Abstract
Biallelic mutations in XPNPEP3 gene, encoding a mitochondrial peptidase, mainly cause nephronophthisis, but associated muscle involvement remains poorly described. We report here a 44-year-old male presenting since childhood with exercise intolerance and recurrent rhabdomyolysis. Electroneuromyography revealed a sensory axonal neuropathy and brain MRI showed white matter lesions in the posterior cranial fossa. Muscle biopsy revealed ragged-red fibers, COX negative fibers and abnormal mitochondria in electron microscopy. Whole genome sequencing identified a homozygous frameshift variant in the XPNPEP3 gene. Our results expand the spectrum associated with XPNPEP3 variants, including metabolic myopathy with subclinical central and peripheral nervous system involvement.
Introduction
X-prolyl aminopeptidase 3 (XPNPEP3), a nuclear gene located on chromosome 22q13, encodes a metallo-exopeptidase that cleaves the N-terminal amino acid from peptides containing a proline residue in the second position.1,2 It is expressed in the mitochondria of several tissues, including the kidneys, heart, pancreas, skeletal muscles and testes, where it acts on proteins that have been imported into the mitochondrial matrix, contributing to their stabilization.1,3 Pathogenic biallelic variants of the XPNPEP3 gene are known to cause nephronophthisis (NPHP), a progressive cystic kidney disorder that leads to end-stage renal disease (OMIM#61359).2,4–7 This condition is considered a ciliopathy, as the majority of proteins implicated in NPHP localize to primary cilia, basal bodies and centrosomes. 8 Interestingly, XPNPEP3 is expressed in mitochondria rather than renal cilia, but it appears to play a role in ciliary function by interacting with proteins that are primarily involved in this process.7,8 This was demonstrated in zebrafish models, where suppression of XPNPEP3 resulted in phenotypes consistent with those observed in ciliopathies. 2 Biallelic XPNPEP3 variants have also been associated with extra-renal manifestations such as various neurological manifestations including tremor, dystonia, peripheral neuropathy, sensorineural hearing loss, epilepsy and cardiomyopathy.2,5 However, muscle involvement is poorly described to date. 5 Here we report for the first time a novel homozygous XPNPEP3 variant causing primary mitochondrial myopathy presenting as a metabolic myopathy with exertional symptoms and recurrent rhabdomyolysis as the predominant manifestation.
Methods
The patient was examined by one of the authors in the Neuromuscular Reference Center in Pitié Salpêtrière Hospital in Paris, France. The patient gave his informed consent for publication. The study was conducted in accordance with European Union and French bioethics laws and with the Convention of Helsinki.
Electroneuromyography (ENMG), grip test, blood tests, brain and muscle magnetic resonance imaging (MRI) and electroencephalography (EEG) results were analyzed. An open muscle biopsy of the deltoid muscle was performed, the samples were frozen in isopentane cooled in liquid nitrogen (−180°) and cryosections were processed according to recommended standard histological and histochemical techniques: Hematoxylin and Eosin (H&E), and Gömöri Trichrome (GT) to assess the morphology of the muscle sample and detect the presence of ragged red fibers (RRF), Oil Red O (ORO) to determine the content in lipids and Periodic acid-Schiff (PAS) for the content in glycogen. 9 Histochemical studies consisted in the following stains: reduced nicotinamide adenine dinucleotide-tetrazolium reductase (NADH-TR) that allows the detection of anomalies of the myofibrils and the distribution of the mitochondria, succinate dehydrogenase (SDH) to check for the distribution of the mitochondria and cytochrome c oxidase (COX) to identify fibers devoid of COX activity or with abnormal mitochondrial distribution. 9 We also performed ultrastructural analysis of a muscle fragment fixated in glutaraldehyde and included in epon. DNA was extracted from peripheral blood for genetic analysis. Next generation sequencing (NGS) was initially performed, followed by a whole genome sequencing (WGS), and putative variant was confirmed with Sanger sequencing. The WGS was performed at the SeqOIA laboratory (https://laboratoire-seqoia.fr/). The library was prepared using the NEB Next Ultra II End repair/A-tailing DNA Library Prep Kit (New England Biolab, Ipswich, MA, USA) and sequenced in paired ends (2 × 150 bp) using an Illumina Novaseq6000 platform. The reads were aligned to the reference human genome (GRCh38.92) using the BWA- MEM 0.7.15 software package. ClinSV 10 was used to annotate and prioritize structural variants, including CNVs. The resulting variants were then annotated with AnnotSVv2.5.1 in an in-house developed workflow (SeqOIA-IT platform).
