Studying nuclear protein import in yeastlab.rockefeller.edu/rout/assets/file/9_Leslie2006.pdfD.M. Leslie et al. / Methods 39 (2006) 291–308 293 oocytes, and also have a tough cell

Methods 39 (2006) 291–308

www.elsevier.com/locate/ymeth

Studying nuclear protein import in yeast

Deena M. Leslie a,1, Benjamin Timney b,1, Michael P. Rout b, John D. Aitchison a,¤

a Institute for Systems Biology, 1141 N 34th St., Seattle, WA 98103, USAb The Rockefeller University, 1230 York Ave., New York, NY 10021, USA

Accepted 15 July 2006

Abstract

The yeast Saccharomyces cerevisiae is a common model organism for biological discovery. It has become popularized primarilybecause it is biochemically and genetically amenable for many fundamental studies on eukaryotic cells. These features, as well as thedevelopment of a number of procedures and reagents for isolating protein complexes, and for following macromolecules in vivo, have alsofueled studies on nucleo-cytoplasmic transport in yeast. One limitation of using yeast to study transport has been the absence of a recon-stituted in vitro system that yields quantitative data. However, advances in microscopy and data analysis have recently enabled quantita-tive nuclear import studies, which, when coupled with the signiWcant advantages of yeast, promise to yield new fundamental insights intothe mechanisms of nucleo-cytoplasmic transport.© 2006 Elsevier Inc. All rights reserved.

Keywords: Nuclear transport; Karyopherins; Import; Yeast; Saccharomyces cerevisiae; QuantiWcation; Transport assay; Importin; Nucleus

1. Introduction

The hallmark feature of eukaryotic cells is the presenceof a nucleus, which is deWned by the nuclear envelope (NE).Operationally, the NE physically separates nuclear DNAfrom the cytoplasm; segregating the sites of gene transcrip-tion and ribosome biogenesis from the site of protein syn-thesis. This compartmentalization allows the cell to strictlycoordinate numerous key cellular processes, but it alsodemands that an astonishing number of proteins andRNAs move between the nucleus and cytoplasm. As aresult, eukaryotic cell survival is dependent upon bi-direc-tional nucleo-cytoplasmic transport pathways. To under-stand the mechanisms that drive nucleo-cytoplasmictransport pathways, and how they inXuence cell growth,numerous studies have focused on unraveling the web ofphysical interactions that facilitate transport [reviewed in1–3]. Since the late 1980s our knowledge of the molecularmachinery that drives this process and the roles individual

* Corresponding author.E-mail address: [email protected] (J.D. Aitchison).

1 These authors contributed equally to this study.

1046-2023/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.ymeth.2006.07.013

components play have increased in parallel with the devel-opment of experimental techniques that have permitted thegenetic identiWcation, biochemical puriWcation, and visuali-zation of these cellular factors. Collectively, these studieshave revealed that, operationally, nuclear transport path-ways can be divided into two phases: a stationary phase,comprised of the NE and the macromolecular protein com-plexes, termed nuclear pore complexes (or NPCs), that areembedded in it; and a soluble (or mobile) phase, whichincludes nuclear transport receptors, their regulators, andthe cargoes they translocate.

NPCs are central to all nucleo-cytoplasmic exchange asthey are the conduits through which all communicationbetween the nucleoplasm and cytoplasm occurs. NPCs arehighly organized, evolutionarily conserved macromolecularprotein assemblies »45–60 MDa in size, »100 nm in diame-ter [4,5], and composed of »30 distinct proteins, termednucleoporins or Nups [6,7]. Nups can be divided into threeclasses [reviewed in detail in [8–11]. Pore membrane pro-teins (or poms) anchor the NPC to the NE. Non-FG Nupsthat are thought to primarily provide the positioningscaVold for the third class of nucleoporins, the FG Nups.FG-Nups contain degenerate repeats of the dipeptide

mailto: [email protected]

mailto: [email protected]

292 D.M. Leslie et al. / Methods 39 (2006) 291–308

phenylalanine–glycine (FG) and, with a few exceptions, aresymmetrically distributed on both the cytoplasmic andnuclear faces of the NPC [6,7]. These FG repeat containingNups are thought to facilitate active transport by providingthe binding sites for transport complexes traversing theNPC and, thus, directing the rapid accumulation of pro-teins and other macromolecules in the nucleus orcytoplasm.

Signal-mediated nuclear transport is dependent on therecognition of signal sequences present in the cargo mole-cule. Proteins are marked for nuclear import and nuclearexport by the presence of nuclear localization signals(NLSs) [12] or nuclear export signals (NESs) [13,14],respectively. These targeting signals are recognized by solu-ble transport receptors termed karyopherins, or Kaps (theyare also known as importins, transportins, and exportins)[reviewed in 1, reviewed in 2]. Eukaryotic cells contain twostructurally related families of Kaps: the �-karyopherins(�-Kaps) and the �-karyopherins (�-Kaps). There are 14 �-Kaps in S. cerevisiae and more than 20 �-Kaps in highereukaryotes that can be divided into three categoriesdepending on the direction in which they transport cargo:import �-Kaps, export �-Kaps and �-Kaps that are capableof both importing and exporting cargoes [reviewed in 1,reviewed in 2]. In general, a given �-Kap recognizes andinteracts directly with its cargo and facilitates the cargo’stranslocation across the NE by interacting directly withFG-Nups [reviewed in 1,2, reviewed in [15–19]. The excep-tion to this trend comes from the Wrst import pathway char-acterized, the Kap �1/Kap � (Kap95p/Kap60p in yeast)transport pathway [20]. In this instance, the �-Kap/Kap60pacts as an adapter, bridging the interaction between thecargo and Kap �1/Kap95p, which mediates the movementof this trimeric import complex through the NPC.

The small GTPase Ran is also a key regulator of trans-port that provides directionality to nucleo-cytoplasmictransport, in part by regulating the formation of Kap-cargocomplexes [2,10,11, reviewed in [18,21–24]. In eukaryoticcells, Ran cycles between a GTP- and GDP-bound state.The cytoplasmic localization of RanGTPase-activatingprotein [25,26] and the nuclear localization of it’s guanosinenucleotide exchange factor (RanGEF) [27,28], generates agradient of RanGTP across the NE. This gradient isthought to control the association and dissociation of Kap-cargo complexes [24,29]. Accordingly, import complexesform in the cytoplasm where the concentration of RanGTPis low. Once on the nucleoplasmic face of the NPC, importcomplexes encounter an environment rich in RanGTP.Here, RanGTP binds the import �-Kap, stimulating Kap-cargo complex dissociation and terminating the importcycle. Conversely, export karyopherins bind their cargoescooperatively with Ran-GTP in the nucleus. These nuclearexport complexes dissociate once they reach the cytoplasm,where RanGAP induces GTP hydrolysis [reviewed in2,10,11, reviewed in 18,23,24].

Now that the roles of many of the soluble transport fac-tors have been deWned, the remaining questions about

nucleo-cytoplasmic transport are rather more subtle, con-cerning how exactly individual transport reactions proceed,how they determine transport rates and how they are regu-lated. Biochemical methods have generated much usefuldata using puriWed proteins with in vitro solution bindingassays [reviewed in [8–11]. Hypotheses based on these bio-chemical Wndings have been tested in in vitro nuclearimport assays. These assays permeabilize the plasma mem-brane of tissue culture cells with digitonin [30], then re-sup-ply the system with puriWed transport factors toreconstitute nucleo-cytoplasmic transport in a semi-cell-free environment. Using this technique, a number of the keysoluble cellular factors and metabolites required for nucleartransport in vitro were characterized [21,31–42]. Thisin vitro system has also been used to measure the importrates of transport factors with or without their cargoes[43,44], and has been coupled with single molecule micros-copy methods to measure the millisecond dwell times ofNTF2 [45] and Kap-cargo import complexes at the NPC[44]. In vitro techniques have also been employed to studyimport in Xenopus oocytes, either using intact nuclei [46] orby adhering nuclear envelopes over microscopic wells, suchthat import into these wells recorded the transport of singleNPCs [47–49]. These studies have provided informationabout the rates of individual transport reactions, but it isunclear whether these transport rates relate to the situationin vivo, where the import of a particular cargo likely com-petes with that of many other cargoes and transport path-ways. Microinjection of labeled cargoes into intact tissueculture cells [48,95,94] or Xenopus oocytes [98] have thusprovided more physiologically relevant import rates, thatoften diVer from their in vitro counterparts.

While the techniques described above were developed tostudy nucleo-cytoplasmic transport in metazoan cells,S. cerevisiae is the most genetically tractable nucleatedmodel organism, in which it is possible to make systematicalterations to components of its nuclear translocationmachinery. Many of the techniques available to performsuch manipulations in yeast are either completely unavail-able or prohibitively time-consuming in metazoan cells. Inyeast, proteins of interest can be expressed from eithergenomically integrated cassettes or from autonomouslyreplicating plasmids [50], often as Xuorescent fusion proteinchimeras (FPs) attached to the Green Fluorescent Protein(GFP) from Aequorea victoria, or one of its derivatives[99,100]. The distribution of these proteins can be analyzedin batteries of systematic knockout or temperature-sensi-tive (ts) strains. Directed hypotheses about nucleo-cytoplas-mic transport can, therefore, be tested by studying theeVects that these genomic modiWcations have on both cellviability and gross nuclear transport [reviewed in 1–3].Until recently however, our ability to study the more subtleaspects of nucleo-cytoplasmic transport in yeast wererestricted by the experimental tools available, which wererelatively limited compared to those available for use withmetazoan cells. This was because yeast cells are muchsmaller than mammalian tissue culture cells or Xenopus

D.M. Leslie et al. / Methods 39 (2006) 291–308 293

oocytes, and also have a tough cell wall. Thus attempts toadapt the in vitro nuclear import techniques previouslydescribed for metazoan cells to yeast cells have been onlymildly successful [101]. Furthermore, the study of nucleo-cytoplasmic transport of cargoes in an in vivo context canoften be complicated by the biological processes that thesecargoes are involved in and aVected by. The recent develop-ment of alternative nuclear import assays has addressedmany of these problems and, as a result, has opened thedoor to quantitatively studying nucleo-cytoplasmic trans-port in the most genetically tractable model organism.

