Numerous diseases are linked to mutations in the actin-binding domains (ABDs) of conserved cytoskeletal proteins, including β-III-spectrin, α-actinin, filamin, and dystrophin. A β-III-spectrin ABD mutation (L253P) linked to spinocerebellar ataxia type 5 (SCA5) causes a dramatic increase in actin binding. Reducing actin binding of L253P is thus a potential therapeutic approach for SCA5 pathogenesis. Here, we validate a high-throughput screening (HTS) assay to discover potential disrupters of the interaction between the mutant β-III-spectrin ABD and actin in live cells. This assay monitors FRET between fluorescent proteins fused to the mutant ABD and the actin-binding peptide Lifeact, in HEK293-6E cells. Using a specific and high-affinity actin-binding tool compound, swinholide A, we demonstrate HTS compatibility with an excellent Z’-factor of 0.67 ± 0.03. Screening a library of 1280 pharmacologically active compounds in 1536-well plates to determine assay robustness, we demonstrate high reproducibility across plates and across days. We identified nine Hits that reduced FRET between Lifeact and ABD. Four of those Hits were found to reduce Lifeact cosedimentation with actin, thus establishing the potential of our assay for detection of actin-binding modulators. Concurrent to our primary FRET assay, we also developed a high-throughput compatible counter screen to remove undesirable FRET Hits. Using the FRET Hits, we show that our counter screen is sensitive to undesirable compounds that cause cell toxicity or ABD aggregation. Overall, our FRET-based HTS platform sets the stage to screen large compound libraries for modulators of β-III-spectrin, or disease-linked spectrin-related proteins, for therapeutic development.

Recent structural studies of β-III-spectrin and related cytoskeletal proteins revealed N-terminal sequences that directly bind actin. These sequences are variable in structure, and immediately precede a conserved actin-binding domain composed of tandem calponin homology domains (CH1 and CH2). Here we investigated in Drosophila the significance of the β-spectrin N-terminus, and explored its functional interaction with a CH2-localized L253P mutation that underlies the neurodegenerative disease spinocerebellar ataxia type 5 (SCA5). We report that pan-neuronal expression of an N-terminally truncated β-spectrin fails to rescue lethality resulting from a β-spectrin loss-of-function allele, indicating that the N-terminus is essential to β-spectrin function in vivo. Significantly, N-terminal truncation rescues neurotoxicity and defects in dendritic arborization caused by L253P. In vitro studies show that N-terminal truncation eliminates L253P-induced high-affinity actin binding, providing a mechanistic basis for rescue. These data suggest that N-terminal sequences may be useful therapeutic targets for small molecule modulation of the aberrant actin binding associated with SCA5 β-spectrin and spectrin-related disease proteins.

No abstract

Spinocerebellar ataxia type 5 (SCA5) is a neurodegenerative disease caused by mutations in the SPTBN2 gene encoding the cytoskeletal protein β-III-spectrin. Previously, we demonstrated that a L253P missense mutation, localizing to the β-III-spectrin actin-binding domain (ABD), causes increased actin-binding affinity. Here we investigate the molecular consequences of nine additional ABD-localized, SCA5 missense mutations: V58M, K61E, T62I, K65E, F160C, D255G, T271I, Y272H, and H278R. We show that all of the mutations, similar to L253P, are positioned at or near the interface of the two calponin homology subdomains (CH1 and CH2) comprising the ABD. Using biochemical and biophysical approaches, we demonstrate that the mutant ABD proteins can attain a well-folded state. However, thermal denaturation studies show that all nine mutations are destabilizing, suggesting a structural disruption at the CH1-CH2 interface. Importantly, all nine mutations cause increased actin binding. The mutant actin-binding affinities vary greatly, and none of the nine mutations increase actin-binding affinity as much as L253P. ABD mutations causing high-affinity actin binding, with the notable exception of L253P, appear to be associated with early age of symptom onset. Altogether, the data indicate increased actin-binding affinity is a shared molecular consequence of numerous SCA5 mutations, which has important therapeutic implications.

β-III-spectrin is a key cytoskeletal protein that localizes to the soma and dendrites of cerebellar Purkinje cells, and is required for dendritic arborization and signaling. A spinocerebellar ataxia type 5 (SCA5) L253P mutation in the cytoskeletal protein β-III-spectrin causes high-affinity actin binding. Previously we reported a cell-based fluorescence assay for identification of small molecule actin-binding modulators of the L253P mutant β-III-spectrin. Here we describe a complementary, in vitro, fluorescence resonance energy transfer (FRET) assay that uses purified L253P β-III-spectrin actin-binding domain (ABD) and F-actin. To validate the assay, we screened a 2,684-compound library of FDA-approved drugs. Importantly, the screening identified numerous compounds that decreased FRET between fluorescently labeled L253P ABD and F-actin. The activity and target of multiple Hit compounds were confirmed in orthologous co-sedimentation actin-binding assays. Through future medicinal chemistry, the Hit compounds can potentially be developed into a SCA5-specific therapeutic. Furthermore, our validated FRET-based in vitro HTS platform is poised for screening large compound libraries for β-III-spectrin ABD modulators.