Conserved genes act as modifiers of invertebrate SMN loss of function defects

Abstract

Spinal Muscular Atrophy (SMA) is caused by diminished function of the Survival of Motor Neuron (SMN) protein, but the molecular pathways critical for SMA pathology remain elusive. We have used genetic approaches in invertebrate models to identify conserved SMN loss of function modifier genes. Drosophila melanogaster and Caenorhabditis elegans each have a single gene encoding a protein orthologous to human SMN; diminished function of these invertebrate genes causes lethality and neuromuscular defects. To find genes that modulate SMN function defects across species, two approaches were used. First, a genome-wide RNAi screen for C. elegans SMN modifier genes was undertaken, yielding four genes. Second, we tested the conservation of modifier gene function across species; genes identified in one invertebrate model were tested for function in the other invertebrate model. Drosophila orthologs of two genes, which were identified originally in C. elegans, modified Drosophila SMN loss of function defects. C. elegans orthologs of twelve genes, which were originally identified in a previous Drosophila screen, modified C. elegans SMN loss of function defects. Bioinformatic analysis of the conserved, cross-species, modifier genes suggests that conserved cellular pathways, specifically endocytosis and mRNA regulation, act as critical genetic modifiers of SMN loss of function defects across species.

Figure 1. Survival and average length of Cesmn-1(lf) animals is decreased.

Homozygous loss of Cesmn-1(lf) results in slower growth and most animals die during larval stages. Animals that reach the adult stage are short-lived and sterile. A) Image of age-matched heterozygous hT2(lethal)[myo-2p::GFP]/Cesmn-1(lf) (+/Cesmn-1(lf), described in text) and homozygous Cesmn-1(lf) individuals (below). Scale bar indicates 150 microns. B) To test the impact of candidate modifier genes on growth rates and size, a C. elegans growth assay was established. +/Cesmn-1(lf) animals were collected as unhatched eggs, reared for 5 days/2 generations on standard C. elegans plates upon RNAi bacterial strains, as illustrated in the flow chart. Bacterial cultures expressed dsRNA corresponding to the gene of interest for RNAi knockdown; bacterial cultures containing vector with no insert were used as a control (‘empty’). C) Using an automated system, the length and fluorescence of animals was determined ; the smallest larvae in each culture were indistinguishable from debris and were excluded from the analysis as illustrated in the graph. Genotypes were discriminated by GFP fluorescence and length was determined as ‘time-of-flight’ through the laser chamber; each dot in the graph represents an individual animal. The disparate growth/survival rates of Cesmn-1(lf) and +/Cesmn-1(lf) animals alters the percentage of ‘large’ animals in these mixed stage cultures. As illustrated graphically, +/Cesmn-1(lf) and Cesmn-1(lf) animals can be sorted into blue and red boxes, respectively; animals designated as ‘large’ fall into shaded boxes. The fraction of large animals was calculated (# animals in shaded box/# animals in both shaded and unshaded box for each genotype) and is reported as % large with standard error of the mean (S.E.M.); significance was determined by Mann-Whitney U two-tailed test with p≤0.05. At least three independent cultures of more than 200 animals were scored for each targeted gene.

Figure 2. Genome-wide RNAi screen for Cesmn-1(lf) modifier genes.

A) To identify genes whose knockdown modifies the growth of Cesmn-1(lf) animals, the progeny of hT2(lethal)[myo-2p::GFP]/Cesmn-1(lf) (+/Cesmn-1(lf), described in text) were reared for more than 2 generations in 96-well, liquid culture format on bacteria expressing dsRNA corresponding to over 16,500 C. elegans genes. Modifier genes identified in this assay were tested subsequently in neuromuscular assays in C. elegans and Drosophila. B) Cesmn-1(lf) and +/Cesmn-1(lf) length was measured for each RNAi clone and a ‘growth’ ratio of large∶small animals was determined for each genotype. Representative graphs illustrate the distribution of RNAi clone growth ratios. Candidate enhancer genes were those with a growth ratio more than 2 standard deviations above the mean for Cesmn-1(lf) (shaded in right graph) and within 0.7 standard deviations of the mean for +/Cesmn-1(lf) (shaded in left graph) for each 96-well plate. No suppressors were identified using similar criteria. Two independent determinations were made for each clone in the original screen. Candidate genes were retested in at least quadruplicate and enhanced growth in at least 40% of trials before designation as growth modifier genes; average growth ratios of enhancers from the C. elegans screen are indicated.

