Special Pathologic Features of Adolescent Idiopathic Scoliosis: Could There be a New Type of Muscular Dystrophinopathy?

Background: Adolescent idiopathic scoliosis (AIS) is characterized by vertebral rotation and lateral curvature of the spine and affects 2-4% of the population throughout the world and the cause of the disease is still controversial. Recent researches suggest that there is an internal correlation between certain neuromuscular diseases and AIS. This study aims to characterize the paraspinal muscles of AIS patients, and to further explore the its etiology. Methods: Eighteen AIS patients treated with posterior scoliosis correction surgery were included and had biopsies taken from the paraspinal muscle at the apex vertebra region. Serial sections with conventional H&E staining and histochemical staining were obtained, and immunohistochemical examinations were employed to detect Dystrophin-1, -2, -3, Myosin, MHC-1, CD4, CD8, CD20, and CD68 or CD163 antibodies. Biopsy samples were grouped according to the subjects’ Cobb angle and Nash-Moe’s classication respectively, and the corresponding pathological changes were compared between groups. Results: The immunohistochemical staining results showed signicant differences in the expression pattern of Dystrophin-2 (P=0.023) and Dystrophin-total (P=0.018) between the mild and severe scoliosis groups, with the expression of Dystrophin more abnormal or absent in the severe scoliosis group. There were also signicant differences in the expression pattern of Dystrophin-2 (P=0.035) between the Nash-Moe classication subgroups. The expression of Dystrophin-3 was absent to various extent in all patients. Besides, we observed an inltration of CD4+ and CD8+ cells in the paraspinal muscles and tendons. In all patients, the expression of MHC-1 on myolemma was present in some muscle bers. Conclusions: The expression of dystrophin protein was signicantly reduced and was correlated with the severity of scoliosis, suggesting that dystrophin protein dysfunctions contribute to the development of scoliosis. Meanwhile, the inammatory changes of AIS mainly manifested in T cell activation and inltration, and there seemed to protein in the paraspinal muscles of patients with idiopathic scoliosis. And compared to patients with mild scoliosis, those with severe scoliosis have signicantly more total protein loss as revealed by Dystrophin rod-domain and C-terminal immunostaining. With Dystrophin N-terminal immunostaining, all the biopsy samples showed discontinuously linear expression or even loss of expression, especially in atrophic muscle bers. These results lead to the conclusion that sarcolemma integrity and stability can be damaged by the dysfunction of dystrophin protein in the paraspinal muscles of AIS patients.


Background
Adolescent idiopathic scoliosis (AIS) is a disease characterized by vertebral rotation and lateral curvature of the spine, which affects the development of children during their adolescence (1). It is a very common type of spinal deformity with a prevalence of 2-4% worldwide among children younger than 16 years old (2). But the cause of the disease has not yet been clari ed.
As early as the mid-1970s, the proportion and size of type 1 and type 2 skeletal muscle bers in AIS were measured by several researchers (3,4) and paraspinal muscle imbalance was reported to be one of the key features of AIS. In the 1980s, ultrastructural study of multi dus muscle in progressive idiopathic scoliosis showed structural changes of the sarcolemma and at the myotendinous junction (5,6). According to the previous literature, AIS is often accompanied by certain but unspeci ed pathological changes in paraspinal nerves or muscles (7). In 2015, Luciano et al. (8) reported cases of AIS patients who didn't show the typical symptoms of decreased limb muscle strength and of limited respiratory muscle function; however, the results of the paraspinal muscle biopsies of two patients with AIS indicated a high likelihood of core myopathy, but the patients did not develop core myopathy eventually. A hypothetical explanation is that this sort of AIS patients may be a special manifestation of the core myopathy. On the other hand, some neuromuscular diseases, such as Duchenne muscular dystrophy can also affect the paraspinal muscles in the late stage and cause scoliosis (9). Therefore, whether there is an internal correlation between certain neuromuscular diseases and AIS is still a problem that calls for further etiological studies, and paraspinal myopathy is one of the key potential factors for the initial pathogenesis of idiopathic scoliosis.
Based on the hypothesis that the onset and clinical progression of AIS may be associated with certain neuromuscular diseases, we used pathological methods to study the paraspinal muscle changes in AIS patients, and introduced immunohistochemical antibody markers that are commonly used in the diagnosis of neuromuscular diseases through routine morphology, which enabled further analysis of and discussions on the relationship between AIS and neuromuscular diseases.

