Pathogenic variants in NUBPL result in failure to assemble the matrix arm of complex I and cause a complex leukoencephalopathy with thalamic involvement

Abstract

Disorders of the white matter are genetically very heterogeneous including several genes involved in mitochondrial bioenergetics. Diagnosis of the underlying cause is aided by pattern recognition on neuroimaging and by next-generation sequencing. Recently, genetic changes in the complex I assembly factor NUBPL have been characterized by a consistent recognizable pattern of leukoencephalopathy affecting deep white matter including the corpus callosum and cerebellum.

Here, we report twin boys with biallelic variants in NUBPL, an unreported c.351 G>A; p.(Met117Ile) and a previously reported pathological variant c. 693+1 G>A. Brain magnetic resonance imaging showed abnormal T2 hyperintense signal involving the periventricular white matter, external capsule, corpus callosum, and, prominently, the bilateral thalami. The neuroimaging pattern evolved over 18 months with marked diffuse white matter signal abnormality, volume loss, and new areas of signal abnormality in the cerebellar folia and vermis. Magnetic resonance spectroscopy showed elevated lactate.

Functional studies in cultured fibroblasts confirmed pathogenicity of the genetic variants. Complex I activity of the respiratory chain was deficient spectrophotometrically and on blue native gel with in-gel activity staining. There was absent assembly and loss of proteins of the matrix arm of complex I when traced with an antibody to NDUFS2, and incomplete assembly of the membrane arm when traced with an NDUFB6 antibody. There was decreased NUBPL protein on Western blot in patient fibroblasts compared to controls.

Compromised NUBPL activity impairs assembly of the matrix arm of complex I and produces a severe, rapidly-progressive leukoencephalopathy with thalamic involvement on MRI, further expanding the neuroimaging phenotype.

1. Introduction:

Primary mitochondrial disease is highly heterogeneous with multiple genetic causes including pathogenic variants in mitochondrial or nuclear DNA and diverse clinical presentations. Neurological symptoms are the most common presentation in mitochondrial diseases. They most often involve the basal ganglia, thalamus, and brain stem nuclei as in Leigh disease. A less common presentation of mitochondrial disease is a leukoencephalopathy, which has been described in a number of mitochondrial conditions, most commonly related to deficient activity of complex I or lipoate disorders [1–3].

Leukodystrophies and leukoencephalopathies are a complex group of white matter disorders some of which have associated grey matter involvement, and are due to a large number of genetic etiologies [4,5]. Diagnostic recognition in the leukodystrophies and leukoencephalopathies has been greatly aided by recognizing particular patterns of brain injury including the affected regions on brain imaging such as brain magnetic resonance imaging (MRI) [6,7]. For example, a specific neuroimaging pattern has recently been described for patients in association with pathogenic variants in the gene NUBPL, a gene involved in mitochondrial complex I assembly [8]. These patients were recognized by an early MRI pattern of abnormal T2 signal hyperintensity predominantly affecting the frontoparietal deep white matter with sparing of the U-fibers, internal and external capsules, the central corona radiata, and the genu and splenium of the corpus callosum with relative sparing of the isthmus. On late follow-up MRI, patients demonstrated progressively abnormal T2 signal hyperintensity in the cerebellar cortex and cerebellar subcortical white matter with sparing of the deep central white matter and dentate nucleus, and abnormal signal in the pons and medulla. Although there was progressive cerebellar involvement, there is some improvement in the cerebral white matter signal abnormality [8,9]. Early MRI showed areas of associated restricted diffusion indicating cytotoxic edema [8]. The basal ganglia, thalamus, and cortex were not involved. This neuroimaging pattern appeared highly specific and recognizable, and its identification should prompt specific molecular investigations of NUBPL [8]. However, in addition to this highly specific pattern, other neuroimaging phenotypes have recently been described in patients with NUBPL pathogenic variants. Bilateral putaminal T2 hyperintensity with putaminal and cerebellar atrophy was described in a family with dystonia due to NUBPL pathogenic variants [10].