Figure 3. Loss of C. elegans PLS3 ortholog enhances the pharyngeal pumping defects of Cesmn-1(lf) animals.

During feeding the C. elegans pharynx contracts at over 200 times per minute to capture and grind bacteria. A) Pharyngeal pumping rates can be determined by videotaping feeding and counting contractions as movement of the grinder (indicated by arrow) at slower playback speeds; scale bar indicates 40 microns. B) The progeny of hT2(lethal)[myo-2p::GFP]/Cesmn-1(lf) (+/Cesmn-1(lf)) animals were allowed to hatch on bacterial strains on standard C. elegans plates. Pumping rates of +/Cesmn-1(lf) heterozygous and Cesmn-1(lf) homozygous animals were determined on different days (day 1–4) on a control ‘empty vector’ bacterial strain. Decreased locomotion (sinusoidal movement) was also scored as uncoordinated (Unc). At day 1, the pumping rates and locomotion of Cesmn-1(lf) animals are identical to +/Cesmn-1(lf) or wild type animals. By day 2, the average pumping rates dropped dramatically (as previously reported [44]) and roughly 20% of animals were uncoordinated (column on right). By day 3 and 4, the pumping rates and locomotion of the majority of surviving animals are defective. Expressing Cesmn-1 in neurons is sufficient to restore pumping rates to near normal levels . The behavior of Cesmn-1(lf) animals is initially normal due to maternal loading of Cesmn-1 gene products, but as the maternal contribution is depleted, loss of Cesmn-1 causes progressive defects in neuromuscular function. C) Pumping rates of +/Cesmn-1(lf) and Cesmn-1(lf) animals were determined at 3 days post-hatching on bacterial feeding strains expressing dsRNA corresponding to Cesmn-1 and plst-1 genes; ‘empty vector’ bacterial cultures were used as a control (‘empty’). plst-1(RNAi) or Cesmn-1(RNAi) significantly descreased Cesmn-1(lf) pumping rates, but not +/Cesmn-1(lf) pharyngeal pumps. D) Decreasing plst-1 function has no effect on pumping rates in heterozygous +/Cesmn-1(lf) animals, but pumping rates were significantly decreased in Cesmn-1(lf); plst-1(tm4255) double mutant animals. At least 20 animals were scored for each genotype in at least 2 independent trials. Pumping rates reported are the average of all animals in all trials. Because of the known impact of chromosomal pairing on histone methylation and gene expression , all animals tested herein were the progeny of animals carrying the hT2 balancer chromosome that prevents recombination. S.E.M. is reported. Significance of p≤0.05 is indicated with an asterisk and was determined using either an unpaired two-sample t-test or a Mann-Whitney U two-tailed test according to sample-specific parameters .

Figure 4. C. elegans RNAi screen identifies conserved SMN modifier genes.

A) Four genes were identified as modifiers of Cesmn-1(lf) growth defects in the C. elegans RNAi screen: ncbp-2, T02G5.3, grk-2 and flp-4. The pharyngeal pumping rates of homozygous mutant Cesmn-1(lf) and heterozygous control +/Cesmn-1(lf) animals were determined at day 3 as described in Figure 3. B) For two C. elegans candidate modifier genes, flp-4 and grk-2, loss of function alleles were available. grk-2(rt97) significantly decreased the pumping rates of Cesmn-1(lf) animals, while flp-4(yn35) lowered Cesmn-1(lf) pumping rates, but did not reach statistical significance (p = 0.012 and p = 0.236 by two-tailed Mann-Whitney U test, respectively). Pumping rates were determined on day 3 post-hatching; at least 25 animals were scored for each genotype in at least 3 independent trials. Pumping rates reported are the average of all animals in all trials. Significance versus Cesmn-1(lf) is indicated with an asterisk. To control for genetic background, all animals tested were the progeny of animals carrying the balancer chromosome hT2[myo-2p::GFP]; control animals are hT2/Cesmn-1(lf) siblings.

