Nucl ear Protei n Import Is Inh i b i ted by an Ant i bodyto a Lumenal Ep i tope of a Nucl ear Pore Comp l ex G l ycoprotei nUrs F Greber and Larry GeraceDepartments of Cel l and Mo l ecu l ar Bi o l ogy , The Scr i pps Research Inst i tute, La Jo l l a, Cal i forn i a 92037

Abstract . Gp210 i s a maj or transmembrane g l ycopro -tei n associ ated wi th the nucl ear pore comp l ex that i ssuggested to be i mportant for organ i z i ng pore comp l exarch i tecture and assemb l y. A mouse monocl onal IgGd i rected agai nst an ep i tope i n the l umenal domai n ofrat gp210 was expressed i n cu l tured rat cel l s bymi cro i n j ect i on of mRNA prepared f rom a hybr i domacel l l i ne. The expressed IgG , wh i ch becomes assem-b l ed i nto a funct i onal ant i body i n the l umen of the en -dop l asmi c ret i cu l um, bound to the nucl ear envel ope i nv i vo . Express i on of ant i - gp210 ant i body i n i nterphasecel l s speci f i cal l y reduced approx i matel y four fo l d themed i ated nucl ear i mport of a mi cro i n j ected nucl earprotei n (nucl eop l asmi n) coup l ed to go l d part i cl es . The

T HE nucl ear envel ope, wh i ch forms the boundary of thenucl ear compartment i n eukaryotes , cons i sts of i nnerand outer membranes j o i ned at nucl ear pore com-p l exes . The pore comp l ex i s a l arge supramo l ecu l ar structurethat spans the nucl ear envel ope and prov i des a channel forthe transport of mo l ecu l es between the cytop l asm and nu -cl eus ( for rev i ews see Franke et al . , 1981 ; D i ngwal l andLaskey , 1986 ; Gerace and Burke, 1988 ; Go l dfarb , 1989 ;Burke, 1990 ; S i l ver , 1991) . Regu l ated transport of macro -mo l ecu l es across the pore comp l ex cou l d be of maj or i mpor -tance for the contro l of gene express i on , and has been i mp l i -cated i n both growth and devel opmental contro l mechan i sms(Gerace and Burke, 1988 ; Burke, 1990) .

EM of u l trath i n sect i ons of cel l s and of negat i vel y stai nednucl ear envel opes reveal s that the pore comp l ex has promi -nent ei ght fo l d rotat i onal symmetry when v i ewed al ong anax i s perpend i cu l ar to the p l ane of the membrane (e. g . , Un -wi n and Mi l l i gan , 1982 ; Akey , 1989 ; Rei chel t et al . , 1990) .I t appears as a tr i part i te structure, composed of two coax i al"r i ngs" (wi th a d i ameter of - 120 mn) on the cytop l asmi c andnucl eop l asmi c s i des of the pore comp l ex that are connectedto a central "p l ug - spoke" comp l ex (Rei chel t et al . , 1990) .The "p l ug - spoke comp l ex prov i des the maj or spaci al con -str i ct i on for mo l ecu l ar movement across the pore comp l exand contai ns an expandab l e "central channel " (Akey , 1989) .

Dr. Grebees present address i s Yal e Un i vers i ty Schoo l of Med i ci ne, Sect i onCel l Bi o l ogy , 333 Cedar St . , New Haven , CT 06510.

® The Rockefel l er Un i vers i ty Press , 0021 - 9525 / 92 / 01 / 15 / 16 $2 . 00The Journal of Cel l Bi o l ogy , Vo l ume 116 , Number 1 , January 1992 15 - 30

ant i body al so s i gn i f i cant l y decreased nucl ear i nf l ux ofa 10 - kD dextran by pass i ve d i f fus i on . Th i s transporti nh i b i t i on d i d not resu l t f rom removal of pore com-p l exes f rom nucl ear membranes or f rom gross al tera-t i ons i n pore comp l ex structure, as shown by EM andi mmunocytochemi stry. A phys i o l og i cal consequence of

th i s transport i nh i b i t i on was i nh i b i t i on of cel l progres -s i on f rom G2 i nto M phase. Hence, b i nd i ng of th i sant i body to the l umenal s i de of gp210 must have atransmembrane ef fect on the structure and funct i ons ofthe pore comp l ex . These data argue that gp210 i sd i rect l y or i nd i rect l y connected to pore comp l ex con -st i tutents i nvo l ved i n med i ated i mport and pass i ved i f fus i on .

