Eukaryotic cells store their DNA in the nucleus, cordoned off from the cytoplasm by the nuclear envelope. Made up of two lipid bilayers called the inner and outer nuclear membrane, the nuclear envelope protects DNA from damage by reactive by-products and intermediates of cellular metabolism. It also serves as a critical regulator of gene expression, restricting access to the genome and dictating which transcripts can exit the nucleus. This regulatory responsibility ultimately belongs to the thousands of massive molecular machines that penetrate both nuclear membranes to form gateways between the nucleus and cytoplasm.
Each nuclear pore complex (NPC) consists of more than 1,000 individual protein subunits with a total molecular mass of approximately 120 million daltons—the equivalent of more than 6.5 million water molecules. The size and complexity of the NPC has prevented any single technique or experiment from revealing its entire structure in detail. Over the last decade, however, technological advances have spurred an explosion in data on the structure of the NPC. Improvements in electron microscopy have led to snapshots of NPCs in their native environment at moderate, but improving resolution. Simultaneously, X-ray crystallographic analyses of the individual protein subunits and their interactions with one another at the atomic level have given us high-resolution details of how the pieces fit together. And this year, the powerful combination of these two approaches finally revealed the structure of the NPC’s symmetric core.1,2 Continued interrogation of the structure and function of the NPC will open the door to a deeper understanding of one of the cell’s most important machines.
conserved in eukaryotes, suggesting that NPCs emerged in a proto-eukaryotic ancestor as it first developed nuclei.
The impressive size of NPCs arises as a consequence of extensive symmetry. Early electron micrographs showed that NPCs possessed eightfold rotational symmetry around the central transport channel. In addition, the cytoplasmic and nuclear halves of the interior core of the NPC are identical but rotated 180 degrees relative to each other. (See illustration here.) The interior core is therefore known as the symmetric core.
The symmetric core is composed of three major ring structures: an inner ring that surrounds the central transport channel and two outer rings that sit on either side of the nuclear envelope. Inside the nucleus, nucleoporins specific to the nuclear side of the NPC attach to the symmetric core and form a nuclear basket, which interfaces with chromatin and the transcription machinery. On the cytoplasmic side, a different set of nucleoporins form flexible structures known as the cytoplasmic filaments, which are involved in facilitating transport through the NPC.
Each nuclear pore complex consists of more than 1,000 individual protein subunits with a total molecular mass of approximately 120?million daltons—the equivalent of more than 6.5 million water molecules.
This year, we and our colleagues at the California Institute of Technology published our work detailing the molecular architecture of the symmetric core.1 In the same issue of Science, an international team led by Martin Beck and colleagues at the European Molecular Biology Laboratory (EMBL) in Heidelberg, Germany, described its own efforts to elucidate the NPC’s structure.2 Electron microscopy reconstructions done by Beck’s group and high-resolution crystal structures determined by our group and others provided the frame and jigsaw pieces, respectively. But to confidently piece together the structural puzzle, we also needed a detailed understanding of the biochemistry of nucleoporins and the network of protein-protein interactions that link them. Beck and colleagues used crosslinking reagents to create covalent bonds between protein atoms that sit directly adjacent to each other in the NPC, and they performed mass spectrometry to identify the crosslinked nucleoporins. Our group tackled the problem by purifying the individual nucleoporins and building up the components of the NPC piece by piece to understand the rules of how they assembled. As a result of these complementary approaches, we now know the relative orientation and position of essentially all 17 nucleoporins that constitute the symmetric core in the human NPC.
One critical role of this architecture is stabilizing the extreme curvature of the nuclear membranes in the nuclear pores. In order to form a pore through the nuclear envelope, the nuclear envelope’s inner and outer lipid bilayers must fuse together in a U-shape. Because lipid bilayers prefer to be flat at microscopic scales, the membrane curvature needs to be stabilized by protein complexes, the most well-studied of which exist in the membrane coats that encapsulate small vesicles involved in protein trafficking between intracellular membranous organelles. Atomic-resolution crystal structures of nucleoporins have revealed similarities to complexes found in vesicle coats, including proteins such as Sec13 that are present in both. And, just like proteins in vesicle coats, the majority of nucleoporins are not transmembrane proteins and do not make contact with the nuclear envelope.
