Among the three major Dst isoforms (Figure 1A), the Dst-a and Dst-b isoforms share the most exons, and the Dst-b isoform has five additional isoform-specific exons (Figure 1B). We generated a novel Dst-b mutant allele in which a nonsense mutation was introduced into a Dst-b-specific exon using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) genome editing method. The Dst-b mutant allele harbors a nonsense mutation followed by a XhoI site (Figure 1B). Restriction fragment length polymorphism (RFLP) analysis of PCR products was used to genotype mice (Figure 1C). Hereafter, this Dst-b mutant allele is referred to as the Dst-b^E2610Ter allele. In the heterozygous cross, approximately one-fourth of the progeny were Dst-b homozygous (Dst-b^{E2610Ter/E2610Ter}) mice, according to the Mendelian distribution. Western blotting was performed using an anti-Dst antibody that recognizes both Dst-a and Dst-b proteins. Dst-b bands were detected in tissue extracts from heart and skeletal muscle of wild-type (WT) mice, whereas shifted bands were observed in those of Dst-b^E2610Ter homozygous mice (Figure 1D). We analyzed three independent Dst-b^E2610Ter mouse lines, and similar truncations of the Dst protein were observed in the homozygotes (Figure 1—figure supplement 1). In brain tissue extracts, Dst-a bands were equally detected between WT and Dst-b^{E2610Ter/E2610Ter} mice. Quantitative PCR (qPCR) analyses indicated that Dst-b mRNA was significantly reduced by approximately half in the cardiac tissues of Dst-b^{E2610Ter/E2610Ter} mice compared with those of WT mice, while Dst-a mRNA expression was unchanged (Figure 1E). Among three distinct isoforms of Dst-b (Dst-b1, Dst-b2, and Dst-b3) differing in their N-terminus (Jefferson et al., 2006), the expression level of Dst isoform1 mRNA was remarkably reduced in the cardiac tissue of Dst-b^{E2610Ter/E2610Ter} mice, whereas there was no significant change in the amounts of isoform 2 and isoform 3 mRNAs (Figure 1—figure supplement 2A). Similar changes in mRNA expression of Dst isoforms were observed in the soleus (Figure 1—figure supplement 3A). The frequency of exon usage was compared between cardiac and brain tissues (Figure 1—figure supplement 2B). Dst-b-specific exons were exclusively used in the heart and scarcely used in the brain. In the heart, the first exons of Dst isoform 1 and isoform 3 were more frequently used than those in the brain. However, Dst isoform 2 was predominantly used in the brain and weakly expressed in the heart. Thus, these findings suggest that Dst isoforms are differentially expressed among tissues and that the Dst-b^E2610Ter allele affects each Dst isoform to a different degree.

Dst-b^{E2610Ter/E2610Ter} mice had a normal appearance and seemed to have a normal life span. In the tail suspension test, 1-month-old Dst-b^{E2610Ter/E2610Ter} mice maintained normal postures, while gene trap mutant (Dst^Gt/Gt) mice with the dt phenotype displayed hindlimb clasping and dystonic movement (Figure 2A). Next, histological analysis was performed on the muscle spindles and dorsal root ganglia of 1-month-old mice, which are the main affected structures in dt mice. Dst-b^{E2610Ter/E2610Ter} mice exhibited normal muscle spindle structure, although the muscle spindles were markedly atrophied in Dst^Gt/Gt mice (Figure 2B). Similarly, neurofilament (NF) accumulation and induction of ATF3, a neural injury marker, were observed in the dorsal root ganglia of Dst^Gt/Gt mice but not in those of WT and Dst-b^{E2610Ter/E2610Ter} mice (Figure 2C–E). We also found that parvalbumin (Pvalb)-positive proprioceptive neurons were normally observed in Dst-b^{E2610Ter/E2610Ter} mice (Figure 2F and G) but were markedly decreased in Dst^Gt/Gt mice as previously described in other dt mouse lines (Carlsten et al., 2001). We next investigated the gross phenotypes of Dst-b^{E2610Ter/E2610Ter} mice over an extended time. Dst-b^{E2610Ter/E2610Ter} mice normally gained body weight until 1 year of age; then, they became lower body weight compared with WT mice after 1 year of age (Figure 2H). Some Dst-b^{E2610Ter/E2610Ter} mice exhibited kyphosis (data not shown). We have reported that dt mice display impairment of motor coordination as assessed by the rotarod test and wire hang test (Horie et al., 2020); however, WT and Dst-b^{E2610Ter/E2610Ter} mice older than 1 year exhibited normal motor coordination (Figure 2I). These data also support the idea that deficiency of the Dst-a isoform, but not the Dst-b isoform, is causative of dt phenotypes, including sensory neuron degeneration, abnormal movements, and postnatal lethality.

To further examine the skeletal muscle phenotype in aged Dst-b^{E2610Ter/E2610Ter} mice, we performed histological analyses. In the soleus muscle of Dst-b^{E2610Ter/E2610Ter} mice, small-caliber muscle fibers and centrally nucleated fibers (CNFs), which indicate muscle degeneration and regeneration, were frequently observed in mice at 16–23 months of age (Figure 3A) but were rare in 3–4-month-old Dst-b^{E2610Ter/E2610Ter} mice (Figure 3—figure supplement 1A). The percentage of CNFs was significantly increased in the soleus, gastrocnemius, and erector spinae muscles, with different severities (Figure 3B, Figure 3—figure supplement 2). Muscle mass of soleus normalized by body weight was not significantly different between control and Dst-b^{E2610Ter/E2610Ter} mice (control: 0.258 ± 0.012 [mg/g body weight] vs. Dst-b^{E2610Ter/E2610Ter}: 0.279 ± 0.015 [mg/g body weight]; control, n = 6 muscles from three mice; Dst-b^{E2610Ter/E2610Ter}, n = 6 muscles from three mice; p=0.3056, Student’s t-test). Distribution of cross-sectional area in the soleus showed that small-caliber myofibers were abundant in Dst-b^{E2610Ter/E2610Ter} mice compared with WT mice (Figure 3C). Masson’s trichrome staining indicated skeletal muscle fibrosis and, particularly, a marked increase in the connective tissue surrounding small-caliber muscle fibers (Figure 3D and E) in the soleus muscle of Dst-b^{E2610Ter/E2610Ter} mice. Cardiac fibrosis was also observed in Dst-b^{E2610Ter/E2610Ter} mouse hearts (Figure 3D and E) but was not observed in the hearts of 3–4-month-old Dst-b^{E2610Ter/E2610Ter} mice (Figure 3—figure supplement 1B).