The total mass of the pore comp l ex determi ned by scann i ngtransmi ss i on EM i s - 124 x 106 dal tons (Rei chel t et al . ,1990) .

The general features of pore comp l ex permeab i l i ty havebeen determi ned by a var i ety of phys i o l og i cal stud i es . Thepore comp l ex contai ns an aqueous channel of a funct i onald i ameter of - 10 run that permi ts rap i d nonsel ect i ve d i f fus i onof mo l ecu l es smal l er than - 20 - 40 kD between nucl eus andcytop l asm (Bonner , 1975 ; Pai ne et al . , 1975 ; Peters , 1986) .Med i ated mechan i sms are thought to be respons i b l e fortransport of most macromo l ecu l es across the pore comp l ex ,s i nce d i f fus i on cannot account for thei r rap i d trans l ocat i onseen i n v i vo. Med i ated i mport of protei ns i s speci f i ed byshort ami no aci d reg i ons cal l ed nucl ear l ocat i on sequences(NLS) ' ( for rev i ew, see Gerace and Burke, 1988) , wh i ch

appear to i nteract wi th speci f i c cel l u l ar receptors (Go l dfarbet al . , 1986 ; Adam and Gerace, 1991) . NLS - med i ated nu -cl ear i mport requ i res ATP (Newmeyer and Forbes , 1988 ;R i chardson et al . , 1988) and appears to i nvo l ve expans i onof the central channel of the pore comp l ex to al l ow transportof part i cl es up to at l east 26 nm i n d i ameter (Dworetzky andFel dherr , 1988 ; Fel dherr and Aki n , 1990) . However , l i gandscan b i nd to the pore comp l ex i n the absence of transport atreduced temperature or under cond i t i ons of ATP dep l et i on

1 . Abbrev i at i ons used i n th i s paper: N / C, nucl ear / cytop l asmi c ; NLS , nu -cl ear l ocat i on sequence .

15

on June 15, 2006 www.jcb.org

Downloaded from

(Newmeyer and Forbes, 1988 ; Richardson et al . , 1988) ,

Materials and Methodspossibly ref lecting transport intermediates.

Understanding the process of protein transport across thenuclear envelope wi l l require a substantial amount of infor-mation on pore complex biochemistry . Since it has not beenpossible to isolate highly puri f ied pore complexes by bio-chemical procedures, posit ive identi f ication of pore complexcomponents has rel ied on the use of electron microscope im-munocytochemical local ization . By this approach, a rela-t ively smal l number of pore complex proteins have beenidenti f ied in vertebrate cel ls. These include a group of pe-r ipheral membrane proteins containing O- l inked N-acetyl -glucosamine (Snow et al . , 1987 ; Davis and Blobel , 1987) ,some homologues of which have been identi f ied recently inbudding yeast (Davis and Fink, 1990 ; Nehrbass et al . ,1990) . A second pore complex component present in a widerange of higher eukaryotes is a major integral membrane gly-coprotein with N- l inked carbohydrate cal led gp210 (Geraceet al . , 1982 ; Berrios et al . , 1983 ; Wozniak et al . , 1989) .

Interest has focused on gp210 as a protein that may anchorthe pore complex to nuclear membranes and participate inpore complex assembly. Topological studies involving site-speci f ic antibodies and protease digestions have demon-strated that gp210 has a cytoplasmic tai l of -58 residues anda single transmembrane region, with the remaining mass lo-cated in the lumen of the perinuclear space as a single con-t inuous segment containing the N- l inked carbohydrate (Gre-ber et al . , 1990) . The cytoplasmic tai l , at least in part, mayinteract with other components of the pore complex in amanner important for pore complex structure and assembly.

