Skip to main content
. 2021 May 4;10:e66165. doi: 10.7554/eLife.66165

Figure 1. Small-angle X-ray scattering from ASD-associated SHANK3 mutants shows changes in protein folding and topology.

(A) Size-exclusion chromatograms of Ni2+-IDA pre-purified His6-SUMO-SHANK3(1-676) variants are shown (elution peak at ~12.0 mL). The individual elution peaks were used for SAXS. (B) Rg values derived from Guinier approximation in the range of sRg <1.3 of SAXS profiles measured at different protein concentrations. The WT and R12C mutant show a linear increase of Rg with protein concentration, which suggests the presence of attractive interparticle interactions. (C) Kratky plots from zero-extrapolated SAXS profiles of the WT and R12C mutant resemble the profile of a folded multidomain protein with flexible linkers, while the L68P mutant appears to be partially unfolded. (D) Distance distribution curves were computed with GNOM using zero-extrapolated SAXS profiles as input data. Particles were assumed to be arbitrary monodisperse. The curves indicate a maximum particle diameter (Dmax) of ~14–15 nm for monomeric SHANK3(1-676) without significant alterations due to ASD-associated mutations. (E) Rigid-body CORAL models of the SHANK3 complex topology in solution. High-resolution structures of individual SHANK3 fragments were fitted against zero-extrapolated SAXS profiles without assuming any higher order symmetry (space group P1). The SPN and ARR domains were treated as a single rigid body (PDB 5G4X). The models indicate distal effects of both mutations on the position of the SH3 domain (orange) and PDZ domain (magenta) relative to the ARR domain (green) as well as changes in the orientation of linker regions. (F) CORAL models of monomeric His6-SUMO-SHANK3(1-676) variants with split SPN/ARR domains, where the linker region has been replaced by flexible dummy residues, capture mutation-induced perturbations of the SPN/ARR domain interface. Perspectives of visualized structures were chosen to facilitate highest visibility of structural regions of interest. All acquired SAXS data including fits and models were deposited to SASBDB (WT: SASDLJ3, R12C: SASDLL3 and L68P: SASDLK3). SAXS = small-angle X-ray scattering, ASD = autism spectrum disorders, Rg = radius of gyration, WT = wild type, SPN = SHANK/ProSAP N-terminal, ARR = ankyrin repeat. 

Figure 1.

Figure 1—figure supplement 1. Localization of ASD-associated mutation sites in the SHANK3(1-346) input topology of molecular dynamics simulations.

Figure 1—figure supplement 1.

(A) The two mutated residues R12 and L68 of SHANK3 are highlighted in red. Both residues are located within the N-terminal SPN domain of SHANK3. (B) The R12C mutation is shown to be localized within an antiparallel beta sheet directly at the SHANK3 N-terminus. (C) The L68P mutation is found in a locally disordered region within the SPN domain. In comparison, the non-mutated L68 residue is located in close vicinity to an antiparallel beta sheet as seen in A. The structures were generated as described in the 'Molecular dynamics' section and were visualized with VMD 1.9.3. ASD = autism spectrum disorders, SPN = SHANK/ProSAP N-terminal.
Figure 1—figure supplement 2. Schematic overview of protein purification steps involved in the preparation of His6-SUMO-SHANK3(1-676) and SHANK3(1-676) variants.

Figure 1—figure supplement 2.