Here, we will Wrst describe traditional methods used tostudy nucleo-cytoplasmic transport of full-length Xuores-cent fusion proteins in vivo. We will also highlight and focuson a technique that uses galactose-inducible expression ofXuorescent cargoes to characterize the molecular mecha-nisms that mediate the nuclear import of cellular factorsthat are bound and retained in the nucleus after transport[51]. To further study cargo-import, independent fromother biological processes involving the cargoes, we providean outline of how to deWne nuclear localization signalswithin proteins of interest. We then present three methodol-ogies that can be used to examine the import rates of Xuo-rescent reporter proteins bearing such NLSs. Each of thesemethods requires a diVerent investment of eVort to returnmeasurements of an increasingly quantitative nature. In theWrst case, relative import rates can be obtained from mea-surements of the NLS-GFP protein’s steady-state nuclearversus cytoplasmic concentration ratio, measured directlyfrom growing cells. The second technique can measuremore subtle changes in import by using a variety of meta-bolic energy poisons to inhibit receptor-mediated nucleartransport; this treatment prevents the formation ofRanGTP, which is essential to most signal-mediatedimport. Re-import of the cargo is then studied in the popu-lation of treated cells, after they have been relieved of thistransport-poison, thus providing relative measurements ofXuorescent cargo import rates [52]. The Wnal technique is ahigh-resolution quantitative single cell assay, whose centralmethodology is directly based on this poison-based assay,but involves substantial modiWcations to provide necessaryquantitative information to measure the import eYciencyof the pathway(s) importing a Xuorescent model cargo(Timney et al., manuscript in preparation).

2. Overview

The simplest measurement of import or export of acargo is to monitor its accumulation in the nucleus (N) rel-ative to the cytoplasm (C) or vice versa—typicallyexpressed as an N/C ratio. For example, when deWning the�-Kap that mediates the nuclear import of the cargo pro-tein of interest a fusion protein containing the cargo and areporter protein is expressed in wild-type and mutant yeastcells. If the mutant strain encodes a protein that is requiredfor, or enhances, the nuclear import of the cargo, then itwill be imported at a rate diVerent from that observed in

wild-type cells. As a result, the respective N/C distributionsobserved in these strains will proportionally reXect theseimport-rate diVerences. Fluorescence microscopy can beused to monitor these changes in subcellular localization.

Indirect immunoXuorescence and in vivo Xuorescencemicroscopy techniques could be used to accomplish thistask, of monitoring a protein’s subcellular distribution.However, in vivo studies are simpler and allow researchersto obtain rapid and direct results from complex experi-ments, which can include extended time course experi-ments, multiple treatment conditions, or both. As a result,in addition to being able to rapidly study numerous consti-tutive nuclear transport events, it is possible to study thedynamic aspects of the system. In vivo transport assays havealso paved the way for single cell studies.

2.1. Yeast mutant strains

The nuclear transport assays described here will workequally well regardless of whether yeast knockout or tem-perature-sensitive mutants strains are used. Deletion ortemperature-sensitive mutant strains have been constructedfor virtually all yeast Nups, as well as the �-Kaps, yeastRan (Gsp1p) and its regulators, providing an almost com-plete repertoire of mutant strains for all known yeastnuclear transport factors. Although most of the centralplayers in nucleo-cytoplasmic transport have been identi-Wed, the generation of yeast deletion and titratable pro-moter strain libraries has made it possible to rapidly screenfor additional factors involved in nucleo-cytoplasmic trans-port, the list of which is constantly growing [53,54]. TheSaccharomyces Genome Deletion Project contains a set ofyeast deletions strains, in which >95% of all yeast ORFs inthe Saccharomyces Genome Database were disrupted [55].In addition, Mnaimneh et al. [56] recently constructed alibrary of promoter-shutoV strains for over two-thirds of allessential yeast genes. This collection includes 800 strains,each with a kanR-tetO7-TATA cassette integrated in placeof the endogenous promoter driving the expression of anessential yeast gene. Thus, the expression of these genes canbe turned oV when the cells are cultured in the presence ofdoxycycline, allowing for the down regulation of the pro-moter until the gene of interest is no longer expressed atdetectable levels [56]. Both the Yeast Genome Deletion col-lection and the Tet-promoter library are commerciallyavailable from Open Biosystems (Huntsville, AL).

2.2. Construction of fusion proteins

The ability of these imaging experiments to reveal a con-nection between a nuclear transport factor (Nup, Kap, ortransport cofactor) and the transport pathway or individ-ual cargo in question is dependent on the experimenter’sability to visualize a change in the subcellular localizationof a Xuorescent reporter protein. These proteins are typi-cally constructed by fusing the gene encoding the protein inquestion in frame at its 3� end to a cassette encoding a


Xuorescent tag and a selectable marker. The resultingrecombinant DNA cassettes can then be expressed eitherexogenously from an autonomously replicating expressionvector or incorporated directly into the genome of thestrain(s) of interest via homologous recombination. Expres-sion of these FP gene chimeras can be optimized for Xuo-rescence microscopy analysis by placing them under thecontrol of diVerent regulatory elements [50]. For example,genomic integration of the FP chimeric gene or a combina-tion of CEN vectors and endogenous regulatory elementscan be used to closely replicate endogenous expression lev-els, while constitutive promoter elements (i.e., ADH1,PGK1, TPI1 or PRK1) can be used to increase synthesis ofthe protein chimera to optimal levels for visualization.Alternatively, inducible promoter elements (i.e., GAL1 orCUP1) can be used to speciWcally control expression of therecombinant gene, thus allowing their expression to beturned on and oV when desired [51,57]. However, it isimportant that initial experiments are performed underconditions where endogenous expression levels are repli-cated as closely as possible because altering the nativeexpression patterns of any cellular factor could produce anartiWcial environment that is not physiologically relevant.Regardless of the expression system used, all constructsshould be rigorously tested to ensure that the fusion pro-teins behave in the same manner as the untagged targetprotein.

While the brightness of an FP fusion protein is primarilydependent on its expression level, the intrinsic brightness ofthe FP tag, as well as the folding and translation eYciencyof the FP also inXuence its overall brightness and visibility.Folding eYciency of the FP is primarily dependent on envi-ronmental conditions [58,59], but it can also be inXuencedby the target protein. Therefore, bridging the intersectionbetween the target protein and the FP with a short linkerpeptide has been shown to enhance the maturation of thefusion protein [59]. In addition, translation eYciency isaVected by codon usage, therefore researchers have con-structed yeast-optimized FPs to improve the overall expres-sion of these tags in yeast [60,61]. Moreover, for imagingexperiments that require long exposures, one must alsoconsider the rate at which the FP photobleaches [59].

GFP is the most commonly used Xuorophore for theexperiments described here, but a number of GFP variantsare now available, including yeGFP [60], EGFP, EBFP,ECFP, EYFP (Clontech, Mountain View, CA), DsRed [62],hcRED [63], and RedStar [64]. Janke et al. recently con-structed a new toolbox of tagging vectors [65], whichincludes 37 new cassettes for C-terminal epitope taggingusing a variety of diVerent Xuorescent proteins in all colorsavailable, new selection markers, and a variety of epitopetags. There are also 37 N-terminal tagging cassettes, whichnot only facilitate the N-terminal tagging of the target butalso allow the replacement of the endogenous promoterwith an inducible one. Each of these GFP derivatives hasbeen successfully used in yeast and can theoretically be usedto monitor the localization of any yeast protein. The suit-

ability of each FP must be tested for each target proteinbeing analyzed.

2.3. Hints for troubleshooting

It is important to note that for most experiments, theabsolute brightness of the FP does not appear to be nearlyas important as reducing or combating the inherent auto-Xuorescence of yeast cells. Using an FP that Xuoresces in aregion of the spectrum with low yeast autoXuorescence,even if the FP is less bright than others, is often advanta-geous. But, this may not always be possible as the autoXuo-rescent spectra are generally broad and overlap with theemission wavelengths of GFP and many of its derivatives.Many cellular components, as well as components of theculturing medium, exhibit autoXuorescence [58]. It is there-fore important that the Wlter sets used in these experimentsare closely tailored to the targeted emission spectrum. Vari-ous software tools can also be used to remove the autoXuo-rescence signal and enhance FP visualization in theresulting images. Another point to consider is that someyeast strains also produce more autoXuorescent metabo-lites than others. For instance, the W303 strain can be con-siderably more autoXuorescent than strains such as DF5,particularly when approaching stationary growth or whengrown on plates. Finally, studies have demonstrated thatsome yeast media, such as yeast extract, emits a broadbandXuorescence that may combine with poor expression andweak emission of the FP to completely mask its signal [58].Therefore, for the imaging experiments described here,yeast cells are grown in complete synthetic media (CSM—0.17% YNB-AA/AS,2 0.5% (NH4)2SO4, CSM powderaccording to the manufacturer (Difco)) supplemented withglucose or an alternative carbon source. While commer-cially available yeast nitrogen base is most commonly used,some researchers have found that it can also be highly Xuo-rescent and prepare custom low Xuorescence YNB [66].

3. Subcellular localization studies of full-length import-cargoes in yeast

3.1. Microscopy

In vivo Xuorescence localization studies and changes inthe steady-state distributions of FP chimeras and nucleartransport factors as a result of a genetic or environmentalperturbation are often suYcient to qualitatively evaluatenucleo-cytoplasmic exchange. We have used both laser-scanning confocal microscopy (LSM 510 NLO; Carl Zeiss,Inc.) and conventional microscopy (Axioskop2 or Axiop-hot; Carl Zeiss, Inc.), and have found that both systemswork well. Oil immersion 63£ and 100£ lenses (numericalaperturesD1.4 or 1.25, respectively) have typically been

2 YNB-AA/AS, yeast nitrogen base minus amino acids and ammoniumsulfate.


used for visualization and image acquisition. The imagespresented here were typically captured using either a Spot(Diagnostic Instruments, Inc.) or a CoolSNAP (Photomet-rics) digital microscope camera. On average, at least 10images, from multiple areas of the slide, are taken at 63£magniWcation. Higher magniWcations can then be used toobtain images with higher spatial resolution. Prior to dataacquisition, the Xuorescent channel being imaged shouldalso be calibrated to reproducibly capture the Xuorescencesignal of the target protein in wild-type cells. This is accom-plished by visually scanning and acquiring images fromvarious regions of the slide, the detector gain and/or expo-sure times are then adjusted such that few pixels within theWeld of view reach saturation. While most cameras areequipped with an auto-expose function, it is possible toincrease the quality of the images by adjusting these param-eters manually. Once the exposure time and gain have beenoptimized, they must be maintained throughout the courseof the experiment.