In this report, we have used a microinjection approach toanalyze a possible interaction of the pore complex glycopro-tein gp210 with the transport apparatus of the pore complex .A mouse monoclonal IgG directed to a lumenal epitope ofgp210 was expressed in normal rat kidney (NRK) cel ls bymicroinjecting mRNA from IgG-secreting hybridoma cel ls.Heavy and l ight chains of immunoglobul ins are secretoryproteins that are translocated into and assembled inside thelumen of the endoplasmic reticulum fol lowing injection ofhybridoma mRNA (Burke and Warren, 1984 ; Hurtley andHelenius, 1989) . The assembled antibody has access to theperinuclear space between inner and outer nuclear mem-branes, since the ER is continuous with the nuclear enve-lope, and in addit ion the outer nuclear membrane functionsas rough ER (Franke, 1974) .

We observed that the expressed anti -gp210 antibodybound to the nuclear envelope in vivo . As a consequence,both mediated transport and passive di ffusion of proteinsfrom the cytoplasm to the nucleus was severely reduced, and

progression of cel ls from G2 into M phase was effectivelyinhibited . This indicates that binding of the antibody to thelumenal domain of gp210 causes mal functioning of the nu-

clear pore complex by a mechanism that involves transmis-sion of a structural change from the lumen to the cytoplasmicside of nuclear membranes where the major components ofthe transport apparatus are located . These results suggestthat gp210 is directly or indirectly in contact with the ap-paratus important for mediated protein import and passivedi ffusion . In addit ion, these results suggest a novel potentialmechanism by which the cel l may regulate pore complexpermeabi l ity from the lumenal side of the nuclear envelope .

The Journal of Cel l Biology, Volume 116, 1992

1 6

Cel l Cultures and MicroinjectionsNRK cel ls were maintained in a humidi f ied incubator at 37°C under a 5 %CO2 atmosphere in DME containing high glucose (Gibco Laboratories,Grand Island, NY) that was supplemented with 10% FBS (Hyclone Labora-tories, Logan, UT) and penici l l in /streptomycin . Cel ls were removed fromnearly conf luent plastic dishes 24 to 48 h before microinjection by trypsini -zation and were replated on glass coversl ips in growth medium sup-plemented with 15 mM Hepes-NaOH, pH 7.4, at a density of Nl x 105cel ls per 9 .5-cm2 petri dish . For experiments involving sequential microin-jections, cel ls were plated on scored glass coversl ips obtained from Bel lcoGlass, Inc . (Vineland, NJ) (for f luorescence analysis) or on scratched plasticdishes (for electron microscopic analysis) , which faci l itated subsequentidenti f ication of the injected cel ls . Coversl ips and plastic dishes werepretreated with Cel l -Tak (Col laborative Research Inc. , Lexington MA) toenhance cel l attachment. Populations enriched in G2 cel ls were obtainedby a double thymidine block, involving two sequential single thymidineblock (Greber et al . , 1990) separated by a growth interval of 12 h in mediumlacking added thymidine . Mouse hybridoma cel ls were grown in DME con-taining high glucose and supplemented with 20% FBS, nonessential aminoacids, and penici l l in /streptomycin .

The fol lowing f luorescent samples and concentrations were used inmicroinjection experiments : rhodamine-conjugated nucleoplasmin (1 mg /ni l ; prepared as in Dingwal l et al . , 1982) , rhodamine-conjugated goat anti -rabbit IgG (5 mg /ml ; Pierce Chemical Co. , Rockford, IL) , FITC-con-jugated bovine serum albumin (5 mg /ml ; Cappel Laboratories, Cochran-vi l le, PA) , 10 kDa FITC-dextran (2 mg /ml ; Sigma Chemical Co . , St. Louis,MO) , and 150 kD FITC-dextran (5 mg /ml ; Sigma Chemical Co . ) . Al l theabove microinjection samples were brief ly dialysed against a solution con-taining 10 mM Hepes-KOH, pH 7.4, 120 mM KC1 before microinjection .Samples of mRNA were prepared for injection by adding 1 /5 vol of stocksolutions of f luorescently labeled dextran or IgG to solutions containing 1 .2mg /ml mRNA . To prepare stock solutions of f luorescent materials for di lu-t ion into mRNA solutions, samples were concentrated in a col lodium bagsystem using a 25-kD cut off membrane (Schleicher & Schuel l , Keene,NH) . Al l samples, except the nucleoplasmin-gold conjugates, were cen-tri fuged for 10 min at 13,000 g before microinjection .