Taken together, the structural details of the NPC provide us with a much better understanding of the molecules and forces that hold such a crucial machine together and how it is able to operate a portal between the nucleus and cytoplasm. Importantly, we are now able to design targeted studies to probe how the NPC interacts with other cellular machineries and processes.
Molecular transport through the NPC
In contrast to transport into the endoplasmic reticulum or the mitochondrion, which requires that proteins be unfolded to pass through the organelle’s membrane translocation machinery, macromolecules are transported through the NPC in their native state. Proteins called karyopherins recognize specific molecules and also bind to the FG repeats, allowing them to shuffle the oversized molecules across the nuclear envelope.6 The karyopherins are a large protein family, including at least 19 known members in humans and 14 in budding yeast. Each family member recognizes its own set of cargoes, which can include proteins, tRNAs, and even pre-ribosomal subunits. Most karyopherins specialize in either import or export, though some karyopherins can mediate both.
The best-characterized method of cargo recognition involves short amino acid motifs known as nuclear localization signals (NLSs). First identified in viral proteins, NLSs are also used by cellular proteins. The presence of an NLS in a protein is a strong predictor for nuclear localization, and synthetic sequences are now commonly used to ensure the nuclear localization of engineered proteins. Cargoes can also be linked to karyopherins through adaptor proteins, and many macromolecules are recognized by their three-dimensional folds. Thus, while several canonical pathways for regulating nuclear import or export are now well understood, the specific mechanisms that control the localization of many proteins remain unknown.
Scientists do understand what dictates the directionality of transport, however. For most proteins and some small RNAs that depend on karyopherin-mediated transport, their release on the appropriate side of the NPC is determined by what is known as the Ran gradient. Ran is a small GTPase protein that binds and hydrolyzes GTP to GDP and adopts different conformations in its GTP- and GDP-bound states. The Ran-activating protein RanGAP is localized to the cytoplasmic filaments on the outside of the NPC. In contrast, the guanine exchange factor called RCC1, which facilitates the exchange of GDP for GTP in Ran, localizes to chromatin in the nucleus. As a result, Ran is activated almost exclusively in the cytoplasm, and thus exists there in the GDP-bound form (RanGDP), while it remains in its GTP-bound conformation (RanGTP) in the nucleus. These two conformations interact with karyopherins to regulate the uptake and release of nuclear pore cargo.
Unlike many channels and transporters, the 100-nanometer-wide NPC does not open and close to regulate transport.
When a karyopherin shuttles cargo from the cytoplasm into the nucleus, it encounters RanGTP, which binds to the karyopherins in a manner that is mutually exclusive with cargo binding. Because of the high concentrations of RanGTP, imported cargoes are released into the nucleus. In contrast, export complexes require RanGTP for their assembly. But, the complex of karyopherin, cargo, and RanGTP is in a strained conformation, as if spring-loaded. After exiting the nucleus, export karyopherins encounter RanGAP, which stimulates Ran’s GTPase activity. The bound GTP is hydrolyzed into GDP, causing Ran to adopt a conformation incompatible with karyopherin binding, which in turn triggers the karyopherin to release its cargo into the cytoplasm.
In contrast to molecules that are transported by karyopherins, the export of mRNAs occurs independently of the Ran gradient. Prior to export, mRNAs undergo several processing steps and are loaded with many proteins to form export-competent messenger ribonucleoproteins (mRNPs). One set of added proteins is a transport factor composed of the proteins Nxf1 and Nxt1. Similar to karyopherins, Nxf1/Nxt1 bind to FG repeats and pass through the diffusion barrier, shuttling the mRNP through the NPC. Nucleoporins on the cytoplasmic side of the pore recruit and activate an ATPase to remove Nxf1/Nxt1, freeing the mRNA to be translated by the ribosome.