To assess the influence of the Dst-b mutation on cardiomyocytes, expression of the heart failure markers natriuretic peptide A (Nppa) and natriuretic peptide B (Nppb) (Sergeeva and Christoffels, 2013) was investigated. Nppa mRNA was markedly upregulated in the myocardium of the left ventricle in 16-month-old Dst-b^{E2610Ter/E2610Ter} mice and was not increased in that of WT mice (Figure 3F). Nppb mRNA was also upregulated in the left ventricle myocardium in Dst-b^{E2610Ter/E2610Ter} mice (Figure 3F). qPCR analysis also demonstrated increased expression of Nppa and Nppb mRNAs in the cardiac tissue of Dst-b^{E2610Ter/E2610Ter} mice (Figure 3G). Expression of profibrotic cytokines, including transforming growth factor beta-2 (Tgfb2) and connective tissue growth factor (Ctgf) mRNAs, was also remarkably increased in Dst-b^{E2610Ter/E2610Ter} mice (Figure 3H). At 3–4 months of age, the Nppa transcript was slightly upregulated in Dst-b^{E2610Ter/E2610Ter} mouse hearts without obvious cardiac fibrosis (Figure 3—figure supplement 1B). To assess the cardiac function, ECGs were recorded from anesthetized WT and Dst-b^{E2610Ter/E2610Ter} mice at 18–25 months of age (Figure 3I). Dst-b^{E2610Ter/E2610Ter} mice exhibited remarkably prolonged QT intervals compared with those of WT mice (Figure 3J), which suggests abnormal myocardial repolarization. However, the RR interval and QRS duration were not different between WT and Dst-b^{E2610Ter/E2610Ter} mice (Figure 3J). In addition, premature ventricular contractions (PVCs) were recorded in two of nine Dst-b^{E2610Ter/E2610Ter} mice analyzed (Figure 3K). Individual Dst-b^{E2610Ter/E2610Ter} mice harboring PVCs showed severe cardiac fibrosis and remarkable upregulation of Nppa mRNA in the ventricular myocardium (data not shown). Such ECG abnormalities were not observed in 3-month-old Dst-b^{E2610Ter/E2610Ter} mice (RR of WT mice: 134.2 ± 26.6 ms vs. Dst-b^{E2610Ter/E2610Ter} mice: 138.8 ± 33.5 ms; p=0.8234, Student’s t-test; QRS of WT mice: 17.3 ± 2.0 ms vs. Dst-b^{E2610Ter/E2610Ter} mice: 17.4 ± 1.2 ms; p=0.9514, Student’s t-test; QT of WT mice: 31.7 ± 4.3 ms vs. Dst-b^{E2610Ter/E2610Ter} mice: 31.2 ± 2.3 ms; p=0.8599, Student’s t-test; WT, n = 4 mice; Dst-b^{E2610Ter/E2610Ter}, n = 5 mice). These data indicate that the Dst-b isoform-specific mutation causes late-onset cardiomyopathy and cardiac conduction disturbance.

MFM is a type of hereditary myopathy that is defined on the basis of common pathological features, such as myofibril disorganization beginning at the Z-disks and ectopic protein aggregation in myocytes (De Bleecker et al., 1996; Nakano et al., 1996). Desmin is an intermediate filament protein located in Z-disks, and aggregation of desmin and other cytoplasmic proteins is a pathological hallmark of MFM (De Bleecker et al., 1996; Clemen et al., 2013; Batonnet-Pichon et al., 2017). First, we performed desmin immunohistochemistry (IHC) and found abnormal desmin aggregation in the skeletal muscle of Dst-b^{E2610Ter/E2610Ter} mice at 16–23 months of age (Figure 4A and B). In the soleus of Dst-b^{E2610Ter/E2610Ter} mice, desmin massively accumulated in the subsarcolemmal region, whereas in that of WT mice, desmin protein was predominantly located underneath the sarcolemma (Figure 4A). Desmin aggregates were also observed in the cardiomyocytes of Dst-b^{E2610Ter/E2610Ter} mice older than 16 months (Figure 4—figure supplement 1). In longitudinal sections of Dst-b^{E2610Ter/E2610Ter} soleus muscle, localization of desmin to the Z-disks was mostly preserved; however, displacement of Z-disks was occasionally observed in the muscle fibers bearing subsarcolemmal desmin aggregates (Figure 4C). At 3–4 months of age, desmin aggregates were scarcely observed in Dst-b^{E2610Ter/E2610Ter} mouse soleus (Figure 3—figure supplement 1A).

Next, we investigated the molecular composition of the protein aggregates in Dst-b^{E2610Ter/E2610Ter} mouse muscle fibers. We found that αB-crystallin was co-aggregated with desmin in the subsarcolemmal regions of the Dst-b^{E2610Ter/E2610Ter} soleus (Figure 4D). αB-crystallin is a small heat shock protein that binds to desmin and actin and is known to accumulate in muscles of patients with MFMs (Clemen et al., 2013). In addition, plectin had also accumulated in desmin-positive aggregates of the Dst-b^{E2610Ter/E2610Ter} soleus (Figure 4E). Plectin is a cytoskeletal linker protein that is normally located in the Z-disks. Because protein aggregates in muscles affected by MFM often contain the Z-disk protein encoded by the causative gene itself, we examined subcellular distribution of myotilin that is a Z-disk component encoded by a causative gene for MFM. We found that myotilin was accumulated in some myofibers of the Dst-b^{E2610Ter/E2610Ter} soleus (Figure 4—figure supplement 2). Furthermore, we evaluated the subcellular distribution of truncated Dst proteins in the Dst-b^{E2610Ter/E2610Ter} soleus. As expected, truncated Dst proteins accumulated in subsarcolemmal aggregates of desmin in the Dst-b^{E2610Ter/E2610Ter} soleus (Figure 4F). In addition, some small-caliber myofibers exhibited cytoplasmic accumulation of truncated Dst proteins surrounding nuclei. Dst proteins were also distributed on Z-disks (Figure 4G), as reported in previous studies (Boyer et al., 2010a; Steiner-Champliaud et al., 2010). Truncated Dst proteins were also localized in normally shaped Z-disks in the Dst-b^{E2610Ter/E2610Ter} soleus and were dispersed from disorganized Z-disks (Figure 4G). Taken together, these data indicate that protein aggregates in Dst-b^{E2610Ter/E2610Ter} mouse muscles are composed of a variety of proteins, similar to aggregates found in muscles affected by other MFMs.