Case presentation
A French 44-year-old man with a history of exercise intolerance and myalgia since childhood was referred to our neuromuscular reference center due to the suspicion of a metabolic myopathy. There was no consanguinity, nor family history of neuromuscular diseases. Motor development was normal. He reported transient episodes of intense post-exercise myalgia and muscle stiffness lasting several days, associated with rhabdomyolysis. Creatine kinase (CK) levels were variably raised, with a peak value of 24,000 U/L (normal <190 U/L) at first rhabdomyolysis episode.
The neurological examination was normal aside from a mild action tremor of the upper limbs. There was no muscle weakness and deep tendon reflexes were normal in four limbs.
Blood tests at rest revealed normal baseline CK levels and no thyroid dysfunction. Grip test showed mild baseline hyperlactatemia (1,8 mmol/L, N < 1,8 mmol/L) and significantly raised CK levels up to 5692 UI/L at 48 h post-exercise. Carnitine levels and acylcarnitine profile were in normal range. Lumbar puncture showed slightly elevated protein (730 mg/l, N < 450 mg/l) and lactate (2 mmol/L, N < 2 mmol/L). ENMG revealed an infraclinical axonal sensory neuropathy, with sural nerve amplitudes of 6.4 µV on the right and 5.4 µV on the left side (N ≥ 10 µV), and superficial fibular nerve amplitudes of 7.9 µV on the right and 5.8 µV on the left side (N ≥ 10 µV). Motor and sensory nerve conduction velocities and needle electromyography were normal in all four limbs. Brain MRI showed T2- and FLAIR hyperintensities affecting pons and middle cerebellar peduncles (Figure 1) whereas MRI spectroscopy was normal. EEG was normal. Lower limbs muscle MRI was normal. Deltoid muscle biopsy at 44 years old (Figure 2) showed the presence of RRF in GT stain (1–3 per section). SDH stain revealed 6 ragged blue fibers per section, and COX stain showed 6 COX negative fibers per section. In view of the age of the patient at the time of the biopsy these changes point towards a mitochondrial disorder. No abnormal accumulation of lipid droplets was seen outside the RRF. Activities of respiratory chain complex were normal. On electron microscopy, there was a slight increase of free glycogen and accumulations of mitochondria in the subsarcolemmal area. The mitochondria showed variable sizes, including small, normal, and large ones. In some fibers, certain mitochondria had an abnormal morphology with a very round aspect and crista disposed in a concentric or circular way, others had circular inclusions similar to myelin body type lesions.

Cerebral MRI of this man with homozygous XPNPEP3 variant.

Histopathologic and electron microscopy findings of the deltoid muscle biopsy done at 44 years of age.
A NGS panel analysis including 138 genes involved in mitochondrial disorders did not show any nuclear or mitochondrial DNA mutations. NGS panel for limb-girdle muscular dystrophies and leukodystrophies revealed no pathogenic variants. Moreover, presence of tremor, in addition to white matter lesions on MRI, prompted us to look for a Fragile X-associated tremor/ataxia syndrome (FXTAS), but no abnormal expansion was evidenced.
