Transportin regulates major mitotic assembly events: from spindle to nuclear pore assembly.

Mitosis in higher eukaryotes is marked by the sequential assembly of two massive structures: the mitotic spindle and the nucleus. Nuclear assembly itself requires the precise formation of both nuclear membranes and nuclear pore complexes. Previously, importin alpha/beta and RanGTP were shown to act as dueling regulators to ensure that these assembly processes occur only in the vicinity of the mitotic chromosomes. We now find that the distantly related karyopherin, transportin, negatively regulates nuclear envelope fusion and nuclear pore assembly in Xenopus egg extracts. We show that transportin-and importin beta-initiate their regulation as early as the first known step of nuclear pore assembly: recruitment of the critical pore-targeting nucleoporin ELYS/MEL-28 to chromatin. Indeed, each karyopherin can interact directly with ELYS. We further define the nucleoporin subunit targets for transportin and importin beta and find them to be largely the same: ELYS, the Nup107/160 complex, Nup53, and the FG nucleoporins. Equally importantly, we find that transportin negatively regulates mitotic spindle assembly. These negative regulatory events are counteracted by RanGTP. We conclude that the interplay of the two negative regulators, transportin and importin beta, along with the positive regulator RanGTP, allows precise choreography of multiple cell cycle assembly events.

Xenopus importin beta validates human importin beta as a cell cycle negative regulator.

BACKGROUND: Human importin beta has been used in all Xenopus laevis in vitro nuclear assembly and spindle assembly studies. This disconnect between species raised the question for us as to whether importin beta was an authentic negative regulator of cell cycle events, or a dominant negative regulator due to a difference between the human and Xenopus importin beta sequences. No Xenopus importin beta gene was yet identified at the time of those studies. Thus, we first cloned, identified, and tested the Xenopus importin beta gene to address this important mechanistic difference. If human importin beta is an authentic negative regulator then we would expect human and Xenopus importin beta to have identical negative regulatory effects on nuclear membrane fusion and pore assembly. If human importin beta acts instead as a dominant negative mutant inhibitor, we should then see no inhibitory effect when we added the Xenopus homologue. RESULTS: We found that Xenopus importin beta acts identically to its human counterpart. It negatively regulates both nuclear membrane fusion and pore assembly. Human importin beta inhibition was previously found to be reversible by Ran for mitotic spindle assembly and nuclear membrane fusion, but not nuclear pore assembly. During the present study, we observed that this differing reversibility varied depending on the presence or absence of a tag on importin beta. Indeed, when untagged importin beta, either human or Xenopus, was used, inhibition of nuclear pore assembly proved to be Ran-reversible. CONCLUSION: We conclude that importin beta, human or Xenopus, is an authentic negative regulator of nuclear assembly and, presumably, spindle assembly. A difference in the Ran sensitivity between tagged and untagged importin beta in pore assembly gives us mechanistic insight into nuclear pore formation.

Global histone acetylation induces functional genomic reorganization at mammalian nuclear pore complexes.

The nuclear localization of genes is intimately tied to their transcriptional status in Saccharomyces cerevisiae, with populations of both active and silent genes interacting with components of the nuclear envelope. We investigated the relationship between the mammalian nuclear pore and the human genome by generating high-resolution, chromosome-wide binding maps of human nucleoporin 93 (Nup93) in the presence and absence of a potent histone deacetylase inhibitor (HDACI). Here, we report extensive genomic reorganization with respect to the nuclear pore following HDACI treatment, including the recruitment of promoter regions, euchromatin-rich domains, and differentially expressed genes. In addition to biochemical mapping, we visually demonstrate the physical relocalization of several genomic loci with respect to the nuclear periphery. Our studies show that inhibiting HDACs leads to significant changes in genomic organization, recruiting regions of transcriptional regulation to mammalian nuclear pore complexes.

Topology of yeast Ndc1p: predictions for the human NDC1/NET3 homologue.

The nuclear pore complex is the predominant structure in the nuclear envelope that spans the double nuclear membranes of all eukaryotes. Yeasts have one additional organelle that is also embedded in the nuclear envelope: the spindle pole body, which functions as the microtubule organizing center. The only protein known to localize to and be important in the assembly of both of these yeast structures is the integral membrane protein, Ndc1p. However, no homologues of Ndc1p had been characterized in metazoa. Here, we identify and analyze NDC1 homologues that are conserved throughout evolution. We show that the overall topology of these homologues is conserved. Each contains six transmembrane segments in its N-terminal half and has a large soluble C-terminal half of approximately 300 amino acids. Charge distribution analysis infers that the N- and C-termini are exposed to the cytoplasm. Limited proteolysis of yeast Ndc1p in cellular membranes confirms the orientation of its C-terminus. Although it is not known whether vertebrate NDC1 protein localizes to nuclear pores like its yeast counterpart, the human homologue contains three FG repeats in the C-terminus, a feature of many nuclear pore proteins. Moreover, a small region containing mutations that affect assembly of the nuclear pore in yeast is highly conserved throughout evolution. Lastly, we bring together data from another study to demonstrate that the human homologue of NDC1 is the known inner nuclear membrane protein, NET3.

The periplasmic serine protease inhibitor ecotin protects bacteria against neutrophil elastase.

Ecotin is a dimeric periplasmic protein from Escherichia coli that has been shown to inhibit potently many trypsin-fold serine proteases of widely varying substrate specificity. To help elucidate the physiological function of ecotin, we examined the family of ecotin orthologues, which are present in a subset of Gram-negative bacteria. Phylogenetic analysis suggested that ecotin has an exogenous target, possibly neutrophil elastase. Recombinant protein was expressed and purified from E. coli, Yersinia pestis and Pseudomonas aeruginosa, all species that encounter the mammalian immune system, and also from the plant pathogen Pantoea citrea. Notably, the Pa. citrea variant inhibits neutrophil elastase 1000-fold less potently than the other orthologues. All four orthologues are dimeric proteins that potently inhibit (<10 pM) the pancreatic digestive proteases trypsin and chymotrypsin, while showing more variable inhibition (5 pM to 24 microM) of the blood proteases Factor Xa, thrombin and urokinase-type plasminogen activator. To test whether ecotin does, in fact, protect bacteria from neutrophil elastase, an ecotin-deficient strain was generated in E. coli. This strain is significantly more sensitive in cell-killing assays to human neutrophil elastase, which causes increased permeability of the outer membrane that persists even during renewed bacterial growth. Ecotin affects primarily the ability of E. coli to recover and grow following treatment with neutrophil elastase, rather than the actual rate of killing. This suggests that an important part of the antimicrobial mechanism of neutrophil elastase may be a periplasmic bacteriostatic effect of protease that has translocated across the damaged outer membrane.