Mitochondrial abnormalities have also been reported in muscles of patients with MFMs and animal models (Joshi et al., 2014; Bührdel et al., 2015; Winter et al., 2016). We investigated the mitochondrial distribution using cytochrome C and Tom20 IHC and found abnormal accumulation of mitochondria underneath the sarcolemma in some fibers of Dst-b^{E2610Ter/E2610Ter} soleus muscles (Figure 5A). Double staining with cytochrome C and desmin revealed that mitochondrial accumulation occurred in the same region as desmin aggregates (Figure 5C). We also found a strong phosphorylated PERK signal, an organelle stress sensor, in the subsarcolemmal spaces with accumulated mitochondria (Figure 5D).

Next, we quantified the mRNA expression levels of several genes responsible for oxidative phosphorylation (Ndufb8, Sdha, Uqcrc2, Cox4i1, and Atp5a1) in the soleus (Figure 5E). The expression of several genes, including Ndufb8, Sdha, and Cox4i1, was significantly reduced in Dst-b^{E2610Ter/E2610Ter} mice, suggesting mitochondrial dysfunction in the skeletal muscle of Dst-b^{E2610Ter/E2610Ter} mice. Changes in genes responsible for oxidative phosphorylation suggest the possibility of muscle fiber type switch and/or oxidative stress (Bonnard et al., 2008; Zhang et al., 2017). However, quantification of expression levels of muscle fiber-type marker genes and oxidative stress-related genes showed no changes in the soleus of Dst-b^{E2610Ter/E2610Ter} mice (Figure 1—figure supplement 3B and C). At 3–4 months of age, mitochondrial accumulation was already observed in soleus of Dst-b^{E2610Ter/E2610Ter} mice before the appearance of CNF and desmin aggregates (Figure 3—figure supplement 1A). Therefore, mitochondrial abnormalities may precede protein aggregation and muscle degeneration.

Transmission electron microscope (TEM) analysis was performed to further investigate the ultrastructural features of skeletal and cardiac muscles of Dst-b^{E2610Ter/E2610Ter} mice. In the Dst-b^{E2610Ter/E2610Ter} soleus, focal dissolution of myofibrils and Z-disk streaming were observed at 22 months of age (Figure 6A). Subsarcolemmal regions of the Dst-b^{E2610Ter/E2610Ter} soleus were often filled with accumulated mitochondria and granulofilamentous materials including focal accumulation of glycogen granules, electron-dense disorganized Z-disks, and electron-pale filamentous materials (Figure 6B). These observations have been reported as typical features of MFMs (Batonnet-Pichon et al., 2017). Furthermore, we found dysmorphic nuclei in the soleus of Dst-b^{E2610Ter/E2610Ter} mice (Figure 6C).

Focal myofibrillar dissolution and mitochondrial accumulation were observed in Dst-b^{E2610Ter/E2610Ter} mouse cardiomyocytes (Figure 6D). We often observed an invaginated nuclear envelope and a structure similar to the nucleoplasmic reticulum (Bezin et al., 2008; Malhas et al., 2011) with cytoplasmic components in the nucleus (Figure 6E). Interestingly, inclusions containing crystalline structures were frequently observed in dysmorphic nuclei, which were a unique feature of Dst-b^{E2610Ter/E2610Ter} cardiomyocytes. Only a limited number of reports have described myopathy with intranuclear inclusions (Oteruelo, 1976; Oyer et al., 1991; Weeks et al., 2003; Ogasawara et al., 2020).

The observation of unique nuclear inclusions in dysmorphic nuclei in heart tissue prompted us to perform detailed histological analyses of Dst-b^{E2610Ter/E2610Ter} mouse cardiomyocytes. Upon H&E staining, eosinophilic structures were observed in Dst-b^{E2610Ter/E2610Ter} cardiomyocyte nuclei (Figure 7A). Dysmorphic nuclei were clearly observed by lamin A/C IHC, showing a deeply invaginated nuclear envelope devoid of DNA (Figure 7B). In TEM observations, nuclear crystalline inclusions and cytoplasmic organelles, such as lysosomes surrounded by the nuclear envelope, were sometimes observed within one nucleus (Figure 7C). Crystalline inclusions were observed as lattice structures (Figure 7D). Some subcellular organelles, such as mitochondria, were frequently observed within the nucleus of Dst-b^{E2610Ter/E2610Ter} cardiomyocytes as a result of nuclear invaginations (Figure 7E and F) because they were surrounded by the nuclear membrane. Moreover, vacuoles that seemed to contain liquid were observed in the nucleus. In some cases, endoplasmic reticulum and ribosomes were observed at higher densities within the nucleus of Dst-b^{E2610Ter/E2610Ter} cardiomyocytes (Figure 7G), which is similar to the findings for the nucleoplasmic reticulum (Bezin et al., 2008; Malhas et al., 2011). Next, we investigated the molecular components of nuclear inclusions and found that the autophagy-associated protein p62 was deposited in Dst-b^{E2610Ter/E2610Ter} cardiomyocyte nuclei (Figure 7H). The percentage of cell nuclei harboring p62-positive structures was remarkably increased in Dst-b^{E2610Ter/E2610Ter} cardiomyocytes compared with that in WT cardiomyocytes (Figure 7I, WT mice: 0.55 ± 0.25% vs. Dst-b^{E2610Ter/E2610Ter} mice: 14.3% ± 1.9%; n = 3 WT mice: 871 total nuclei, n = 4 Dst-b^{E2610Ter/E2610Ter} mice: 860 total nuclei). We observed co-localization of p62 and ubiquitinated protein inside the nucleus using super-resolution microscopy (Figure 7J and Video 1). Small ubiquitin-related modifier (SUMO) proteins are major components of intranuclear inclusions in some neurological diseases, including neuronal intranuclear inclusion disease (NIID) (Mori et al., 2012; Pountney et al., 2003). SUMO proteins were also detected in the intranuclear structures of Dst-b^{E2610Ter/E2610Ter} cardiomyocytes using anti-SUMO-1 and anti-SUMO-2/3 antibodies (Figure 7—figure supplement 1A and B). Because desmin and αB-crystallin were absent from the intranuclear structures, protein aggregates in the nuclei were distinct from those in some molecular components of the cytoplasm (Figure 7—figure supplement 1C and D). The p62-positive structures inside nuclei were also devoid of LC3, which is an autophagy-associated molecule, whereas p62 and LC3 were co-localized in the cytoplasm (Figure 7—figure supplement 1E). Lamin A/C IHC also confirmed that the p62- and ubiquitin-positive structures were surrounded by or adjacent to lamin A/C-positive lamina (Figure 7—figure supplement 1F and G). Desmin IHC also confirmed cytoplasmic desmin in the space surrounded by highly invaginated nuclear membrane (Figure 7—figure supplement 1H). These results revealed that Dst-b gene mutations lead to dysmorphic nuclei and formation of nuclear inclusions containing p62, ubiquitinated proteins, and SUMO proteins in the Dst-b^{E2610Ter/E2610Ter} myocardium.