Many viruses have developed ways to subvert regulated nucleocytoplasmic transport, either by interfering with bulk mRNA export or by developing novel ways to hijack existing transport pathways. HIV-1 genomes, for example, hijack karyopherin-mediated transport by encoding a specific RNA sequence that is recognized by the HIV-1–encoded protein Rev, which serves as an adaptor to karyopherins. Together, the genomic RNA and Rev are exported to the cytoplasm like any other karyopherin-dependent cargo. An alternative strategy is exploited by some simian retroviruses, which encode a specific RNA element that is recognized by Nxf1/Nxt1, so that their RNA genomes are exported through the canonical mRNA export pathway. Both approaches help the virus bypass the regulatory steps in the nucleus that normally ensure only the correct mRNAs are exported from the nucleus.
Other roles and links to disease
Many nucleoporins and factors involved in nucleocytoplasmic transport also appear to have functions outside the nucleus. For example, the nucleoporin Sec13 is an essential component of vesicle coats and also a component of the amino acid–sensing TORC1 pathway. And, components of the Ran cycle appear to play a role in cell division. RCC1, the Ran guanine-exchange factor, remains localized to chromatin during mitosis, maintaining a higher concentration of RanGTP near chromosomes as microtubules segregate chromosomes into two daughter cells. Several proteins that regulate microtubules are recognized by karyopherins and kept in an inactive state, but are released by the karyopherins near chromosomes because of the elevated levels of RanGTP. Thus, Ran provides a signal for spatially controlled activation of microtubule formation near chromosomes. Many nucleoporins are also localized at important mitotic structures, suggesting that the transport machinery may have an even larger role in the progression of the cell cycle.
Given the NPC’s role in ensuring the flow genetic information from the nucleus to the translational machinery in the cytoplasm and in protecting the genome, and given the many additional functions of nucleoporins in the cell, it is not surprising that a wide variety of human diseases have been associated with the dysfunction of the NPC or its component parts. Several cancers involve genomic rearrangements that have resulted in nucleoporin genes fused to other genes, for example.8 And several mutations in nucleoporins have been linked to heritable diseases that result in lethal developmental defects.9
Researchers have also recently linked NPC function to neurodegenerative diseases including amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Recent animal studies found that nucleoporins are extremely long-lived proteins in nondividing cells such as neurons, with half-lives on the order of months.10 Thus, unlike most cellular proteins, damaged NPC proteins are not replenished by protein turnover in these cells, but instead would accumulate over time, meaning aging NPCs may have an important role in human disease. Taken together, it is clear that building on our current understanding of NPC structure and function could lead to novel ways to combat a wide array of human diseases.
Daniel H. Lin is a biochemistry and molecular biophysics graduate student at the California Institute of Technology in Pasadena. His advisor, André Hoelz, is a professor of chemistry at the California Institute of Technology, a Heritage Principal Investigator, and a Faculty Scholar of the Howard Hughes Medical Institute.
- D.H. Lin et al., “Architecture of the symmetric core of the nuclear pore,” Science, 352:aaf1015, 2016.
- J. Kosinski et al., “Molecular architecture of the inner ring scaffold of the human nuclear pore complex,” Science, 352:363-65, 2016.
- H.G. Callan, S.G. Tomlin, “Experimental studies on amphibian oocyte nuclei. I. Investigation of the structure of the nuclear membrane by means of the electron microscope,” Proc R Soc Lond B Biol Sci, 137:367-78, 1950.
- M.L. Watson, “Further observations on the nuclear envelope of the animal cell,” J Biophys Biochem Cytol, 6:147-56, 1959.
- A. Hoelz et al., “The structure of the nuclear pore complex,” Annu Rev Biochem, 80:613-43, 2011.
- A. Cook et al., “Structural biology of nucleocytoplasmic transport,” Annu Rev Biochem, 76;647-71, 2007.
- A. Ibarra, M. W. Hetzer, “Nuclear pore proteins and the control of genome functions,” Genes Dev, 29:337-49, 2015.
- A. Kohler, E. Hurt, “Gene regulation by nucleoporins and links to cancer,” Mol Cell, 38:6-15, 2010.
- H.O. Nousiainen et al., “Mutations in mRNA export mediator GLE1 result in a fetal motoneuron disease,” Nat Genet, 40:155-57, 2008.
- J.N. Savas et al., “Extremely long-lived nuclear pore proteins in the rat brain,” Science, 335:942, 2012.