Subjects
A total of 18 AIS patients were included, all of whom received posterior scoliosis correction surgery. The subjects were followed up in our hospital from November 2018 to August 2019, and their average age was 16.11 ± 3.77 years old (Table 1). All subjects had main thoracic curves with a Cobb angle of above 40° by the standing posteroanterior radiographs, and they had normal BMI and no related complications. Patients diagnosed of neuromuscular diseases by the neurology department were excluded from the study. This study received approval from the institutional reviewing board. All patients voluntarily provided informed consent and were protected by law.

Clinical group and classi cation
Based on the Cobb Angle, the patients in this study were divided into two groups: severe scoliosis group and mild scoliosis group.
According to the Nash-Moe classi cation standards, the patients in this study were also divided into three groups: Nash-Moe type , Nash-Moe type , and Nash-Moe type .

Interpretation of pathological results
The following items were observed and evaluated respectively: 1) degrees of muscle atrophy, hypertrophy, fatty in ltration, and muscle degeneration; 2) the presence of whorled bers or internal nuclei; 3) the proportion and distribution of muscle ber types; 4) the expression of Dystrophin and MHC-1 proteins, 5) the interstitial condition of muscle tissues, and 6) the expression and distribution of in ammatory cells in tendons and muscles. Two pathologists independently analyzed each subject, and there were no inconsistent judgments.

Statistical analysis
The pathological changes of patients in different Cobb Angle groups and Nash-Moe classi cation groups were analyzed with χ 2 test, and the correlation between clinical classi cation and the pathological changes was also examined. All statistical analyses were performed using SPSS 24.0 (IBM Corporation, Armonk, New York, USA). And the statistical signi cance was de ned as a P value less than 0.05 in a two-sided hypothesis test.