To determine the molecular etiology of pathophysiological features of Dst-b^{E2610Ter/E2610Ter} mouse hearts, we performed RNA-seq analysis of transcripts obtained from the ventricular myocardium of 14–19-month-old Dst-b^{E2610Ter/E2610Ter} mice and age-matched WT mice. Principal component analysis (PCA) and hierarchical clustering of RNA-seq data showed that transcriptomic characteristics between WT and Dst-b^{E2610Ter/E2610Ter} mouse hearts were different (Figure 8—figure supplement 1A and B). RNA-seq identified 728 differentially expressed genes with a false discovery rate (FDR) < 0.1 and p<0.001 (Figure 8A), including 481 upregulated genes and 247 downregulated genes in Dst-b^{E2610Ter/E2610Ter} mouse hearts compared with those from WT mice. A volcano plot demonstrated that upregulated genes were associated with pro-fibrotic pathways (e.g., Ctgf, Tgfb2, Crlf1), cytoskeletal regulation (e.g., Acta2, Cenpf, Ankrd1, Mybpc2, Myom2, Nefm), and metabolic control (Ucp2, Nmrk2, Tbx15) and that downregulated genes were associated with transmembrane ion transport (e.g., Scn4b, Cacng6, Ano5, Tmem150c, Cacna1s, Scn4a). The reliability of RNA-seq data was validated by using qPCR analysis (Figure 8—figure supplement 1C). The differentially expressed genes included several causative genes of congenital heart defects, hypertrophic cardiomyopathy, congenital long-QT syndrome, muscular dystrophy, and dilated cardiomyopathy (e.g., Acta2, Ankrd1, Scn4b, Ano5, Rpl3l, Cenpf) (Figure 8B). Gene Ontology (GO) analysis confirmed the following terms specific to the skeletal muscle and heart: biological process (e.g., upregulated: muscle contraction, heart development, skeletal muscle tissue development, sarcomere organization, and regulation of heart rate; downregulated: muscle organ development) and cellular components (e.g., upregulated: Z-disc, sarcolemma, myofibril, neuromuscular junctions, sarcoplasmic reticulum) (Figure 8C and D, Supplementary file 1A–E). Kyoto Encyclopedia of Genes and Genomics (KEGG) pathway analysis also indicated abnormal muscle pathways (e.g., upregulated: dilated cardiomyopathy, hypertrophic cardiomyopathy; downregulated: cardiac muscle contraction) (Figure 8E, Supplementary file 1C and F). These data provide a comprehensive landscape for cardiomyopathy in Dst-b^{E2610Ter/E2610Ter} mice.

Because truncated Dst-b proteins expressed from the Dst-b^E2610Ter allele harbor an actin-binding domain (ABD) and a plakin domain, it is possible that they have a gain-of-function effect on the pathogenesis of Dst-b^{E2610Ter/E2610Ter} mice. Therefore, we analyzed conditional Dst^Gt mice (Figure 9A), in which both Dst-a and Dst-b are trapped within the N-terminal ABD and lose their function in a genetically mosaic manner. We crossed female Dst^{Gt-inv/Gt-inv} mice with male β-actin (Actb)-iCre; Dst^Gt-DO/wt mice to generate Actb-iCre; Dst^Gt-DO/Gt-inv mice (mosaic mice, Figure 9B). More than half of the mosaic mice survived over several months and showed mild dt phenotypes, such as smaller body size and impairment of motor coordination (Figure 9C). Western blotting analysis indicated that the Dst-b protein was almost absent in cardiac extract from surviving mosaic mice; however, the Dst-a protein was detected at variable levels in brain extract (Figure 9D). Cardiac fibrosis was observed in hearts of 10-month-old mosaic mice (Figure 9E). In addition, increases in Nppa mRNA expression and p62 deposition were evident in the left ventricular myocardium of mosaic mice compared with control mice (Figure 9F). These data suggest that myopathy in Dst-b^{E2610Ter/E2610Ter} mice is caused by a loss-of-function mutation of Dst-b.

Next, we performed in silico screening for DST mutant alleles with nonsense mutations within DST-b-specific exons in a database containing normal human genomes. In a search of the dbSNP database, which contains information on single-nucleotide variations in healthy humans, we identified 58 different types of nonsense mutations (94 alleles total) in all five DST-b-specific exons (Table 1, Figure 10A). The alleles with nonsense mutations were found in different populations (European, Asian, African, and American, Figure 10B). In Japan, four alleles with nonsense mutations were found among 16,760 alleles in the ToMMo 8.3KJPN database (Tadaka et al., 2021), suggesting that homozygous or compound heterozygous mutants may exist in one per approximately 16 million people. We further investigated DST-b mutations in Japanese patients diagnosed with myopathy (Nishikawa et al., 2017). Although nonsense mutations were not identified in those patients, we identified two patients with myopathy who harbored compound heterozygous DST variants (Supplementary file 1G). All variants were predicted to be variants of uncertain significance (VUS); therefore, the involvement of these variants in myopathic manifestations needs to be carefully interpreted. Taken together, these data suggest that unidentified familial myopathies and/or cardiomyopathies caused by DST-b mutant alleles exist in all populations worldwide.

