Phenotype Genotype Characterization of FKRP-related Muscular Dystrophy among Indian Patients

INTRODUCTION

Fukutin-related protein (FKRP) is a ribitol phosphate transferase involved in the glycosylation of dystroglycans on the sarcolemma, which is an important step that helps in its attachment to the laminin proteins in the extracellular matrix [1]. The gene encoding FKRP is located at chromosome 19q13.3 and spans about 12.5kb [2]. Mutations in the FKRP gene result in highly variable phenotypes and comprise of autosomal recessive LGMD 2I [3], currently renamed as LMGD R9- FKRP related [4], autosomal recessive FKRP related congenital muscular dystrophy, also called merosin-deficient congenital muscular dystrophy type 1C (MDC1C) [2], and the severe phenotype of Walker-Warburg syndrome and Muscle-Eye-Brain disease [5, 6]. Here we describe the clinical, genetic, and laboratory findings in nine patients with FKRP gene mutations from India. For this study patients who have not attained the ability to walk independently till 2 years of age and till the last available follow-up have been classified as having a congenital muscular dystrophy phenotype and others as having a LGMD phenotype in accordance to the 229th ENMC international workshop proposal for Limb girdle muscular dystrophies [4].

MATERIALS AND METHODS

This retrospective study includes all patients with a genetically confirmed FKRP mutation between 2016 and 2020. An institutional ethics committee approval was obtained for the study (NIMHANS/IEC/2020-21). Among 345 genetically confirmed cases of autosomal recessive muscular dystrophies and 73 cases of congenital muscular dystrophy (CMD), nine unrelated patients with FKRP mutations were identified and included in the current study. Clinical characteristics and laboratory results, including histopathological findings and mutational characteristics, were extracted from archived records. The data includes a pre-designed elaborate proforma maintained in each patient’s records.

Histopathology

Seven patients had undergone open biopsy from either the quadriceps, biceps or tibialis anterior muscle. The biopsied muscle was subjected to snap (Flash) freezing using isopentane precooled (chilled) in liquid nitrogen, after which the frozen muscle was cryosectioned using a cryomicrotome. The transverse sections of the muscle thus obtained after cryosectioning, were subjected to non-enzyme histochemical staining that included Haematoxylin Eosin (HE) stain, modified Gomori trichrome (MGT) along with periodic stain with and without diastase stain. Also, the enzyme histochemical stains were performed on the cryosections which included succinic dehydrogenase(SDH), Nicotinamide adenine dinucleotide Tetrazolium Reductase (NADH -TR), adenosine tri phosphatase (ATPase) pH 9.4, 4.6 and succinic dehydrogenase–cytochrome C oxidase (SDH-COX). The cryosections were subjected to immunohistochemistry, the steps of which includes peroxidase blocking (NCL), protein blocking, thorough washing with tris-buffered saline (TBS), incubation with the primary, post primary (polymer) and secondary antibodies and finally application of DAB (DiaminobenzidineHCl, NCL-1:20) chromogen. Immunohistochemistry for dystrophins (Dy4/6D3, Dy8/6C5, Dy10/12B2, NovocastraTM Leica Biosystems) sarcoglycans (alpha, Beta, Gamma, Delta: Ad1/20A6, βSarc1/5B1, 35DAG/21B5, δSarc3/12C1 NovocastraTM Leica Biosystems), and dysferlin (Ham3/17B2, NovocastraTM Leica Biosystems, Germany) proteins was done in four samples (patients: 2,5,6, and 8). Immunohistochemistry for merosin (Mer3/22B2 NovocastraTM Leica Biosystems, Germany) was done in five patients (patients 2,6,7,8 and 9). Western blotting for alpha-dystroglycan (IIH6C4, Anti-α-Dystroglycan Antibody 05-593, Merck Millipore) was done in one patient (patient 7) using the anti-α-Dystroglycan antibody.

RESULTS

Investigations

The mean creatine kinase (CK) level was 12,120 U/L (range 2793 to 32,396 U/L, lab reference value 20–200 U/L). Muscle biopsy was done in 7 patients, and histology revealed a variable degree of dystrophic changes with evolving fibrosis. Immunohistochemistry for dystrophin, sarcoglycans, and dysferlin proteins was done in four samples (patients: 2,5,6, and 8), which showed normal expression. An absent or reduced staining for merosin was noted in three patients (patients 7,8,9): and normal expression in two of them (patients 2 and 6). Western blotting for alpha-dystroglycan was done in one patient (patient 7) and showed an absence of the 156kDa band. Figure 3 shows representative microphotographs of the histopathological changes.

MRI of the brain was done on two patients (patient 2 and 9) both of whom had a delay in speech development (Fig. 1c and 1d). Both showed periventricular white matter T2 hyperintensities. CT scan of the brain done in one patient (patient 7) who had not attained ambulation till 3 years of age was normal.