WGS allowed us to identify a homozygous frameshift variant NM_022098.4 (XPNPEP3): c.1153dup p.(Tyr385LeufsTer13) in the XPNPEP3 gene. This variant is located in exon 8/10 of the gene and has a very low allelic frequency (GnomADv4, 1/152130 alleles, Figure 3, Table 1). This duplication results in a reading frame shift with the creation of a premature STOP codon. It is not reported in the Leiden Open Variation Database (LOVD), Human Gene Mutation Database Professional (HGMD Pro) or Clinvar databases. Its MPA (MoBiDiC Prioritization Algorithm) score is 10/10.11,12 At the protein level, the variant is located in the peptidase/creatinase/aminopeptidase-like structural domain and considered to be of uncertain significance (ACMG class 3). Despite the fact that the patient reported no familiar consanguinity, it is worth noting that the variant is located within a 3 Mb region of homozygosity, which is indicative of consanguinity. No other candidate genes were identified by WGS. The sister of the patient is a heterozygous carrier of the variant and does not exhibit any symptoms suggestive of a neuromuscular disorder. DNA from the parents was not available for analysis.

XPNPEP3 pathogenic variants.
XPNPEP3 variants and phenotypes described in the literature.
Characteristics of XPNPEP3 variants and phenotypes described in the literature. The symbols * and # indicate variants present in a heterozygous combined state in a given patient.
Abbreviations: C.het: compound heterozygous; CK: creatine kinase; GT: Gomori trichrome; Hom: homozygous; HGMD Pro: Human Gene Mutation Database Professional; LOVD: Leiden Open Variation Database; NPHP: nephronophthisis; NPHPL: nephronophthisis-like nephropathy; NR: not reported; P: pathogenic; RRF: ragged-red fibers; y: years; 5'UTR: 5’ Untranslated Transcribed Region.
The disease evolution was marked by recurrent episodes of rhabdomyolysis with CK elevation up to 20,000 UI/L which resolved with Coenzyme Q10 (300 mg/day) and Levocarnitine (300 mg/day). At age 45 the patient developed a chronic kidney disease with a glomerular filtration rate reduced to 57 ml/min/1.73 m2 (N ≥ 60 ml/min/1.73 m2). Kidney ultrasound showed no morphological abnormalities and normal-sized kidneys. No kidney biopsy was performed. At last examination (age 56), there was no muscle weakness and the patient was fully ambulant but continued to experience exercise myalgia on prolonged walks.
Discussion
We report for the first time a case of mitochondrial myopathy presenting with a purely metabolic phenotype, manifesting as exertional symptoms and rhabdomyolysis associated with a homozygous XPNPEP3 variant.2,4,6,13
Muscular involvement associated with XPNPEP3 variants has been infrequently documented and inadequately characterized, and instances of rhabdomyolysis have not been reported. To our knowledge, only 6 patients from 3 families (from Finland, Turkey and Sweden) have been reported with neurological features and amongst them only two patients had neuromuscular involvement in a context of pathogenic homozygous XPNPEP3 variants (Table 1).2,5 All patients suffered from kidney disease with different severity. Regarding extra-renal involvement the Finnish family presented essential tremor and hearing loss, while in the Turkish family the two affected siblings presented with seizures, mental or developmental delay and dilated cardiomyopathy. Skeletal muscle involvement was only described in one patient with muscle fatigue and revealed complex I deficiency and decreased NADH-CoQ oxidoreductase activity. Ben-Shabat et al. reported a chronic kidney disease patient with various neurological issues including cerebellar ataxia, dysarthria, tremor, ptosis, axonal sensorimotor neuropathy, myopathy along with COX negative fibers, and RRF on muscle biopsy. 5
Interestingly, mitochondrial myopathies can mimic a metabolic myopathy with exercise intolerance and episodes of rhabdomyolysis, although it is rather an uncommon presentation.14–18 This has been described in cases of mitochondrial DNA mutations, as in mitochondrial oxidative phosphorylation (OXPHOS) deficiencies due to cytochrome b 14 or cytochrome c oxydase,15,16 but also with nuclear DNA mutations such as in ISCU, FDX2, TANGO2 genes,14,17,18 but has never been associated with XPNPEP3 mutations.2,5 In patients with biallelic XPNPEP3 variants, given the risk of rhabdomyolysis episodes that may secondarily lead to renal dysfunction, monitoring renal function and CK levels is mandatory.