Isoform-specific Dst-b mutant mice exhibited late-onset protein aggregate myopathy with cardiomyopathy, which could be termed ‘dystoninopathy.’ Protein aggregates observed in the striated muscles of Dst-b mutant mice were composed of desmin, αB-crystallin, plectin, and truncated Dst-b. MFMs are major groups of protein aggregate myopathies and are caused by mutations of genes encoding Z-disk-associated proteins (e.g., DES, CRYAB, MYOT, ZASP, FLNC, BAG3, PLEC) (Keduka et al., 2012; Clemen et al., 2013; Batonnet-Pichon et al., 2017). It is plausible that Dst proteins are also aggregated in muscles affected by other MFM types. Because Dst-b encodes cytoskeletal linker proteins that are localized in the Z-disks, it is possible that mutations in Dst-b lead to Z-disk fragility, and repeated contraction may lead to myofibril disruption and induction of muscle regeneration. In line with this notion, we observed Z-disk streaming and CNFs, which reflect muscle degeneration and regeneration.

It is possible that protein quality control impairment through the unfolded protein response, autophagy, and the ubiquitin proteasome system is involved in protein aggregation in Dst-b mutant mouse muscles. Moreover, several genes responsible for the unfolded protein response, including heat shock proteins, were upregulated in Dst-b mutant mouse hearts. We also observed the accumulation of the autophagy-associated molecules p62 and ubiquitinated proteins in the muscles of Dst-b mutant mice. Co-aggregation of p62 and ubiquitinated proteins has been reported in autophagy-deficient neurons of Atg7 conditional knockout mice (Komatsu et al., 2007). In the sensory neurons of dt mice, loss of Dst was reported to lead to disruption of the autophagic process (Ferrier et al., 2015; Lynch-Godrei et al., 2021). Autophagy was also reported to be involved in the pathogenesis of other types of MFMs; the CryAB R120G mutation (CryAB^R120G) increases autophagic activity, and genetic inactivation of autophagy aggravates heart failure in CryAB^R120G mice (Tannous et al., 2008), while enhanced autophagy ameliorates protein aggregation in CryAB^R120G mice (Bhuiyan et al., 2013). It would be interesting to investigate the involvement of autophagy and other protein degradation machinery in the process of protein aggregation myopathy in Dst-b^{E2610Ter/E2610Ter} mutant mice.

Accumulated evidence strongly suggests that myofibrillar integrity is essential for the maintenance of mitochondrial function (Vincent et al., 2016). Mitochondrial dysfunctions have also been well recognized in some animal models of MFMs, such as desmin (Des), αB-crystallin (Cryab), and plectin (Plec) mutants (Winter et al., 2015; Winter et al., 2016; Diokmetzidou et al., 2016; Alam et al., 2020). Abnormal accumulation of mitochondria in the subsarcolemmal space was observed in Dst-b^{E2610Ter/E2610Ter} mutant muscles. Mitochondrial dysfunction in Dst-b^{E2610Ter/E2610Ter} mutant mouse cardiomyocytes was also suggested by RNA-seq data, including altered expression of genes responsible for oxidative phosphorylation and other metabolic processes. The altered heart functions characterized by long QT intervals and PVCs in Dst-b mutant mice may be explained by mitochondrial dysfunction, as well as altered expression of genes involved in transmembrane ion transport and muscle contraction.

Deep invaginations of the nuclear membrane were also a characteristic feature of Dst-b mutant muscle fibers. The linker of nucleoskeleton and cytoskeleton (LINC) complex is a protein complex that spans the inner and outer nuclear membrane to bridge the cytoskeleton in the cytoplasm to the nuclear lamina underlying the inner nuclear membrane (Stroud et al., 2014). Mutations in genes encoding LINC complex proteins are known to cause skeletal myopathy and cardiomyopathy, called Emery–Dreifuss muscular dystrophy (Heller et al., 2020), and defects in nuclear structure are often observed in mice with mutations of LINC and LINC-associated molecules such as lamin A (Lmna), nesprin, ezrin, and Des (Nikolova et al., 2004; Banerjee et al., 2014; Wada et al., 2019; Heffler et al., 2020). Because Dst can associate with nesprin-3α via the ABD (Young and Kothary, 2008), and we demonstrated that Dst-b can localize around nuclei of muscle fibers, Dst-b may be involved in the molecular bridge between the LINC complex and cytoskeleton and maintain the shape of nuclei.

Furthermore, we found that nuclear inclusion was a unique hallmark of cardiomyopathy in Dst-b mutant mice. The nuclear inclusions were different from the cytoplasmic protein aggregates in autophagy-deficient mice because nuclear inclusions lacked LC3 protein. Intranuclear inclusions are also observed in NIID, which is a neurological disease caused by GGC repeat expansion of the NOTCH2NLC gene (Sone et al., 2019), which has also been reported to be associated with oculopharyngodistal myopathy with intranuclear inclusions (Ogasawara et al., 2020). The intranuclear inclusions of NIID predominantly include p62, ubiquitinated proteins, and SUMOylated proteins (Mori et al., 2012; Pountney et al., 2003). In our observations, SUMOylated proteins were also present in the nuclear inclusions of Dst-b mutants. Thus, nuclear inclusions of Dst-b mutant cardiomyocytes have common features with those of NIID in terms of molecular components (Sone et al., 2011). It would be interesting to further investigate the molecular components and pathogenesis of nuclear inclusions in the cardiomyopathy of Dst-b mutant mice and compare them with those found in NIID.

A wide variety of genetic diseases have been reported to be caused by DST mutations. Loss-of-function mutations in both DST-a and DST-b result in HSAN-VI, and loss-of-function mutations of DST-e result in the skin blistering disease epidermolysis bullosa simplex (Groves et al., 2010; Edvardson et al., 2012). Recently, DST was reported as a candidate gene of pulmonary atresia, a rare congenital heart defect (Shi et al., 2020). According to this study, it is possible that unidentified hereditary myopathies and heart disease caused by human DST-b-specific mutations exist. Because the causative gene is unidentified in almost half of MFM cases, it would be worthwhile to perform candidate gene screening on DST-b-specific exons in patients with MFMs.

