Introduction

Adenylosuccinate synthase 1 (ADSS1) myopathy is a rare autosomal recessive muscle disease caused by mutations in the ADSS1. The causative gene for this myopathy was previously referred to as ADSSL1 but has been renamed ADSS1 according to the updated gene nomenclature.[1] The ADSS1 encodes adenylosuccinate synthetase 1, an enzyme critical for the purine nucleotide synthesis pathway [2,3], and its dysfunction is thought to impair energy homeostasis in muscle cells, leading to progressive muscular dysfunction [4–6]. First characterized in Korean patients in 2016 [2], ADSS1 myopathy has since been identified in other populations including Japan, India, Europe, China, and the United States [7–11]. To date, over a hundred genetically confirmed cases have been reported worldwide, establishing ADSS1 myopathy as a distinct hereditary myopathy [10].

Initial reports described ADSS1 myopathy as a childhood-onset, slowly progressive, distal myopathy with facial and oropharyngeal involvement [2,3]. Subsequent studies revealed wider phenotypic variability, with both proximal and distal onset presentations and varying degrees of severity across different ethnic groups [7,8,10,12]. Notably, the founder mutation c.781G > A (p.Asp261Asn) has been frequently reported in East Asian populations, suggesting a regional genetic signature with potential implications for diagnosis and screening [3,8,10]. Radiological studies have demonstrated distinctive patterns of muscle involvement predominantly affecting the distal muscles [13,14]. Recent studies have progressively expanded our understanding of the clinical spectrum of the disease, highlighting ADSS1 myopathy as an emerging entity of global significance despite its initial characterization in Korean patients [7–10].

Despite these advances, significant questions remain unresolved regarding ADSS1 myopathy. Current phenotypic classifications are inadequate for capturing the full clinical spectrum, and genotype-phenotype correlations remain unclear [6,10]. Moreover, significant variability in disease severity among patients with identical genotypes suggests the presence of unidentified modifying factors [7,8,10,12]. In other neuromuscular disorders, such as dysferlinopathy, muscle MRI has proven invaluable in redefining clinical subtypes and identifying shared pathological signatures [14–17]. Although ADSS1 myopathy is associated with distinctive progressive proximal and distal muscle weakness [2,3,8], systematic imaging studies remain scarce. In landmark studies of other myopathies, distinct phenotypes converged on common MRI patterns, supporting the conceptual unification of these disorders [15,18]. This suggests that MRI can serve as a powerful tool for phenotype refinement and disease monitoring beyond clinical observation alone [13,14].

To date, however, no study has quantitatively characterized the MRI patterns of ADSS1 myopathy. Our study provides the first systematic radiological evaluation to define its disease-specific MRI signature and to clarify the clinical-radiological relationship. By classifying patients according to the anatomical location of symptom onset (proximal vs. distal) and correlating these classifications with muscle imaging findings and genetic profiles, we assessed whether the current clinical subtypes reflected the underlying pathophysiology.

Methods

Results

Demographic and clinical characteristics (Table 1 and S1 Table)

Our cohort consisted of 30 Korean patients with genetically confirmed ADSS1 myopathy (18 males, 12 females). The median age at symptom onset was 8.0 years [interquartile range (IQR): 7.0–14.0], with patients examined at a median age of 38.0 years [IQR: 25.0–49.8], representing a median disease duration of 24.0 years [IQR: 17.0–34.0], as summarized in Table 1. Based on the anatomical location of the initial weakness, the patients were classified into the proximal-onset (n = 9) and distal-onset (n = 20) groups, with one patient presenting with isolated hyperCKemia without muscle weakness.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Clinical characteristics of 30 patients with ADSS1 myopathy, including a comparison between proximal and distal onset groups (n (%) or median [interquartile range]). Detailed information for the 10 MRI-evaluated patients is provided in S2 and S3 Table.

https://doi.org/10.1371/journal.pone.0324116.t001

The most common initial symptoms were slow running (66.7%), early fatigue (16.7%), gait disturbance (10.0%), and difficulty climbing stairs (10.0%). On examination, most patients (66.7%) exhibited greater weakness in the distal muscles than in the proximal muscles. Notably, facial muscle involvement (facial diplegia) was observed in 80.0% of patients. Laboratory findings revealed a median serum creatine kinase (CK) level of 214.0 IU/L [IQR: 125.0–394.0] and a median uric acid level of 5.6 mg/dL [IQR: 4.7–6.8]. There were no statistically significant differences between proximal- and distal-onset groups in age at onset, CK, urate, or disease duration (all p > 0.05). Other systemic or cranio-bulbar features (facial weakness, dysarthria, respiratory or cardiac involvement) were uncommon and are summarized in S1 Table.