Brain MRI T2 hyperintensities observed in our case in pons and middle cerebellar peduncles could resemble those found in POLG mutations, characterized by cerebellar atrophy and T2 hyperintensities located in dorsal thalami, cerebellar white matter and inferior olivary nucleus. 19 As previously described,2,5 our patient presented also a long-known action tremor in upper limbs which is probably part of the overall clinical phenotype. Indeed, tremor and white matter lesions on brain MRI prompted us to exclude FXTAS. In FXTAS, brain MRI typically shows white matter lesions in middle cerebellar peduncle and/or brainstem, cerebral hemispheres or the splenium of corpus callosum. 20 Our imaging findings deferred to the previous reported case, with c.766C > T XPNPEP3 variant, in whom severe and progressive cerebellar atrophy and predominant frontoparietal lobes atrophy was also observed. Although brain MRI showed significant abnormalities, our patient only exhibited an action tremor, highlighting the discrepancy between radiological findings and clinical presentation.
The XPNPEP3 variants identified in patients with neuromuscular features are exclusively truncated variants, either nonsense, small deletion, duplication nucleotide 5 or aberrant splicing leading to a frameshift. 2 Nonetheless, a truncated variant has been found in a patient with a pure renal phenotype. 4 On the other hand, missense mutations are only involved in pure renal phenotypes. 6 Regarding the variants’ localization, those associated with muscle involvement are located in the peptidase/creatinase/aminopeptidase structure domain in the C-part of the protein suggesting a loss of function effect, whereas in phenotypes with pure renal phenotype, they can affect the entire gene, or even its upstream part (Figure 3). A patient with a splice-site XPNPEP3 variant has been reported with a severe cardiomyopathy and neurological manifestations, including epilepsy and intellectual disability, highlighting extrarenal features of XPNPEP3-related disease. 21
Despite the limited number of pathogenic XPNPEP3 variants reported in the literature, accumulating evidence supports a spectrum of XPNPEP3-associated dysfunction, in which hypomorphic variants primarily cause nephronophthisis, while loss-of-function variants abolish mitochondrial aminopeptidase activity and give rise to frank mitochondrial pathology with neuromuscular and cardiac manifestations. These data suggest a correlation in XPNPEP3 diseases, where residual enzyme activity determines if the condition leads to a primarily renal or a multisystem mitochondrial disorder.2,6,21
Our results expand the clinical spectrum associated with XPNPEP3 variants, including a mitochondrial myopathy with a metabolic presentation, associated with mild renal involvement. Moreover, it highlights the utility of whole genome sequencing in cases where a mitochondrial disease is suspected and standard genetic analyses are inconclusive.
Footnotes
Ethical considerations
The study was conducted in accordance with European Union and French bioethics laws and with the Convention of Helsinki.
Consent to participate
The patient gave his informed consent to participate.
Consent for publication
The patient gave his consent for publication.
Author contributions
Katia Staedler (Conceptualization; Methodology; Investigation; Writing - Original Draft); Juliette Nectoux (Investigation; Writing - Review & Editing); Corinne Metay (Investigation; Writing - Review & Editing); Alban Lermine (Investigation); Rocio Nur Villar-Quiles (Investigation; Writing - Review & Editing); Teresinha Evangelista (Investigation; Writing - Review & Editing); Clemence Labasse (Investigation); Emmanuelle Lacène (Investigation); Tanya Stojkovic (Conceptualization; Methodology; Investigation; Writing - Review & Editing).
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability
The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
