The insertion of fusion proteins has enabled the crystallization of a wide range of G-protein-coupled receptors. Here, we adapted this engineering strategy to cryo-electron microscopy (cryo-EM). We inserted the soluble protein AmpC (3-lactamase into the third intracellular loop (ICL3) of ultra-thermostable (31-adrenoceptor ((31AR) via chimeric helix fusions. Biochemical and biophysical characterization showed that the resulting fusion protein after expression, solubilization and purification was monodisperse and able to bind the known (31AR weak partial agonist cyanopindolol, and the antagonist propranolol. The protein particles comprised sufficient mass and discernable structural features to elucidate its cryo-EM structure in complex with cyanopindolol without any natural (G-proteins, arrestins) or artificial (Nanobodies, DARPins) binding partners, to an overall resolution of 4.2 & Aring;. The seven-helix architecture and helix eight, as well as both GPCR - AmpC (3-lactamase connections are clearly resolved. (31AR is in an inactive-like conformation. 3D variability analysis revealed significant flexibility between the two protein domains and within the GPCR helices. The map contains clear density for the cyanopindolol. The fusion protein geometry is expected to be compatible with a subset of other class A GPCRs exhibiting suitable architecture. For receptors meeting these geometric requirements, this approach may facilitate cryo-EM structure determination of GPCR-ligand complexes in an inactive-like state. In addition, it could support structural studies of GPCRs in the absence of ligands.
The mitochondrial processing peptidase (MPP) is an essential heterodimeric protease responsible for cleavage of N-terminal mitochondrial targeting sequences (N-MTSs) from most nuclear-encoded mitochondrial proteins following import into the mitochondrial matrix. Pathogenic variants in the α-subunit of MPP (PMPCA) cause a spectrum of mitochondrial disorders, including spinocerebellar ataxia type 2 (SCAR2), encephalopathy, and severe neurodevelopmental disease; however, the molecular mechanisms underlying these phenotypes remain poorly understood. Here, we combined biochemical and structural approaches to characterize five disease-associated PMPCA variants (S96L, R185W, R268W, A377T, and G515R). Enzymatic analyses revealed variant-specific defects in substrate processing, with most mutations severely impairing cleavage of the rapidly processed substrate malate dehydrogenase 2. In contrast, several variants exhibited only modest impairment in the processing of the slowly cleaved substrate PINK1. Mass photometry demonstrated reduced heterodimerization of all variants, whereas circular dichroism identified varying degrees of structural destabilization, particularly for S96L and G515R. To define the structural basis of these defects, hydrogen–deuterium exchange mass spectrometry was performed on wild type MPP and selected variants. Pathogenic mutations altered substrate-induced conformational changes within previously characterized substrate-recognition elements, including the glycine-rich loop. Furthermore, we identified a previously unrecognized region, termed the INS loop, that exhibited substrate-dependent protection uniquely in the wild type enzyme but was consistently disrupted across pathogenic variants. Together, these findings demonstrate that PMPCA mutations impair MPP function through diverse mechanisms involving protein stability, dimerization, substrate selectivity, and allosteric conformational dynamics. Our work provides a mechanistic framework for understanding genotype–phenotype relationships in PMPCA-associated disease.
The tandem BRCT domains in the SMC5/6 complex localization factor 1 (SLF1, SLF1tBRCT) mediate interactions with a phosphorylated region in Rad18 and interaction targets the SMC5/6 complex to DNA damage sites for repair. Whether SLF1tBRCT mediate interactions with additional proteins is unclear. In this study, we searched for novel SLF1tBRCT binding partners and found that RNA polymerase II subunit 2 (RPB2) may contain a SLF1tBRCT binding site. The crystal structure of SLF1tBRCT in complex with a RPB2 peptide containing the predicted binding site revealed a SLF1tBRCT-RPB2 interface that bares similarity to the previously reported SLF1tBRCT-Rad18 interface but contains several unique features. Importantly, the SLF1tBRCT BRCT2 domain undergoes conformational changes to accommodate the C-terminus of the RPB2 peptide. Interestingly, this domain also mediates domain-swapping interactions between neighboring SLF1tBRCT molecules in the crystal. Our study suggests that RPB2 may be a potential SLF1 binding partner and provides insights into the structural flexibility of SLF1tBRCT.