Recently, HSAN-VI has been proposed to be a spectrum disease because of its varying disease severity and the diverse deficiency of DST isoforms (Lynch-Godrei and Kothary, 2020). DST-a2 seems to be a crucial isoform for HSAN-VI pathogenesis because DST-a2-specific mutations result in adult-onset HSAN-VI (Manganelli et al., 2017; Fortugno et al., 2019), and neuronal expression of Dst-a2 partially rescues the dt phenotype (Ferrier et al., 2014). We suggest that Dst-a1 and Dst-a2 play redundant roles because Dst^Gt mice, in which both Dst-a1 and Dst-a2 are trapped, show a severe phenotype (Horie et al., 2014). Among three Dst-b isoforms, remarkable reduction of Dst-b1, but not Dst-b2 or Dst-b3, was observed in Dst-b^{E2610Ter/E2610Ter} mice, suggesting that reduced Dst-b1 expression is involved in myopathic phenotypes in Dst-b^{E2610Ter/E2610Ter} mice. It is possible that DST isoforms 1 and 2 may be important to different extents in neural and muscular tissues. Regarding this point of view, it would be interesting to compare muscle phenotypes between patients harboring DST-a1, -a2, -b1, and -b2 isoform mutations and patients with DST-a2 and -b2 isoform-specific mutations (Manganelli et al., 2017; Fortugno et al., 2019; Motley et al., 2020).

Members of the plakin protein family other than Dst, such as plectin and desmoplakin, are known to be expressed and play crucial roles in maintaining muscle integrity (Leung et al., 2002; Boyer et al., 2010b; Horie et al., 2017). Plectin is a Dst-associated protein and one of the most investigated members of the plakin protein family (Castañón et al., 2013). There are four alternative splicing isoforms of plectin in striated muscle fibers, which are localized in differential subcellular compartments: plectin 1 in nuclei, plectin 1b in mitochondria, plectin 1d in Z-disks, and plectin 1f in the sarcolemma (Mihailovska et al., 2014; Staszewska et al., 2015; Winter et al., 2015). Plec deficiency is lethal in mice at the neonatal stage, and these mice exhibit skin blistering and muscle abnormalities (Andrä et al., 1997). Conditional deletion of Plec in muscle fibers leads to progressive pathological alterations, such as aggregation of desmin and chaperon protein, subsarcolemmal accumulation of mitochondria, and an abnormal nuclear shape (Konieczny et al., 2008; Staszewska et al., 2015; Winter et al., 2014; Winter et al., 2015), some of which are also observed in the muscles of aged Dst-b mutant mice.

Dst and plectin also have common features in other cell types. In keratinocytes, plectin and Dst-e localize to the inner plaque of hemidesmosomes and anchor keratin intermediate fibers to hemidesmosomes (Künzli et al., 2016). Conditional deletion of Plec from epidermal cells causes epidermal barrier defects and skin blistering (Ackerl et al., 2007), both of which is more severe than that observed in dt mice carrying Dst-e mutations (Guo et al., 1995; Yoshioka et al., 2020). Furthermore, we recently demonstrated that conditional deletion of Dst from Schwann cells in the peripheral nervous system leads to disorganization of the myelin sheath (Horie et al., 2020), similar to that observed for Plec-deficient Schwann cells (Walko et al., 2013). These studies suggest that Dst and plectin, which are plakin family proteins, have overlapping roles in different cell types.

Dst-b^E2610Ter mice, Dst^Gt(E182H05) mice (MGI number: 3917429; Horie et al., 2014), and Actb-iCre mice (Zhou et al., 2018) were used in this study. Dst^Gt(E182H05) allele was abbreviated as Dst^Gt. Homozygous Dst-b^E2610Ter and Dst^Gt mice were obtained by heterozygous mating. For mosaic analysis, female Dst^{Gt-inv/Gt-inv} mice were crossed with male Actb-iCre;Dst^Gt-DO/wt mice. Dst-b^E2610Ter mutant line was C57BL/6J. The Dst^Gt mutant line was backcrossed to C57BL/6NCrj at least 10 generations. Mice were maintained in groups at 23°C ± 3°C, 50% ± 10% humidity, 12  hr light/dark cycles, and food/water availability ad libitum. Both male and female mice were analyzed in this study. Genotyping PCR for the Dst^Gt allele was performed as previous described (Horie et al., 2014). For genotyping PCR of the Dst-b^E2610Ter allele, the following primer set was used to amplify 377 bp fragments: Dst-b forward: 5′-TGA GCG ATG GTA GCG ACT TG-3′ and Dst-b reverse: 5′-GCG ACA CAC CTT TAG TTG CC-3′. PCR was performed using Quick Taq HS DyeMix (Toyobo, Osaka, Japan) and a PCR Thermal Cycler Dice (TP650; Takara Bio Inc, Shiga, Japan), with the following cycling conditions: 94°C for 2 min, followed by 32 cycles of 94°C for 20 s, 61°C for 30 s, and 72°C for 30 s. PCR products were cut with XhoI, which produced two fragments (220 bp and 157 bp) for the mutant Dst-b^E2610Ter allele. For genotyping PCR for Actb-iCre knockin allele, the following primer set was used to amplify 540 bp fragments: iCre-F: CTC AAC ATG CTG CAC AGG AGA T-3′ and iCre-R: 5′-ACC ATA GAT CAG GCG GTG GGT-3′.

We attempted to induce G-to-T (Glu to Stop) point mutation in the Dst-b/Bpag1b using a previously described procedure (Sato et al., 2018). The sequence (5′-GCT ATC AGG AAA GAA CAC GG-3′) was selected as guide RNA (gRNA) target. The gRNA was synthesized and purified by GeneArt Precision gRNA Synthesis Kit (Thermo Fisher Scientific, MA) and dissolved in Opti-MEM (Thermo Fisher Scientific). In addition, we designed a 102-nt single-stranded DNA oligodeoxynucleotide (ssODN) donor for inducing c.8058 G>T of the Dst-b1 (accession # NM_134448.4); the nucleotide T was placed between 5′- and 3′-homology arms derived from positions 8013–8057 and 8066–8114 of the Dst-b1 coding sequence, respectively. This ssODN was ordered as Ultramer DNA oligos from Integrated DNA Technologies (IA, USA) and dissolved in Opti-MEM. The pregnant mare serum gonadotropin (five units) and the human chorionic gonadotropin (five units) were intraperitoneally injected into female C57BL/6J mice (Charles River Laboratories, Kanagawa, Japan) with a 48 hr interval, and unfertilized oocyte were collected from their oviducts. We then performed in vitro fertilization with these oocytes and sperm from male C57BL/6J mice (Charles River Laboratories, Kanagawa, Japan) according to standard protocols. 5 hr later, the gRNA (5 ng/μl), ssODN (100 ng/μl), and GeneArt Platinum Cas9 Nuclease (Thermo Fisher Scientific) (100 ng/μl) were electroplated to zygotes by using NEPA21 electroplater (Nepa Gene, Chiba, Japan) as previously reported (Sato et al., 2018). After electroporation, fertilized eggs that had developed to the two-cell stage were transferred into oviducts in pseudopregnant ICR female and newborns were obtained. To confirm the G-to-T point mutation induced by CRISPR/ Cas9, we amplified genomic region including target sites by PCR with the same primers used for PCR-RFLP genotyping. The PCR products were cut with XhoI for genome editing validation, and then, sequenced by using BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific). We analyzed three independent Dst-b lines in this study: line numbers #1, # 6, and #7.

