Comprehensive Modeling of Spinal Muscular Atrophy in Drosophila melanogaster

Abstract

Spinal muscular atrophy (SMA) is a neurodegenerative disorder that affects motor neurons, primarily in young children. SMA is caused by mutations in the Survival Motor Neuron 1 (SMN1) gene. SMN functions in the assembly of spliceosomal RNPs and is well conserved in many model systems including mouse, zebrafish, fruit fly, nematode, and fission yeast. Work in Drosophila has focused on the loss of SMN function during larval stages, primarily using null alleles or strong hypomorphs. A systematic analysis of SMA-related phenotypes in the context of moderate alleles that more closely mimic the genetics of SMA has not been performed in the fly, leading to debate over the validity and translational value of this model. We, therefore, examined 14 Drosophila lines expressing SMA patient-derived missense mutations in Smn, with a focus on neuromuscular phenotypes in the adult stage. Animals were evaluated on the basis of organismal viability and longevity, locomotor function, neuromuscular junction structure, and muscle health. In all cases, we observed phenotypes similar to those of SMA patients, including progressive loss of adult motor function. The severity of these defects is variable and forms a broad spectrum across the 14 lines examined, recapitulating the full range of phenotypic severity observed in human SMA. This includes late-onset models of SMA, which have been difficult to produce in other model systems. The results provide direct evidence that SMA-related locomotor decline can be reproduced in the fly and support the use of patient-derived SMN missense mutations as a comprehensive system for modeling SMA.

Keywords: SMA; SMN1; SMN2; invertebrate models; neuromuscular disease; spinal muscular atrophy.

Figure 1

Survival MotorNeuron (Smn) missense mutations produce a spectrum of viability defects. (A,B) Developmental viability of animals expressing Smn missense mutations at the pupal stage (A) and the adult stage (B). % Viability is the proportion of animals that survive to the pupal/adult stage relative to the number of larvae initially collected. (C) Viability of animals carrying Smn missense mutations through pupal development. In the case, ns indicates a p-value of 0.99 and *** indicates a p-value of exactly 0.0001. (D) Representative images of the pupal stages assessed in (C). Data: bars show average. Error bars show standard error. Data points represent biological replicates of 35–50 animals each, n-values (shown as numbers in parentheses next to genotypes) reflect the number of individual animals counted. n-values are the same for panels (A–C). Statistical analysis: values above the data indicate significance vs. WT from one-way ANOVA using the Dunnet correction for multiple comparisons. ns: not significant (p > 0.05), *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2

Smn missense mutations or knockdown reduce larval locomotion. (A) Larval crawling speed, measured in body lengths/s, in early third instar larvae. (B) Representative traces for the data shown in (A). (C) Larval size for the same animals measured in the locomotion assay in (A). (D) Larval crawling speed, measured in body lengths per second, in wandering third instar larvae. (E) Representative traces for the data shown in (D). (F) Larval size for the same animals measured in the locomotion assay in (D). Data: bars show averages, points represent individual larvae, error bars represent standard error, n-values (# individual larvae) are shown in parentheses adjacent to genotype on x-axis. Statistical significance was determined by ANOVA (ns = not significant, *p < 0.05, **p < 0.01, ***p < 0.001).

Figure 3

Neuromuscular Smn knockdown reduces viability and larval locomotion. (A) Representative western blot of SMN levels and total protein (loading control). (B) Quantification of SMN levels on western blot replicates. SMN levels are divided by total protein levels to control for loading and normalized to OR for comparison between blots. (C,D) Developmental viability at the pupal stage (C) and adult stage (D) for animals expressing Smn missense mutations. % Viability is the proportion of animals the survive to the pupal/adult stage relative to the number of larvae initially collected. (E) Larval crawling speed, measured in body lengths/second, in early third instar larvae. Data: bars show average. Error bars show standard error. For (A,B), data points represent biological replicates of 30–50 animals each, n-values (shown in parentheses next to genotypes) reflect the number of individual animals counted. For (C), data points and n-values (shown in parentheses next to genotypes) reflect the number of individual animals assayed. Statistical analysis: values above the data indicate significance vs. WT from one-way ANOVA using the Dunnet correction for multiple comparisons. ns: not significant (p > 0.05), *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Smn missense mutations and Smn knockdown reduce lifespan. (A–C) Survival curves for adult flies expressing Smn missense mutations. Data is split into three graphs for visual clarity. Colors correspond to phenotypic Class. (D) Lifespan measured as the age at which 10% of animals remain alive for each missense mutation. (E,F) Survival curves for adult flies expressing Smn-RNAi in all neurons and muscles. (G) Lifespan measured as the age at which 10% of animals remain alive for each RNAi condition and control. Shown next to genotypes, n-value (# of individual flies), *** indicates p < 0.0001 by Chi square analysis using the logrank rank/Mantel-Cox test.

Figure 5

Smn missense mutations or knockdown reduces free moving adult locomotion. (A) Adult walking speed, measured in mm/s, in adults 1 day after eclosion for animals expressing Smn missense mutations. (B) Representative traces for the data shown in (A). (C,D) Adult waking speed, measured in mm/s, in adults 1 day after eclosion for animals expressing Smn-RNAi either ubiquitously with the da-Gal4 driver (C) or in both neurons and muscle using the C15 driver line (D). (E) Representative traces for the data shown in (D). Data: bars show average. Error bars show standard error. Data points and n-values (shown in parentheses next to genotypes) reflect the number of individual animals assayed. Statistical analysis: values above the data indicate significance vs. WT from one-way ANOVA using the Dunnet correction for multiple comparisons. ns: not significant (p > 0.05), *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

Animals expressing mild Smn missense mutations develop late-onset locomotor defects as adults. (A) Adult walking speed, measured in mm/s, in adults ranging from 1 day of age (0 weeks) through 6 weeks of age for animals expressing the WT transgene or the D20V, G73R, I93F Smn missense mutations. (B,C) Adult walking speed (mm/s) for the same genotypes shown in (A) at either 5 weeks of age (B) or 6 weeks of age (C). (D) Representative traces for the data shown in (B). (E) Representative traces for the data shown in (C). Data: in (A), points show averages in (B,C), bars show average. Error bars show standard error in all cases. Data points and n-values (shown in parentheses next to genotypes) reflect the number of individual animals assayed. Statistical analysis: values above the data indicate significance vs. WT from one-way ANOVA using the Dunnet correction for multiple comparisons. ns: not significant (p > 0.05), *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 7

Progressive loss of motor function is observed in the legs of partially eclosed Smn missense mutants and in animals expressing neuromuscular Smn knockdown. (A) Qualitative leg function scores (average or all three leg pairs) over time for partially eclosed animals expressing the wild type Smn transgene (n = 11) or the Y208C (n = 10) or G210V (n = 7) Smn missense mutations. (B–D) Qualitative leg function scores for each leg pair over time for same animals as in (A). (E) Qualitative leg function scores (average or all three leg pairs) over time for animals expressing neuromuscular Smn knockdown. (F,G) Qualitative leg function scores for each leg pair over time for the same animals as in (E). Statistical analysis: values above the data indicate significance vs.WT from one-way ANOVA using the Dunnet correction for multiple comparisons. ns: not significant (p > 0.05), p < 0.05, p < 0.01, p < 0.001.