Samples were injected into the cytoplasm of NRK cel ls at room tempera-ture using an Eppendorf microinjector (5242 ; Brinkman Instruments Inc. ,Westbury NY) equipped with an Eppendorf micromanipulator (5170 ;Brinkman Instruments Inc. ) and Eppendorf Femptotips under an invertedmicroscope (Carl Zeiss, Oberkochen, Germany) equipped with an Achro-stigmat 32 x objective . Injections were performed at a rate of about 10 cel lsper minute with a needle pressure of about 100 hPa for the nucleoplasminand dextran preparations and at a pressure of about 150 hPa for mRNAsamples .

Fluorescence MicroscopyTo examine nuclear protein import, microinjected NRK cel ls growing onglass coversl ips were f ixed in 3 % paraformaldehyde in PBS for 7 min atroom temperature and mounted in PBS containing 0.1% para-phenylenedia-mine (Baker Co. , Inc. , Sanford, ME) as an anti -quenching agent. For im-munof luorescence microscopy, formaldehyde f ixed cel ls were permeabi -l ized with 0.2 % Triton X-100 in PBS for 4 min, quenched with 0.2 % gelatin(Sigma Chemical Co . ; 60 bloom) and stained for expressed mouse IgG withrhodamine conjugated goat anti -mouse IgG (2 pg /m ; Cappel Laboratories) .Indirect staining for pore complex antigens was performed using puri f iedmouse monoclonal RLl antibody (5 pg /ml ; Snow et al . , 1987) fol lowed byrhodamine conjugated goat anti -mouse IgG (Pierce Chemical Co . ) . Cel lswere viewed through a Planapochromat 63x immersion objective on anAxiophot microscope (Carl Zeiss) equipped with rhodamine and f luores-cein f luorescence f i lters and photographs were taken on T-Max 400 f i lm(Eastman Kodak Co. , Rochester, NY) .

Electron MicroscopyAfter nuclear import reactions, cel ls attached to the plastic dish were f ixedby adding 1 vol of 4 % glutaraldehyde (Ted Pel la, Inc. , Irvine, CA) in 0.2 Mcacodylate-NaOH, pH 7.3, 0.2 % tannic acid (Mal l incrodt) directly to theculture . After 1 h at room temperature, the cel ls were rinsed in 0 .1 Mcacodylate-NaOH, pH 7.3, and postf ixed for 1 h at 0°C in 1% osmium

on June 15, 2006 www.jcb.org

Downloaded from

tetrox i de (Ted Pel l a, Inc. ) i n veronal acetate buf fer (Farquhar and Pal ade,1955) . Cel l s were r i nsed i n dei on i zed water , dehydrated i n steps of i ncreas -i ng ethano l concentrat i ons , and embedded i n Epon res i n (Po l ysci ence Inc. ,Warr i ngton , PA) . Af ter cur i ng , th i n sect i ons (90 - run go l d sect i ons) were cutf rom the cel l s of i nterest , mounted on gr i ds , stai ned wi th urany l acetate,

