(A) Drosophila females expressing AR in eyes using GMR-GAL4 were raised in medium containing vehicle or DHT and adult eye phenotypes were assessed by light microscopy. (B) Blinded scoring of the external eye phenotypes in (A) using a quantitative scoring system (Pandey et al., 2007). (C) Western blot showing levels of AR expression for AR12Q- and AR52Q-expressing flies shown in (A). (D) Pupae expressing AR in eyes using GMR-GAL4 were raised in medium with or without DHT and whole mount preparations of eye discs were immunostained for lamin (blue) and AR (green). Phalloidin (red) was used to stain F-actin. Samples were examined by confocal microscopy. (E–F) Third instar larvae expressing AR using the salivary gland fkh-GAL4 were dissected and stained with DAPI (blue) and phalloidin (red). Overall gland size shown in (E), cell size shown in (F). Phalloidin staining was used to delineate cell boundaries and determine cell size. Scale bar, 50 μm. (G) Third instar larvae expressing AR12Q or AR52Q using D42-GAL4 were assessed for their ability to travel distances along the surface of an agar plate. (H–I) Larvae expressing AR52Q using OK371-GAL4 were raised in medium containing vehicle or DHT, dissected as third instar wandering larvae, and stained using the post-synaptic marker discs large (DLG, green) and the pre-synaptic marker HRP (red). Type 1B boutons were counted at muscle 4. Scale bar, 10 μm. (J) Female flies expressing AR12Q using GMR-GAL4 were raised on medium containing DHT. Each line shown represents an independent transformant line. (K) Western blot analysis of heads shown in (J). In (A) and (J), left side of each diptych shows light micrograph imaged at 63x, while right side shows increased magnification (~140x) of the posterior region of the eye in which degeneration is concentrated. ** p<0.01 in all panels. Bars, mean + SEM in all panels.

(A) HEK293T cells were transfected with indicated AR constructs together with both the luciferase pARE-E1b-Luc and the β-galactosidase pCMVβ reporter constructs. AR transactivation was measured in the presence and absence of DHT by luciferase assay and normalized to β-galactosidase activity. (B) Salivary glands of Drosophila larvae expressing AR using fkh-GAL4. Larvae were raised were raised in medium containing DHT and processed for immunocytochemistry. AR was detected with anti-AR antibody (green) and nuclei were stained with DAPI (blue). Scale bar, 50 μm. (C) Drosophila females expressing AR in eyes using GMR-GAL4 were raised in medium containing DHT and adult eye phenotypes were assessed by light microscopy. (D) Blinded scoring of the external eye phenotypes in (C) using a quantitative scoring system. (E) Third instar larvae expressing AR using the motor neuron driver D42-GAL4 were assessed for their ability to travel distances along the surface of an agar plate. (F–G) Third instar larvae expressing AR using the salivary gland fkh-GAL4 were dissected and stained with DAPI (blue) and phalloidin (red). Overall gland size shown in (F), cell size shown in (G). Phalloidin staining was used to delineate cell boundaries and determine cell size. Scale bar, 50 μm. ** p<0.01 in all panels. n.s., not significant. Bars, mean + SEM in all panels. See also Figure S2 and S3.

(A) RNA from fly heads expressing AR using GMR-GAL4 was extracted and analyzed using Affymetrix arrays. Using a false discovery rate of 0.1, 149 genes were identified as showing significant changes in AR52Q flies due to DHT treatment and were thereby selected for further analysis. (B) Scoring of the external eye phenotype shows a correlation between the severity of the observable external phenotype and the clustering results. (C) PCA analysis using the 149 genes shown in (A). See also Figure S6 and Table S1.

(1) Ligand binding induces a conformational change in the LBD to create the AF-2 binding surface. Ligand also induces post-translational modifications that are not depicted. (2) Ligand-activated AR translocates to the nucleus. (3) Prior to DNA binding, the AF-2 domain is occupied by the N-terminal FxxLF in an intra- or inter-molecular interaction. (4) Following DNA binding, AF-1 and AF-2 interact with coregulators. In order to initiate pathogenesis, polyglutamine-expanded AR must bind DHT, translocate to the nucleus, bind DNA, and interact with coregulators at AF-2. While interactions at AF-1 modify toxicity, AF-2 function is essential for toxicity.

