Comprehensive phenotypic characterization of an allelic series of zebrafish models of NEB-related nemaline myopathy

Abstract

Nemaline myopathy (NM) is a rare congenital neuromuscular disorder characterized by muscle weakness and hypotonia, slow gross motor development, and decreased respiratory function. Mutations in at least twelve genes, all of each encode proteins that are either components of the muscle thin filament or regulate its length and stability, have been associated with NM. Mutations in Nebulin (NEB), a giant filamentous protein localized in the sarcomere, account for more than 50% of NM cases. At present, there remains a lack of understanding of whether NEB genotype influences nebulin function and NM-patient phenotypes. In addition, there is a lack of therapeutically tractable models that can enable drug discovery and address the current unmet treatment needs of patients. To begin to address these gaps, here we have characterized five new zebrafish models of NEB-related NM. These mutants recapitulate most aspects of NEB-based NM, showing drastically reduced survival, defective muscle structure, reduced contraction force, shorter thin filaments, presence of electron-dense structures in myofibers, and thickening of the Z-disks. This study represents the first extensive investigation of an allelic series of nebulin mutants, and thus provides an initial examination in pre-clinical models of potential genotype-phenotype correlations in human NEB patients. It also represents the first utilization of a set of comprehensive outcome measures in zebrafish, including correlation between molecular analyses, structural and biophysical investigations, and phenotypic outcomes. Therefore, it provides a rich source of data for future studies exploring the NM pathomechanisms, and an ideal springboard for therapy identification and development for NEB-related NM.

Keywords: alternative splicing; muscle; nebulin; nemaline myopathy; zebrafish.

Figure 1

Zebrafish nebulin structure and nebulin mutations. (A) Diagram of zebrafish nebulin with position of mutations. N-terminus of nebulin consists of a glutamic acid rich sequence followed by several distinctive domains that mediate interactions with tropomodulin. Central region of nebulin has a repetitive, modular structure, and, in zebrafish, is organized into 21 super-repeats (SR). The C-terminus of nebulin consists of several linker modules, a serine-rich and an SH3 domains, which mediate interactions with proteins in the Z-disk. (B) Diagram of two SR. Each SR is made up of seven simple repeats (R1-R7) and contains a conserved WLKGIGW motif (troponin/tropomyosin binding site) in R3 (star). Each simple repeat contains an actin-binding motif (SDXXYK) (dots). (C) Percent spliced in index (PSI) showing alternative splicing in nebulin. Skipping of exons 33–37 was observed only in neb mutants, whereas splicing of exons 49–73 and 107 was observed in both wt and mutant zebrafish. Several splicing events were observed at the C-terminus of nebulin. (D–K) Splicing analysis and variants identified by RNAseq in zebrafish splice mutants. Four wild-type and four mutant replicates (rep1-rep4) have been analyzed and the PSI has been calculated for each. Light and darkblue arrowheads indicate the location of the alternative splice site/s used. Exons are represented by grey rectangles, introns by thick black lines, and intron inclusion by black rectangles. (D and E) Diagram (D) and table (E) illustrating alternative splicing events in neb^hu28 mutant. One splice variant was identified in neb^hu28 (alt 1), which was generated by the usage of an alternative splice donor site located in intron 46. This caused a partial intron retention, a frameshift and the formation of a stop codon in exon 47. (F and G) Diagram (F) and table (G) illustrating two alternative splicing events in neb³⁴ mutant (alt 1 and alt 2). In the alt 1 variant, an alternative splice donor site in exon 54 is used, which resulted in skipping the last 18 base pairs from exon 54, and generation of an in-frame transcript. In the alt 2 variant, the full intron was retained. (H and I) Diagram (H) and table (I) illustrating three alternative splicing events in neb²¹ mutant (alt 1, alt 2, and alt 3). In the alt 1 and 2 variants, usage of two different alternative splice acceptor sites located in intron 105 resulted in addition of 34 base pairs and out-of-frame transcript (alt 1, light blue arrowhead), or in addition of 3 base pairs and in-frame transcript (alt 2, dark blue arrowhead). In the alt 3 variant, the full intron was retained. (J and K) Diagram (J) and table (K) illustrating two alternative splicing events in neb³⁰ mutant (alt 1 and alt 2). A new splice acceptor site was generated by the mutation, located just one base pair downstream of the original site (alt 1). This caused a frameshift and the formation of a premature stop in exon 135, which is the last exon in zebrafish nebulin. In the alt 2 variant, the full intron was retained.

