This study investigates the nuclear basket, a peripheral structure of the nuclear pore complex (NPC), using in-cell cryo-electron tomography and integrative structural modeling across species. It reveals that a hub of nucleoporins (Nups) anchors the nuclear basket to a double nuclear ring, with Mlps/Tprs forming struts and distal densities that potentially dock mRNA. The findings suggest a 20-nm exclusion zone around the basket, indicating its role in chromatin organization. Variations in the stoichiometry of the outer rings across species highlight structural adaptability, providing insights into the basket’s roles in mRNA surveillance and chromatin dynamics during nucleocytoplasmic transport.

This study reveals an unexpected enzymatic function of the scaffold protein 14-3-3ζ, traditionally known for phosphoprotein binding. Using computational modeling and structural comparison approaches, we mapped two ATP-binding sites within 14-3-3ζ: a catalytic site in the amphipathic groove and a second site at the dimer interface. Our analysis predicted key catalytic residues (E131 and E180), which were experimentally validated as essential for ATP hydrolysis. We further uncovered a dual allosteric mechanism in which ATP binding at the dimer interface enhances ATPase activity at the catalytic groove while selectively inhibiting the binding of certain non-phosphorylated ligands such as ExoS. Together, these findings redefine 14-3-3ζ as not only a phosphoprotein scaffold but also an allosterically regulated ATPase, expanding its functional landscape in cellular signaling.

This study investigates the molecular rules that dictate the strength and orientation (parallel or antiparallel) of coiled-coil helices in protein-protein interactions. The research highlights the role of hydrophobic core densities in stabilizing specific dimer conformations and their importance in structural specificity. A computational platform, COiled-COil aNalysis UTility (COCONUT), was developed and validated for predictive modeling of coiled-coil structures. COCONUT successfully predicted substitution-sensitive orientations, identified residue pairings in non-canonical heterodimers, and constructed multi-stranded coiled-coil models, showcasing its utility for analyzing and designing coiled-coil-based interactions and assemblies.

This article reviews computational methods for predicting 3D structures of protein interactions, categorized into template-based modeling, protein-protein docking, and hybrid/integrative modeling. Advances in sampling techniques, such as fast Fourier transforms and spherical harmonics, have enhanced the accuracy of docking methods, while scoring binding free energy remains a key challenge. Hybrid modeling, which integrates experimental data like electron microscopy with computational approaches, has made significant strides in resolving large and complex protein assemblies. Despite progress, modeling conformational flexibility remains a limitation, but ongoing developments suggest the potential to model molecular-level structures of entire organelles or cells in the future.