Genetic analysis (Fig 1–2, and Table 2)

Genetic analysis of 30 patients (60 alleles) revealed five distinct pathogenic variants of ADSS1 (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Genotype distribution in 30 Korean patients with ADSS1 myopathy.

https://doi.org/10.1371/journal.pone.0324116.t002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Pathogenic variants in the ADSS1 gene identified in Korean patients with ADSS1 myopathy.

Schematic representation of the ADSS1 gene showing the distribution and frequency of the five identified variants. Numbered boxes represent exons [1–13]. Dots represent individual alleles from 30 patients (60 alleles total), with variants color-coded by mutation type: missense (green), frameshift (purple), and splicing (red). The most common variants were c.781G > A (51.7%) and c.919del (40.0%). Asterisk indicates the novel splice-site variant c.667-2A > T.

https://doi.org/10.1371/journal.pone.0324116.g001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Comparison of ADSS1 variant distributions across different populations.

Bar graph showing the allele counts of ADSS1 pathogenic variants in Korean (blue), Japanese (red), and other populations (green). Note the predominance of c.781G > A and c.919del variants in both Korean and Japanese populations. Greater variation is observed in other populations(India, Turkey, and USA) with variants not detected in East Asian patients.

https://doi.org/10.1371/journal.pone.0324116.g002

The most prevalent variant was c.781G > A (p.Asp261Asn), which accounted for 51.7% (31/60) of all alleles, followed by c.919del (p.Ile307Serfs25), which accounted for 40.0% (24/60). The less common variants included c.233_234del (p.Lys78Argfs33) at 5.0% (3/60), c.1091T > C (p.Leu364Pro) at 1.7% (1/60), and a novel splice-site variant c.667-2A > T (p.?) also at 1.7% (1/60). Zygosity analysis revealed that most patients (73.3%, 22/30) were compound heterozygotes for c.781G > A and c.919del (Table 2). Other genotypes included homozygosity of c.781G > A (6.7%, 2/30), compound heterozygosity of c.781G > A and c.233_234del (10.0%, 3/30), homozygosity of c.919del (3.3%, 1/30), and compound heterozygosity of c.781G > A with either c.667-2A > T or c.1091T > C (3.3% each, 1/30 each). All variants except c.667-2A > T have been previously reported.

An international comparison of variant distributions demonstrated similarities between Korean and Japanese patients, with c.781G > A and c.919del as the predominant variants in both populations (Figure 2).

In our Korean cohort, these two variants accounted for more than 90% of all disease alleles. In contrast, patients from other countries showed different variant patterns with additional variants (c.769G > A and c.794G > A) that were not observed in our Korean cohort. The Japanese cohort exhibited the highest numbers of c.781G > A and c.919del alleles among the reported populations. The c.667-2A > T variant identified in one patient represents a previously unreported splice-site variation in ADSS1, affecting the canonical acceptor splice site of intron 6.

Muscle MRI findings (Figs 3,4)

Muscle MRI examinations were performed in 10 patients (two with proximal onset and eight with distal onset) to evaluate the patterns of muscle involvement. Neurological examinations performed around the time of MRI confirmed that the distribution of muscle weakness at imaging was consistent with the clinical phenotype, allowing direct comparison between strength patterns and imaging findings. The severity of fatty infiltration in the lower limb muscles was assessed using the Mercuri scale and visualized as a heat map (Fig 3). Heat map analysis (Fig 3) demonstrated that the most severely affected muscles were predominantly located in the distal compartments of the lower limbs.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Fatty infiltration pattern in lower limb muscles of ADSS1 myopathy patients assessed using T1-weighted MRI.

Heat map displaying the severity of fatty infiltration in 22 individual muscles of both lower limbs from 10 patients with ADSS1 myopathy, graded according to Mercuri#39;s scale from 0 (blue, normal) to 4 (red, severe fatty replacement). Patients are arranged in data collection order (not ordered by total Mercuri score) and grouped by their clinical onset classification. Note the consistent pattern of predominant involvement of distal muscles, particularly gastrocnemius and soleus, across all tested patients. Left and right sides are displayed separately for completeness, although asymmetry was rare (observed in one patient).

https://doi.org/10.1371/journal.pone.0324116.g003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Representative T1-weighted MRI images from proximal-onset and distal-onset ADSS1 myopathy patients.