Greber and Gerace Inh i b i t i on of Nucl ear Import by Ant i bod i es to gp210

Fi gure 1. The mAb RL27 spe-ci f i cal l y recogn i zes a l umenalep i tope of gp210. (A and B)Exponent i al l y growi ng NRKcel l s were so l ub i l i zed and theant i gen recogn i zed by RL27was i mmunoadsorbed to i m-mob i l i zed ant i body and sub -sequent l y anal yzed by SDS -PAGE ( l anes 2) . As a speci f i -ci ty contro l , cel l l ysates werei mmunoadsorbed wi th RL23 ,a mouse mAb that reacts wi than uncharacter i zed protei nof the endop l asmi c ret i cu l um( l anes 1) . A shows a Coomas -s i e b l ue- stai ned gel and Bshows an i mmunob l ot probedwi th RL16 , wh i ch i s speci f i cfor gp210 (Greber et al . , 1990) .The arrows po i nt to gp210 .Note that the RL27 - adsorbedsamp l es ( l anes 2) contai n ami nor band mi grat i ng mores l owl y than gp210 and reac-t i ve wi th RL16 . Th i s band i sal so seen i n gel anal ys i s ofl ect i n - i so l ated gp210 (Greberet al . , 1990) , and i s l i kel y torepresent a h i gher mo l ecu l arwei ght aggregate of gp210.Mo l ecu l ar wei ght markers areel ectrophoresed i n l anes M.The heav i l y l abel ed bandsmi grat i ng at - 50 and 30 kDi n l ane 2 represent ant i bodyheavy and l i ght chai ns that re-act strong l y wi th protei n Aand are el uted f rom the Seph -arose beads wi th SDS . Theheavy chai n of IgG i s the heav -i l y l abel ed band seen i n l ane 1.(C and D) Iso l ated rat l i vernucl ear envel opes were d i gest -ed wi th 10 pg / ml papai n , sep -arated i nto a so l ub l e superna-tant ( l anes 1) and a membranepel l et ( l anes 2) by centr i fuga-t i on and f ract i onated by SDS -PAGE (Greber et al . , 1990) . Cshows an i mmunob l ot wi thpur i f i ed RL27 and D i s an i m-munob l ot wi th the po l ycl onalant i body Rb - 68 , wh i ch i s d i -rected agai nst a COOH - ter -mi nal pept i de of gp210 (Gre-ber et al . , 1990) .

and l ead ci trate and exami ned i n an el ectron mi croscope (CM12 ; Ph i l i psE l ectron i c Instruments , Inc . , Mahwah , NJ) at 100 kV.

For determi n i ng the f requency of pore comp l exes i n mi cro i n j ected NRKcel l s , nucl ei were randoml y sel ected and i mages were photographed at amagn i f i cat i on of 6 , 300x . Mi crographs were pr i nted and anal yzed b l i nd l y

1 7

on June 15, 2006 www.jcb.org

Downloaded from

to determi ne the number of pores per urn of nucl ear envel ope contourl ength . On l y reg i ons where the i nner and outer membranes were cl ear l yconnected were scored as nucl ear pores .

For quant i tat i on of the i mport of nucl eop l asmi n - go l d con j ugates i ntonucl ei of mRNA- i n j ected cel l s (Tab l e I ) , sect i ons through doub l e- i n j ectedcel l s hav i ng a comparab l e rat i o of cytop l asmi c to nucl ear area (2 . 33 + / -0. 21) were sel ected for anal ys i s . One sect i on per cel l was anal yzed . The go l dgrai ns i n the ent i re cytop l asm and the nucl eop l asm were counted i n photo -graph i c pr i nts f rom el ectron mi croscope negat i ves taken at a magn i f i cat i onof 9 , 600x .

Preparat i on of mRNAHybr i doma cel l s (one l i ter ) were grown i n ro l l er bott l es (1 , 750 cm2 ; Fi sherLaborator i es , Pi ttsburgh , PA) under CO2 atmosphere to a dens i ty of - 8 x10 5 / ml . Cel l s were co l l ected by centr i fugat i on at 500 g , washed br i ef l y i nPBS , and l ysed i n a guan i d i n i um hydroch l or i de based buf fer (Sambrook etal . , 1989) . Total RNA was obtai ned af ter centr i fugat i on through 4 Mces i um ch l or i de and po l y A ' RNA was i so l ated by two subsequent chro -matography steps over o l i go dT cel l u l ose (Co l l aborat i ve Research Inc. ,Lex i ngton , MA) accord i ng to estab l i shed protoco l s (Sambrook et al . ,1989) . mRNA was stored i n 70% ethano l at - 70°C at a concentrat i on ofabout 0. 1 mg / ml . Before a mi cro i n j ect i on exper i ment , 3 M sod i um acetate,pH 5. 2 , was added to 101+g mRNA to g i ve a f i nal concentrat i on of 0. 3 M .mRNA was preci p i tated at - 70 ' C for several hours , co l l ected by centr i fu -gat i on , washed twi ce i n 70% ethano l , br i ef l y dr i ed , and resuspended i nmi cro i n j ect i on buf fer .

The i so l ated mRNAs were shown to be funct i onal for d i rect i ng synthes i sof IgG by i n v i tro trans l at i on i n the presence of dog pancreas mi crosomes(Promega Bi otec, Mad i son , WI ) and [35S]meth i on i ne / cystei ne (AmershamCorp. ) , s i nce l abel ed po l ypept i des wi th an Mr of 150 , 000 (nonreduced)and an M, of - 53 , 000 and 25 , 000 ( reduced) were i so l ated f rom trans l at i onmi xes by b i nd i ng to Protei n G - Sepharose (Pharmaci a Fi ne Chemi cal s , Pi s -cataway , NJ) .