3.2. Subcellular localization assays using constitutively or endogenously expressed FP fusion proteins

Initially wild-type strains are used to determine the natu-ral cellular localization of the cargo-protein in question andto deWne the cellular and/or environmental conditions thatnormally lead to its nuclear transport. These initial experi-ments should address whether the protein of interest consti-tutively localizes to the nucleus or cytoplasm with anapparently steady-state distribution, or transiently movesbetween these two compartments of the cell. When transientimport or export events are being studied, the conditions thatinduce translocation should also be deWned. Finally, whenmonitoring nuclear import, it is also important to establish ifthe cargo is also actively exported from the nucleus anddetermine which transport, binding or processing events aremost important in determining the steady-state or Xuctuatingcellular localization(s) of that protein. Once these conditionshave been deWned, simple side-by-side visual comparison ofthe images obtained from wild-type and mutant yeast strains,either knockout or ts mutants (Fig. 1), should reveal whetherthe nuclear transport factor encoded by the mutant allele isrequired to maintain the wild-type cellular distribution of theprotein of interest, be it a cargo molecule, karyopherin or sol-uble transport cofactor.

When ts mutants are being studied, starting cultures ofthe parental and mutant strains are grown at 23 °C and thesubcellular localization of the FP fusion protein determinedat the permissive temperature. Typically, at this point in theexperiment, the localization of the protein of interestshould be the same in both strains (Fig. 1). The cultures arethen shifted to the non-permissive temperature (37 °C) and,typically, re-examined after the shift. Cultures are imagedperiodically over several hours. The extent to which a cargomislocalizes varies, and is unique to each cargo and strainbeing studied (see Fig. 4C). These results may also varyfrom experiment to experiment depending on factors such

as culture density. Therefore, even subtle changes in locali-zation patterns can be indicative of a functional role incargo movement and additional experiments should be per-formed to precisely deWne its requirement.

3.3. Inducible nuclear import assays

The success of the subcellular localization assays onmutant strains described above are occasionally dependenton whether the biology of a given cargo allows its distribu-tion to perceptibly change, during the time course of such anexperiment. For instance, a protein that is not actively shut-tled into and out of the nucleus may not change its distribu-tion rapidly in response to a mutation in even its cognateKap. Also, many proteins are imported and localized to dis-tinct subnuclear domains, like the nucleolus, immediatelyfollowing translocation and this remains their Wnal destina-tion [51]. In these cases, it becomes diYcult to characterizethe nuclear transport pathway that mediates the import ofsuch cargoes, as they fail to accumulate in the cytoplasmupon �-Kap inactivation; this may be particularly problem-atic for some ts strains that can often only kept viable for afew hours at restrictive temperatures. Given that mostnuclear import cargoes display some degree of pleiotropicmislocalization under these conditions, one could be con-vinced that another member of the �-karyopherin proteinfamily is importing these cargoes. However, it is also possiblethe alternative mechanisms such as binding and retentionwithin a subnuclear domain tether these cargoes in thenucleus. To distinguish between these two possibilities, Leslieet al. [51] developed an inducible in vivo Xuorescence nuclearimport assay. This modiWcation was successfully used toidentify the transport receptor that mediates the import ofnumerous nucleolar proteins whose post-import nuclearretention made it technically impossible to characterize thesetransport mechanisms using conventional nuclear importassays (Fig. 2) [51].

Fig. 1. Subcellular localization studies. Ste12p was expressed as a GFPchimera in the indicated kap121-ts haploid cells, as well as kap121-34 cellsexpressing a wild-type copy of KAP121. Cultures were grown to mid-loga-rithmic phase at 23 °C and GFP detected using direct Xuorescence micros-copy. In each case, the GFP fusion localized predominantly to the nucleus.Each culture was then shifted to 37 °C for 3 h prior to microscopy analy-sis. Note that at 37 °C this chimera was primarily mislocalized to the cyto-plasm (images originally published in 82).


This assay is based on the assumption that if the insolublenuclear protein being studied is a bona Wde import cargo ofthe �-Kap in question, then inducing its expression after �-Kap inactivation will reveal whether this transport pathwayis required for its nuclear import. Furthermore, while thisassay has only been used when studying conditional yeastmutant strains, it could also be used to deWne primary andsecondary transport pathways when multiple �-Kaps trans-port one particular cargo protein. To this end, cargo-FPfusion genes are placed under the control of an inducible pro-moter (e.g., GAL1 promoter in the yeast 2�m expression vec-tor pYES2; Invitrogen, Carlsbad, CA) transformed into thestrains of interest and cultured at 23°C under non-inducingconditions (i.e., in media containing glucose or raYnose as acarbon source). To inactivate the temperature-sensitive �-Kap of interest, the cultures are then divided in half and incu-bated at permissive and restrictive temperatures (e.g., 23 or37°C for 2 h). Following inactivation, cells from each cultureare quickly harvested via centrifugation, washed with sterilewater and resuspended in preheated induction medium (e.g.,CSM supplemented with 2% galactose) and incubated at 23or 37°C for an additional 2–3 h to induce expression of thecargo-FP gene fusions. The subcellular localization of the FPfusion protein is then established using direct Xuorescencemicroscopy (Section 3.1). If the insoluble nuclear protein istransported by the �-Kap in question then it should accumu-late in the cytoplasm of the mutant cells when its expressionis induced at 37°C, but be transported to its appropriatenuclear or subnuclear domain at 23°C or when a wild-typecopy of the transport receptor is present (Fig. 2).

3.4. Controls

Before one can conclude that the transport factor beingstudied (be it a Nup, �-Kap or some other accessory fac-

tor), is required for a nuclear transport event to proceed inan eYcient manner a number of controls must be per-formed. In addition to parallel experiments being carriedout in parental and mutant strains, the localization of theFP fusion protein should also be assessed after a functionalcopy of the ts or knocked-out allele has been added back tothe mutant strain. If multiple ts strains are available, thelocalization of the fusion should be monitored in each ofthese strains. This type of control experiment can demon-strate whether the subcellular localization of the protein ofinterest is aVected by allele or strain-speciWc variations[67,68]. Controls assessing the speciWcity of the transportblock being studied should also be performed. To this end,the cellular localization of a cargo molecule that is nottransported by the import or export pathway being studiedshould also be monitored.


Many ts alleles are naturally sensitive to temperatureXuctuations that could potentially aVect the outcome of anexperiment, even if those changes are slight. Hence, waterbath incubators are preferable for ts cell growth, as they areless susceptible to temperature Xuctuations than air incuba-tors. Moreover, it is also important that the temperature ofthe cultures be maintained throughout the imaging step ofthe experiment. We have found that using a heated micro-scope stage helps maintain the desired culture temperature.

It should also be noted that growth at 37 °C is stressfulfor yeast cells. The heat shock response pathway is acti-vated to circumvent many of the secondary instabilitiesassociated with growth at elevated temperatures. As aresult, the targeting of many macromolecules to the nucleusceases under these conditions [69–73], while the import ofsome proteins is facilitated by heat shock protein expres-

Fig. 2. Inducible subcellular localization studies. The distribution of endogenously tagged NOP1 (eNop1p-GFP) and galactose-inducible NOP1GFP (GAL1-Nop1p-GFP) were monitored by direct Xuorescence microscopy in kap121-34 cells. NOP1 GFP;kap121-34 cells were visualized after growth to mid-logarith-mic phase at the permissive temperature (23 °C) and after 3 h of growth at the restrictive temperature (37 °C). Although the galactose-inducible NOP1GFPchimera was expressed in kap121-34 cells at 23 or 37 °C, eNop1p-GFP remains nucleolar in hap121-34 cells after 3 h of growth at 37 °C. However, GAL1-Nop1p-GFP accumulates in the cytoplasm of kap121-34 cells when expressed at the non-permissive temperature (images originally published in 51).


sion [52]. Moreover, ts mutants are by nature unhealthyand often grow poorly even at permissive temperatures. Itis advised that the cells be evaluated at a range of tempera-tures, paying particular attention to the morphology of thenucleus and the ability of cells to recover following theshift.

3.6. Limitations

As mentioned above, several members of the �-karyoph-erin protein family can recognize and import common car-goes [51,74–79]. These redundancies, which are collectivelyreferred to as nucleo-cytoplasmic transport networks, addan additional level of complexity to the characterization ofKap-cargo interactions and complicate the interpretationof the data acquired from the localization studies describedhere. These imaging experiments simply answer the ques-tion: Does a particular perturbation alter the subcellularlocalization of the protein in question relative to thatobserved in wild-type cells? Therefore, they are most easilyand accurately interpreted when dramatic mislocalizationphenotypes are observed. But when multiple �-Kaps trans-port an individual cargo molecule it is unlikely that loss ofa single transport pathway in the network will result in aprofound mislocalization phenotype. It is more probablethat the mutation will decrease the rate at which the cargois imported and that the magnitude of the mislocalizationwill directly reXect the signiWcance of the role that transportfactor plays in the import of that cargo molecule. Forexample, loss of function of the primary transport receptorof a cargo should result in a “strong” cargo mislocalizationphenotype, while loss of a secondary transporter is morelikely to result in more a subtle mislocalization phenotype[74,75,80]. This type of analysis can be used to assign a sig-niWcance order or rank to each transport pathway import-ing the common cargo. These hypotheses can then be testedusing the semi-quantitative and quantitative microscopymethodologies described below.

4. DeWning nuclear localization signals

The observed subcellular distributions of full-lengthcargo proteins, as discussed in the previous section, are fre-quently complicated by subsequent cellular processes inwhich those cargoes may be involved. This is a natural con-sequence of examining functional proteins in their in vivoenvironment. Ribosomal proteins are a good example ofthe complexity of the combined interactions that can giverise to the steady-state distribution of a full-length cargo.Ribosomal proteins are translated in the cytoplasm,imported into the nucleus and assembled into ribosomalsubunit precursors in the nucleolus. These subunits are thenexported back to the cytoplasm where they function[reviewed in 102]. Thus, the steady-state distribution ofeach ribosomal protein is a complicated combination ofcytoplasmic, nuclear and nucleolar localizations. Therefore,it is diYcult to study the import or export of a ribosomal

protein by examining its steady-state distribution as a full-length ribosomal protein-FP in wild-type or mutant strains.Experiments like those described in the previous section,using galactose-inducible expression of cargo molecules,can in some cases be used to overcome these technical diY-culties. However, it is often necessary, and informative, toidentify the NLS or NES-containing regions of the cargomolecule. Once identiWed, NLS-FP or NES-FP fusion pro-teins can be used to characterize the import or export path-way that is required for the nuclear import of the full-length protein independent of the other functional domainsof that protein.

