You have full access to this article via Jozef Stefan Institute.

Eukaryotic genomes are segregated from the cytoplasm by the nuclear envelope, thus necessitating a robust molecular exchange across this barrier that is essential for life. Indeed, it has been estimated that a kilogram of molecules transits across the nuclear envelope every second in the human body. What is perhaps most remarkable is that this bulk flow of material is not indiscriminate; its specificity is critical for the discrete segregation of the transcriptional and translational machineries and the biochemical makeup and function of the nucleoplasm and cytoplasm. Central to achieving this compartmentalization is the nuclear transport machinery, which comprises shuttling nuclear transport receptors (NTRs, also known as karyopherins, importins, exportins and biportins), a physical and biochemical gradient of the GTPase Ran (bound to guanosine triphosphate (GTP) or guanosine diphosphate), and nuclear pore complexes (NPCs) (Fig. 1 and Box 1).

NPCs provide the physical channel for molecular exchange between the nucleoplasm and cytoplasm. They are enormous ~100-megadalton (MDa) protein assemblies that are embedded into non-randomly distributed nuclear pores across a given nucleus (Fig. 1 and Box 1). NPCs are massive out of necessity: they must allow a bulk exchange of myriad molecules while accommodating the transport of some of the largest macromolecular complexes in a cell, which transit NPCs in intact, folded states. For example, NPCs transport several MDa-scale assemblies, including ribosomal subunits, proteasomes and messenger mRNA-containing ribonucleoprotein particles (mRNPs). NPCs also provide access to viral capsids as large as HIV-1 (65-nm wide) to enter the nucleus. It is remarkable that they can achieve the translocation of these large and chemically distinct macromolecules while also maintaining an effective diffusion barrier to others smaller than ~5 nm (~40 kDa). Indeed, NPCs may even help control the flow of water across the nuclear envelope (Fig. 1).

Our understanding of the structure of NPCs has evolved over the past few decades, driven by advances in imaging technology, most notably in cryo-electron microscopy (cryo-EM) and cryo-electron tomography (cryo-ET). Indeed, the highest-resolution cryo-EM and cryo-ET maps of yeast, Xenopus and human NPCs allow for the confident placement of crystal structures and AI-generated atomic models of virtually all the 'scaffold' nucleoporins (nups). In addition, structures of the asymmetric 'peripheral' aspects of the NPC, the cytoplasmic filaments and nuclear basket that extend into the cytoplasm and nucleoplasm, respectively, have also recently been revealed. Indeed, visualizing more true-to-life snapshots of in cellulo NPCs by cryo-ET of focused-ion-beam milled lamella from frozen cells to preserve their in vivo state revealed that NPCs are not all the same. Instead, NPCs exhibit structural and compositional plasticity both within and among individual cells, in different cell types and across organisms (Fig. 2). The functional importance of this plasticity is poorly understood, although it is interesting to consider that it could impact the selective transport properties of the NPC. Here we will discuss the current state of our understanding of structural and compositional changes to NPCs. Based on these findings, we propose a central role for cellular mechanics in driving NPC plasticity and suggest a rationale for tissue-specific diseases related to nup dysfunction.