Fig. 1

Clinical photographs and MRI images of patients with congenital muscular dystrophy phenotype. a and b: Calf muscle hypertrophy and scapular winging in patient 2 with congenital muscular dystrophy phenotype c and d: Brain MRI showing periventricular white matter T2 hyperintensities in patient 9 (figure c) and patient 2 (figure d) with congenital muscular dystrophy phenotype.

Fig. 2

Schematic representation of the FKRP gene. The exons are represented as rectangular boxes with respective exonic numbers. 5^′ UTR is blackened and 3^′ UTR is shown in grey. The four novel mutations identified in this study are marked in red rectangular boxes.

Fig. 3

Representative microphotographs showing the histopathological changes. A to D: Representative microphotographs showing variation in fibre size with variable atrophic and scattered hypertrophic fibres with regenerating fibres and evidence of interstitial (endomysial) fibrosis (Transverse section, H & E, X 200). E: Microphotograph showing muscle fibres with reduced expression of merosin (Inset shows control staining) (IHC, X 200). The perimysium is thickened and endomysial fibrosis is observed. F: Representative images of muscle immunohistochemistry (Alpha, Beta, Gamma and Delta Sarcoglycan; IHC x 200).

DISCUSSION

This is the first study examining the spectrum of genetically confirmed FKRP-related muscular dystrophy from India and highlights the clinical and genetic heterogeneity of the disorder. LGMD R9 phenotype seen in 6 patients was the most common presentation in this short series, while a congenital muscular dystrophy presentation was noted in only three patients. All the patients with LGMD R9 had a Duchenne-like phenotype, however in all these patients the first symptom was noted before two years of age which is earlier than the average age of onset of Duchenne muscular dystrophy reported from India [9]. Notably the age of onset of our patients is much earlier as compared to prior reports of patients with LGMD R9 [10–13]. The mean age of onset of the Western blot confirmed LGMD2I/R9 patients from our previously published cohort of autosomal recessive LGMD was 15.4 years [14]. This wide range of presentations of the FKRP-mediated LGMD phenotype is well established. The Duchenne-like presentation was first described by Brockington et al., in 5 out of 17 families with LGMD 2I, while a later age at presentation and slower progression was seen in the other 12 families [3]. The high frequency of the c.826C>A mutation in the western population, which has a milder phenotype could explain this phenotypic difference [13]. Table 2 summarizes the age of onset and mutations reported in various case series of the FKRP related muscular dystrophy published in literature previously.

Calf muscle hypertrophy, tendo-achilles contractures, scapular winging, and macroglossia seen in our patients have previously been reported in patients with LGMD R9 [10, 15].

We found a selective weakness of the quadriceps and hip extensors in our patients where a formal assessment of the muscle power was possible. Selective weakness of the quadriceps among FKRP patients is previously described [10]. However, in contrast to observations by Poppe et al., we noted a selective weakness of the hip extensors [10]. This is in keeping with muscle MRI observation for early involvement of gluteus maximus previously reported [15].

Unassisted ambulation was lost in 3 out of 6 patients with LGMD R9 phenotype at a median age of 7 years; however, the age at last follow-up of the patients with retained ambulation was between 3 and 7 years. In their Danish cohort, Sveen et al., reported the loss of ambulation at a mean age of 44 years in patients with c.826C>A homozygous mutation and 20 years with other mutations [13]. Mercuri et al., in their study, report loss of independent ambulation in teenage with a Duchenne like phenotype and preserved ambulation beyond the second decade in the less severe group [16]. The tendency for an earlier loss of ambulation in our patients is in keeping with the Duchenne-like phenotype.

None of our patients had symptoms suggestive of cardiac dysfunction or respiratory involvement; however, none had entered into the second decade of life at last follow-up. Cardiac and respiratory involvement is well described in LGMD R9, but the age of onset is quite variable [11, 13, 15, 16]. Sveen et al., have reported a reduced ejection fraction in 10 out of 39 patients with LGMD 2I, while four in their study were on continuous ventilator support and six were using other respiratory assistive devices intermittently [13].