4.1. DeWning candidate NLSs

NLSs and NESs are typically short peptides that do notWt a well-deWned consensus sequence and can be found any-where within the sequence of a protein. Nevertheless, stud-ies have demonstrated that these signals are generallyrecognized by speciWc Kaps [see examples in 1,81]. The lackof a simple, reliable primary sequence consensus has madeit diYcult to predict which amino acids will be required fortransport. As a result, the most straightforward way toidentify these signal sequences is to construct a series offragments of the original cargo molecule as FP fusions andmonitor their subcellular localizations in vivo, essentially aspreviously described (Section 3). We focus from this pointon the identiWcation of NLSs, although complementarymethods could be adapted to identify NESs.

A less than random process for the generation of puta-tive NLS-FP fusion proteins can be adopted if there is aknown or putative b-Kap for the import of the cargo inquestion. By applying the generalities in basicity andhydrophobicity of NLS sequences recognized by the puta-tive �-Kap, the number of cargo deletion fragments to bemade can be reduced, and the truncation of potential nucle-ophilic sequences avoided. Amino acid sequence alignmentsoftware packages, such as MegAlign (DNAStar, Inc.,Madison, WI) and ClustalW (http://www.ebi.ac.uk/clu-stalw/, European Bioinformatics institute, Cambridge,UK), can be used to identify regions that share some degreeof identity or similarity with previously characterized NLSsrecognized by the �-Kap [51,82,83]. For example, whencharacterizing the NLS of Sof1p, the full-length amino acidsequence of Sof1p was independently aligned with each ofthe previously identiWed lysine-rich NLSs of the putative �-Kap for this cargo, Kap121p, (Table 1) using MegAlign(Lipman-Pearson: kputle, 2; gap penalty, 4; gap length pen-alty, 12). Random sequence similarities and aligned NLSfeatures were discriminated by visually analyzing alignedsequences and identifying domains that could potentiallyfunction as NLSs (Fig. 3A). Using these Wndings as a tem-plate, four deletion fragments fused to GFP were generatedand the potential of each cargo fragment to be importedinto nuclei in vivo assessed in wild-type yeast cells (Fig. 3B).FP fusion proteins containing amino acids 381–489 ofSof1p localized predominantly to the nucleus while

http://www.ebi.ac.uk/clustalw/




fragments containing amino acids 1–380 diVusely localizedthroughout the cell (Fig. 3B), indicating that the NLS ofSof1p was likely localized between amino acids 381–489.

Other studies have also used consensus guided tech-niques to identify the NLSs in 5 proteins found to interactwith the ribosomal import-Kap, Kap123p (Timney et al.,manuscript in preparation), namely Mak16p, Nug1p,Yil096cp, and Yra1p. In this case, the authors developed alist of criteria to identify putative NLSs based on theapproximate ribosomal NLS consensus sequence previ-ously suggested [84]. The candidate NLS sequence criteriawere: 1, length more than 20 amino acids; 2, rich in lysineand arginine; 3, could include few acidic residues; and 4,were terminated at one or both ends by prolines or glycines.While this consensus can typically be found in several loca-tions within the sequences of larger proteins, in this case allthe proteins in question were reasonably small and, there-fore, there were only a few occurrences of this consensusthat were identiWed simply by visual observation. GFPfusion proteins containing all of the domains of these fourproteins that conformed to this consensus were then sub-cloned for expression in yeast and their localization moni-tored in vivo (Fig. 3C). For each of the cargoes assayed, onepotential NLS meeting the above criteria proved to facili-tate the nuclear accumulation of the GFP fusion proteinand was, therefore, deWned as the cargo’s putative NLS(Table 2). Indeed, independent research on YRA1 con-Wrmed the region selected by these criteria to be its singleNLS [103].

Such putative NLS-GFP fusion proteins can be found tolocalize to the nucleus for reasons other than being kap-dependent signal sequences, for example, by being able to

diVuse passively through NPCs and bind to nuclear targetsites. One can test whether the nuclear localizations of theseputative NLS-GFP cargoes are sensitive to the transportassay poison mixture (see Section 6.4), which destroys theRanGTP/GDP gradient and stops most Kap-dependenttransport. Remember though that even true NLS-GFP FPswill not rapidly redistribute in transport poison if they aresigniWcantly larger than the 40 kDa limit for rapid passivediVusion through NPCs. Experimenters can also testwhether the localization of putative NLS-GFP cargoes aredependent on a particular �-Kap, by looking at the distri-bution of those Xuorescent FPs in appropriate Kap dele-tion or ts strains as described in the previous section.


While application of a consensus-based approach canexpedite the identiWcation of NLSs, this process is naturallylimited by the small number of NLSs that have been char-acterized for each �-Kap and by the assumption that theputative �-Kap is of primary biological importance. Thus,for some proteins such an approach may preclude the dis-covery of NLSs recognized by unexpected �-Kaps or non-canonical NLS sequences. Thus, in some cases a more ran-dom, or shotgun approach, to creating cargo fragmentclones should be taken and the functionality of all possibleNLS sequences analyzed, regardless of whether they belongto the class that is typically recognized by the �-Kap inquestion. Following these initial identiWcations, moreextensive deletion analysis can be used to precisely deWnethe minimal amino acid sequence that is both necessary andsuYcient for NLS activity [for examples refer to 85,86].

Table 1NLSs recognized by Kap121p

The NLS sequences present of previously characterized Kap121p nuclear import cargoes are presented here. These amino acid residue sequences are writ-ten using the standard single letter abbreviations.

Cargo Amino acid residues

Sequence Reference

Pho4p 140–166 SANKVTKNKSNSSPYLNKRRGKPGPDS [96]Spo12p 76–143 KKSTSNLKSSHTTSNLVKKTMFKRDLLKQDPKRKLQLQQRFASPTDRLVSPCSLKLNEHKVK

MFGKKK[85]

Pdr1p 729-769 WTDMNKILLDFDNDYSVYRSFAHYSISCIILVSQAFSVAEF [97]Rpl25p 1–62 MAPSAKATAAKKAVVKGTNGKKALKVRTSATFRLPKTLKLARAPKYASKAVPHYNRLDSY

KV[77]

Yap1p 5–59 TAKRSLDVVSPGSLAEFEGSKSRHDEIENEHRRTGTRDGEDSEQPKKKGSKTSKK [80]Histone H2A 1–46 MSGGKGGKAGSAAKASQSRSAKAGLTFPVGRVHRLLRRGNYAQRIG [75]Histone H2B 1–52 MSAKAEKKPASKAPAEKKPAAKKTSTSTDGKKRSKARKETYSSYIYKVLKQT [75]Histone H3 1–28 MARTKQTARKSTGGKAPRKQLASKAARK [74]Histone H4 1–42 MSGRGKGGKGLGKGGAKRHRKILRDNIQGITKPAIRRLARRGG [74]Aft1p 198–226 and

332–365TSSIKPKKKRCVSRFNNCPFRVRATYSL and SKRPCLPSVNNTGSINTNNVRKPKSQCKNKDTLL

[86]

Ste12p 494–688 NNMLYPQTATSWNVLPPQAMQPAPTYVGRPYTPNYRSTPGSAMFPYMQSSNSMQWNTAVSPYSSRAPSTTAKNYPPSTFYSQNINQYPRRRTVGMKSSQGNVPTGNKQSVGKSAKISKPLHIKTSAYQKQYKINLETKARPSAGDEDSAHPDKNKEISMPTPDSNTLVVQSEEGGAHSLEVDTNRRSDKNLPDAT

[82]

Nop1p 1–90 MSFRPGSRGGSRGGSRGGFGGRGGSRGGARGGSRGGFGGRGGSRGGARGGSRGGFGGRGGSRGGARGGSRGGRGGAAGGARGGAKVVIEP

[51]

Sof1p 381–489 ERSNVKTTREKNKLEYDEKLKERFRHMPEIKRISRHRHVPQVIKKAQEIKNIELSSIKRREANERRTRKDMPYISERKKQIVGTVHKYEDSGRDRKRRKEDDKRDTQEK

[51]


5. Import rates of NLS-GFP cargoes can be estimated from their N/C ratios in growing cells

NLS-GFP reporter proteins as those deWned above areused as models to study the nuclear import of many cargomolecules. These model-cargoes are free of many of theother functional domains of a protein that may otherwisestrongly bind to target sites (such as for Sof1p) or cause itto be processed or regulated in ways that can confound theanalysis of its import as a full-length protein (such as for

Rpl25p). Second, full-length FP fusions are often signiW-cantly larger than the »40 kDa diVusion limit of the NPC,whereas NLS-GFP constructs are typically smaller thanthis. This means that NLS-GFP constructs can leak out ofthe nucleus within relatively short time-scales of importbeing perturbed (say by a speciWc ts mutation), whereas afull-length protein larger than 40 kDa may be unable toredistribute during a physiologically useful time scale [87].For these model-cargoes, the simplest evaluation of theirimport rates derives simply from a measurement of their

Fig. 3. DeWning nuclear localization signals (NLSs). (A) Schematic diagram of Sof1p. The NLS sequences of previously characterized Kap121p nuclearimport substrates (Table 1) were compared to full-length Sof1p by using MegAlign (Lipman-Pearson: ktuple, 2; gap penalty, 4; gap length penalty, 12). (B)Galactose-inducible GFP chimeric proteins containing the four fragments of Sof1p were synthesized at 23 °C in kap121-34 cells and analyzed by directXuorescence microscopy. Note the nuclear localization of Sof1p(aa381–489)-GFP and Sof1p(aa411–450)-GFP (images originally published in 51). (C)NLSs were identiWed in several proteins that bound Kap123p in biochemical assays, and were compared with the import of the NLS of the previouslydescribed Kap123p cargo Rpl25p [77]. Potential NLSs were located by visual inspection of each proteins sequence, compared with the proposed ribosomalNLS consensus [84]. Potential NLS sequences (Table 2) were then expressed as GFP fusion proteins from 2 � plasmids under the TPI1 promotor. Imageswere collected using spinning disk confocal microscopy, from which serial image sections have been combined as a maximum projection. These domainswere deWned as putative NLS if they localized to the nucleus in wild-type cells and if that localization was sensitive to metabolic energy poisons (notshown) (images are from Timney et al., manuscript in preparation).


accumulation in the nucleus (N), relative to the cytoplasm(C)—typically expressed as an N/C ratio.