Axial T1-weighted MRI sections at thigh (upper panels) and calf (lower panels) levels from (A) a proximal-onset patient (G-01) and (B) a distal-onset patient (L-01). Both patients show similar patterns of muscle involvement with predominant fatty replacement in distal muscles, particularly the gastrocnemius and soleus, while proximal muscles are relatively less affected.

https://doi.org/10.1371/journal.pone.0324116.g004

Specifically, the medial and lateral gastrocnemius muscles showed the most extensive fatty replacement (grades 3−4) in the majority of patients, followed by the soleus muscle. The tibialis anterior, popliteus, extensor digitorum longus, peroneus longus, and tibialis posterior muscles also showed moderate-to-severe involvement. In contrast, the proximal muscles generally displayed less severe fatty infiltration, except for the vastus lateralis and sartorius, which were more frequently and severely affected than other proximal muscle groups. Representative MRI images from a proximal-onset patient (G-01) and a distal-onset patient (L-01) are shown in Fig 4.

Both patients exhibited predominant fatty replacement in the distal lower limb muscles, particularly in the posterior compartment, whereas the proximal muscles were relatively spared.

Discussion

This study represents the largest single-center cohort analysis of ADSS1 myopathy in Korea and provides comprehensive insights into its clinical, genetic, and radiological features. Our findings highlight a striking contrast between the clinical heterogeneity of the initial presentations and the remarkably consistent pattern of muscle involvement observed in imaging studies. One of the most significant findings of our study was the discordance between clinical phenotypes and radiological manifestations of ADSS1 myopathy. Clinically, we observed considerable heterogeneity in the anatomical sites of the initial symptoms, allowing us to classify the patients into proximal-onset (n = 9) and distal-onset (n = 20) groups. These groups exhibited differences in their clinical presentations. Patients with proximal onset exhibited higher rates of early fatigue and facial weakness, whereas those with distal onset more frequently presented with slow running and gait disturbances. However, none of these clinical distinctions were statistically significant, suggesting a substantial overlap between the phenotypic subgroups. This clinical variability is consistent with the results of previous reports. While ADSS1 myopathy was initially described as a predominantly distal myopathy in Korean patients [2,3], the phenotypic spectrum has since expanded. Recent studies have documented both distal and proximal muscle weakness in non-Korean patients and variable muscle symptoms affecting the proximal and/or distal leg muscles, tongue, masseter, diaphragm, and paraspinal muscles in Japanese cohorts [7,8]. Additional clinical features, including significant seasonal fluctuations and decremental responses to repetitive nerve stimulation, have been reported in patients from other populations [12]. In contrast to this clinical variability, muscle MRI revealed a remarkably uniform pattern of muscle involvement across all examined patients, regardless of their initial clinical presentation. Both proximal- and distal-onset patients demonstrated predominant fatty replacement in the distal lower limb muscles, particularly the gastrocnemius and soleus, with relative sparing of the proximal muscles (except for the vastus lateralis and sartorius, which showed earlier involvement than the other proximal muscles). This consistent distal-dominant pattern on MRI suggests that the underlying pathological process follows a predictable anatomical distribution, independent of the initial symptomatic muscle groups. This discrepancy between clinical presentation and imaging findings raises fundamental questions regarding the pathophysiology of ADSS1 myopathy. A comparable clinical–radiological dissociation has been reported in dysferlinopathy, where patients classified by initial presentation as either limb-girdle (LGMD2B) or distal onset (Miyoshi myopathy) exhibit identical MRI patterns of muscle involvement (Paradas et al., Neurology, 2010). In that study, as in ours, classification was based on the initial site of weakness, as standardized strength scores concurrent with imaging were not uniformly available. Nevertheless, the MRI findings revealed a consistent underlying pattern that redefined the disease as a single pathological spectrum. Similarly, the present study demonstrates that despite variable clinical onsets, ADSS1 myopathy follows a uniform distal-dominant MRI distribution, supporting its conceptualization as a unified disorder rather than distinct proximal and distal forms. This suggests that factors beyond the underlying pattern of muscle involvement, such as individual compensatory mechanisms, daily activity patterns, or genetic modifiers, may influence the muscle groups that first manifest clinically detectable weakness. However, consistent MRI patterns may reflect the true biological vulnerability of specific muscle groups to ADSS1 deficiency [8,14]. Our findings strongly support the utility of muscle MRI as a valuable diagnostic tool for ADSS1 myopathy. The characteristic pattern of distal dominant muscle involvement, particularly the gastrocnemius, soleus, and tibialis anterior muscles, appears to be consistent with the radiological signature of this disorder. This pattern was observed in all examined patients, suggesting high sensitivity for detecting ADSS1 myopathy. This radiological signature may be particularly valuable in diagnostically challenging cases, such as those presenting with isolated facial weakness or minimal limb symptoms, where the characteristic distal-dominant MRI changes help confirm the diagnosis, especially when initial genetic testing is inconclusive. In these scenarios, recognition of the characteristic MRI pattern can guide targeted genetic analysis and potentially expedite diagnosis [13,14]. Furthermore, in the context of the East Asian founder mutations identified in our cohort (c.781G > A and c.919del), the combination of this distinctive MRI pattern with targeted genetic testing for these common variants may represent an effective diagnostic algorithm for suspected ADSS1 myopathy in Korean patients. This observation parallels findings in other neuromuscular disorders, where muscle MRI has significantly contributed to phenotype refinement and diagnostic accuracy. Previous studies have demonstrated distinct muscle imaging patterns in various myopathies that help separate genetically heterogeneous subtypes and reveal specific patterns that aid in diagnosis [16,18]. Notably, muscle MRI has been used to redefine the phenotypes of other neuromuscular disorders, demonstrating that patients with clinically distinct presentations may show identical patterns of muscle involvement in imaging [15]. Beyond its diagnostic utility, muscle MRI provides unique insights into the natural history and pathophysiology of ADSS1 myopathy, which cannot be obtained through clinical examination alone. The ability to visualize and quantify the extent of muscle involvement across different muscle groups allows for a more precise assessment of disease severity and progression than is possible with clinical strength testing, particularly in muscles that are difficult to isolate on examination [13]. Despite clinical heterogeneity, the consistent distal dominant pattern observed on MRI supports the conceptualization of ADSS1 myopathy as a unified disease entity with a common underlying pathophysiological mechanism. This perspective parallels the evolution of our understanding of other neuromuscular disorders, in which initially distinct clinical syndromes were later recognized as different manifestations of the same underlying pathological process [15]. The preferential involvement of distal muscles, particularly those with high energy demands such as the gastrocnemius and soleus, may reflect their heightened vulnerability to disruptions in energy metabolism caused by ADSS1 deficiency [4,19]. These muscles, which are continuously engaged in postural maintenance and locomotion, may be particularly dependent on efficient purine nucleotide cycling and ATP regeneration [4,5]. The purine nucleotide cycle, comprising adenylosuccinate synthetase (ADSS1), adenylosuccinate lyase, and adenosine monophosphate (AMP) deaminase, plays a crucial role in muscle energy metabolism during intense exercise, contributing to ATP regeneration and providing Krebs cycle intermediates [4]. The selective vulnerability of specific proximal muscles (vastus lateralis and sartorius) can be explained by their metabolic profiles or fiber-type composition, which may render them more susceptible to the metabolic consequences of ADSS1 dysfunction than other proximal muscles. This hypothesis is supported by transcriptome analysis demonstrating that genes involved in purine nucleotide cycling (ADSS1, ADSL, and AMPD1) are significantly downregulated in the muscle tissues of patients with ADSS1 myopathy [6]. This unified model of ADSS1 myopathy, which is characterized by selective muscle vulnerability based on metabolic requirements rather than anatomical location, has important implications for understanding disease progression and developing therapeutic strategies. This suggests that treatments targeting the fundamental metabolic disruption caused by ADSS1 deficiency may be effective regardless of the clinical phenotype.

