performed genetic analyses of two infants with COXPD24 and identified three novel NARS2 mutations. Both patients showed early-onset mitochondrial encephalopathy with EPC, bilateral sensorineural deafness, hypotonia, and growth retardation. With the wide application of genetic testing in neurogenetic diseases, the novel clinical phenotype of COXPD24 would be increasingly reported, suggesting that clinicians should pay more attention to the characteristics of aaRS. The non-coding sequence mutation found in this study was confirmed to be a biohazard in the minigene experiment. The structural model analysis suggested that two point mutations might affect protein dimerization. However, this study had some limitations. First, we could not explore the mitochondrial respiratory chain function in the patients’ tissue samples, thereby leading to the lack of a correlation analysis between the mutation and cell function. Second, we could not perform subcellular localization experiments for the novel mutations or verify the interference of the mutation with protein dimerization. Nevertheless, the findings of this study improve our understanding of the NARS2 mutation profile and phenotype.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

The studies involving human participants were reviewed and approved by the Ethics Committee of Hunan Children’s Hospital. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin. Written informed consent was obtained from the individual(s), and minor(s)’ legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

WH and HF carried out the data curation and formal analysis, and wrote the original draft of the manuscript. YP, LL, QL, and WQ carried out the data curation, visualized the data, did the conceptualization, and performed the methodology. DG, JT, and JY carried out the software and experimental validation. LW and ZN supervised the data and reviewed and edited the manuscript. All authors contributed to the article and approved the submitted version.