Ant i bod i es and Nucl eop l asmi n - Go l dThe mAbs RL24 (ant i - l ami n B2) and RL27 (ant i - gp210) were obtai ned af -ter i mmun i z i ng mi ce wi th rat l i ver nucl ear envel opes (Snow et al . , 1987) .Product i on of asci tes was ach i eved as descr i bed (Greber et al . , 1990) .Pur i f i ed IgG was obtai ned f rom asci te f l u i ds or cu l ture supernatants of hy -br i domas grown i n serum- f ree Ex - Cel l - 300 med i um (J . R . Sci ent i f i c,Wood l and , CA) . IgG was i so l ated by b i nd i ng to Protei n G - Sepharose (Phar -maci a Fi ne Chemi cal s) and el ut i ng wi th 0. 1 M Tr i s - G l yci ne, pH 3. 0 . Contro lmRNA used for cel l mi cro i n j ect i on exper i ments was prepared f rom hybr i d -omas secret i ng RL24 , and f rom the CRL- 1713 and TIB- 131 cel l l i nes . Themouse hybr i doma cel l l i ne CRL4713 (secret i ng IgG , ant i - DNA po l ymeraseI I I , Escher i ch i a co l i ) and TIB- 131 (secret i ng IgG , ant i - i ntermed i ate f i l a-ment protei ns) were purchased f rom Amer i can Type Cu l ture Co l l ect i on(Rockv i l l e, MD) . They had been estab l i shed by fus i on of mouse sp l een cel l swi th the same mouse P3 cel l l i ne that we have used to create the RL24 ( IgG ,ant i - l ami n B2) and RL27 ( IgG , ant i - gp210) l i nes . The po l ycl onal rabb i tant i - gp210 ant i body Rb - 68 was obtai ned as descr i bed (Greber et al . , 1990) .Nucl eop l asmi n - coated go l d comp l exes were prepared f rom pur i f i ed Xeno -pus egg - nucl eop l asmi n and 10 - nn co l l o i dal go l d part i cl es (Janssen Li feSci ences Products , Pi scataway , NJ) accord i ng to pub l i shed procedures(Dworetzky and Fel dherr , 1988 ; Leun i ssen and DeMey , 1989) .

Immunoadsorpt i on and Immunob l ott i ngFor i mmunoadsorpt i ons , 101 NRK cel l s were l ysed i n 2 ml of PBS (G i bcoLaborator i es) contai n i ng 1% Emp i gen - BB (Cal b i ochem- Behr i ng Corp. , In -d i anapo l i s , IN) , 1 mM EDTA, 1 mM PMSF, 1 kg / ml l eupept i n and pepsta-t i n , and 2 Ag / ml aprot i n i n . The samp l e was cl eared by centr i fugat i on for15 mi n at 13 , 000 g and gp210 and i mmunoadsorbed to pur i f i ed RL27 - IgGi mmob i l i zed on CNBr - act i vated Sepharose (Pharmaci a Fi ne Chemi cal ) . Acontro l adsorpt i on i nvo l ved i mmob i l i zed RL23 IgG , an ant i body d i rectedagai nst a membrane protei n of the endop l asmi c ret i cu l um (L . Gerace, un -pub l i shed observat i on) . Immunocomp l exes were washed four t i mes wi th l y -s i s buf fer , once wi th PBS , d i sso l ved i n SDS samp l e buf fer (Gerace andBl obel , 1980) , and separated by SDS - PAGE (8% acry l ami de ; Laemml i ,1970) . Mo l ecu l ar wei ght markers (S i gma Chemi cal Co . ) compr i sed myos i n(205 kD) , beta- gal actos i dase (116 kD) , phosphory l ase (96 kD) , BSA (66kD) , oval bumi n (45 kD) , and carbon i c anhydrase (30 kD) . Immunob l ott i ngwas per formed as descr i bed (Greber et al . , 1990) and ant i gen was detected

The Journal of Cel l Bi o l ogy , Vo l ume 116 , 1992

us i ng rabb i t ant i - mouse IgG (2 i 4g / ml ; Cappel Laborator i es) and 125I -

protei n A (10$ cpm / pg) at 106 cpm / ml (Featherstone et al . , 1988) .