Patients with NUBPL pathogenic variants have decreased complex I activity likely due to a defect in the assembly of complex I [8, 18, 19]. Complex I is a large multiprotein complex composed of 44 subunits, which are integrated in a membrane arm and a matrix arm. The matrix arm contains eight iron-sulfur clusters integrated within five subunits (NDUFS1, NDUFS7, NDUFS8, NDUFV1, and NDUFV2). The assembly of this large protein complex occurs in a multistep process with various intermediates recognized [11–14]. Various assembly proteins assist with this assembly process, including the NUBPL protein (originally called Ind1 protein) encoded by the NUBPL gene [15,16]. In the absence of NUBPL protein, complex I formation is incomplete with a 450 kDa subcomplex present. Because NUBPL protein contains a CxxC iron-sulfur cluster binding motif and a [4Fe-4S] cluster, it was suggested to be involved in the assembly of the iron-sulfur clusters on the matrix arm of complex I [15,16]. A role for NUBPL in mitochondrial protein translation has also been proposed in the plant Arabidopsis thaliana [17]. NUBPL is expressed in mitochondria with the highest levels present in liver, kidney, small intestine and brain [16]. In patients with pathogenic variants in NUBPL, complex I activity is decreased [8, 18, 19], with reduced fully assembled complex I with absence of assembly intermediates on native gel when probed with an NDUFS3 antibody [8].

We describe identical twins with pathogenic variants in NUBPL who presented with leukoencephalopathy and a novel neuroimaging phenotype distinct from previously described cases [8]. The diagnosis was functionally confirmed and delineated in fibroblasts.

2. Methods

Patients provided informed consent on an IRB approved study (COMIRB# 16–0146). Brain MRI was obtained as part of routine clinical care at age 4 months and 22 months and reviewed by a pediatric neuroradiologist (FP).

Mitochondrial respiratory chain enzyme assays were assayed in post-600 g homogenates spectrophotometrically as described [20,21]. The results were normalized for protein, and ratios over the activity of citrate synthase and complex II were calculated and compared to the normal range derived from control samples, which after log transformation are normally distributed. The results of the patient were then described as Z-scores of this distribution. Separation of mitochondrial complexes from solubilized inner mitochondrial membrane on blue native polyacrylamide gel electrophoresis (PAGE) with in-gel activity staining was performed as previously described [20,21].

The assembly of complex I was evaluated by separation of the solubilized inner mitochondrial membrane fraction on a blue native gel and after blotting probed with an antibody against NDUFS2, a subunit that is incorporated early during complex I assembly, as previously described [22]. In addition, the assembly of complex I was also evaluated using clear native gel electrophoresis as described [16] and probed with an antibody against NDUFB6, which is part of an early subcomplex of the membrane component [12,13]. We evaluated by Western blot the amount of protein of subunits from each respiratory chain complex (total OXPHOS human antibodies) and for complex I from a protein pertinent to the membrane arm (NDUFB8) or to the matrix arm (NDUFS3).

The amount of NUBPL protein was evaluated by SDS-PAGE followed by western blot and detected with an antibody against NUBPL protein in isolated mitochondrial fractions derived from both patient and control fibroblasts obtained through differential centrifugation. The sources and dilution of each antibody are listed in Supplementary Table 1.

High resolution respirometry was done on an Oroboros Oxygraph 2k system following a substrate inhibitor (SUIT) protocol as described comparing six replicates of the patient cells with 41 control fibroblast lines each done at least in duplicate [22]. Populations were described as mean and standard deviation, and comparisons between these two populations were done by Student t-test.

Mitochondrial translation was assessed by incubation of fibroblasts in culture medium with ³⁵S-methionine and ³⁵S-cysteine in the presence of the cytosolic ribosome inhibitor emetine, and after harvesting the protein products are separated on an SDS-PAGE followed by autoradiography as described [23].

3. Case Report

Identical twin brothers were delivered at 38 weeks gestational age by C-section to unrelated parents of primarily northern European descent. Father has psoriasis and maternal grandmother has migraine headaches, but no family history of neuromuscular disorders. Mother noted vigorous in utero movements similar to their older healthy brothers (pedigree provided as Supplementary Figure 1). Twin B had mild hyperbilirubinemia requiring hydration without phototherapy, and both boys were discharged to home after several days and had normal initial development.

At 4 months of age both boys developed multiple episodes of projectile vomiting, metabolic acidosis, and elevated levels of lactate (both > 7 mmol/L, normal < 2.1), plasma alanine > 950 μmol/L (normal < 495). Urine organic acids showed increased excretion of lactate, pyruvate and the citric acid cycle intermediate fumarate (in both boys >212 mg/gm creatinine; normal < 40); a pattern suggestive of an underlying oxidative phosphorylation disorder (Supplementary Table 2).