Clinical data of the enrolled patients
The subjects included three males and 15 females, with an average age of 16 years old. The average Cobb curve angle of the thoracic segment was 51.04° and the median Cobb angle in scoliosis was 55.9°. Thus, we adopted 55° as the grouping criteria for the degree of deformity among the patients. There were ten patients with severe scoliosis (Cobb angle > 55°) and eight with mild scoliosis (Cobb angle < 55°). In the Nash-Moe type classi cation, ten patients were classi ed as type , seven as type , and three as type . (Table 1) (Table 2) Muscle atrophy was found in all the enrolled patients, and the proportion of moderate and severe atrophy accounted for 94.4% of the subjects. The percentage of severe atrophy was 70.83%, suggesting that severe paraspinal atrophy was prevalent.
Compensatory hypertrophy of paraspinal muscle was identi ed in 88.89% of the enrolled patients, and in 33.33% of the patients, the paraspinal muscle was accompanied by several whorled bersin. Meanwhile, internal nuclei was observed in muscle bers in 88.89% of the patients ( Figure. 1A). The paraspinal muscles of all patients were subject to varying degrees of muscle ber degeneration, with 55.56% of the patients presented edema and 22.22% presented extensive edema. Besides, 27.8% of the patients had minor myo ber destruction and myolysis in the biopsy tissue, indicating that myogenic lesions are more predominant than single neurogenic muscular atrophy in pathological changes.
Moth-eaten bers were observed in the NADH-TR staining of muscle biopsy in 70% of the patients with severe scoliosis and 25% of the patients with mild scoliosis (Figure.1B). Although there was no statistically signi cant difference between the two groups (P = 0.053), it shows a trend that the severer the scoliosis is, the more moth-eaten bers may appear.
As for Myosin staining, type 2 bers were positive in the muscles of AIS patients, but the staining was uneven to various degrees.
There was a small amount of uneven staining in 80% of the severe scoliosis group and 50% of the mild scoliosis group, though there was no statistically signi cant difference between the two groups (P = 0.327). In addition, myosin staining suggested that there were slightly more type 2 muscle bers than type 1 in the biopsy muscle tissues (Fig. 1C).
Dystrophin protein immunostaining showed that Dystrophin-2 (C-terminal epitope) and Dystrophin-total were both absent in most of the atrophic muscle bers and some of the non-atrophic muscle bers, and the differences between the two groups were statistically signi cant. In the severe scoliosis group, the expression of anti-atrophic protein was more abnormal or absent. There were signi cant differences in the expression patterns of both Dystrophin-2(P = 0.023) and Dystrophin-total(P = 0.018) between the two groups ( Fig. 2), and the expression of dystrophin was subject to increased abnormities or absence in the severe scoliosis group. As for Dystrophin-1 (rod domain epitope, P = 0.054) and Dystrophin-3 (N-terminal epitope, P = 0.411), the sarcolemma of dystrophic and degenerated myo bers faded in all the cases involved ( Fig. 3), even though there was no signi cant difference between the mild and severe scoliosis groups. 3.3 In ammatory cell in ltration of paraspinal muscle biopsy of AIS and comparisons between the scoliosis groups (Table 3) MHC-1 immunostaining showed weakly positive results of some sarcolemma of dystrophic and degenerated myo bers in almost all AIS cases. The in ltrating in ammation cells were dominantly CD4 + T cells and CD8 + T cells (Fig. 4), and the in ammation cells distributed around small vessels or scattered among the stroma in muscle bundles. Muscle tissues and the tendons of paraspinal muscle were both affected and the level of in ammation was almost the same. Few CD20 + B cells were observed to in ltrate in muscle bundles or tendons. The presence of CD68/CD163 + cells indicates that the monocyte-macrophage system may play a role in the progression of AIS (Fig. 5). Most of the pathological changes in the muscle biopsies of AIS patients did not show any statistically signi cant differences among different Nash-Moe subgroups, and the only exception was the expression of dystrophin protein. Immunostaining showed signi cantly different expression patterns of Dystrophin-2 among the subgroups, and the higher the group type was, the more abnormal or absent the Dystrophin-2 expression was (P = 0.035). However, although the expression of Dystrophin-1 showed statistically signi cant differences among the three Nash-Moe classi cation groups (P = 0.034), it was not related to the severity of the disease, with the Nash-Moe patients showing more Dystrophin-1 light staining and more discontinuous muscle bers in muscle biopsies compared with the other two groups. Adolescent idiopathic scoliosis (AIS) is a common three-dimensional spine deformity characterized by a lateral spinal curvature of more than 10° (2). Although the etiopathogenesis of scoliosis is still unclear, various factors that may contribute to the etiology of scoliosis have been identi ed, such as genetic factors, structural factors, and environmental factors. The progression of the patient's side curvature is particularly evident during puberty, and this stage is also the period when the paraspinal muscles grow rapidly (11). Rehabilitation exercises for paraspinal muscles, such as Schroth gymnastics, also have a therapeutic effect on scoliosis (12). Besides, scoliosis most often occurs in the late stages of neuromuscular diseases and may continue to progress with sagittal deformities (13). This research on the correlation between paraspinal muscle lesions and scoliosis is particularly important for understanding the pathogenesis of the disease, for which muscle biopsy is the most effective research method.