Frozen heart, gastrocnemius muscle, and brain tissues were homogenized using a Teflon-glass homogenizer in ice-cold homogenization buffer (0.32 M sucrose, 5 mM EDTA, 10 mM Tris-HCl, and pH 7.4, phosphatase inhibitor cocktail tablet; Roche, Mannheim, Germany), centrifuged at 4500 rpm for 10 min at 4°C and the supernatants were collected. The protein concentration was determined using the bicinchoninic acid Protein Assay Reagent (Thermo Fisher Scientific). Lysates were mixed with an equal volume of 2× sodium dodecyl sulfate (SDS) sample buffer (125 mM Tris-HCl, pH 6.8, 4% SDS, 20% glycerol, and 0.002% bromophenol blue) for a final protein concentration of 2 μg/μl and denatured in the presence of 100 mM dithiothreitol at 100°C for 5 min. SDS-polyacrylamide gel electrophoresis (PAGE) was performed with 20 μg per lane on 5–20% gradient gels (197-15011; SuperSep Ace; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) running at 10–20 mA for 150 min. The gels were blotted onto an Immobilon-P transfer membrane (Millipore, Billerica, MA). After blocking with 10% skim milk for 3 hr, blotted membranes were incubated with the following primary antibodies: rabbit polyclonal anti-Dst antibody (gifted from Dr. Ronald K Leim; 1:4000; Goryunov et al., 2007) that recognizes the plakin domain of Dst, and mouse monoclonal anti-glyceraldehyde-3-phosphate dehydrogenase (Gapdh) antibody (016-25523; clone 5A12, 1:10,000, Wako). Each primary antibody was incubated overnight at 4°C. Then membranes were incubated with peroxidase-conjugated secondary antibodies for 1 hr at room temperature: anti-rabbit immunoglobulin G (IgG) (AB_2099233; Cat# 7074, 1:2000, Cell Signaling Technology), anti-mouse IgG (AB_330924; Cat# 7076, 1:2000, Cell Signaling Technology). Tris-buffered saline (10 mM Tris-HCl, pH 7.5, 150 mM NaCl) containing 0.1% Tween-20 and 10% skim milk was used for the dilution of primary and secondary antibodies, and Tris-buffered saline containing 0.1% Tween-20 was used as the washing buffer. Immunoreactions were visualized with ECL (GE Healthcare, Piscataway Township, NJ) or ImmunoStar LD (FUJIFILM Wako Pure Chemical) and a Chemiluminescent Western Blot Scanner (C-Digit, LI-COR, Lincoln, NE). The signal intensity of each band was quantified using Image Studio software version 5.2 (LI-COR).

Total RNA was extracted from the heart, soleus, and brain using the RNeasy Mini Kit (QIAGEN, Hilden, Germany) including DNase digestion. 100 ng of RNA template was used for cDNA synthesis with oligo (dT) primers. Real-time PCR was performed using a StepOnePlus Real-Time PCR system (Thermo Fisher Scientific) and the following cycling conditions: 95°C for 2 min, followed by 40 cycles of 95°C for 15 s, 60°C for 40 s, and 95°C for 15 s. Gene expression levels were analyzed using the ΔΔCT method. Actb or Gapdh were used as internal controls to normalize the variability of expression levels. The primers used for real-time PCR are listed in Table 2.

RNA-seq analysis was performed as described in the previous study with slight modification (Bizen et al., 2022; Hayakawa-Yano et al., 2017; Hayakawa-Yano and Yano, 2019). mRNA libraries were generated from total RNA 5 μg/sample extracted from the heart using illumine TruSeq protocols for poly-A selection, fragmentation, and adaptor ligation. Multiplexed libraries were sequenced as 150-nt paired-end runs on an NovaSeq6000 system. Sequence reads were mapped to the reference mouse genome (GRCm38/mm10) with Olego. Expression and alternative splicing events were quantified with Quantas tool. Integrative Genomics Viewer (IGV) was used as visualization of alignments in mouse genomic regions (Thorvaldsdóttir et al., 2013). Statics of differential expression was determined with edseR (Robinson et al., 2010). Differentially expressed genes were corrected with threshold set at p<0.01 and FDR < 0.1. GO analysis and KEGG pathway analysis were performed using DAVID Bioinformatics Resources (Huang et al., 2009) (https://david.ncifcrf.gov/summary.jsp). The threshold of GO analysis and pathway analysis was set at p<0.01. Three highly changed genes, Xist, Tsix, and Bmp10, were excluded because the differences were due to extreme outliers in only one individual. In GO analysis, the terms of cytoplasm and membrane that contained more than 190 genes were excluded from the list. PCA and hierarchical clustering were performed and visualized with R software using the RNA-seq datasets obtained from three independent WT and Dst-b^E2610Ter. PCA was conducted in the prcomp function in RStudio (version 2021.09.1 build 372). RNA-seq data can be accessed from the Gene Expression Omnibus under accession, GSE184101.

The rotarod test and wire hang test (O’Hara & Co., Tokyo, Japan) were performed to evaluate motor coordination as described in the previous study with slight modification (Horie et al., 2020). In the rotarod test, the latency to fall from a rotating rod (30 mm diameter) with an acceleration from 10 to 150 rpm was measured. Each trial was conducted for a period of 3 min. In each mouse, two trials were conducted in a day. In the wire hang test, mouse was placed on the top of the wire lid. The lid was slightly shaken several times to force the mouse to grip the wires. The lid was turned upside down. The latency to fall was measured.