Figure 2

Nebulin expression is reduced in neb mutants. (A) Nebulin transcript levels, determined from the RNAseq data, were drastically reduced in the nonsense and neb^hu28 mutants, whereas the levels in the splice mutants were consistent with those in the wt siblings (numbers indicate the padj values). (B) SDS-agarose gel stained with Coomassie blue showing reduced levels of nebulin in all mutants. (C) Confocal micrographs of myofibers stained with antibodies against Neb-N and α-Actinin. The nonsense, neb²¹, and neb^hu28 mutants show altered localization and drastically reduced levels of nebulin. The localization pattern of nebulin in neb³⁴ and neb³⁰ splice mutants is similar to that in the wt myofibers; however, nebulin levels are significantly reduced. To better illustrate the altered localization of nebulin, the brightness in the region marked by the square in left column panels was increased to 150% in panels showing high magnification detail of Neb-N staining. Scale bars = 10 μm. (D) Quantification of fluorescence intensity of anti-Neb-N antibody staining in myofibers (mean ± SEM). Violin plot showing that fluorescence intensity is significantly reduced in neb³⁴ and neb³⁰ mutants compared to wt (56.5% in neb³⁴ and 92.7% in neb³⁰, respectively). n = 56 measurements/genotype. Asterisks indicate p-values (** ≤ 0.01; **** ≤ 0.0001). See Supplemental Table 1 for fluorescence intensity value measurements and statistics.

Figure 3

Morphological defects and early mortality in neb mutants. (A) Brightfield images of wt and neb mutant embryos at 6 dpf. All the neb mutants, except neb³⁰, fail to inflate their swim bladder by 6 dpf, show defects in the lower jaw, and some mutants have bent body. (B) Polarized stereomicroscope images showing birefringence of skeletal muscle in zebrafish embryos at 6 dpf. All the neb mutants, except neb³⁰, have reduced birefringence, indicating that muscle integrity and structure are compromised. (C) Quantification of birefringence intensity (mean ± SEM). All the neb mutants, except neb³⁰, showed significantly reduced birefringence. Asterisks indicate p-values (* ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001). (n = 8 for wt; n = 5 for each mutant). See Supplemental Table 2 for intensity value measurements and statistics. (D) Survival curve indicating drastically shorter life span in all the neb mutants, except neb³⁰. (n = 20 for wt, neb¹⁵, neb¹¹, neb²¹; n = 10 for neb³⁴; n = 21 for neb³⁰). Median survival: 11 days (neb¹⁵, neb¹¹), 12 days (neb³⁴), and 13.5 days (neb²¹).

Figure 4

Swimming behaviour is defective in neb mutants. Muscle performance and function in neb mutants was assessed by tracking optovin-induced swimming of embryos at 7 dpf. Graphs showing time spent moving, distance travelled, and speed for each zebrafish line. The nonsense mutants neb¹⁵ and neb¹¹ showed significantly shorter time swimming, distance travelled, and lower speed compared to their siblings. The splice mutants neb³⁴ and neb²¹, with mutations in SR region, also showed significantly shorter time spent swimming. The motor behaviour of the splice mutant neb³⁰ is not significantly different from that recorded for its siblings. Graphs display values normalized to wt (mean ± SEM). Asterisks indicate p-values (* ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001). See Supplemental Table 3 for descriptive statistics.

