A Targeted, Low-Throughput Compound Screen in a Drosophila Model of Neurofibromatosis Type 1 Identifies Simvastatin and BMS-204352 as Potential Therapies for Autism Spectrum Disorder (ASD)

Abstract

Autism spectrum disorder (ASD) is a common neurodevelopmental condition for which there are no pharmacological therapies that effectively target its core symptomatology. Animal models of syndromic forms of ASD, such as neurofibromatosis type 1, may be of use in screening for such treatments. Drosophila larvae lacking Nf1 expression exhibit tactile hypersensitivity following mechanical stimulation, proposed to mirror the sensory sensitivity issues comprising part of the ASD diagnostic criteria. Such behavior is associated with synaptic dysfunction at the neuromuscular junction (NMJ). Both phenotypes may thus provide tractable outputs with which to screen for potential ASD therapies. In this study, we demonstrate that, while loss of Nf1 expression within the embryo is sufficient to impair NMJ synaptic transmission in the larva, constitutive Nf1 knock-down is required to induce tactile hypersensitivity, suggesting that a compound must be administered throughout development to rescue this behavior. With such a feeding regime, we identify two compounds from a targeted, low-throughput screen that significantly and consistently reduce, but do not fully rescue, tactile hypersensitivity in Nf1^P1 larvae. These are the HMG CoA-reductase inhibitor simvastatin, and the BK_Ca channel activator BMS-204352. At the NMJ, both compounds induce a significant reduction in the enhanced spontaneous transmission frequency of Nf1^P1 larvae, though again not to the level of vehicle-treated controls. However, both compounds fully rescue the increased quantal size of Nf1^P1 mutants, with simvastatin also fully rescuing their reduced quantal content. Thus, the further study of both compounds as potential ASD interventions is warranted.

Keywords: Drosophila; Nf1; autism spectrum disorder; drug screening; neurofibromatosis type 1.

Figure 1.

Constitutive knock-down of Nf1 is required to induce tactile hypersensitivity in third instar larvae. Abbreviated genotypes for the lines tested are given in the figure panels. GFP^RNAi (green) and Nf1^RNAi (orange) refer to lines in which GAL4 is expressed under the control of elav to drive expression of either UAS-GFP^RNAi or UAS-Nf1^RNAi, respectively, and UAS-Dicer2, with GAL80^ts expressed under the control of the tubulin promoter. All transgenic constructs are hemizygous or heterozygous in the larvae tested. A, Constitutive knock-down of Nf1 expression (Nf1^RNAi) throughout all life stages results in larval tactile hypersensitivity, as indicated by a significantly greater number of larvae responding to a mechanical stimulus compared with GFP^RNAi. B, Knock-down of Nf1 only within the embryo has no significant impact on the number of responding larvae (p = 0.0580), nor does (C) knock-down of Nf1 within the larval stages (p > 0.9999). D–F, K33 and Nf1^P1 larvae were also subjected to the same shifts in temperature as those required for constitutive, embryonic, and larval knock-down of Nf1, respectively. Regardless of the temperature paradigm, Nf1^P1 larvae always demonstrated a significantly greater likelihood of displaying the nocifensive response following stimulation. Numbers within each bar represent the sample size for that group. For the ease of comparing groups in which different sample sizes were used, raw data have been presented as percentages within the figure. Statistical comparisons via Fisher’s exact test were nonetheless conducted on raw data before normalization.

Figure 2.

Knock-down of Nf1 in the embryo is sufficient to induce synaptic transmission deficits in the third instar larval stage. Full genotypes for those abbreviated in the figure (i.e., GFP^RNAi and Nf1^RNAi) are explained in the legend for Figure 1. A–F, Raising Nf1^RNAi larvae at 30°C, sufficient to ensure constitutive knock-down of Nf1, mimics the Nf1^−/− larval phenotype (Dyson et al., 2022), in that it does not affect EJP amplitude (p = 0.5334) but increases mEJP frequency and amplitude, and reduces quantal content. G–L, Knock-down of Nf1 within the embryonic period has a similar effect on synaptic transmission as constitutive knock-down on all parameters examined, including no significant impact on (G) EJP amplitude (p = 0.8967). M–R, In contrast, knock-down of Nf1 only during larval stages has no effect on (M) EJP amplitude (p = 0.3010), (N) mEJP frequency (p = 0.3230), (O) mEJP amplitude (p = 0.3694), or (P) quantal content (p = 0.9998). Data in each panel were analyzed via unpaired, two-tailed Student’s t test. For each experiment, n = 13. Data are presented as mean ± SEM.

Figure 3.