NLS-GFP cargoes are imported by their cognate Kap(or Kaps) at a particular rate, but are also small enough toleak back out of the nucleus by simple diVusion throughNPCs. Thus, cells in culture have a distribution of theseNLS-GFP fusion-proteins that has reached a steady-state,with their rate of NLS-GFP import matching the rate ofNLS-GFP leak out of the nucleus. It can be shown that ifimport of cargo is a kinetically Wrst order process, with rateconstant k, and the leak of naked cargo is through passivediVusion, with NPC permeability constant p, then this N/Cratio corresponds simply to the relationship k/p + 1 [87]. If acargo is not actively imported, k is 0, therefore its steady-state N/C ratio will be 1 (i.e., evenly distributed), while twocargoes with similar passive NPC permeabilities (i.e., car-goes of similar sizes) that are imported at diVerent rates willproportionally reXect these rates in their respective N/Cvalues. Since subcellular Xuorescence is proportional to theconcentration of the Xuorophore, this measurement of N/Cratio of NLS-GFP cargoes can be made directly from back-ground-subtracted microscopy images; cells are simplygrown to logarithmic growth phase and imaged (see Sec-tions 3.1 and 6.3 for descriptions of how such images can beacquired). Most imaging software packages provide toolsto measure the pixel intensities within selected regions ofinterest, that can be created either by hand, or using soft-ware speciWc tools such as thresholding or edge detectionalgorithms. It is possible to measure cellular N/C ratiosfrom microscopy images using a variety of software pack-ages, including but not limited to, Metamorph (UniversalImaging Corporation), Imaris (Bitplane AG), Openlab(Improvision) and the custom software developed using theImage Analysis Toolpack of MatLab (The Mathworks).Each work equally well at capturing pixel intensities ofnuclei and cytoplasms, but vary in their subjective ease ofuse and their available tools, processing functions andautomatability. This last point is often important, sincemeasurements of N/C ratio need to be taken from manycells in a population to provide statistically reliable mea-surements.

This ratio is simple to derive, and can be obtained fromcells that have not been treated with any potentially harm-ful chemicals. Furthermore, it is underappreciated that thiscargo N/C ratio can provide researchers with important

information about the relative import rates of their cargoes,simply by comparing these ratios between diVerent strains.For example, a 5-fold decrease in the N/C ratio of a givenNLS-GFP cargo, resulting from say the deletion of a par-ticular �-Kap, probably indicates a 5-fold decrease in theimport-rate of that cargo. But, the relationship that deter-mines the steady-state N/C ratio of a cargo also includesthe term for the passive NPC permeability of that cargo, soanalysis of such experiments needs to be cautious. Forinstance, imagine comparing the N/C distribution of thesame cargo in two cell types whose NPC compositions werediVerent. We would be unable to distinguish whether achange in cargo N/C ratio had resulted from disparateimport capacities of the altered NPCs, or from a change inthe leakiness of these pores. Similarly, the absence of adetectable change in N/C ratio could either indicate thatthere was no functional diVerence between strains or thatthere were equal changes to both the active import and pas-sive permeability components of accumulation. Thus, N/Cratio can only be used as a quantitative import measurewhen the passive permeability of a cargo is already known,and can be used for relative comparisons only where cargopermeability can be assumed to be constant betweenconditions.

6. Metabolic-poisoning import-assays

6.1. Rationale

To measure active import, separate from passive diVu-sion, measurements of nuclear accumulation of cargomust be taken when the system is far from these steady-state N/C ratios. As discussed earlier, various kineticassays have been developed to monitor nuclear transportin particular metazoan cell types. Yeast cells are, how-ever, much smaller than mammalian tissue culture cellsor Xenopus oocytes. In addition, they have a tough cellwall. These obstacles have meant that the techniques usedto measure nucleo-cytoplasmic transport in metazoancells could not be readily adapted for use with yeast. AsigniWcant advance in measuring nuclear transport kinet-ics in yeast was made by Goldfarb and colleagues, whointroduced a method that uses metabolic energy poisonsto temporarily stop import of a NLS-GFP cargo, bydestroying the Ran gradient [52]. Poison-treated yeast

Table 2Putative NLS sequences of various Kap123p-binding proteins

a Manuscript in preparation.

Protein Function NLS sequence Reference

Mak16p Ribosome assembly NKNSAKRRKKGTSAKTKRPKV(aa 267–287) Timney et al. (2006)a

Nug1p Ribosome assembly MRVRKRQSRRTSTKLKEGIKKKASAHRKKEKKMAKKDVTWRSRSKKDPGIPSNFPYKAKIL(aa 1–61)

Timney et al. (2006)a

Yil096cp Ribosome assembly MARKLKGKIGSKGLKGALLRHKAKVKLVRNIESKQKHELRKKNSSANNKTVKRNQEFQKLNQGKVMP(aa 1–67)

Timney et al. (2006)a

Yra1p mRNA export GSNKAGSNRARVGGTRGNGPRRVGKQVGSQRRSLPNRRGPIRKNTRAPPNAVARVAK(aa 14–70)

[103] Timney et al. (2006)a


cells revive and restart transport shortly after the poisonmixture is removed; import can then be quantiWed byscoring the percentage of cells that, over time, have visu-ally recovered a signiWcant nuclear signal of the cargo(Fig. 4). This technique has been successfully used tomonitor the relative eVect upon import of genetic altera-tions that would have been infeasible in other organisms.Such studies have examined deletions of structurallyimportant Nups [87] and FG-domains [89], and have alsolooked for Xuctuations in import during phases of thecell cycle [88].

Facilitated by modern microscopy and data processingtechniques, existing yeast nuclear import assays [52] havebeen modiWed, introducing methods for rapid, semi-auto-mated cell-by-cell quantitation of import rates. The modi-Wed method provides more sensitive detection of subtleperturbations to the import machinery and allows for quan-titative import-rate measurements to be made in single yeastcell. As in the original method, transport is temporarilystopped with metabolic energy poisons, which destroy theRanGTP/GDP gradient [91]; re-import of NLS-GFP car-goes is measured after these poisons are removed (Fig. 6).

Fig. 4. Overview of nuclear import assay. Illustration of import in cells treated with metabolic energy poisons for an import assay. For explanation seeWgure annotations.


Two key improvements were made to data collection andanalysis so that import rates in single cells could be quanti-Wed and represented as units of cargo molecules/NPC/s.First, advanced automated microscopy was employed toobserve import in single cells over time, while 3-D recon-struction of confocal images was used to measure cellularmorphometry. Second, quantitative Western blotting wasused to calibrate Xuorescence measurements to subcellularconcentration of the NLS-GFP cargo. Import measure-ments are thus calibrated for cargo concentration and nor-malized for cell-to-cell variations in intracellular volume andNPC number. Because measurements of nuclear accumula-tion of cargo are taken when the N/C ratio is almost 1, netpassive Xux of cargo is close to zero, therefore these initialaccumulation rates are a direct measure of active import.

6.2. Plasmids and strains

NLS-GFP model cargoes are expressed constitutivelyunder the TPI1 promotor in the 2 � multi-copy pYX242shuttle plasmid (Novagen discontinued). Alternatively, anyconstitutive expression plasmid could be substituted.

For these assays the following are necessary: a strainwith Xuorescently tagged proteins that would mark boththe nuclear and cytoplasmic volumes for measurement; anda labeled protein of known abundance, to calibrate theabundance of the NLS-GFP cargoes to this internal stan-dard. Therefore, import assays are performed in a strainbackground containing a Fluorescent Protein (CFP)-tagged Htb2p (a nuclear histone protein) and Tpi1p (aubiquitous glycolysis enzyme), which mark the nucleus andcytoplasm, respectively, while Htb2-CFPp also serves as aninternal calibrant for NLS-GFP concentration.

6.3. Microscopy

Images are collected at room temperature (air condi-tioned to »23 °C) with a spinning disk confocal micro-scope [for e.g., Hamamatsu Orca ER cooled CCD cameraattached to a Zeiss Axiovert 200 microscope Wtted with aPerkin-Elmer UltraView spinning disk confocal imaginghead and using a 100£ objective lens (NA 1.5)]. CFP andGFP are excited with HeCad 442 nm or Argon 514 nmlasers, respectively, and separate images of each areobtained using a standard CFP/YFP dichroic with sepa-rate excitation/emission Wlter sets (Chroma). These CFP/YFP optics suYce to image NLS-GFP along with Htb2-CFPp and Tpi1-CFPp with no bleed through of Xuores-cence seen between GFP and CFP image channels. Thesystem described above is controlled with MetaMorphimaging software (Universal Imaging Corporation).Exposure settings are chosen during test runs, to maxi-mize use of the 12 bit gray scale of the detector, withoutsaturation, and these settings should be maintained for allsubsequent images of the strain. Control images of ablank slide without laser excitation are taken for back-ground subtraction of detector noise. Areas around the

edge of an image Weld can be considerably less bright thanthose in the center, but this can be compensated for byacquiring a control image of a Xuorescent glass slide(Chroma). This control image is normalized, so that itsmaximum pixel intensity is set to a value of 1. Thus, if aperipheral area of the Weld is 60% as bright as the centerthen the normalized control image would have values of»0.6 in this area; dividing all images of cell Welds by thiscontrol image will hence provide the appropriate compen-sation for such uneven Weld illumination. Some micros-copy setups, such as Applied Precisions DeltaVisionsystems, routinely perform this image correction automat-ically.

6.4. Metabolic energy poisons temporarily halt transport

Yeast cells are typically grown at 30 °C in 15 mL cul-tures to mid-log phase in selective minimal media. Foreach import assay a 1 mL sample of cell culture is har-vested and frozen for later NLS-GFP concentration cali-bration (see Section 6.8). Poison treatment and release isperformed essentially as described in the original method[52]. Cells are washed with water and resuspended in500 �L of the transport poison (SC media without a car-bon source, containing 10 mM 2-deoxyglucose and 10 mMsodium azide). With transport stopped, after 15 min atroom temperature the NLS-GFP cargoes will have equili-brated between the nucleus and cytoplasm—alternativemethods of slowing transport, such as using only deoxy-glucose or cold treating the cells [87], can be used, but arenot always eVective in fully equilibrating all import car-goes (see Section 6.11 for discussion). Over the course ofthe next hour samples are taken from this pool of poi-soned cells for import assays. After that hour the poi-soned cells are discarded and a new batch prepared.Despite this precaution import rates do not appear tochange even after up to 75 min of incubation in thepoison.

