siRNA screen identifies QPCT as a druggable target for Huntington's disease

Abstract

Huntington's disease (HD) is a currently incurable neurodegenerative condition caused by an abnormally expanded polyglutamine tract in huntingtin (HTT). We identified new modifiers of mutant HTT toxicity by performing a large-scale 'druggable genome' siRNA screen in human cultured cells, followed by hit validation in Drosophila. We focused on glutaminyl cyclase (QPCT), which had one of the strongest effects on mutant HTT-induced toxicity and aggregation in the cell-based siRNA screen and also rescued these phenotypes in Drosophila. We found that QPCT inhibition induced the levels of the molecular chaperone αB-crystallin and reduced the aggregation of diverse proteins. We generated new QPCT inhibitors using in silico methods followed by in vitro screening, which rescued the HD-related phenotypes in cell, Drosophila and zebrafish HD models. Our data reveal a new HD druggable target affecting mutant HTT aggregation and provide proof of principle for a discovery pipeline from druggable genome screen to drug development.

Figure 1. Downregulation of QPCT in flies rescues HD toxicity

a. The eye phenotype of flies that express Q48 crossed to w¹¹¹⁸ (VDRC stock number 60000) is rescued upon downregulation of Drosophila Glutaminyl cyclase (QC^GD38277, VDRC GD-RNAi line 38277). Representative images of eye pigmentation rescue are shown. F=female; M=male. b. Downregulation of QPCT fly orthologs QC and isoQC using KK-RNAi lines (lines QC^KK106341 and isoQC^KK101533) reduced the number of black necrotic-like spots on Q48 flies (see Supplementary Fig.5a for quantification). Fisher’s exact test was applied for statistical comparison between control and test genotypes. Females: isoQC^KK101533 p=2.42 E-14; QC^KK106341 p= 3.05 E-12; males: isoQC^KK101533 p=3.53 E-0.8; QC^KK106341 p= 1.72 E-0.9 c. Loss of rhabdomeres due to expression of expanded huntingtin exon1 (elav-Gal4; GMRHTT.Q120) in the eye was significantly rescued upon downregulation of QPCT fly orthologues QC or isoQC (GD- or KK-RNAi lines as indicated). Graph shows the mean ± SEM of the average number of rhabdomeres per eye from 4 independent experiments; one-tailed paired t-test was used to test significance. d. The number of aggregates in the eyes of flies expressing expanded huntingtin HTTex1-Q46-eGFP using GMR-GAL4 was reduced by downregulating QPCT fly orthologs QC and isoQC (RNAi lines isoQC^KK101533, QC^KK10634, QC^GD38277). Graph shows mean ± SEM of the number of aggregates from 4 independent crosses for each genotype with control levels set at 100%. One-tailed paired t-test was used for comparison between control and test genotypes (n = 4). In all panels, * p<0.05, ** p<0.01 and *** p<0.001. Scale bars represent 200 μm.

Figure 2. QPCT modulates HTT toxicity and aggregation in mammalian cell lines and primary neurons

a. The percentage of cells with apoptotic nuclei or HTT(Q74) aggregates is reduced in HEK293 cells transiently expressing EGFP-HTT(Q74) and treated with QPCT siRNA. Representative images are shown in supplementary figure 6a. b. QPCT shRNA significantly reduced the number of aggregates in mouse primary cortical neurons expressing Q80-EGFP. Scale bar represents 10 μm. The mean of 3 independent experiments in triplicate is represented in the graph. Significance was analysed by two-tailed paired Student’s t-test. c,d. Overexpression of QPCT (pCMV6-QPCT) together with EGFP-HTT(Q74) in HeLa cells for 48h increased the percentage of cells with apoptotic nuclear morphology and aggregates (c), this effect is not observed with a catalytically inactive QPCT (QPCT(E201Q)-Flag) (d). e. The percentage of HeLa cells expressing EGFP-HTT(Q74), EGFP-Q57 or EGFP-Q81 with aggregates is enhanced upon QPCT-Flag overexpression for 48 h. f. QPCT siRNA reduces the percentage EGFP-Q81 or EGFP-A37 with aggregates in HEK293. g. Overexpression of QPCT enhanced the amount of mutant HTT(1-548)-Flag co-immunoprecipitating with HTT(1-588)-GFP. Levels of Flag-HTT(1-588) co-immunoprecipitated relative to total lysates from 5 independent experiments are represented in the graph. Data were analyzed by two-tailed paired Student’s t-test (n= 5 experiments). Full blot images are shown in Supplementary Fig. 17a. In all panels, unless indicated, graphs show mean values with control conditions set to 100 and error bars represent standard deviation from a triplicate experiment representative of at least three independent experiments. Statistical analyses were performed by two-tailed unpaired Student’s t-test: ***p<0.001, **p<0.01; *p<0.05; NS, not significant