Figure 5

Muscle and sarcomere organization is defective in neb mutants. (A) Confocal micrographs of wholemount preparations of embryos at 6 dpf stained with antibody against α-Actinin and Rhodamine Phalloidin. Nonsense mutants and the splice mutant neb²¹ show disorganized, misaligned or broken muscle fibers (arrows), defects in the MTJ regions, with partially detached fibers and accumulations of aggregates of actin filaments (arrowheads), and aggregates of actin/α-Actinin (open arrowheads). Scale bar = 20 μm. (B) Transmission electron micrographs of zebrafish embryos at 6 dpf showing defects in the MTJ regions (dotted line), with disorganized muscle fibers (arrows) and misaligned sarcomeres (arrowheads) in the nonsense, neb³⁴, and neb²¹ splice mutants. Scale bars = 2 μm.

Figure 6

Details of ultrastructural defects in neb mutants. (A–U) Transmission electron micrographs of zebrafish embryos at 6 dpf. (A) Aligned sarcomeres in wt muscle. (B–E) Increased vacuolation of the sarcoplasmic reticulum in neb¹⁵, neb¹¹, neb³⁴, neb²¹ mutants (arrowheads). (F) Z-disks in wt muscle (arrowheads). (G–J) Significantly thicker Z-disks in neb¹⁵, neb¹¹, neb³⁴, neb²¹ mutants (arrowheads). (K) No abnormal protein aggregates have been observed in wt muscle. (L, M, and O) Electron-dense structures with a fibrous composition were observed in neb¹¹, neb¹⁵, neb²¹ (arrowheads). (L and N) Accumulation of ribosomes were observed in neb¹⁵ and neb³⁴ mutants (arrows). (Q) Normal triad structure and organization in wt muscle (arrowhead). (R–U) Triad structure and organization did not seem to be defective in neb mutants (arrowheads), except in splice mutant neb³⁴, where enlarged terminal cisternae were observed (arrows). Scale bars A-O = 500 nm. Scale bars Q-U = 100 nm.

Figure 7

Force production is reduced in neb mutants. (A–F) Plots of force production at optimal length (mean ± SEM). Across all frequency range of stimulation (1–200 Hz), all of the mutants showed lower force production than the wt controls at optimal sarcomere length. The reduction in force production is significant in both nonsense mutants and the splice mutants neb^hu28 and neb³⁴. Asterisks indicate p-values (* ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001). See Supplemental Table 4 for measurements and statistics.

Figure 8

Thin filaments are shorter neb mutants. (A) Linear regression graphs of thin filament lengths measured at 1.6–2.6 μm sarcomere lengths. Thin filament length was reduced in all neb mutants compared to wt siblings. (B) Graphs of thin filament lengths measured at 1.85–2.2 μm sarcomere length (mean ± SEM). Thin filament length was significantly reduced in all neb mutants. Asterisks indicate p-values (* ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001). See Supplemental Table 6 for statistics.

Figure 9

neb mutants show altered transcriptome profile. (A–D) Dendrograms and heatmaps from Metascape enrichment analysis illustrating selective (unique to each comparison in the study) and shared (common to all the comparisons in the study) enriched terms. (A) Selective downregulated pathways. (B) Shared downregulated pathways. Many of the downregulated genes in the nonsense mutants and the splice mutants neb^hu28 and neb²¹ are involved in muscle contraction, muscle tissue development, and intermediate filament-based processes. (C) Selective upregulated pathways. (D) Shared upregulated pathways. Highly upregulated genes in the nonsense mutants are mostly involved in protein degradation processes and response to stress, and cluster under the proteasome and stress response pathways. The heatmap cells are coloured by their p-values; white cells indicate the lack of enrichment for that term in the corresponding gene list. p-value cutoff = 0.05; minimum enrichment = 1.5; minimum overlap = 3; kappa score = 0.3. (E) UpSet plot illustrating number of DEGs unique and shared between the data sets. The nonsense mutants and neb^hu28 cluster together. Only three DEGs are shared by all five neb mutants depicted in the plot. (F) Dendrogram and heatmap illustrating expression of nemaline myopathy genes in nebulin mutants. Noticeable, acta1 and tnnt1 are downregulated in several of the mutants investigated in this study. See Supplemental Table 7 for recorded values.