Three of our patients had a CMD-like presentation with language delay in two, and none achieved independent ambulation. None of our patients had the severe phenotype of Walker-Warburg syndrome or Muscle-Eye-Brain disease. The inability to walk independently after 2 years of age and till the last available follow-up was used as the criteria for classification as congenital muscular dystrophy in this study. One of the 6 patients (patient 6) classified as having LGMD R9 had a documented history of reduced fetal movements antenatally and poor sucking after birth. However, this patient has been classified as LGMD R9 as he eventually attained independent ambulation. None of the other LGMD R9 patients had a documented hypotonia at birth. Four of the seven different mutations identified were previously unreported, of which the three missense variants are currently classified as variants of uncertain significance (VUS) as per ACMG guidelines [7]. However, the c.650C>A (P217Q) in patient 2 is adjacent to the previously reported c.646C>T (R216W) [17] observed in patient 1, suggesting a possible hotspot location in exon 4. Compared to patient 2, who never attained independent ambulation and had additional language delay and MRI brain abnormalities, patient 1 had a milder phenotype having attained independent ambulation at 1 year and remaining ambulant up until 9.5 years of age. In patient 9, who like patient 2 had a CMD phenotype with language delay and brain MRI abnormalities, the two novel missense changes, c.1060_1061delGGinsCC (G354P) and c.1136G>C (R379P), were confirmed to be present in trans based on parental segregation. Based on the above findings and clinical correlation, we posit that these three FKRP missense changes can be considered disease-causing. However, further functional evidence and identification of more patients might be required to establish the pathogenicity.

The most common FKRP mutation reported from most of the studies from Europe and Brazil is the c.826C>A mutation [3, 11, 13, 15, 18] which typically has a milder presentation. In contrast, c.948delC and c.169G>A is the most common FKRP mutation reported in studies from Taiwan and Japan, respectively [12, 19]. Table 2 summarizes the mutations described in the various prior studies on FKRP.

The significant finding in our series is the recurrence of c.1343C>T(P448L) in five patients which was first reported as homozygous in a Libyan patient from a consanguineous family [2]. This patient was reported to have congenital onset hypotonia with feeding difficulties and motor delay evidenced by lack of ambulation at 16 months of age. Calf hypertrophy and facial weakness were reported with elevated CK. However, brain MRI was normal with normal intelligence suggesting a CMD phenotype without brain involvement. Few other studies have also reported the c.1343C>T(P448L) mutation in patients with LGMD phenotype and Duchenne like phenotype [20–22]. Further animal studies showed that the homozygous (P448L) mouse model exhibited a moderate to severe form of dystrophic pattern with a marked reduction in functional glycosylation of α-DG in skeletal muscles but without any brain defects consistent with severe LGMD2I (R9) or MDC1C phenotype observed in the Libyan patient [23]. In our cohort, the P448L mutation is present as homozygous in four patients (patients 3, 4, 6, 7) and as likely compound heterozygous with a novel frameshift c.935_936del(R312Lfs*77) mutation in patient 8. The P448L variant was found to be in heterozygous state in unaffected parents tested and all five families with the mutation hailed from south India. GnomAD population database reported P448L variant in 2 south Asian controls in heterozygous. Further haplotype studies can be crucial in identifying any possible founder effect. Three of the four patients with homozygous P448L had LGMD2I with a Duchenne-like phenotype. They had motor development delay, with independent ambulation attained between 20 and 24 months and loss of independent ambulation being recorded in one at the age of 6.5 years. Calf hypertrophy was seen in all 4, and scapular winging was absent. The mean CK of these patients was 10,924 U/L. One had a CMD phenotype. However, in comparison to the other two patients who had a CMD phenotype, this patient had no delay in non-motor developmental realms and had a normal brain CT scan similar to the first report of this mutation [2]. He was considered to have a CMD phenotype rather than LGMD as he was unable to stand even with support till the last follow-up, which was at three years of age [4].

One patient (patient 8) had the mutation in a compound heterozygous state. She also presented with an LGMD R9 –Duchenne like phenotype, attaining independent ambulation by 3 years, having calf hypertrophy and highly elevated CK level of 13,030 U/L. Four of the five patients harboring the P448L mutation had undergone a muscle biopsy, exhibiting a dystrophic picture of variable degree and evolving fibrosis. Among the three patients who had immunohistochemistry stains done, two of them displayed a reduced or deficient staining for merosin, including one with the P448L mutation in a homozygous state (patient 7) and one in a compound heterozygous state (patient 8). This is in contrast to the muscle biopsy findings reported previously in a patient harboring the same mutation who had normal staining for merosin [2]. Merosin deficiency in patients with FKRP-mediated muscular dystrophies is thought to be secondary to the loss of alpha-dystroglycan in the muscle membrane [2]. The reduced merosin staining in our patients with the P448L mutation could represent a more advanced stage of the dystrophic process.

CONCLUSIONS

This is the first description of the clinical and mutational spectrum of genetically confirmed cases of FKRP mutations from India. LGMD R9 with a Duchenne-like phenotype was more common than a congenital muscular dystrophy phenotype in our cohort. We report c.1343C>T(P448L) as a common recurrent mutation in Indian FKRP patients but further larger studies are required to determine the frequency of this mutation and a possible founder effect. Awareness of a Duchenne muscular dystrophy mimicker in terms of the clinical presentation, progression of symptoms, and creatine kinase levels is critical among clinicians, especially for guiding genetic testing.

CONFLICTS OF INTEREST

No potential conflicts of interest to declare.