Crystallographic home-source X-ray data for the atomic-resolution experimental phasing of the Shank3 SH3 domain structure from pseudomerohedrally twinned crystals.

By far most macromolecular crystallographic data collection and experimental phasing is nowadays carried out using synchrotron radiation. Here, we present two crystallographic datasets collected on a home-source X-ray diffractometer, which can per se be use to experimentally solve the atomic-resolution crystal structure of the Src homology 3(SH3)-like domain from the postsynaptic protein Shank3. The refined structure was described in the article "Structure of an unconventional SH3 domain from the postsynaptic density protein Shank3 at ultrahigh resolution" (Ponna et al., 2017) [1]. Crystals of the Shank3 SH3 domain were derivatized through soaking in 1 M sodium iodide prior to diffraction data collection at a wavelength of 1.54 Å. High-resolution data are reported for a native crystal to 1.01 Šand an iodide-derivatized one to 1.60 Å. The crystals suffered from several anomalies affecting experimental phasing: a high fraction (34-40%) of pseudomerohedral twinning, significant pseudotranslational symmetry (> 15%) with the operator 0.5,0,0.5, and a low solvent content. Twinning with the operator h,-k,-l is made possible by the space group P21 coupled with a unit cell ß angle of 90.0°. The data can be used to repeat and optimize derivatization and phasing procedures, to understand halide interactions with protein surfaces, to promote the use of home X-ray sources for protein structure determination, as well as for educational purposes and protocol development.

Structural basis for PDZ domain interactions in the post-synaptic density scaffolding protein Shank3.

The Shank proteins are crucial scaffolding elements of the post-synaptic density (PSD). One of the best-characterized domains in Shank is the PDZ domain, which binds to C-terminal segments of several other PSD proteins. We carried out a detailed structural analysis of Shank3 PDZ domain-peptide complexes, to understand determinants of binding affinity towards different ligand proteins. Ligand peptides from four different proteins were cocrystallized with the Shank3 PDZ domain, and binding affinities were determined calorimetrically. In addition to conserved class I interactions between the first and third C-terminal peptide residue and Shank3, side chain interactions of other residues in the peptide with the PDZ domain are important factors in defining affinity. Structural conservation suggests that the binding specificities of the PDZ domains from different Shanks are similar. Two conserved buried water molecules in PDZ domains may affect correct local folding of ligand recognition determinants. The solution structure of a tandem Shank3 construct containing the SH3 and PDZ domains showed that the two domains are close to each other, which could be of relevance, when recognizing and binding full target proteins. The SH3 domain did not affect the affinity of the PDZ domain towards short target peptides, and the schizophrenia-linked Shank3 mutation R536W in the linker between the domains had no effect on the structure or peptide interactions of the Shank3 SH3-PDZ unit. Our data show the spatial arrangement of two adjacent Shank domains and pinpoint affinity determinants for short PDZ domain ligands with limited sequence homology.

Structure of an unconventional SH3 domain from the postsynaptic density protein Shank3 at ultrahigh resolution.

The Shank family comprises three large multi-domain proteins playing central roles as protein scaffolds in the neuronal postsynaptic density. The Shank proteins are closely linked to neuropsychiatric diseases, such as autism spectrum disorders. One characteristic domain in the Shank family is the SH3 domain, assumed to play a role in protein-protein interactions; however, no specific ligand binding to any Shank SH3 domain has been described. We solved the crystal structure of the SH3 domain from Shank3 at sub-atomic resolution. While the structure presents the canonical SH3 domain fold, the binding site for proline-rich peptides is not conserved. In line with this, no binding of Pro-rich sequences by the Shank3 SH3 domain was observed. Sequence comparisons indicate that all Shank isoforms have similarly lost the classical Pro-rich peptide binding site from the SH3 domain. Whether the corresponding site in the Shank SH3 domains has evolved to bind a non-poly-Pro target sequence is currently not known. Our work provides an intriguing example of the evolution of a well-characterized protein-protein interaction domain within the context of multi-domain protein scaffolds, allowing the conservation of structural features, but losing canonical functional sites. The data are further discussed in light of known mutations in the SH3 domain or its vicinity in the different Shank isoforms.

Parkin cooperates with GDNF/RET signaling to prevent dopaminergic neuron degeneration.

Parkin and the glial cell line-derived neurotrophic factor (GDNF) receptor RET have both been independently linked to the dopaminergic neuron degeneration that underlies Parkinson's disease (PD). In the present study, we demonstrate that there is genetic crosstalk between parkin and the receptor tyrosine kinase RET in two different mouse models of PD. Mice lacking both parkin and RET exhibited accelerated dopaminergic cell and axonal loss compared with parkin-deficient animals, which showed none, and RET-deficient mice, in which we found moderate degeneration. Transgenic expression of parkin protected the dopaminergic systems of aged RET-deficient mice. Downregulation of either parkin or RET in neuronal cells impaired mitochondrial function and morphology. Parkin expression restored mitochondrial function in GDNF/RET-deficient cells, while GDNF stimulation rescued mitochondrial defects in parkin-deficient cells. In both cases, improved mitochondrial function was the result of activation of the prosurvival NF-κB pathway, which was mediated by RET through the phosphoinositide-3-kinase (PI3K) pathway. Taken together, these observations indicate that parkin and the RET signaling cascade converge to control mitochondrial integrity and thereby properly maintain substantia nigra pars compacta dopaminergic neurons and their innervation in the striatum. The demonstration of crosstalk between parkin and RET highlights the interplay in the protein network that is altered in PD and suggests potential therapeutic targets and strategies to treat PD.