For tissue preparations, mice were euthanized via intraperitoneal injection with pentobarbital sodium (100 mg/kg body weight), and then perfused with 4% paraformaldehyde (PFA) in 0.1 M phosphate-buffered (PB) solution (pH 7.4). The tissues were fixed by cardiac perfusion with 0.01 M phosphate-buffered saline (PBS) followed by ice-cold 4% PFA in 0.1 M PB (pH 7.4). Dissected tissues were immersed in the same fixative overnight. To cut spinal cord and DRG sections, the specimens were rinsed with water for 10 min and decalcified in Morse solution (135-17071; Wako, Osaka, Japan) overnight. Tissues were then dehydrated using an ascending series of ethanol and xylene washes, and then embedded in paraffin (P3683; Paraplast Plus; Sigma-Aldrich, St. Louis, MO). Consecutive 10-μm-thick paraffin sections were cut on a rotary microtome (HM325; Thermo Fisher Scientific), mounted on MAS-coated glass slides (Matsunami Glass, Osaka, Japan), and air-dried on a hot plate overnight at 37°C. Paraffin sections were deparaffinized in xylene, rehydrated using a descending series of ethanol washes, and then rinsed in distilled water. H&E staining and Masson’s trichrome staining were performed by standard protocols.

For IHC, deparaffinized sections were treated with microwave irradiation in 10 mM citric acid buffer, pH 6.0 for 5 min, and incubated overnight at 4°C with the primaries listed in Table 3. All primary antibodies were diluted in 0.1 M PBS with 0.01% Triton X-100 (PBST) containing 0.5% skim milk. Sections were then incubated in horseradish peroxidase-conjugated secondary antibody (1:200; MBL, Nagoya, Japan) diluted in PBST containing 0.5% skim milk for 60 min at 37°C. Between each step, sections were rinsed in PBST for 15 min. After rinsing sections in distilled water, immunoreactivity was visualized in 50 mM Tris buffer (pH 7.4) containing 0.01% diaminobenzidine tetrahydrochloride and 0.01% hydrogen peroxide at 37°C for 5 min. Sections were then dehydrated through an ethanol-xylene solution and placed on coverslips with Bioleit (23-1002; Okenshoji, Tokyo, Japan). Digital images were taken with a microscope (BX53; Olympus, Tokyo, Japan) equipped with a digital camera (DP74, Olympus), and the TIF files were processed with Photoshop software (Adobe, San Jose, USA).

For immunofluorescent staining, sections were incubated in mixtures of Alexa488- or Alexa594-conjugated antibodies (1:200; Invitrogen, CA) for 60 min at 37°C. Cell nuclei were counterstained with DAPI (1:2000; Dojindo, Kumamoto, Japan) for 10 min at room temperature. Mounted sections were air-dried and coverslipped. Sections were observed and digital images were taken using a confocal laser scanning microscopy (FV-1200, Olympus). TIF files were processed with Adobe Photoshop software. Super-resolution images were recorded on the confocal laser scanning microscopy LSM 980 equipped with Airyscan 2 (Leica, Germany).

ISH was performed on paraffin sections as described in previous studies (Takebayashi et al., 2000) using following mouse probes: PV, also known as parvalbumin (Pvalb, GenBank accession, NM_013645, nt 92-885), Nppa, also known as atrial natriuretic peptide (ANP, GenBank accession, BC089615, nt 124-529), and Nppb, also known as brain natriuretic peptide (BNP, GenBank accession, D16497, nt 42-752, without intron sequence) were used.

The detailed procedure for TEM analysis was described previously (Shibata et al., 2015). Briefly, the muscle samples for TEM were primary fixed with 2.5% glutaraldehyde for 24 hr at 4°C. Samples were washed in 50 mM HEPES buffer (pH 7.4) and were postfixed with 1.0% osmium tetroxide (TAAB Laboratories, England, UK) for 2 hr at 4°C. Samples were dehydrated with a series of increasing concentrations of ethanol (two times of 50, 70, 80, 90, 100% EtOH for 20 min each), soaked with acetone (Sigma-Aldrich) for 0.5 hr, and with n-butyl glycidyl ether (QY-1, Okenshoji Co. Ltd., Tokyo, Japan) two times for 0.5 hr, graded concentration of epoxy resin with QY-1 for 1 hr, and with 100% epoxy resin (100 g Epon was composed of 27.0 g MNA, 51.3 g EPOK-812, 21.9 g DDSA, and 1.1 ml DMP-30, all from Okenshoji Co. Ltd.) for 48 hr at 4°C, and were polymerized with pure epoxy resin for 72 hr at 60°C. Resin blocks with tissues were trimmed, semi-thin sliced at 350-nm-thickness stained with toluidine blue, and were ultrathin-sectioned at 80-nm thickness with ultramicrotome (UC7, Leica) by diamond knife (Ultra, DiATOME, Switzerland). The ultrathin sections were collected on the copper grids and were stained with uranyl acetate and lead citrate. The sections were imaged with TEM (JEM-1400 Plus, JEOL, Japan) at 100 keV.

Under anesthesia with 2–3% isoflurane (Pfizer Inc, NY), ECG signals were recorded. Three needle electrodes were inserted into right and left forelimbs as recording, and right hindlimb as grounding. The ECG signals were amplified using AC amplifier (band pass: 0.1–1 kHz), and the signals were digitized with A/D converter (Power 1401, Cambridge Electronic Design Ltd., Cambridge, UK).

Morphometric analysis was performed at least three sections per mouse. Quantifications of cross-sectional area and fibrosis were performed with MetaMorph software (Meta Series Software version 7.10.2; Molecular Devices, San Jose, CA). Unless otherwise noted, sample size (n) is the number of animals in each genotype. For statistical analysis, Student’s t-test and ANOVA were carried out. ANOVA was performed using ANOVA4 on the Web (https://www.hju.ac.jp/~kiriki/anova4/).

In silico screening was performed using dbSNP database (Sherry et al., 2001) (https://www.ncbi.nlm.nih.gov/snp/?cmd=search). After downloading the results of a search using the keyword "DST" in dbSNP in TSV format, we extracted the entries whose "function_class" was "stop_gained".

All animal experiments were performed in accordance with the guidelines of the Ministry of Education, Culture, Sports, Science and Technology of Japan and were approved by the Institutional Animal Care and Use Committees of Niigata University (permit number: SA00521 and SA00621) and Tsukuba University (permit number: 17-078). Human study was approved by the ethical committees of the NCNP (permit number: A2019-123). The human materials used in this study were obtained for diagnostic purposes. The patients or their parents provided written informed consent for use of the samples for research.