Our study has several limitations. First, despite being the largest Korean cohort with ADSS1 myopathy to date, our sample size was relatively small, particularly for subgroup analyses. Second, muscle MRI was performed in only a subset of the patients (10 of 30), potentially limiting the generalizability of our radiological findings. Third, the cross-sectional nature of our imaging data precludes definitive conclusions regarding the temporal progression of muscle involvement. Additionally, quantitative MRI fat-fraction or biochemical correlations and RNA-based analyses were not performed due to lack of available data and IRB restrictions. Future research should address these limitations through longitudinal studies that track both clinical progression and changes in muscle MRI findings over time. It is particularly important to investigate the temporal relationship between the initial onset of clinical symptoms and the development of characteristic MRI patterns. Such studies could clarify whether radiological changes precede clinical manifestations, potentially identifying the pre-symptomatic phase during which therapeutic interventions might be most effective. Additionally, correlation studies between muscle MRI findings and functional measures of ADSS1 enzyme activity or markers of purine metabolism could provide deeper insights into the pathophysiological mechanisms underlying the selective muscle vulnerability observed in this disorder [6,20]. Recent pilot investigations using electrical impedance myography have demonstrated the potential for noninvasive characterization of muscle integrity in patients with ADSS1 myopathy and could complement MRI findings in future studies [21]. Further exploration of the potential genetic modifiers influencing the clinical manifestations of ADSS1 myopathy in the setting of identical pathogenic variants would also be valuable [10].

Supporting information