Mutations outside the N-terminal part of RBCK1 may cause polyglucosan body myopathy with immunological dysfunction: expanding the genotype–phenotype spectrum

Some glycogen storage diseases are characterized by the pathological accumulation of polyglucosan bodies, which consist of unspecific polysaccharides resistant to alpha-amylase digestion as a consequence of a defective glycogen metabolism [1]. Clinically, heart and skeletal muscles are predominantly affected due to their high physiological glycogen turnover [2, 3]. Homozygous or compound heterozygous mutations in the RBCK1 gene were recently discovered to underlie a few cases with primary muscular involvement, a condition that was subsequently termed polyglucosan body myopathy 1 (PGMB1, MIM#615895) [4, 5]. From a functional point of view, the corresponding protein product HOIL-1 plays a crucial role in myogenesis and is enriched in fast-twitch glycolytic muscle fibres along with its interaction partners [6]. Accordingly, subjects with biallelic loss-of-function mutations resulting in HOIL-1 deficiency usually suffer from progressive muscular weakness and childhood- or juvenile-onset dilated cardiomyopathy, often necessitating heart transplantation at a young age [4, 5]. In addition to its function in muscle cells, HOIL-1 also constitutes an essential part of the so-called linear ubiquitination chain assembly complex (LUBAC), which regulates a variety of important NF-κB-dependent immune response mechanisms [7]. Affected subjects may simultaneously suffer from both chronic autoinflammation and immunodeficiency including recurring septicaemia [8]. The patients with RBCK1 mutations reported so far vary considerably with respect to their leading clinical presentation (i.e., skeletal muscle, heart muscle, autoinflammation or immunodeficiency). The reason for this individual variability remains unclear, though it was hypothesized that the exact location of the variant within the gene might be a predictor for the predominant phenotype, with mutations in the N-terminal region of RBCK1 primarily leading to immunological dysfunction and mutations in the middle or C-terminal parts rather resulting in a (cardio)myopathy phenotype [4].

Here, we report two unrelated individuals carrying the same homozygous mutation in the middle part of the RBCK1 gene. Both presented clinically with a severe (cardio)myopathy and concomitantly with a comparatively mild, but clearly detectable immunological dysfunction. Our report sheds further light on the genotype–phenotype correlations of RBCK1-related diseases.

In this report, we describe two new patients with pathogenic mutations in the RBCK1 gene and thus expand the total number of known families with this condition in the literature from 12 to 14. We can confirm the few previous reports insofar as both a myopathy and an immunological phenotype are part of the clinical spectrum of RBCK1-associated disease.

Previous reports presented cases with a strong preponderance of either the immunological or the myopathic phenotype and it still remains unclear what exactly determines the clinical picture.

The patients described by Boisson et al. (all harbouring mutations in the N-terminal part of RBCK1) suffered from severe, childhood-onset immunological dysfunction leading to early death in infancy due to septicaemia [8]. In contrast, the majority of patients published by Nilsson et al. and Wang et al. carried mutations in the middle or C-terminal part of the gene and showed a later developing neuromuscular and cardiac involvement [4, 5].

It was subsequently hypothesized that the nature and localization of the underlying mutation might predict the phenotype, with N-terminal mutations mainly causing immunological dysfunction. In contrast, variants in the middle- or C-terminal regions were presumed to predominantly cause cardiomyopathy and neuromuscular symptoms. Further, it was suggested that truncating variants might generally result in more severe phenotypes than missense mutations (see Table 2 and Fig. 3 for all reported families including main phenotype and variant localization) [4].

The present characterization of two subjects with a homozygous mutation in the middle part (exon 7, RING domain) of the gene expands our still scanty knowledge on the genotype–phenotype correlation insofar as both patients developed both an immunological and a myopathic phenotype, although (cardio)myopathy was the dominant clinical feature in both cases. Additionally, the two (unrelated) individuals also suffered from recurrent bacterial and viral infections including severe and prolonged manifestations such as pneumonia. Patient II developed a dermatologic condition called Sweet’s Syndrome later in the disease course, which is currently considered to be an autoinflammatory disease [11]. Generalized erythema was previously described for N-terminal RBCK1 mutations with immunological phenotypes by Boisson et al. [8]. Further, Patient II showed a non-necrotizing granulomatous reaction pattern of the tonsils, an uncommon, but unspecific histological finding suggestive of an exaggerated immune response [12, 13]. Interestingly, this phenomenon has previously been associated with an N-terminal deletion and a compound heterozygous frameshift variant in the middle part of RBCK1 by Nilsson et al. (Family 3). Another individual (Family 13) suffered from sarcoidosis, which is also characterized by granulomatous inflammation. Further, type 1 diabetes (Family 13) and gluten intolerance (Family 9) were mentioned in the same publication [4]. We suggest that all these conditions may potentially represent a manifestation of autoimmune pathology related to mutations in RBCK1. In contrast, the patients of Nilsson et al. carrying the same variant as our patients (c.896_899del) were not reported to display any autoinflammation or immunodeficiency phenotype. However, we cannot exclude that subtle immunological symptoms either occurred later in the disease course or remained unreported due to a mild phenotype in these cases. In comparison, the disease onset in our cases was at a higher age (12 and 14 vs. 5 and 6 years) and rather dominated by cardiac symptoms than by muscular weakness. Ultimately, all four patients with this truncating mutation developed dilated cardiomyopathy either resulting in heart transplantation (3 subjects) or death due to heart failure (1 subject).

Considering all cases reported to date, it appears that pathogenic RBCK1 variants appear to invariably cause a myopathy, but not necessarily immunological symptoms. So far, severe immunological phenotypes have only been reported for protein damaging N-terminal mutations, but milder immunological dysfunction is apparently not limited to this gene region (Fig. 3). A detailed review of the literature alongside with our newly reported cases shows that frameshift mutations beyond the N-terminus of RBCK1 may lead to a combined phenotype including both myopathy and immunological dysfunction in single families. We are aware that truncating mutations may potentially be associated with nonsense-mediated decay, eventually resulting in the complete absence of a functioning protein [14]. This obviously limits any conclusions of the within-gene location of mutations on the phenotype. However, the fact that some reported frameshift mutations outside the N-terminal region (as in our patients) do not cause a predominant, severe immunological phenotype, suggests that some protein function is remaining in these cases, which would support the view that the phenotype is at least partly explained by the within-gene location of the mutation.

As our cases demonstrate, a precise molecular diagnosis might provide supporting information prior to invasive treatments such as heart transplant surgery or stem cell transplantation, which was reported as a therapeutic option by Boisson et al. in RBCK1-related cases with severe immunodeficiency [8]. For example, in Patient II, neuromuscular symptoms had first been attributed to a treatment with corticosteroids. Knowing the underlying genotype, the myopathy was much more likely to be part of the genetic syndrome rather than a side effect of medication. This again highlights the clinical usefulness of next-generation sequencing techniques in complex neurogenetic diseases [15‐18].