Genotype-phenotype correlation in recessive DNAJB4 myopathy

Abstract

Protein aggregate myopathies can result from pathogenic variants in genes encoding protein chaperones. DNAJB4 is a cochaperone belonging to the heat shock protein-40 (HSP40) family and plays a vital role in cellular proteostasis. Recessive loss-of-function variants in DNAJB4 cause myopathy with early respiratory failure and spinal rigidity, presenting from infancy to adulthood. This study investigated the broader clinical and genetic spectrum of DNAJB4 myopathy. In this study, we performed whole-exome sequencing on seven patients with early respiratory failure of unknown genetic etiology. We identified five distinct pathogenic variants in DNAJB4 in five unrelated families of diverse ethnic backgrounds: three loss-of-function variants (c.547C > T, p.R183*; c.775C > T, p.R259*; an exon 2 deletion) and two missense variants (c.105G > C, p.K35N; c.181A > G, p.R61G). All patients were homozygous. All affected individuals exhibited early respiratory failure, and patients from three families had rigid spine syndrome with axial weakness in proportion to appendicular weakness. Additional symptoms included dysphagia, ankle contractures, scoliosis, neck stiffness, and cardiac dysfunction. Notably, J-domain missense variants were associated with a more severe phenotype, including an earlier age of onset and a higher mortality rate, suggesting a strong genotype-phenotype correlation. Consistent with a loss of function, the nonsense variants presented decreased stability. In contrast, the missense variants exhibited normal or increased stability but behaved as loss-of-function variants in yeast complementation and TDP-43 disaggregation assays. Our findings suggest that DNAJB4 is an emerging cause of myopathy with rigid spine syndrome of variable age of onset and severity. This diagnosis should be considered in individuals presenting with suggestive symptoms, particularly if they exhibit neck stiffness during infancy or experience respiratory failure in adults without significant limb muscle weakness. Missense variants in the J-domain may predict a more severe phenotype.

Keywords: Chaperonopathy; DNAJB4; Heat shock proteins; Protein aggregate myopathy; Respiratory failure; Rigid spine syndrome.

Figure 1

Pedigrees of patients with DNAJB4 variants. (a) Pedigrees of five families demonstrating consanguinity and recessively inherited myopathy. The genotype of the patient(s) is represented with the pedigree. Below each symbol, the genotype is annotated: “del” indicates a deletion, and “ref” denotes the reference sequence. (b) X-ray image of P4 at 15 years of age, showing thoracic lumbar scoliosis. (c) Muscle magnetic resonance images of P3 at 14 years of age. Axial thigh muscle images of T1-weighted image (T1WI, upper image) and T2-weighted fat-saturated image (T2FS, lower image). High intensity in the semimembranosus muscle (arrows) is observed in both T1WI and T2FS, whereas the adductor longus shows high signal intensity only in T2FS (arrow).

Figure 2

Myopathological findings from patients with DNAJB4 variants. (a) H&E staining from P1 at 4 years of age, demonstrating marked variation in fiber size, prominent endomysial fibrosis, and adipose tissue infiltration. Scale bar: 100 μm. (b-d) Quadriceps biopsy from P2 at 5 years of age. (b) H&E staining showing moderate variation in fiber size and a few small angular fibers. Scale bar: 50 μm. (c) NADH-tetrazolium reductase (NADH-TR) staining highlighting small rubbed out areas (arrows). Scale bar: 50 μm. (d) ATPase staining at pH 9.4 demonstrating mild type 2 fiber atrophy. Scale bar: 50 μm.

Figure 3

Schematic representation of DNAJB4 domains and variant locations. The schematic illustration of DNAJB4 shows the J-domain, glycine-phenylalanine rich (G/F) domain, and C-terminal domain. The locations of identified pathogenic variants are marked in black, whereas previously reported variants are indicated in blue.

Figure 4

DNAJB4 variants have a loss-of-function effect. (a) Tetracycline-regulated isogenic 293 cell lines were developed to express V5-tagged DNAJB4 (wild type, R183*, K35N, R259*, R61G). After induction with tetracycline for 24 hours, the cells were harvested at various times postinduction. Protein lysates were analyzed by immunoblotting for V5 and GAPDH, with GAPDH used as a loading control. (b) Densitometric analysis of V5-DNAJB4 normalized to GAPDH from three independent experiments, with the initial day’s value as a reference. Statistical significance was determined by comparing each variant to the WT on each day via a t test with Bonferroni correction for multiple comparisons (*P ≤ 0.0125, **P < 0.0025). (c) Representative fluorescence microscopy images displaying mCherry-tagged TDP-43 in HeLa cells post heat shock. (d) Quantification of cells displaying TDP-43 nuclear inclusions. n = 200–300cells analyzed per condition; the experiment was conducted 3 times. Statistical analysis was conducted by comparing the percentage of cells with TDP-43 aggregates after 1 hour at 42°C to that of WT cells via a t test with Bonferroni correction for multiple comparisons (**P < 0.002, ***P < 0.0002). (e) Yeast colonies lacking Sis1 were supplemented with either an empty vector (EV), wild-type DNAJB1 (DNAJB1-WT), DNAJB1-K35N, Sis1-WT, or the Sis1-K37N variant and then spotted on FOA media (left panel). The corresponding colonies on full media (YPD) are shown in the right panel. Data from five independent experiments were collected for each condition. EV, empty vector; WT, wild type.