6.5. Original poison assay data collection and analysis

In the original application of this method, a sample ofpoisoned cells of 100–200 �L are washed several timeswith water, then the cells resuspended in a similar volumeof SC media containing glucose, to restart import of theNLS-GFP. From this pool of released cells, 3 �L samplesare taken at 5 min intervals for image acquisition usingstandard epiXuorescence microscopy, as described in Sec-tion 3.1. At least 30 cells from each time-point arecounted, with a particular cell scored as having a nuclearsignal if its nucleus is brighter than the surrounding cyto-plasm with a clear nuclear to cytoplasmic boundary. Thepercentages of cells at each time-point so judged to be“nuclear” are then plotted (Fig. 5), and the relative ratesof nuclear import are calculated from linear regressionlines drawn through the linear portions of the import timecourse.


These measurements of import can thus show relativechanges in import rates between strains on a populationlevel, such as is illustrated in Fig. 5 as an example of theother important studies over the past 10 years successfullyusing this assay [87–89]. However, the values obtainedusing this technique are not actual import rates such asthose that can be obtained from single metazoan cells,where importing cargo molecules are counted. Recentlythough, this assay has been modiWed in order to providequantitative measurements of cargo-import over time insingle-cells (Timney et al., 2006, manuscript in prepara-tion). While the original analytical method is more rapidand will still be of use in cases where relative, yet rapid,measurements are acceptable, these modiWcations cannow provide quantitative data about comparatively sub-tle changes in import rates.

For the modiWed import assay the pool of poisoned cellsare prepared as in Section 6.4 and each quantitative importassay is performed as follows.

6.6. Automated microscopy visualizes import in single cells

For each import assay, a 50–100 �L sample of the poi-soned cells is harvested, washed once with water, resus-pended in 10–20 �L of water to give an appropriatedensity of cells for microscopy. 1.5 �L of this cell suspen-sion are mixed on a poly-lysine coated microscope slidewith 1.5 �L of synthetic minimal media, containing glu-cose. Once released from poison re-import of NLS-GFPquickly restarts, so minimum time should lapse (less than60 s) before the commencement of image acquisition.Time points are collected at 15 s intervals using an auto-mated spinning disk confocal microscope (see Section6.3). At each time point images of the NLS-GFP and theCFP-tagged subcellular markers are taken; 3 each atfocal-plane separated by 1 �m increments up and downthrough the cells. Thus, focused images of most cells’nuclei are acquired at each time-point, in both the NLS-GFP and Htb2-CFPp/Tpi1-CFPp Xuorescence channels

Fig. 5. Metabolic-poisoning assay. (A) Intracellular localization of cNLS-GFP Xuorescence before and after treatment with sodium azide and deoxyglu-cose. Logarithmically growing wild-type cells expressing the cNLS-GFP fusion protein before (right panel), immediately following azide/deoxyglucosetreatment (middle panel), and 15 min after recovery in glucose-containing CSM medium. (B) cNLS-GFP import kinetics in SRP1 and srp1-31 cells at35 °C. srp1-31 (Wlled squares) and SRP1 (Wlled circles) cells were grown in CSM medium at 30 °C to an OD600 D 0.1–0.3. The cells were then treated with10 mM sodium azide and 10 mM deoxyglucose for 45 min at 30 °C. After equilibration, nuclear import kinetics were assayed at 35 °C (non-permissive tem-perature) over a 30 min time course as described in the text and plotted as a function of time (images were adapted from those originally published in [52]).


(Fig. 6A). After 10 min the NLS-GFP will have generallyreturned to its steady-state distribution in all cells.

6.7. Cell morphometry calculated from 3-D confocal image serial section

It is important to normalize for the diVerent sizes andshapes of the assayed cells. Hence, a complete confocalimage series should be acquired up and down through thecells, at focal-planes separated by 0.4 �m intervals (or theapproximate z-axis resolution limit of the objective lens),after the last timepoints of import have been recorded(Fig. 6B). The volume of a voxel is calculated by calibratingX and Y distances with a reference slide. Thus, nuclear andcytoplasmic volumes could be measured directly from the3-D confocal image series, by summing the number of vox-els in each. The area of the nuclear envelope is calculatedfrom a rendering of the nuclear surface, made by applyingthe MatLab (The Mathworks) isosurface locating algo-rithms to the 3-D data—similar rendering options areavailable in some commercial image analysis programs,

such as Imaris (Bitplane AG) and Volocity (Improvision).This surface area is converted to an estimate of the numberof NPCs using a Wgure of 12 NPCs/�m2 [104], a densityshown to be relatively constant throughout the yeast cellcycle. The morphological measurements taken from cells inthese import assays are consistent with the values measuredby Winey et al. [104]. Notably, this technique also yieldsaccurate measurements of commercially prepared sphericalXuorescent beads with known 2.5 �m diameters (MolecularProbes).

6.8. NLS-GFP cargo concentration calibrated using quantitative Western blots

The NLS-GFP concentration in a population of import-ing cells is calibrated using quantitative Western blotting,which compares the abundance of NLS-GFP to that of aninternal standard, Htb2-CFPp (Fig. 6C).

The average amount of Htb2-CFPp per cell in log-phasepopulations had been determined by quantitatively com-paring its signal from Western blots with that of a puriWed

Fig. 6. Flowchart of how nuclear import rates are measured in vivo for single yeast cells. (Data are from Timney et al., manuscript in preparation). Datafrom a single example cell importing the Nab2NLS-GFP model cargo [67,109] illustrate how measurements are taken and combined to measure import-rate. (A) Images of the import time course were analyzed to provide subcellular quantities of the NLS-GFP cargo in Xuorescence units. Htb2-CFPp andTpi1-CFPp images were used to demark the nuclear and cytoplasmic regions for measurement. Scale bars are 1 �m. (B) Nuclear and cytoplasmic volumes,as well as the surface area of the nuclear envelope, were derived from a three dimensional rendering of the yeast cell, using images from a complete confo-cal volume sectioning, collected subsequent to the time course. (C) A sample of cells from each assay was taken to calibrate the amount of NLS-GFP percell. This was done by quantitatively comparing the Western signal from NLS-GFP with that of an intracellular calibrant of known abundance, Htb2-CFPp. Thus, knowing the average abundance of NLS-GFP per cell, we were then able to convert individual cellular Xuorescence to number of cargo mol-ecules. (D) These measurements were combined to give a single-cell import curve in units of micromolar nuclear concentration of the NLS-GFP cargoover time. These data were Wtted to a simple single exponential curve, from which we derived a measure of initial import-rate in units of cargo molecules,per NPC, per second for each cell in each assay.


recombinant GFP standard [105], essentially as describedpreviously [6,106]. The monoclonal anti-GFP antibodyused for these experiments (Roche, Cat# 11 814 460 001)recognized GFP indistinguishable from CFP or YFP (notshown). Concentration of recombinant GFP has beendetermined by the Coomassie-based Bradford method [107]and amino acid analysis of triplicate hydrolyses [108]. Thecell-sample and standard are mixed together and run in thesame lanes to eliminate liquid handling uncertainties(150,000 cell equivalents gave a signal comparable to »1 ngof puriWed GFP). Western blots are quantitated usingOpenLab image analysis software (Improvision), by mea-suring the total pixel intensity of each band (above back-ground) in the linear detection range. The signal fromHtb2-CFPp is compared to the GFP standard calibrationcurve from the same blot, thus calculating an average num-ber of copies per cell for Htb2-CFPp. The population aver-age concentration of Htb2-CFPp, by this analysis, is240,000 copies per cell (§18% SD); a relatively constantvalue between strains, making it a good choice for a con-centration calibrant.

For each import assay, NLS-GFP and Htb2-CFPpbands on Western blots are quantitatively compared witheach other (as above; Fig. 6C), to calculate the abundanceof the cargo; for instance, a population with exactly twicethe NLS-GFP signal of Htb2-CFPp would thus have anaverage abundance of 480,000 cargo molecules per cell.Following the assays of each day, using the same micros-copy settings, we acquire 3-D confocal image series ofseveral Welds of unperturbed cells from the same cultureof cells used in the assays of that day. From these data wemeasure the total Xuorescence of NLS-GFP in >200 ofthese cells, by summing the values of all voxels for eachcell, which gave us a reliable measurement of the averagetotal NLS-GFP Xuorescence of our population ofimporting cells. Since the quantitative blotting yields theaverage NLS-GFP molar amount for cells in an importassay, and microscopy yields the average total Xuores-cence of those cells, and the total Xuorescence should beproportional with total Xuorophore quantity. Thereforedividing these two averages yields a calibration coeY-

cient that can be used to convert a given cell’s Xuores-cence to the abundance of NLS-GFP cargo in that cell.For example, a single cell whose total NLS-GFP Xuores-cence was twice that of the population average wouldhave twice the average abundance of NLS-GFP (or960,000 copies, using the example population frombefore).

6.9. Analysis provides measurement of imported molecules per NPC per second for single cells

For each condition, import assays should be repeateduntil import data from at least »100 cells have been col-lected. The CFP images of Htb2-CFPp and Tpi1-CFPpassist in segmenting each cell’s image-data into nuclear andcytoplasmic compartments, respectively, from which NLS-

GFP Xuorescence measurements are taken at each timepoint as described above.

Photobleaching can be minimized by choice of equip-ment and exposure conditions, but residual Xuorescenceloss and general signal Xuctuations can also be normalizedmathematically, since the total amount of NLS-GFP in acell should be constant over the timescale of an importassay. Total cellular Xuorescence is calculated at each time-point as the average Xuorescence of each compartment,multiplied by that compartment’s volume and summed forall compartments. The Xuorescence values of each time-point are then multiplied by the necessary factor such thatthe total cellular Xuorescence at each time-point is main-tained at a constant value. The constant value, to which thedata are normalized, is the maximum total cellular Xuores-cence calculated at any time-point during the assay for thatcell.

For each cell, the abundance of NLS-GFP therein isestimated, its distribution between the nucleus and cyto-plasm at each time point is measured, and the volumes ofthe nucleus and cytoplasm are calculated. Therefore, cal-culating the import rates in these cells is a simple matterof combining these measurements and calculating thenuclear NLS-GFP concentration at each point of the timecourse (Fig. 6). The resulting plots of NLS-GFP concen-tration over time Wt well to single exponential relation-ships (using MatLab’s Curve Fitting toolbox; average r-square of all single-cell Wts should be »0.95). Cells whosedata deviate signiWcantly from this relationship typicallyhave either aberrant morphology, very low Xuorescencesignal, or suVer from excessive movement, and exclusionfrom the analysis is warranted. Import rates can either becalculated from the exponential Wt coeYcients of thesecurves or from a linear regression of the initial, approxi-mately linear, time-points. Linear regression is preferredbecause this measure is less prone to the large Xuctuationsthat result from slight changes in the goodness of theexponential Wts. The import rates from these initial time-points are active import measurements for individualcells, free of passive diVusion, and are represented in unitsof cargo molecules/NPC/s.