The growth rate of length and weight of both twins was similar, and significantly decreased compared to healthy children. At nine months of age, weights were less than the second percentile (Z scores −2.9) and at 22 months of age were at the 12th to 15th percentile. Length was more affected and over the same period, remained below the second percentile with Z-scores −4.5 at 9 months and −3 at 22 months. Head circumference was dissimilar, with twin A at the 12th percentile at 9 months and decreased to the 2.6 percentile at 22 months, whereas twin B remained at the 50th percentile.

Twin A had a left eye gaze preference with occasional random nystagmoid movements. Pupils were sluggishly reactive to light, but there was no blink to bright light or confrontation. Optic nerves were pale bilaterally; all other cranial nerves were grossly normal. There was a mild axial hypotonia but appendicular structures were hypertonic, lower extremities more involved than upper extremities. Over the 22-month period, increased spasticity in the extremities became more apparent. Patellar cross adduction was present without clonus at the ankles. Muscle stretch reflexes were increased in the upper and lower extremities. Twin B had a similar clinical exam over the course of the 22 months except for persistent leftward eye movements at age 22 months and constant mixed nystagmoid movements.

Both boys had electroencephalograms (EEG) at 6 months of age due to Twin A having what was thought to be a clinical seizure. Both EEG studies were similar demonstrating multifocal spikes over the left and right hemisphere, and background slowing for age. This EEG finding was compatible with the most common EEG abnormalities seen in early onset epilepsy in patient with mitochondrial disease [24,25]. Based on the EEG findings, both boys started levetiracetam, and were subsequently placed on a ketogenic diet which, after several months, was discontinued due to persistent vomiting. Soon thereafter, levetiracetam was also stopped as no further clinical events had occurred.

Twins A and B had similar findings on two serial brain MRI and magnetic resonance spectroscopy (MRS) scans obtained at 4 months and 22 months of age (Figure 1). At 4 months of age, MRI images showed abnormal T2 signal hyperintensity diffusely involving the periventricular white matter, external capsule, corpus callosum, and thalami bilaterally which were enlarged and edematous, swollen. There were areas of associated restricted diffusion. Eighteen months later, there was evolution of diffuse supratentorial high T2-signal involving the cortex, white matter, and thalami with extensive associated volume loss and ex vacuo dilation of the lateral and third ventricles. New multifocal T2-signal hyperintensity was noted in the vermis. There was relative sparing of the basal ganglia, medial occipital lobes, hippocampus and brainstem. Brain MRS in the centrum semiovale showed a large lactate peak and reduced N-acetyl-aspartate (NAA) peak, consistent with mitochondrial dysfunction with disruption of ATP production and neuronal injury (Supplementary Figure 2).

Figure 1: Brain magnetic resonance imaging of patients with NUBPL pathogenic variants.

Legend: Brain MRI of Twin A (A-D) and Twin B (E-H) at age 4 months (A, B, E, and F) and at age 22 months (C, D, G, and H). At 4 months of age, T2 images (A and E) show similar diffuse hyperintense signal in the white matter (black asterisks) and enlarged thalami (white arrowheads) in both twins. Diffusion-weighted imaging (B and F) demonstrates hypointense signal on ADC maps in the thalami (white arrowheads) indicating cytotoxic edema. Abnormal cortical T2 hyperintense signal (A) and restriction diffusion (B) is present at 4 months in Twin A (white arrows). At 22 months of age, T2 images (C, D, G, and H) show progressive T2 hyperintense signal involving the supratentorial white matter (black asterisks) and cortex (white arrows) with new T2 signal hyperintensity in the cerebellar vermis (dashed white arrows in D and H). There is interval thalamic atrophy (white arrowheads in C and G) with relative sparing of the basal ganglia (black arrows in C and G), hippocampus (black arrowheads in D and H), and occipital lobes (black dashed arrows in D and H).