Figure 5. Loss of PLS3 orthologs enhances SMN loss of function defects in invertebrates.

Previous studies in vertebrates suggested that PLS3 orthologs in C. elegans and Drosophila might modify SMN loss of function defects . A) A genetic interaction was found between the Drosophila PLS3 ortholog (Fimbrin or Fim) and DmSmn using previously described DmSmn RNAi knockdown lines and Fim loss of function alleles. Percentage early or late larval lethality is reported. Ubiquitous RNAi knockdown of DmSmn using the tubulinGAL4 (TubGAL4) driver results in pupal death; modifier genes alter the percentage of animals that die at early versus late pupal stages (day 7 versus day 9 [26]). Loss of Fim function significantly increased the percentage of animals that died as early pupae (p≤0.05 by Chi-square analysis); Fim is an enhancer of DmSmn loss of function growth/survival defects. At least 100 animals of each genotype were scored in four replicates for the lethality assay. No significant variation was observed between control and experimental crosses with each independent trial. B) The Drosophila d02114 allele likely eliminates Fim function and in homozygous animals modestly perturbs NMJ morphology at larval muscle 4. Fim^d02114 enhances DmSmn loss of function NMJ defects consistent with studies in vertebrates . All strains carry the ubiquitous TubGAL4 driver. Significance of p≤0.05 versus single mutant strains was determined by ANOVA and is indicated with asterisk; S.E.M. is shown. In representative images, red corresponds to anti-Discs Large (DLG), green is anti-synaptotagmin and blue is DAPI; scale bar indicates 15 microns.

Figure 6. Loss of Drosophila Cbp20 or FMRF enhances DmSmn loss of function neuromuscular defects.

For two C. elegans modifier genes, loss of function alleles were available for orthologous Drosophila genes: Cbp20 and FMRF –. Loss of either Drosophila gene enhanced the NMJ defects of DmSmn animals. The number of synaptic boutons in the A2 segment of muscles 6 and 7 was counted in third instar larvae (visualized using Discs large and synaptotagmin immunoreactivity, shown in red and green, respectively). Homozygous loss of DmSmn in DmSmn^f1109/DmSmn^73Ao animals dramatically decreases bouton number but loss of one copy of DmSmn, Cbp20 or FMRF had little effect. Loss of one copy of either Cbp20 or FMRF in animals heterozygous for either DmSmn allele significantly decreased bouton numbers. This nonallelic noncomplementation suggests a strong genetic interaction between these two conserved modifier genes and DmSmn. Scale bar indicates 10 microns in representative images; S.E.M. is shown in graph; all transheterozygous combinations are significantly different from the corresponding single heterozygous controls with p≤0.05 by ANOVA as indicated with asterisk.

Figure 7. Interaction map of SMN modifier genes.

A) Numerous published physical and/or functional interactions were found in the literature that connect many invertebrate SMN modifier genes. The names of vertebrate genes are used in the figure, but interactions are drawn from literature in any animal species. Each type of interaction is represented by different colors of connecting lines and references are provided in the table in part B. Not all interactions are indicated including interactions of FGF, TGF-beta and neuropeptide signaling pathways with endocytosis/cytoskeletal pathways, dynein interactions with APP and possible connection between VAPB, the ALS8 locus, and SK channels through riluzole –. The interaction analysis here is not exhaustive and additional interactions may exist. B) The type of interaction is indicated in the table –: P pull-down, Y yeast 2-hybrid, F functional/genetic, D direct physical interaction. Genes shaded in blue are cross-species invertebrate modifiers; genes shaded in yellow are mammalian genes whose perturbation can cause neurodegeneration; genes outlined in grey are SMA modifiers in drug studies or in patients. Genes pertinent to endocytosis include CIN85, EndoA, Alix, PLS3, CAV1, GPRK, Ataxin2, Profilin, EPS15, Phocein, Dynamin, and Striatin; genes pertinent to RNA processing include PABP, CBC20, DcpS, SMN, and FMRP. Combined, our results and published studies support a model in which endocytosis and RNA translational control pathways are physically and functionally coupled , . In normal animals, coupling of these seemingly disparate pathways may help coordinate synaptic activity and protein synthesis in the neuromuscular system.