Resu l ts

Express i on of a Funct i onal Ant i body to a LumenalEp i tope ofGp210 by Mi cro i n j ect i on of mRNAWe have obtai ned a number of mAbs that react wi th the l u -menal domai n of gp210 (Greber et al . , 1990) . To i nvest i gatea poss i b l e i nteract i on of th i s protei n wi th the transport ap -paratus of the pore comp l ex , we deci ded to exami ne whetherb i nd i ng of ant i bod i es to the l umenal domai n of gp210i nf l uences pore comp l ex funct i ons . Such an ef fect wou l dnecessar i l y requ i re a transmembrane structural change i nthe pore comp l ex , s i nce protei n transport i s carr i ed out bymach i nery l ocated on the cytop l asmi c s i de of nucl ear mem-branes (Gerace and Burke, 1988) .

Ant i bod i es cannot be del i vered to the l umen of the ER byi n j ect i on of pur i f i ed IgG i nto the cytop l asm . However , th i scan be accomp l i shed i nd i rect l y by mi cro i n j ect i on of mRNAi so l ated f rom hybr i doma l i nes secret i ng the appropr i ate ant i -body , wh i ch l eads to product i on of assemb l ed IgG i n the ERl umen (Burke and Warren , 1984) . In p i l ot stud i es wherepo l yA+ mRNA f rom three ant i - gp210 hybr i doma l i nes wasi n j ected i nto cu l tured rat cel l s , we found that the h i ghest l evelof IgG express i on was obtai ned us i ng a hybr i doma l i nesecret i ng RL27 , an ant i body that was not prev i ous l y charac-ter i zed . Because of th i s ef f i ci ent i n v i vo express i on and thel umenal l y d i sposed ep i tope of RL27 (see bel ow) , we haveper formed the work descr i bed i n th i s study wi th RL27mRNA .

A b i ochemi cal character i zat i on of RL27 IgG i s shown i nFi g . 1 . RL27 speci f i cal l y i mmunopreci p i tated a band ofabout 205 kD f rom detergent - so l ub i l i zed NRK cel l s (Fi g . 1A) . On i mmunob l ots , th i s i mmunoadsorbed protei n strong l yreacted wi th the RL16 mAb (Fi g . 1 B) , wh i ch we prev i ous l yfound to be speci f i c for gp210 (Greber et al . , 1990) . To de-termi ne whether RL27 recogn i zes an ep i tope i n the l umenaldomai n of gp210 , i so l ated nucl ear envel opes f rom rat l i verwere d i gested wi th papai n , and supernatant and pel l et f rac-t i ons were anal yzed by i mmunob l ott i ng (Fi g . 1 , C and D) .Our prev i ous stud i es showed that papai n d i gest i on of i so l atednucl ear envel opes rel eased an - 200 - kD f ragment of the l u -menal domai n of gp210 i nto a supernatant f ract i on (Greberet al . , 1990) . RL27 reacted wi th both the i ntact gp210 i n themembrane f ract i on (Fi g . 1 C, l ane 2) and al so wi th the- 200 - kD l umenal f ragment of gp210 rel eased i nto the su -

pernatant (Fi g. 1 C, l ane 1) . Th i s contrasts wi th the po l y -cl onal ant i - pept i de ant i body Rb - 68 speci f i c for the cytop l as -mi c tai l of gp210 , wh i ch recogn i zes on l y the i ntact gp210present i n the membrane pel l et f ract i on (Fi g . 1 D , l ane 2) .These data demonstrate that gp210 reacts wi th the l umenaldomai n of gp210 , l i ke the RL16 and RL20 ant i - gp210 ant i -bod i es prev i ous l y character i zed (Greber et al . , 1990) . In i m-munof l uorescence stai n i ng of NRK cel l s , cu l ture superna-tants f rom the RL27 cel l l i ne gave a s i mi l ar pattern andi ntens i ty of i mmunof l uorescence stai n i ng as prev i ous l y ob -tai ned wi th stai n i ng of cu l tured rat cel l s wi th RL20 IgG(Greber et al . , 1990 ; data not shown) . Th i s further conf i rmsthat RL27 reacts strong l y wi th nat i ve gp210.