(A) Schematic of AR constructs used. NTD, N-terminal transactivation domain; DBD, DNA-binding domain; H, hinge; LBD, ligand-binding domain; NLS, nuclear localization sequence. (B) Blinded scoring of the external eye phenotypes in (D) and (F) using a quantitative scoring system. Bars, mean + SEM. (C) Salivary glands of Drosophila larvae expressing AR using fkh-GAL4. Larvae were raised in medium containing DHT and processed for immunocytochemistry. AR was detected with anti-AR antibody (green) and nuclei were stained with DAPI (blue). Scale bar, 50 μm. (D) Drosophila females expressing AR in eyes using GMR-GAL4 were raised in medium containing vehicle or DHT and adult eye phenotypes were assessed by light microscopy. (E) Salivary glands of Drosophila larvae expressing AR using fkh-GAL4. Larvae were raised in medium containing ethanol and processed for immunocytochemistry as in (C). Scale bar, 50 μm. (F) Drosophila females expressing AR in eyes using GMR-GAL4 were raised in medium containing vehicle or DHT and adult eye phenotypes were assessed by light microscopy. See also Figure S1 and S2.

(A) Crystal structure of the AR LBD (PDB ID: 2AMA) showing the AF-2 binding surface (gold) and the two charge clamp residues in AF-2, K720 (blue) and E897 (red). Green, FxxLF peptide co-crystallized with AF-2 (PDB ID: 1T7R). Pink, LxxLL peptide co-crystallized with AF-2 (PDB ID: 1T7F). (B) Salivary glands of Drosophila larvae expressing AR using fkh-GAL4. Larvae were raised in medium containing DHT and processed for immunocytochemistry. AR was detected with anti-AR antibody (green) and nuclei were stained with DAPI (blue). Scale bar, 50 μm. (C) HEK293T cells were transfected with indicated AR constructs together with both the luciferase pARE-E1b-Luc and the β-galactosidase pCMV β reporter constructs. AR transactivation was measured in the presence and absence of DHT by luciferase assay and normalized to β-galactosidase activity. (D) Viability assay of Drosophila expressing indicated AR transgenes using elav-GAL4. Crosses were performed in medium containing DHT. The number of pupae from each 1×1 cross was counted and normalized to lacZ. Expression of AR52Q resulted in larval lethality, while E897K or K720A mutations increased survivability to the pupal stage. (E) Drosophila females expressing AR in eyes using GMR-GAL4 were raised in medium containing DHT and adult eye phenotypes were assessed by light microscopy. Left, AR constructs with 12Q. Right, AR constructs with expanded Q (AR52Q wt, AR66Q E897K, AR72Q K720A, AR65Q G21E). See Figure S5E for phenotype severity scores. (F) Third instar larvae expressing AR using the motor neuron driver D42-GAL4 were assessed for their ability to travel distances along the surface of an agar plate. (G–H) Larvae expressing AR using the motor neuron driver OK371-GAL4 were raised in medium containing vehicle or DHT, dissected as third instar wandering larvae, and stained using the post-synaptic marker discs large (DLG, green) and the pre-synaptic marker HRP (red). Type 1B boutons were counted at muscle 4. Scale bar, 10 μm. ** p<0.01 in all panels. Bars, mean + SEM in all panels. See also Figure S2 and S5.

(A–D) RNAi knockdown of limpet (A) and a P-element allele of limpet (limpet^GE27535) (B) enhance the phenotype of AR52Q alone (C). Flies with a chromosomal duplication of a region containing limpet (Dp(3;3)st⁺g18) (D) show suppression of the AR52Q degenerative phenotype. (E–H) Limpet alleles similarly modify the phenotype of AR12Q flies with a strong phenotype. (I–J) Expression of limpet RNAi fails to enhance the phenotype in flies expressing AR66Q E897K. (K–L) Chromosomal duplication of limpet fails to suppress the phenotype in flies expressing pure polyglutamine (127Q). (M) Blinded scoring of the external eye phenotypes in (A–H) using a quantitative scoring system. All crosses performed on medium containing DHT. Bars, mean + SEM. ** p<0.01, * p<0.05. (N) Of 81 genes that changed in a DHT-dependent manner by expression profile analysis, 37 genes showed an opposite change in the presence of limpet duplication. Expression changes of these 37 genes are shown and plotted as relative expression (std dev) to the mean. See also Figure S7.