(A) Expression of His6-SUMO-SHANK3(1-676) variants was induced with 300 μM IPTG in BL21(DE3) E. coli bacteria and continued for 18 hr at 18°C. Bacteria were lysed and proteins were pre-purified over Ni2+-IDA as described in the protein purification section of the materials and methods chapter. For SAXS, proteins were batch eluted from the Ni2+-IDA column (bed volume ~2 mL) with 200 mM Imidazole and were subsequently subjected to size-exclusion chromatography (SEC) using an analytical Superdex 75 10/300 GL column (0.4 mL/min; GE Healthcare). Peak fractions were pooled and concentrated to obtain sufficiently pure His6-SUMO-SHANK3(1-676) variants. To facilitate measurements over a broader concentration range, the His6-SUMO-tag was not removed for SAXS studies. (B) For complementary CD spectroscopy and nDSF measurements, the SUMO-tag was removed by treatment with the home-made SUMO protease His6-SenP2 (~0.3 mg/mL final concentration) followed by dilution to reduce the Imidazole concentration to approximately 30 mM. Removal of the His6-SUMO-tag as well as His6-SenP2 was done by two additional passages over a freshly packed Ni2+-IDA column (bed volume ~2 mL). (C) For buffer exchange, diluted proteins were initially concentrated by ultrafiltration and passed over a PD10 desalting column (GE Healthcare). Final protein re-concentration yielded SHANK3(1-676) variants in sufficient purity for subsequent measurements. SAXS = small-angle X-ray scattering. 
Figure 1—figure supplement 3. Guinier plots and pair distance distribution functions (PDDFs) from SAXS profiles measured at different concentrations.

Figure 1—figure supplement 3.

(A) Guinier plots show protein concentration-dependent changes in the slope of the Guinier fit, suggesting the presence of interparticle effects. A suitable Guinier range was detected by the AUTORG function of the ATSAS 3.0.1 software package. (B) PDDFs indicate a concentration-dependent increase in the maximum particle diameter (Dmax) which corroborates the idea of potential SAM domain independent self-interaction of SHANK3(1-676). All PDDFs were obtained from GNOM analysis using the ATSAS 3.0.1 software package. (C) Monomeric SHANK3 volume fractions were estimated by OLIGOMER analysis using the ATSAS 3.0.1 software package. Thereby, monomeric and dimeric SHANK3 models obtained from CORAL analyses were fitted against SAXS profiles measured at given concentrations of His6-SUMO-SHANK3(1-676) including zero-extrapolated SAXS profiles. SAXS data extrapolated to infinite dilution almost exclusively contains scattering contributions from monomeric His6-SUMO-SHANK3(1-676). However, the amount of monomeric His6-SUMO-SHANK3(1-676) contributing to the experimental scattering profiles decreases with increasing protein concentration, suggesting a dynamic and concentration-dependent monomer-dimer equilibrium. SAXS = small-angle X-ray scattering, WT = wild type.
Figure 1—figure supplement 4. CORAL-derived models of a dimeric SHANK3 protein complex topology in solution.

Figure 1—figure supplement 4.

For the CORAL input, experimental SAXS profiles have been merged between highest and lowest concentrations. (A) The CORAL fit for His6-SUMO-SHANK3(1-676) WT suggests that the protein could partially exist in a homo-dimeric form in solution, which is formed by a dual interface between SH3 and PDZ domains. The two large ankyrin-rich repeat domains (ARR) are oriented in a cis-conformation as is visible from the front and back view. (B) The ASD-associated R12C mutation is predicted to have no major impact on the complex topology of His6-SUMO-SHANK3(1-676). (C) The more dominant L68P mutation is predicted to result in an altered orientation of the ARR domains and the position of the SMT3-SPN-ARR domain cluster is altered with respect to the SH3-PDZ cluster. SAXS = small-angle X-ray scattering, WT = wild type, ASD = autism spectrum disorders.
Figure 1—figure supplement 5. Zero-extrapolated SAXS data fitted with CORAL to derive structural models of monomeric His6-SUMO-SHANK3(1-676) variants in solution.

Figure 1—figure supplement 5.

(A) Individual zero-extrapolated SAXS profiles were fitted with CORAL using high-resolution structures of individual SHANK3 domains. Here, the SPN and ARR domains were treated as a single rigid body (PDB 5G4X). (B) To capture potential mutation-induced disruptions of the SPN/ARR domain interface, the linker region (sequence KRRVYAQNLI) was removed from PDB 5G4X in silico and was replaced by the same number of dummy residues during CORAL fitting to allow flexible reorientation of the SPN domain. The corresponding models are shown in Figure 1E and F. SAXS = small-angle X-ray scattering, ARR = ankyrin repeat, SPN = SHANK/ProSAP N-terminal, WT = wild type.