A targeted pharmacological screen identifies simvastatin and BMS-204352 to improve tactile hypersensitivity in Nf1^P1 larvae. Twenty compounds were screened for their ability to reduce the number of Nf1^P1 larvae responding to a mechanical stimulus, with three proving toxic at the concentration (50 μm) tested. Only administration of simvastatin (50% of vehicle-treated Nf1^P1) and BMS-204352 (61.1% of vehicle-treated Nf1^P1) resulted in a significant decrease in the number of responding larvae. Data are presented as a percentage of the number of Nf1^P1 larvae raised on an equivalent concentration of DMSO (vehicle) that responded to the stimulus (dotted line = 100%). Statistical comparisons were conducted on raw data using Fisher’s exact test between compound-treated and vehicle-treated Nf1^P1 larvae.

Figure 4.

Simvastatin and BMS-204352 consistently improve, but do not fully rescue, tactile hypersensitivity in Nf1^P1 larvae, while having no effect on overall activity. A, Across four independent trials, 50 μm Simvastatin significantly reduces the mean percentage responding larvae, as does (B) 50 μm BMS-204352. C, The presence of simvastatin solely during the embryo has no impact on the percentage of Nf1^P1 larvae responding to stimulation (p = 0.3026). D, Conversely, exposing only larval progeny to simvastatin results in a significant reduction in the mean percentage of responding Nf1^P1 larvae. E, The presence of BMS-204352 solely during the embryo results in a significant reduction in the percentage of Nf1^P1 larvae responding to stimulation. F, The percentage of responding Nf1^P1 larvae is also significantly reduced when only larval progeny are exposed to BMS-204352. G, K33 and Nf1^P1 larvae do not significantly differ in their total distance traveled over a 3-min period at room temperature, nor is this impacted by treatment with 50 μm simvastatin (p = 0.9518). H, 50 μm BMS-204352 treatment also does not alter distance traveled in Nf1^P1 larvae, which again show no significant difference in crawling behavior compared with K33 controls (p = 0.1881). In panels A–F, each data point represents the percentage responding larvae from a single trial, with n = 20 per trial, such that n = 80 larvae overall. Data in these panels were analyzed via unpaired, two-tailed Student’s t test. Comparisons were made between individual trials, such that n = 4 per group. Data in panels G, H were analyzed via one-way ANOVA followed by Tukey’s post hoc test. All data are presented as mean ± SEM.

Figure 5.

Simvastatin and BMS-204352 improve synaptic transmission deficits at the Nf1^P1 larval NMJ, and have no impact on normal transmission in K33 larvae. A, Simvastatin (50 μm) treatment has no effect on EJP amplitude in either Nf1^P1 or K33 larvae (p = 0.2793). B, The increased mEJP frequency of Nf1^P1 larvae is reduced following simvastatin treatment, although this is still significantly greater than that of vehicle-treated K33 larvae. There is no significant difference between vehicle-treated and simvastatin-treated K33 larvae (p = 0.9784). C, Simvastatin rescues the enhanced mEJP amplitude of Nf1^P1 larvae to values indistinguishable from those of vehicle-treated K33 larvae (p = 0.8097), which do not show a significant difference compared with simvastatin-treated K33 larvae (p > 0.9999). D, Simvastatin rescues the reduced quantal content of Nf1^P1 larvae to values indistinguishable from those of vehicle-treated K33 larvae (p = 0.4934), which also do not show a significant difference compared with simvastatin-treated K33 larvae (p = 0.9989). E, F, Representative traces of data presented in panels A–D. G, BMS-204352 (50 μm) treatment has no effect on EJP amplitude in either Nf1^P1 or K33 larvae (p = 0.5878). H, The increase in mEJP frequency of Nf1^P1 larvae is reduced following BMS-204352 treatment, although this is still significantly greater than that of vehicle-treated K33 larvae. There is no significant difference between vehicle-treated and simvastatin-treated K33 larvae (p = 0.9098). I, BMS-204352 rescues the enhanced mEJP amplitude of Nf1^P1 larvae to values indistinguishable from those of vehicle-treated K33 larvae (p = 0.9211), which do not show a significant difference compared with BMS-204352-treated K33 larvae (p = 0.9996). J, The increase in quantal content in Nf1^P1 larvae following BMS-204352 treatment is not significant (p = 0.1075); however, the quantal content of BMS-204352 treated Nf1^P1 larvae also does not differ from that of vehicle-treated K33 larvae (p = 0.8849). BMS-204352 does not significantly alter quantal content in K33 larvae either (p = 0.9991). All statistical comparisons were made via two-way ANOVA followed by Tukey’s post hoc test, in which each genotype + treatment group was compared with all others. n = 13 for each group. All data are presented as mean ± SEM. Although not explicitly stated in the figure or legend, in panels B–D and H–J, mEJP frequency and amplitude were both significantly increased, and quantal content significantly reduced, in vehicle-treated Nf1^P1 larvae relative to vehicle-treated K33 controls, as would be expected in larvae lacking Nf1 expression (Dyson et al., 2022).