Paraspinal muscle atrophy and degeneration
Both the degree of paraspinal atrophy and the atrophy occurance among the subjects in our study were slightly higher than those reported by Wajchenberg M. (7). We suggest that, in addition to the severity of the disease, the degree of atrophy is also related to multiple factors, such as the patient's Cobb angle and the course of disease. In addition, the location of the biopsy sampling will also affect the proportion of atrophy. In this study, the paraspinal muscle was sampled relatively closer to the tendon, which may account for the higher atrophy rate than reported in previous studies.
The paraspinal muscle ber structures of our subjects were observed abnormal to varying degrees and in different forms, including changes in rostral muscle ber, nuclear shift, and abnormality in myosin structure. The rate of nuclear shift in our study was higher than the results of Wajchenberg et al. (7). At the same time, the Myosin staining results suggested that the Type 1 ber content among the patients generally decreased, which was similar to previous results (14,15). Some of the patients experienced a serious decrease in the proportion of Type 1 paraspinal muscle bers, which was lower than that of Type 2 muscle bers. Type 2 muscle bers are fast-shrinking bers and mostly distribute in skeletal muscles of the extremities. In contrast, Type 1 muscle bers are slow-muscle bers and typically distribute in the trunk. Mannion et al. (16) contend that the loss of Type 1 muscle bers may cause the muscles to withstand the long-term rigid contraction, leading to the loss of spinal stability and eventually to risks of scoliosis.
In addition, some patients had paraspinal muscle degeneration to different degrees, such as lysis and destruction, suggesting the existence of more serious myogenic lesions in the skeletal muscles. The NADH-TH staining indicated that most patients had moth-eaten bers, which has also been reported in existing literature. Moth-eaten bers were indicated by the aky and irregular loss of oxidase in staining, and could be present near the center or the edge of muscle bers. Their presence near the center of muscle bers could resemble the presentation of core myopathy, and must be differentiated by clinical manifestations and genetics (7,8). Furthermore, patients with severe scoliosis tended to have more NADH de ciency than patients with mild scoliosis (P = 0.053, although there was no statistically signi cant difference), suggesting the metabolic dysfunction of myo bers. With the increase of Cobb angle, paraspinal muscles are subject to a trend of gradual degeneration and necrosis.

Abnormal expression of dystrophin in paraspinal muscles of AIS
Dystrophin is a 427kD structural protein and a cytoplasmic protein associated with sarcolemma. It links the actin cytoskeleton to the transmembrane dystrophin-glycoprotein complex (17,18) and helps maintain membrane integrity and cellular homeostasis (19).
In our study, three antibodies, whose antigens were respectively Rod-domain, N-terminal, and C-terminal dystrophins, were introduced to conduct immunohistochemical staining and the evaluation of paraspinal musculoskeletal membranes in AIS patients. And our results suggested that there was a signi cant loss of Dystrophin protein in the paraspinal muscles of patients with idiopathic scoliosis. And compared to patients with mild scoliosis, those with severe scoliosis have signi cantly more total protein loss as revealed by Dystrophin rod-domain and C-terminal immunostaining. With Dystrophin N-terminal immunostaining, all the biopsy samples showed discontinuously linear expression or even loss of expression, especially in atrophic muscle bers. These results lead to the conclusion that sarcolemma integrity and stability can be damaged by the dysfunction of dystrophin protein in the paraspinal muscles of AIS patients.
As is widely held, Duchenne muscular dystrophy (DMD) gene mutation disables the expression of dystrophin, which is key to the pathogenesis of DMD. As a fatal X-linked recessive disease, DMD is characterized by skeletal, respiratory, and cardiac muscle deteriorations. In terms of the onset gender, most of the patients are boys because of the genetic character of DMD (20,21). In DMD and Becker muscular dystrophy (BMD) patients, due to mutations in the DMD gene, the synthesis of Dystrophin protein can be reduced or yield abnormal structures, which leads to the destruction of the sarcolemma, cell necrosis and apoptosis, muscle atrophy in the limbs, and severe respiratory muscle and diaphragm muscle weakness; some patients may even have scoliosis due to paraspinal muscle involvement in the later stages (9).
This subjects of this study had no symptoms of weak limb muscles or respiratory and cardiac muscles, and were be excluded if diagnosed with DMD and BMD. However, their paraspinal muscles showed a novel pattern of de ciency in Dystrophin protein, which implies that these AIS patients may represent a special type of muscle dystrophinopathy, just as those patients of BMD and DMD who have decreased or dysfunctional Dystrophin protein. However, unlike DMD or BMD, in this study the pathological changes only existed in the patients' paraspinal muscles, causing asymmetrical traction of the spine and resulting in scoliosis.
As for the dystrophin expression patterns, in DMD, dystrophin is absent except in revertant bers. According to the "read frame" hypothesis in BMD, truncated dystrophin will lead to reduced intensity of some functional proteins at different epitopes such as the N-terminal and C-terminal (22). For the patients in this study, the AIS dystrophin was subject to reduced immunostaining of dystrophin at rod-domain and C-terminal epitopes and to entirely absent immunostaining at the N-terminal epitope of the bundle of atrophic bers. This immune expression pattern is similar to the above myodystrophy, but on the other hand, the abnormal expression of anti-myodystrophy in AIS muscle biopsy tissues with atrophic muscle bers is more obvious, which is different from the expression pattern in DMD or BMD.