Figure 3. Design of QPCT inhibitors that reduce mutant HTT aggregation

a. Chemical structure of compounds designed to inhibit QPCT activity. Table indicating the activity and in vitro ADME properties of the compounds is shown in supplementary fig. 11a. b,c. Treatment of HeLa cells expressing EGFP-HTT(Q74) with SEN177, 817 and 180 (50 μM) for 24h reduced the percentage of cells with aggregates (b) and apoptotic nuclei (c). d. SEN177 reduces the percentage of HEK293 cells with EGFP-HTT(Q74) or EGFP-A37 aggregates in a concentration-dependent manner. e. SEN177 does not further reduce the percentage of EGFP-HTT(Q74) aggregates in QPCT shRNA transfected cells. f. SEN177 reduces the amount of HTT(1-588)-GFP co-immunoprecipitating with HTT(1-548)-Flag in HeLa cells (25 μM SEN177). The amount of GFP-HTT(1-548) immunoprecipitated relative to total lysates was quantifiedand the average of 5 independent experiments is shown in the graph. Data were analyzed by two-tailed paired Student’s t-test (n= 5 experiments). Full blot images are shown in Supplementary information 17b. g. Primary neurons expressing EGFP-Q80 for 3 days were treated with 50 μM of indicated compounds for further 24h. In all panels, unless indicated, graphs show mean values with control conditions set to 100 and error bars represent standard deviation from a triplicate experiment representative of at least three independent experiments. Statistical analyses were performed by two-tailed unpaired Student’s t-test: ***p<0.001, **p<0.01; *p<0.05; NS, not significant.

Figure 4. QPCT inhibition induces alpha B-crystallin levels

a. Alpha B-crystallin (Cryab) protein levels were increased in cells transfected with HTT(Q74)GFP and treated with the indicated compounds at 25 μM for 24 h. Full blot images are shown in Supplementary information 17c. b,c. Knockdown of QPCT for 24 h followed by transfection with HTT(Q74)GFP for another 24h increased protein (b) and mRNA (c) levels of alpha B-crystallin. Fold change in mRNA of QPCT or alpha B-crystallin is represented in the graph with error bars representing standard deviation. The mean of three independent experiments in triplicate was normalized to 1 and significance was calculated by one sample t-test. Full blot images are shown in Supplementary information 17d. d. Overexpression of alpha B-crystallin (CRYAB-Flag) reduced the percentage of cells with HTT(Q74)GFP aggregates. SEN817 decreased aggregation when added at 25 μM for 24h in control but not CRYAB-expressing cells. In all panels, unless indicated, graphs show mean values with control conditions set to 100 or 1, and error bars represent standard deviation from a triplicate experiment representative of at least three independent experiments. Statistical analyses were performed by two-tailed unpaired Student’s t-test: **p<0.01; *p<0.05; NS, not significant.

Figure 5. Pharmacologic inhibition of QPCT in fly

a. Flies that expressed HTTex1-Q46-eGFP in the eye have fewer aggregates after treatment with 50 μM of indicated compounds . Graph represents mean ± SEM from 4 independent crosses for each compound. Statistical analyses were performed by one-tailed unpaired Student’s t-test. Scale bars represent 200 μm. b. Flies expressing HTTEx1-Q120 (GMR-HTT.Q120) show more rhabdomeres after treatment with SEN177 (50 μM). Graph represents the average number of rhabdomeres per eye ±SEM from 3 independent experiments with females and males counted separately, each based on approximately 10 individuals per datapoint, scoring 15 ommatidia from each individual. Statistical analysis was performed using one-tailed paired Student’s t-test.

Figure 6. Pharmacologic inhibition of QPCT in zebrafish

a. Representative sections through the central retina of transgenic HD zebrafish at 7 d.p.f. treated with DMSO, SEN177 (1 mM), SEN817 (100 μM) or SEN180 (100 μM) showing aggregates (arrow) within the rod photoreceptors. Scale bar represents 10 μm. Treatment with QPCT inhibitors resulted in reduction in aggregates (Student’s t-test) for SEN187 and SEN810. b. Representative sections through the central retina of transgenic HD zebrafish at 9 d.p.f. treated with DMSO, SEN177 (1 mM), SEN817 (100 μM) or SEN180 (100 μM). To demonstrate that loss of GFP corresponds to loss of photoreceptors, sections were stained with anti-rhodopsin (1D1) antibody (red). GFP labels the whole rod photoreceptor, whereas rhodopsin is present in the rod outer segment. Merged images show co-localisation of GFP the rhodopsin (red). Photoreceptor degeneration is ameliorated by SEN817 and SEN180. Scale bars, 10 μm. In all panels, **p<0.01; *p<0.05; NS, not significant.