Commercial image analysis software is typically inade-quate to collect and combine so much information from somany cells in a timely manner. But, import assays can beanalyzed using custom software created in MatLab (TheMathworks), developed speciWcally to collect, combine andanalyze these import data as described above. These pro-grams are freely available from the authors upon request,although some signiWcant understanding of MatLab will berequired for their adaptation to, and use on, a particularcomputer platform.

6.10. EVective import rate as a population measure of import, normalized for cargo concentration

Import rates measured for a several cells of a popula-tion are typically spread over wide ranges. This is because


the NLS-GFP expression constructs are cloned in multi-ple copy vectors, conferring a random number of copiesof the gene to each cell with a commensurate randomexpression level of the fusion protein. However, this vari-ability can also be used advantageously, by examiningthe relationship between cargo concentration andimport-rate. The data of all cell’s import-rates and cargoconcentration are split into a number of statistical bins ofincreasing cargo concentration, and the mean import-rate and cargo concentration within each bin can be cal-culated. A linear regression of these averages, Wxedthrough the origin, typically shows a strong linear trend,the slope of which is termed eVective import rate. Animport eYciency value is thus a measurement of howquickly a given cargo is imported by its available trans-port pathway(s), in units of cargo molecules/NPC/s/�Mof cargo.

6.11. Tips and caveats

To obtain reliable data from these assays one needs toexamine the import-rate of an NLS-GFP cargothat reached a high steady-state N/C ratio. Thus, for manysuch import assays, NLS-GFP fusion proteinshad attached to their C-termini a single copy of the 58amino acid Protein A (PrA) repeat sequence. This PrAfragment is not suYcient to bind antibodies in Westernblots (not shown), but simply increases the cargos’nuclear accumulation, presumably because of a decreasedrate of passive leak resulting from the cargoesincreased size, making the resulting data more quantiW-able.

Studies have demonstrated that sodium azide and 2-deoxyglucose have many eVects in the cell beyond theirability to reversibly halt nucleo-cytoplasmic transport inS. cerevisiae (Pan, X. and D.S. Goldfarb, unpublishedresults) [90,91]. For this reason alternative less invasivetechniques have been implemented to stop transport, suchas using only one of the poisons, or chilling the cells to 0 °[87], although we have found that these techniques aresometimes too gentle to cause complete redistribution ofsome import cargoes. Additionally, some researchers haveinhibited transport with various alcohols of increasinglyaliphatic character [92,93]. For now, transport poisonsremain the best tools we have to accurately quantifyin vivo import rates in the most genetically tractablenucleated model organism.

7. Concluding remarks

The development of new methods to study the dynam-ics of cellular processes continues at a fast-pace.Researchers can make directed and testable hypothesesabout molecular and cellular function by combiningthese techniques with the wealth of genetic tools availablefor yeast. We have described how some of these newmethods have been adapted to dissect the processes of

nucleo-cytoplasmic transport in a quantitative manner inliving yeast cells. As researchers, we are continually mon-itoring developing microscopic technologies, includingsuch tools as photoactivatable GFP and FRAP (Xuores-cence recovery after photobleaching), that may provideless invasive methods to study these processes in evengreater detail.

References

[1] H. Fried, U. Kutay, Cell Mol. Life Sci. 60 (8) (2003) 1659–1688.[2] N. Mosammaparast, L.F. Pemberton, Trends Cell Biol. 14 (10)

(2004) 547–556.[3] L.F. Pemberton, B.M. Paschal, TraYc 6 (3) (2005) 187–198.[4] C.W. Akey, M. Radermacher, J. Cell Biol. 122 (1) (1993) 1–19.[5] M.P. Rout, G. Blobel, J. Cell Biol. 123 (4) (1993) 771–783.[6] M.P. Rout, J.D. Aitchison, A. Suprapto, K. Hjertaas, Y. Zhao, B.T.

Chait, J. Cell Biol. 148 (4) (2000) 635–651.[7] J.M. Cronshaw, A.N. Krutchinsky, W. Zhang, B.T. Chait, M.J.

Matunis, J. Cell Biol. 158 (5) (2002) 915–927.[8] J. Bednenko, G. Cingolani, L. Gerace, TraYc 4 (3) (2003) 127–135.[9] M.P. Rout, J.D. Aitchison, J. Biol. Chem. 29 (2001) 29.

[10] M.P. Rout, J.D. Aitchison, M.O. Magnasco, B.T. Chait, Trends CellBiol. 13 (12) (2003) 622–628.

[11] M. Suntharalingam, S.R. Wente, Dev. Cell 4 (6) (2003) 775–789.[12] C. Dingwall, R.A. Laskey, Trends Biochem. Sci. 16 (12) (1991) 478–

481.[13] L. Gerace, Cell 82 (3) (1995) 341–344.[14] U. Fischer, J. Huber, W.C. Boelens, I.W. Mattaj, R. Luhrmann, Cell

82 (3) (1995) 475–483.[15] A.C. Strom, K. Weis, Genome Biol. 2 (6) (2001) 3008.[16] M.S. Moore, J. Biol. Chem. 273 (36) (1998) 22857–22860.[17] I.G. Macara, Microbiol. Mol. Biol. Rev. 65 (4) (2001) 570–594.[18] S. Kuersten, M. Ohno, I.W. Mattaj, Trends Cell Biol. 11 (12) (2001)

497–503.[19] Y.M. Chook, G. Blobel, Curr. Opin. Struct. Biol. 11 (6) (2001) 703–

715.[20] M. Rexach, G. Blobel, Cell 83 (5) (1995) 683–692.[21] F. Melchior, B. Paschal, J. Evans, L. Gerace, J. Cell Biol. 123 (6 Pt. 2)

(1993) 1649–1659.[22] M.S. Moore, G. Blobel, Nature 365 (6447) (1993) 661–663.[23] B.B. Quimby, M. Dasso, Curr. Opin. Cell Biol. 15 (3) (2003)

338–344.[24] K. Weis, Cell 112 (4) (2003) 441–451.[25] F.R. BischoV, C. Klebe, J. Kretschmer, A. Wittinghofer, H. Pon-

stingl, Proc. Natl. Acad. Sci. USA 91 (7) (1994) 2587–2591.[26] C. Klebe, F.R. BischoV, H. Ponstingl, A. Wittinghofer, Biochemistry

34 (2) (1995) 639–647.[27] D.C. Amberg, M. Fleischmann, I. Stagljar, C.N. Cole, M. Aebi,

EMBO J. 12 (1) (1993) 233–241.[28] M. Ohtsubo, H. Okazaki, T. Nishimoto, J. Cell Biol. 109 (4 Pt. 1)

(1989) 1389–1397.[29] P. Kalab, K. Weis, R. Heald, Science 295 (5564) (2002) 2452–2456.[30] S.A. Adam, R.S. Marr, L. Gerace, J. Cell Biol. 111 (3) (1990) 807–816.[31] C.W. Akey, D.S. Goldfarb, J. Cell Biol. 109 (3) (1989) 971–982.[32] A.H. Corbett, D.M. Koepp, G. Schlenstedt, M.S. Lee, A.K. Hopper,

P.A. Silver, J. Cell Biol. 130 (5) (1995) 1017–1026.[33] A.H. Corbett, P.A. Silver, J. Biol. Chem. 271 (31) (1996) 18477–

18484.[34] J. Flach, M. Bossie, J. Vogel, A. Corbett, T. Jinks, D.A. Willins, P.A.

Silver, Mol. Cell. Biol. 14 (12) (1994) 8399–8407.[35] U.F. Greber, L. Gerace, J. Cell Biol. 116 (1) (1992) 15–30.[36] U.F. Greber, L. Gerace, J. Cell Biol. 128 (1–2) (1995) 5–14.[37] N. Imamoto, T. Shimamoto, S. Kose, T. Takao, T. Tachibana, M.

Matsubae, T. Sekimoto, Y. Shimonishi, Y. Yoneda, FEBS Lett. 368(3) (1995) 415–419.


[38] N. Imamoto, T. Shimamoto, T. Takao, T. Tachibana, S. Kose, M.Matsubae, T. Sekimoto, Y. Shimonishi, Y. Yoneda, EMBO J. 14 (15)(1995) 3617–3626.

[39] J.F. Kalinich, M.G. Douglas, J. Biol. Chem. 264 (30) (1989)17979–17989.

[40] F. Melchior, D.J. Sweet, L. Gerace, Methods Enzymol. 257 (1995)279–291.

[41] D.D. Newmeyer, D.J. Forbes, J. Cell Biol. 110 (3) (1990) 547–557.[42] B.M. Paschal, L. Gerace, J. Cell Biol. 129 (4) (1995) 925–937.[43] K. Ribbeck, D. Gorlich, EMBO J. 20 (6) (2001) 1320–1330.[44] W. Yang, J. Gelles, S.M. Musser, Proc. Natl. Acad. Sci. USA 101 (35)

(2004) 12887–12892.[45] U. Kubitscheck, D. Grunwald, A. Hoekstra, D. Rohleder, T. Kues,

J.P. Siebrasse, R. Peters, J. Cell Biol. 168 (2) (2005) 233–243.[46] T. Radtke, D. Schmalz, E. Coutavas, T.M. Soliman, R. Peters, Proc.

Natl. Acad. Sci. USA 98 (5) (2001) 2407–2412.[47] O. Keminer, R. Peters, Biophys. J. 77 (1) (1999) 217–228.[48] O. Keminer, J.P. Siebrasse, K. Zerf, R. Peters, Proc. Natl. Acad. Sci.

USA 96 (21) (1999) 11842–11847.[49] N.I. Kiskin, J.P. Siebrasse, R. Peters, Biophys. J. 85 (4) (2003) 2311–

2322.[50] F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman,

J.A. Smith, K. Struhl (Eds.), Current Protocols in Molecular Biol-ogy, John Wiley & Sons, Hoboken, NJ, 1994.

[51] D.M. Leslie, W. Zhang, B.L. Timney, B.T. Chait, M.P. Rout,R.W. Wozniak, J.D. Aitchison, Mol. Cell. Biol. 24 (19) (2004)8487–8503.