Paraspinal muscle in ammatory cell in ltration
Meanwhile, we also found signs of in ammatory in ltration into the paraspinal muscles in patients with AIS. High expression of MHC-1 was observed, a protein that could release autoantigens to T-lymph receptors on the surface of sensitized CD8 + T cells and activate CD8 + T cells under the action of co-stimulatory factors on the surface of muscle bers. Activated CD8 + T cells release perforin and granzyme, causing muscle ber necrosis. This is also the pathogenesis of polymyositis in in ammatory myopathy (23).
In the biopsies of the paraspinal muscles of AIS patients, we also observed the in ltration of CD4 + and CD8 + cells in the paraspinal muscles and tendons, but no CD20 + cells were present. Therefore, the pathological changes are very similar to the toxic effects of T cells. The killing effect of T cells to paraspinal muscle cells may account for the atrophy and necrosis of the paraspinal muscles as well as the unbalanced tractions of the spine that lead to scoliosis. We did not have a clue as to how the

Limitations
Because of the severe atrophy of the scoliosis and the concave side muscles in some patients, the biopsy areas and locations could be rather limited, which inevitably affect the percentage of atrophy. The paraspinal muscle bers of AIS presented dystrophin dysfunctions, which were closely related to dystrophin-glycoprotein complex and sarcolemma stability. Further studies need to take into account more relevant proteins and explore the mechanisms of AIS.

Conclusion
Paraspinal muscles play an important role in the formation and development of scoliosis. Our study suggests that dystrophin protein may signi cantly decrease or even disappear in patients with scoliosis and that its amount is associated with the severity of scoliosis. Therefore, dystrophin protein dysfunction may contribute to the occurrence and development of scoliosis. Meanwhile, the in ammatory changes mainly manifested in T cell activation and in ltration, and there seemed to be certain correlations between the expression of MHC-1 in in ammatory cell in ltration and the abnormal expression of anti-myoatrophy protein.

Consent for publication
Written consent was obtained from all participants.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
All authors declare no con ict of interest.    Immunohistochemical staining of paraspinal muscle biopsy in AIS patients. Dystrophin-2 and Dystrophin-total proteins were absent in most of the atrophic muscle bers and some of the non-atrophic muscle bers. The expression of dystrophin protein was more abnormal or absent in the severe scoliosis group, and the difference between the two groups was statistically signi cant. In the severe scoliosis group, the expression of Dystrophin protein was subject to signi cantly increased abnormities or absence. (A) Severe scoliosis group: Dystrophin-2 was absent in most atrophic muscle bers and some non-atrophic muscle bers (Dystrophin-2, 40×). (B) Severe scoliosis group: Dystrophin-total was absent in most atrophic muscle bers and some nonatrophic muscle bers (Dystrophin-total, 40×). (C) Mild scoliosis group: Dystrophin-2 was absent in partial atrophic and nonatrophic muscle bers (Dystrophin-2, 40×). (D) Mild scoliosis group: Dystrophin-total was absent in partial atrophic muscle bers and non-atrophic muscle bers (Dystrophin-total, 40×).

Figure 3
Immunohistochemical staining of Dystrophin-1 (rod-like domain epitopes) and Dystrophin-3 (N-terminal epitopes) in AIS paraspinal muscle biopsy. Dystrophin-1 immunostaining showed dizzy or light-colored lineation in atrophic muscle bers, with linear discontinuity in a few cases. Dystrophin-3 immunostaining showed that the myolemma of both atrophic and non-atrophic muscle bers were weakened or even disappeared in all cases. However, there was no statistically signi cant difference between the mild and severe scoliosis groups.