Exploration of cargo spectrum and NES patterns recognized by the exportin CRM1

Abstract

The nuclear envelope (NE) subdivides eukaryotic cells into a nuclear and a cytoplasmic compartment, forcing material exchange between these two compartments to proceed through the nuclear pore complexes (NPCs). While proteins smaller than 30-40 kDa can passively diffuse through the NPCs, larger objects require nuclear transport receptors (NTRs) for efficient transport. NTRs have the privilege of facilitated NPC-passage; they bind transport cargoes and transfer them from one side of the NE to the other. NTRs can act as unidirectional cargo pumps, whereby they utilize the chemical potential of the nucleocytoplasmic RanGTP gradient with high nuclear and low cytoplasmic RanGTP levels. CRM1 is a major, essential and highly conserved nuclear export receptor. It exports a great variety of cargoes from the nucleus to the cytoplasm. CRM1 also keeps e.g. several translation factors and RanGAP cytoplasmic. The latter is required for maintaining the nucleocytoplasmic RanGTP gradient. CRM1 recognizes many cargoes through so-called leucine-rich nuclear export signal (NES), sequences containing 4-5 hydrophobic residues in a 14-15 residues long stretch. Although NESs are described in the context of primary protein structure, a reliable NES prediction has been a challenge and failed, e.g. for eIF2β and Rna1p (S.pombe RanGAP). Here we present a new NES prediction algorithm based on the recent crystal structures of different NES sequences with CRM1. We classified NES two PKI-type and REV-type with two different consensus definitions. PKI-type NES were graded for CRM1 binding strength and additional filtering was applied with disorder prediction. The REV-type NES was a novel classification based on Rev protein NES, and we show that there are several other examples of this type of NES. The estimation power of the new prediction algorithm was shown on prediction of already known NESs as control, and it also was able to predict the NESs of human eIF2β and S.pombe Rna1p, which was also confirmed experimentally. Another challenge had been the question of how many different cargo species are actually transported by CRM1. To address this, we optimized affinity chromatography on immobilized CRM1 and used it to retrieve RanGTP-dependent cargoes from a cytoplasmic HeLa extract. This analysis revealed hundreds of new CRM1 cargo candidates, which were further group into functional protein categories. Most of the ribosomal proteins are found in our dataset. Besides them, we find serine threonine kinases, ATP dependent helicases, spliceosomal proteins, translation initiation factors, actin regulators, and E3 ubiquitin ligases. Proteins of metabolic pathways, cell adhesion, phagosome, and proteasome are excluded from the data set.