[52] N. Shulga, P. Roberts, Z. Gu, L. Spitz, M.M. Tabb, M. Nomura, D.S.Goldfarb, J. Cell Biol. 135 (2) (1996) 329–339.

[53] N. Mosammaparast, B.C. Del Rosario, L.F. Pemberton, Mol. Cell.Biol. 25 (5) (2005) 1764–1778.

[54] N. Mosammaparast, C.S. Ewart, L.F. Pemberton, EMBO J. 21 (23)(2002) 6527–6538.

[55] E.A. Winzeler, D.D. Shoemaker, A. AstromoV, H. Liang, K.Anderson, B. Andre, R. Bangham, R. Benito, J.D. Boeke, H. Bus-sey, A.M. Chu, C. Connelly, K. Davis, F. Dietrich, S.W. Dow,M. ElBakkoury, F. Foury, S.H. Friend, E. Gentalen, G. Giaever,J.H. Hegemann, T. Jones, M. Laub, H. Liao, N. Liebundguth,D.J. Lockhart, A. Lucau-Danila, M. Lussier, N. M’Rabet, P.Menard, M. Mittmann, C. Pai, C. Rebischung, J.L. Revuelta, L.Riles, C.J. Roberts, P. Ross-MacDonald, B. Scherens, M. Snyder,S. Sookhai-Mahadeo, R.K. Storms, S. Veronneau, M. Voet, G.Volckaert, T.R. Ward, R. Wysocki, G.S. Yen, K. Yu, K. Zimmer-mann, P. Philippsen, M. Johnston, R.W. Davis, Science 285(5429) (1999) 901–906.

[56] S. Mnaimneh, A.P. Davierwala, J. Haynes, J. MoVat, W.T. Peng,W. Zhang, X. Yang, J. Pootoolal, G. Chua, A. Lopez, M. Tro-chesset, D. Morse, N.J. Krogan, S.L. Hiley, Z. Li, Q. Morris,J. Grigull, N. Mitsakakis, C.J. Roberts, J.F. Greenblatt, C.Boone, C.A. Kaiser, B.J. Andrews, T.R. Hughes, Cell 118 (1)(2004) 31–44.

[57] M. Marelli, C.P. Lusk, H. Chan, J.D. Aitchison, R.W. Wozniak, J.Cell Biol. 153 (4) (2001) 709–724.

[58] N. Billinton, A.W. Knight, Anal. Biochem. 291 (2) (2001) 175–197.[59] N.C. Shaner, P.A. Steinbach, R.Y. Tsien, Nat. Methods 2 (12) (2005)

905–909.[60] B.P. Cormack, G. Bertram, M. Egerton, N.A. Gow, S. Falkow, A.J.

Brown, Microbiology 143 (Pt 2) (1997) 303–311.[61] A. Crameri, E.A. Whitehorn, E. Tate, W.P. Stemmer, Nat. Biotech-

nol. 14 (3) (1996) 315–319.[62] M.V. Matz, A.F. Fradkov, Y.A. Labas, A.P. Savitsky, A.G. Zarai-

sky, M.L. Markelov, S.A. Lukyanov, Nat. Biotechnol. 17 (10)(1999) 969–973.

[63] N.G. Gurskaya, A.F. Fradkov, A. Terskikh, M.V. Matz, Y.A. Labas,V.I. Martynov, Y.G. Yanushevich, K.A. Lukyanov, S.A. Lukyanov,FEBS Lett. 507 (1) (2001) 16–20.

[64] M. Fischer, I. Haase, E. Simmeth, G. Gerisch, A. Muller-Taubenber-ger, FEBS Lett. 577 (1–2) (2004) 227–232.

[65] C. Janke, M.M. Magiera, N. Rathfelder, C. Taxis, S. Reber, H.Maekawa, A. Moreno-Borchart, G. Doenges, E. Schwob, E. Schie-bel, M. Knop, Yeast 21 (11) (2004) 947–962.

[66] M.A. SheV, K.S. Thorn, Yeast 21 (8) (2004) 661–670.[67] J.D. Aitchison, G. Blobel, M.P. Rout, Science 274 (5287) (1996)

624–627.[68] D. Muhlrad, R. Hunter, R. Parker, Yeast 8 (2) (1992) 79–82.[69] M. Furuta, S. Kose, M. Koike, T. Shimi, Y. Hiraoka, Y. Yoneda, T.

Haraguchi, N. Imamoto, Genes Cells 9 (5) (2004) 429–441.[70] P. Jorgensen, I. Rupes, J.R. Sharom, L. Schneper, J.R. Broach, M.

Tyers, Genes Dev. 18 (20) (2004) 2491–2505.[71] S. Kose, M. Furuta, M. Koike, Y. Yoneda, N. Imamoto, J. Cell Biol.

171 (1) (2005) 19–25.[72] Y. Miyamoto, T. Saiwaki, J. Yamashita, Y. Yasuda, I. Kotera, S. Shi-

bata, M. Shigeta, Y. Hiraoka, T. Haraguchi, Y. Yoneda, J. Cell Biol.165 (5) (2004) 617–623.

[73] U. Stochaj, R. Rassadi, J. Chiu, FASEB J. 14 (14) (2000) 2130–2132.[74] N. Mosammaparast, Y. Guo, J. Shabanowitz, D.F. Hunt, L.F. Pem-

berton, J. Biol. Chem. 277 (1) (2002) 862–868.[75] N. Mosammaparast, K.R. Jackson, Y. Guo, C.J. Brame, J. Shab-

anowitz, D.F. Hunt, L.F. Pemberton, J. Cell Biol. 153 (2) (2001)251–262.

[76] P. Muhlhausser, E.C. Muller, A. Otto, U. Kutay, EMBO Rep. 2 (8)(2001) 690–696.

[77] M.P. Rout, G. Blobel, J.D. Aitchison, Cell 89 (5) (1997) 715–725.[78] Y. Sydorskyy, D.J. Dilworth, E.C. Yi, D.R. Goodlett, R.W. Wozniak,

J.D. Aitchison, Mol. Cell. Biol. 23 (6) (2003) 2042–2054.[79] S. Jakel, D. Gorlich, EMBO J. 17 (15) (1998) 4491–4502.[80] T. Isoyama, A. Murayama, A. Nomoto, S. Kuge, J. Biol. Chem. 276

(24) (2001) 21863–21869.[81] B. Talcott, M.S. Moore, Trends Cell Biol. 9 (8) (1999) 312–318.[82] D.M. Leslie, B. Grill, M.P. Rout, R.W. Wozniak, J.D. Aitchison,

Mol. Cell. Biol. 22 (8) (2002) 2544–2555.[83] C.P. Lusk, T. Makhnevych, M. Marelli, J.D. Aitchison, R.W. Woz-

niak, J. Cell Biol. 159 (2) (2002) 267–278.[84] A.C. Timmers, R. Stuger, P.J. Schaap, J. van ’t Riet, H.A. Raue,

FEBS Lett. 452 (3) (1999) 335–340.[85] S.R. Chaves, G. Blobel, J. Biol. Chem. 276 (21) (2001) 17712–17717.[86] R. Ueta, A. Fukunaka, Y. Yamaguchi-Iwai, J. Biol. Chem. 278 (50)

(2003) 50120–50127.[87] N. Shulga, N. Mosammaparast, R. Wozniak, D.S. Goldfarb, J. Cell

Biol. 149 (5) (2000) 1027–1038.[88] T. Makhnevych, C.P. Lusk, A.M. Anderson, J.D. Aitchison, R.W.

Wozniak, Cell 115 (7) (2003) 813–823.[89] L.A. Strawn, T. Shen, N. Shulga, D.S. Goldfarb, S.R. Wente, Nat.

Cell Biol. 6 (3) (2004) 197–206.[90] A.A. Eddy, K. Backen, G. Watson, Biochem. J. 120 (4) (1970)

853–858.[91] E.D. Schwoebel, T.H. Ho, M.S. Moore, J. Cell Biol. 157 (6) (2002)

963–974.[92] P.W. Piper, FEMS Microbiol. Lett. 134 (2–3) (1995) 121–127.[93] N. Shulga, D.S. Goldfarb, Mol. Cell. Biol. 23 (2) (2003) 534–542.[94] G. Riddick, I.G. Macara, J. Cell Biol. 168 (7) (2005) 1027–1038.[95] A.E. Smith, B.M. Slepchenko, J.C. SchaV, L.M. Loew, I.G. Macara,

Science 295 (5554) (2002) 488–491.[96] A. KaVman, N.M. Rank, E.K. O’Shea, Genes Dev. 12 (17) (1998)

2673–2683.[97] A. Delahodde, R. Pandjaitan, M. Corral-Debrinski, C. Jacq, Mol.

Microbiol. 39 (2) (2001) 304–312.[98] D.S. Goldfarb, J. Gariepy, G. Schoolnik, R.D. Kornberg, Nature 322

(1986) 641–644.[99] M. ChalWe, Y. Tu, G. Euskirchen, W.W. Ward, D.C. Prasher, Science

263 (1994) 802–805.[100] A.B. Cubitt, R. Heim, S.R. Adams, A.E. Boyd, L.A. Gross, R.Y.

Tsien, Trends Biochem. Sci. 20 (1995) 448–455.[101] G. Schlenstedt, E. Hurt, V. Doye, P.A. Silver, J. Cell Biol. 123 (1993)

785–798.[102] J.R. Warner, Trends Biochem. Sci. 24 (1999) 437–440.


[103] D. Zenklusen, P. Vinciguerra, Y. Strahm, F. Stutz, Mol. Cell Biol. 21(2001) 4219–4232.

[104] M. Winey, D. Yarar, T.H. Giddings Jr., D.N. Mastronarde, Mol.Biol. Cell 8 (1997) 2119–2132.

[105] I.M. Cristea, R. Williams, B.T. Chait, M.P. Rout, Mol. Cell Proteo-mics 4 (2005) 1933–1941.

[106] F.R. Cross, V. Archambault, M. Miller, M. Klovstad, Mol. Biol. Cell13 (2002) 52–70.

[107] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.[108] J. Ozols, Methods Enzymol. 182 (1990) 587–601.[109] D.C. Lee, J.D. Aitchison, J. Biol. Chem 274 (1999) 29031–29037.

Documents

Studying nuclear protein import in yeastlab.rockefeller.edu/rout/assets/file/9_Leslie2006.pdfD.M. Leslie et al. / Methods 39 (2006) 291–308 293 oocytes, and also have a tough cell