Clinically obtained whole exome sequencing identified two variants in the NUBPL gene, transcript NM_025152.2, paternal c.693+1G>A and maternal c.351G>A encoding for p.(Met117Ile). The variant c.693+1G>A is predicted to change splicing at intron 8 and has been previously reported [8]. The amino acid Met117 has been completely conserved in evolution (Supplementary Table 3), and prediction programs Mutation Taster, Polyphen-2, and SIFT predict the change p.(Met117Ile) to be disease-causing, probably damaging (score 1.00), and deleterious respectively, with a Grantham score of 10. The variant is not present in GnomAD or 1000 Genomes. A skin biopsy was obtained for fibroblast culture at age 10 months of twin A.

4. Results

A western blot with an antibody against NUBPL showed a double band in control fibroblasts of the post 600 g supernatant as described (Figure 2B) [8], but only a single band in isolated mitochondria (Figure 2C). Both signals were undetectable in the patient’s fibroblasts.

Figure 2: Mitochondrial functional testing on blue native gel with in-gel activity staining and the amount of NUBPL protein.

Legend: (A) The activities of respiratory chain enzyme complexes I, II, IV, and V are visualized by in-gel activity staining after separation of the complexes on blue native polyacrylamide gel electrophoresis from an isolated inner mitochondrial membrane preparation of the fibroblasts from the patient and a control. The assay shows decreased staining of complex I with normal activities of complexes II, IV and V. NUBPL protein: The NUBPL protein is identified on western blot in a post-600 g supernatant (B), and a mitochondrial preparation obtained by differential centrifugation (C). In controls, the NUBPL protein is present in two sizes as a double band in the whole cell preparation (B), but as a single band in the mitochondrial preparation (C), whereas both bands are absent in the patient. Abbreviation: FB = fibroblasts

Respiratory chain enzyme activities in fibroblasts showed strongly reduced activity of complex I without involvement of other complexes (Table 1). On blue native PAGE with in-gel activity staining, the activities of complexes II, IV, and V were normal but the activity of complex I shows strongly reduced staining indicating decreased activity (Figure 2A).

Table 1:

Respiratory chain enzyme activities in fibroblasts of Twin A

	Controls	Patient		Controls	Patient		Controls	Patient
	Activity	Activity	SD	Ratio/CS	Ratio/CS	SD	Ratio/Co II	Ratio/Co II	SD
Complex I	48.89–120.14	21.6	−4.6	151–387	88	−4.1	255–539	134	−3.6
Complex II	144.4–349.2	161.6	−1.0	471–919	657	0.3	NA	NA	NA
Complex III	7.92–30.38	22.1	1.1	21–79	90	1.8	37–118	137	1.7
Complex II-III	102.1–210.6	136.4	0.9	265–571	554	2.3	437–812	844	1.0
Complex IV	2.13–6.84	5.9	0.8	6–25	24	1.3	12–37	37	1.2
Citrate synthase	245.4–540.4	246.1	−1.6	NA	NA	NA	NA	NA	NA

Legend: The activities in fibroblasts of each complex and of combined complex II-III are shown expressed as nmol/min.mg protein and as a ratio over citrate synthase activity (ratio/CS) and as a ratio over the activity of complex II. The values of 25 control fibroblasts are given as a reference normal range. The log transformed values in 25 control fibroblasts are normally distributed, and the patient’s values are expressed as a Z-score in standard deviations (SD) on the distribution in normal controls for comparison to the patient’s results. Activities greater than +2.5 SD or smaller than −2.5 SD are considered abnormal and indicated in bold. Abbreviations: CS = citrate synthase, Co II = complex II, SD = standard deviations.

The assembly of complex I was first traced with an antibody against NDUFS2, which is part of the earliest assembly intermediate of the Q module [12,13] (Figure 3A). This showed markedly reduced amounts of holocomplex I at 1 MDa with very small amounts of subcomplexes around 400 and 460 kDa visible. The assembly of complex I was next traced with an antibody against NDUFB6, which is part of an early subcomplex of the membrane component P₀-a, the ND5 module [12,13] (Figure 3C). This showed again markedly reduced amounts of the holocomplex at 1 MDa, but the same amount of the 460 kDa membrane component as the control fibroblast, of the same size as that noted in HepG2 cells exposed to chloramphenicol. The amount of NDUFS3 protein, a subunit of the matrix arm is strongly reduced (Figure 3B). The amount of NDUFB8 protein, a component of the membrane arm, is reduced but still to a small degree present (Figure 3D). The amount of proteins from other mitochondrial complexes was normal (Figure 3E).

Figure 3: The assembly of complex I is affected in the patient.

