Opposing effects of polyglutamine expansion on native protein complexes contribute to SCA1

Abstract

Spinocerebellar ataxia type 1 (SCA1) is a dominantly inherited neurodegenerative disease caused by expansion of a glutamine-encoding repeat in ataxin 1 (ATXN1). In all known polyglutamine diseases, the glutamine expansion confers toxic functions onto the protein; however, the mechanism by which this occurs remains enigmatic, in light of the fact that the mutant protein apparently maintains interactions with its usual partners. Here we show that the expanded polyglutamine tract differentially affects the function of the host protein in the context of different endogenous protein complexes. Polyglutamine expansion in ATXN1 favours the formation of a particular protein complex containing RBM17, contributing to SCA1 neuropathology by means of a gain-of-function mechanism. Concomitantly, polyglutamine expansion attenuates the formation and function of another protein complex containing ATXN1 and capicua, contributing to SCA1 through a partial loss-of-function mechanism. This model provides mechanistic insight into the molecular pathogenesis of SCA1 as well as other polyglutamine diseases.

Figure 1. ATXN1-S776D but not ATXN1-S776A interacts with RBM17

(a) Schematic representation of the ATXN1 constructs. (b) RBM17 specifically interacted with ATXN1-S776D in the Y2H screen. AD (activation domain) and DB (DNA binding domain) of Gal4 were fused to human ATXN1 or RBM17, respectively. Y2H controls are: lane 1, negative control; lane 2, weak positive control; and lanes 3–5, strong positive controls. (c) ATXN1 interacted with RBM17 in HEK293T cells by co-AP assays. Top panel shows expression of myc-RBM17 after affinity purification on Glutathione-Sepharose 4B beads, demonstrating the ATXN1/RBM17 interaction (arrow). GST-empty vector was used as a control (−). IB, Immunoblot. (d) Co-IP of Atxn1 with RBM17 from wild-type mouse cerebellar extracts. The anti-RBM17 antibody co-immunoprecipitated Atxn1 (arrow), but not the long and short isoforms of Capicua, CIC-L and CIC-S, respectively (bracket).

Figure 2. Enhanced interaction of RBM17 and ATXN1 depends on S776-phosphorylation and polyglutamine tract expansion

(a) Wild-type ATXN1 and ATXN1-S776D, but not ATXN1-S776A, interacted with RBM17. (b) RBM17 bound most strongly to polyglutamine-expanded ATXN1. A bracket showed that there is no obvious difference in S776-phosphorylation depending on polyglutamine expansion in ATXN1 when using a PN1168 antibody (specific to phosphorylated S776 of ATXN1) for IB. (c,d) RBM17 expression in the nuclei of PCs of 18-week-old mice from either wild-type (FVB) (c) or SCA1 transgenic (ATXN1[82Q]) homozygotes (d). RBM17 protein (green) was co-immunostained with calbindin (red, PC marker) and Toto3 (blue, nuclear marker).

Figure 3. RBM17 contributes to polyglutamine-expanded ATXN1 toxicity in the Drosophila eye

Scanning electron microscopy (SEM) of adult Drosophila eyes. Loss of one dRBM17 allele suppressed ATXN1[82Q]-mediated ommatidial disorganization (a,b), while overexpression of hRBM17 worsened abnormalities induced by ATXN1[82Q] but not by ATXN1[30Q] (c–f). Flies were raised at 30°C (a,b) or 25°C (c–f), and genotypes are: (a) GMR-Gal4>UAS-ATXN1[82Q], (b) GMR-Gal4>UAS-ATXN1[82Q]; dRBM17^J23/+, (c) GMR-Gal4>UAS-ATXN1[82Q]; UAS-GFP, (d) GMR-Gal4>UAS-ATXN1[82Q]; UAS-hRBM17, (e) GMR-Gal4>UAS-ATXN1[30Q]; UAS-GFP, and (f) GMR-Gal4>UAS-ATXN1[30Q]; UAS-hRBM17. Magnified images are on the right of each panel. Scale bars are 100μm or 10μm, respectively. Additional data with controls are available in Supplementary Fig. S7.

Figure 4. Polyglutamine expansion enhances RBM17 incorporation into the large ATXN1 native protein complexes

Representative westerns of gel-filtration fractions of (a) wild-type (Atxn1 +/+) and (b) SCA1 knock-in (Atxn1 154Q/+) mouse cerebellar extracts analyzed for CIC, Atxn1, and RBM17. Right top panels show the elution profiles for RBM17 plotted as the average percent protein (± standard error) in each fraction. Right bottom panel shows the relative ratio of RBM17 incorporation into large (red box, fractions 8–10) to small (blue box, fractions 12–13) protein complexes and the dramatic increase in RBM17 incorporation into large protein complexes in Atxn1 154Q/+ mice (*p<0.005, n=4 for Atxn1 +/+ and n=3 for Atxn1 154Q/+). Ex, extract.

Figure 5. RBM17 and CIC form two distinct protein complexes that compete with each other

(a) The CIC anti-serum co-immunoprecipitated Atxn1 (arrow), but not RBM17 (bracket) from mouse cerebellar extracts. (b) ATXN1 co-affinity purified more CIC protein when RBM17 expression was reduced. The interaction of GST-ATXN1[82Q] with endogenous CIC was strongly increased (red arrows, compare lane 2 with lane 3) when endogenous RBM17 expression was decreased (blue arrow) in Neuro-2a cells. Right panel shows the normalized levels of co-affinity purified CIC-L and CIC-S. Mean relative levels (siRNA control [Lamin A/C siRNA]=100%) and standard error are shown (n=4, *P=0.056, **P<0.006).

Figure 6. Loss of wild-type Atxn1 function worsens SCA1 neuropathology in mice

(a,b) Atxn1 154Q/− animals showed a worsened rotarod performance (*P<0.05, **P<0.005) and earlier lethality (P<1×10⁻⁶) than Atxn1 154Q/+ mice. (c) Model for SCA1 neuropathology. In wild-type individuals, there are at least two distinct and mutually exclusive ATXN1-associated endogenous protein complexes in vivo. The formation of one of these complexes (ATXN1/RBM17) appears to be regulated in that it requires phosphorylation of ATXN1. Polyglutamine expansion in ATXN1 favors formation of the RBM17-containing complex, thereby enhancing one endogenous function and contributing to neuropathology via a gain-of-function mechanism. Polyglutamine expansion concomitantly decreases the formation of CIC-containing complex, resulting in a partial loss of function.