Legend: The assembly of complex I on an inner mitochondrial membrane preparation is identified on a blue native gel using an antibody against NDUFS2, a protein of the matrix arm (A) and against NDUFB6 a protein of the membrane arm (C). This shows a reduction in the holocomplex without an intermediate accumulating detectable with the antibody against NDUFS2, a component of the matrix arm (A), but with an intermediate detected around 480 kD with an antibody against NDUFB6, a component of the membrane arm. Chloramphenicol treated HepG2 cells are shown as a reference for the location of abnormal assembly intermediates. The amount of the NDUFS3 subunit, which is present in the matrix arm, is strongly reduced as shown on SDS-PAGE followed by western blot (B), whereas the amount of NDUFB8 protein present in the membrane arm is strongly reduced but not absent (D). Citrate synthase is shown as a loading control. The amount of subunits of other respiratory chain complexes is not different from controls (E). Abbreviation: FB = fibroblasts, CAM = chloramphenicol treated.

High resolution respirometry showed a low oxygen consumption rate using the substrates pyruvate and glutamate with ADP stimulation around the 25th percentile of controls (p<0.001 for both), with normal oxygen consumption rate with succinate following ADP-stimulation (p=0.61), but reduced maximum rate after uncoupling around the 5th percentile for controls (p=0.016) (Figure 4). The ratio of the rate of succinate stimulated over that of glutamate+pyruvate (S/G) was above the 95th percentile of the range in controls (p=0.002) and the rotenone sensitive rate (rate after rotenone inhibition subtracted from the rate after uncoupling) was below the fifth percentile (p<0.001). The azide-sensitive complex IV rate did not significantly differ from controls.

Figure 4: High resolution respirometry.

Legend: The oxygen consumption rate is measured in high resolution respirometry using a substrate-inhibitor protocol, first identifying the ADP-stimulated rate with sequentially added complex I substrates pyruvate (patient 15.0±2.8 pmol/min.10⁶ cells n=6 vs. controls 30.5±14.0 n=41, p<0.001) (blue) and glutamate (patient 16.0±2.9 vs controls 31.5±14.6, p<0.001) (orange), then is added the complex II substrate succinate (patient 46.6±15.6 vs. control 42.9±14.6, p=0.61) (grey), before uncoupling with carbonylcyanide-p-trifluoromethoxy-phenylhydrazone (FCCP) (patient 56.5±18.9 vs control 83.0±23.7, p=0.016) (yellow). The complex I specific activity is reflected in the difference between the uncoupled rate and the rate after complex I inhibition with rotenone (patient 23.9±11.5 vs controls 21.4±15.2, p<0.001) (green). The complex IV activity rate is estimated from the stimulation with N,N,N’,N’-tetramethyl-p-phenylenediamine dihydrochloride (TMPD) and ascorbate after subtraction of the azide inhibited rate (patient 91.0±58.9 vs controls 70.6±28.3, p=0.44) (dark blue). The increase in activity with the complex II substrate over that of the complex I substrates (ratio of the S/G rate, light blue) was greater than the 95th percentile in controls (patient 2.9±0.6 vs controls 1.6±0.6, p=0.002), whereas the rotenone sensitive rate (green) was just above the 5th percentile of controls.

There was no detected difference in mitochondrial translation of proteins in fibroblasts cultured from our patient compared to controls (Supplementary Figure 3).

5. Discussion

A highly specific form of leukoencephalopathy was described on neuroimaging caused by NUBPL pathogenic variants consisting of white matter lesions in the central white matter including the corpus callosum, with relative sparing of the central part [8]. Most notable, the supratentorial cortex and the deep gray nuclei were not involved. This progressed to include the cerebellar cortex, subcortical cerebellar white matter, and brain stem. On follow-up, the corpus callosum and cerebral white matter abnormalities improved, while cerebellar abnormalities worsened. Whereas this pattern has specificity for NUBPL deficiency and has been noted in several patients as described [8], it does not describe the entire neuroimaging phenotype of NUBPL disorder. Here we note a similar leukodystrophy and beginning cerebellar involvement, but with striking involvement of the thalami. In addition, cerebellar involvement with either bilateral striatal necrosis or with pontine involvement has previously been described in the absence of cerebral leukodystrophy [9,10]. Thus, involvement of the central nuclei, including basal ganglia and thalami, and of the brain stem nuclei, typical of Leigh-like lesions in mitochondrial disorders can also be present in NUBPL disorder. In all cases, clinically the phenotype is dominated by pyramidal signs and cerebellar signs of nystagmoid eye movements and ataxia [8–10], and in some cases, dystonia can be present [10]. Biochemically, lactic acid is variably elevated in blood, as in the twin patients reported here, but has been consistently elevated in CSF in the few published patients where it was evaluated [8], and on MRS [9].

Our patients had a previously reported splice site pathogenic variant, and a new missense variant. In addition to the bioinformatic predictions of pathogenicity, the fibroblasts had severely decreased steady state levels of NUBPL protein. Similar to previous studies, we found two bands for the NUBPL protein, one of higher molecular weight localized in mitochondria, and of lower molecular weight which had been attributed before to be likely cytosolic in location [16]. The fact that both bands were severely decreased in the affected patient indicate that both fractions are NUBPL protein, and not a cross-reacting band as had been previously suggested [16]. A complementation study was not performed as the pathogenicity of NUBPL mutants has been well documented before [8,16, 18,19]. A list of currently identified pathogenic variants in patients published with NUBPL disorder is listed in Supplementary table 4. NUBPL has a mitochondrial leader peptide comprising of amino acids 1 to 38, and a single iron transfer P-loop NTPase domain of amino acids 65 to 311.

The NUBPL protein was first identified in a search of proteins with motifs for handling iron-sulfur cluster proteins. It associates with the developing N module [17]. The assembly of the matrix arm of complex I is severely impaired with substantial decrease in multiple matrix arm proteins including NDUFS2 and NDUFS3, which are typically used to trace assembly intermediates for complex I. This is shown here and as previously published in a patient [8] and in HEK293 cells treated with an siRNA [16]. The membrane arm to some extent can still develop, as was first noted with RNAi depletion in HEK293 cells [16], and now also in this patient’s fibroblasts. With decreased development of the membrane arm there is reduced assembled holocomplex amounts resulting in deficient complex I enzyme activity, as was also shown in previously published patients [8,9,18]. This complex I deficiency results in decreased ATP production from complex I substrates [8]. In our patient we show decreased oxygen consumption when only complex I substrates pyruvate and glutamate are provided, but a strong increase when a complex II substrate succinate is provided, resulting in increased succinate/glutamate ratio.

The NUBPL protein has a CxxC motif typical for iron-sulfur cluster containing proteins, and purified protein has an absorbance spectrum indicative of an iron-sulfur cluster containing protein, and takes up radiolabeled iron, with binding of a [4Fe-4S] cluster, with the amino acids Cys242 and Cys245 essential in its binding [16]. Based on the sequence homology to yeast proteins involved in iron-sulfur cluster transfer (Nbp35 and Cfd1), the evidence for the presence of an iron-sulfur cluster in NUBPL, and the association with the N-module, it has been assumed that NUBPL protein is involved in the transfer of iron-sulfur clusters to the apoprotein of the iron-sulfur cluster containing subunits of complex I, namely NDUFS1, NDUFS7, NDUFS8, NDUFV1, and NDUFV2. It is assumed that a decrease in the transfer of iron-sulfur clusters to the apoproteins in the N-module of complex I will result in unstable proteins and subcomplex. Direct evidence included decreased labeling of complex I with ⁵³Fe, and a direct effect on complex I assembly by variants in the cysteine residues involved in Fe-S cluster binding [16]. However, an alternative hypothesis involving a role for NUBPL protein in mitochondrial protein translation of complex I subunits was identified in Arabidopsis thaliana[17]. Similar to the Arabidopsis results, the amount of the existing 460 kDa membrane subcomplex in affected fibroblasts is small, but in contrast to Arabidopsis where absence of NUBPL interfered with mitochondrial translation, such effect could not be discerned in human fibroblasts where mitochondrial translation appeared normal.

In this study, we broaden the neuroimaging phenotype of NUBPL-related leukoencephalopathy to include deep grey nuclei, particularly the thalamus. NUBPL pathogenic variants impair the assembly of the matrix arm of complex I resulting in a severe deficiency of fully assembled I and its enzymatic activity and function. The study shows that fibroblasts can be used effectively to provide functional evidence for pathogenicity in NUBPL without the need for more invasive tissues such as muscle biopsies.