D i rect i mmunof l uorescence mi croscopy was used to ana-

18

on June 15, 2006 www.jcb.org

Downloaded from

Fi gure 2. RL27 and ant i - DNA po l ymerase IgGs are expressed i n v i vo fo l l owi ng i n j ect i on of mouse hybr i doma mRNAs . NRK cel l s weremi cro i n j ected wi th RL27 mRNA (ant i - gp210 , A and B) and contro l mRNA f rom a hybr i doma l i ne secret i ng ant i bod i es to DNA po l ymeraseI I I f rom E. co l i (ant i - Po l , C and D) . Af ter i ncubat i on - for 2 1 / 2 h , cel l s were f i xed i n paraformal dehyde and stai ned wi th rhodami ne-con j ugated goat ant i - mouse IgG (A and C) . Correspond i ng phase contrast i mages are shown i n B and D . Bar , 20 pm .

l yze the express i on of mouse IgGs i n NRK cel l s fo l l owi ngmi cro i n j ect i on of mRNAs i so l ated f rom hybr i domas secret -i ng RL27 and other contro l ant i bod i es (Fi g . 2) . In cel l s thathad been i ncubated for 2 1 / 2 h af ter mRNA i n j ect i on (Fi g .2) , ant i - mouse IgG l abel ed a cytop l asmi c ret i cu l ar compart -ment , presumab l y correspond i ng to the ER / Go l g i systems .Mouse IgG al so was read i l y detectab l e i n th i s ret i cu l ar com-partment 1 h post mRNA i n j ect i on (data not shown) . Mi cro -i n j ected mRNA f rom a cel l l i ne secret i ng an ant i - DNA po l y -

merase ant i body l ed to the product i on of a qual i tat i vel y verys i mi l ar l evel of mouse IgG compared to mRNA f rom the

Omber and Gerace Inh i b i t i on of Nucl ear Import by Ant i bod i es to gp210

RL27 cel l l i ne (Fi g. 2) . Cel l s i n j ected wi th ant i - DNA po l y -merase mRNA are shown as contro l s i n th i s study , al thoughs i mi l ar resu l ts were obtai ned wi th cel l s i n j ected wi th mRNAi so l ated f rom hybr i doma l i nes secret i ng ant i bod i es d i rectedagai nst l ami n BZ (RL24) and cytop l asmi c i ntermed i ate f i l a-ments (TIB- 131 ; see Mater i al s and Methods) .

To determi ne whether the ant i - gp210 IgG cou l d b i nd to i tsant i gen i n v i vo af ter i t was synthes i zed , RL27 mRNA andcontro l ant i - DNA po l ymerase mRNA were i n j ected i nto thecytop l asm of NRK cel l s and cel l s were i ncubated for 2 1 / 2 hat 37 ° C. Subsequent l y , cel l s were extracted wi th buf fer con -

1 9

on June 15, 2006 www.jcb.org

Downloaded from

Figu

re3.

RL27

IgG

spec

ific

ally

bind

sto

then

ucle

aren

velo

peaf

tere

xpre

ssio

nfr

ommi

croi

njec

tedm

RN

A.N

RK

cell

swe

remi

croi

njec

ted

with

hybr

idom

am

RN

As

enco

ding

RL27

IgG

(aru

i-gp

210,

Aan

dB)

and

cont

rol

anti

-DNA

poly

mera

seIg

G(a

nti-

Pol,

Can

dD)

.Aft

erin

cuba

tion

for

21/

2h,

cell

swe

reex

trac

ted

inic

eco

ldbu

ffer

cont

aini

ng0.

1%Tr

iton

X-10

0,0.

02M

Hepe

s-Na

OH,

pH7.

4,0

.02

MNa

Cl,

0.00

1M

MgC1

2for

45s,

imme

diat

ely

fixe

din

para

form

alde

hyde

and

proc

esse

dfo

rfl

uore

scen

cemi

cros

copy

usin

grh

odam

ine

conj

ugat

edgo

atan

ti-

mous

eIg

G(A

and

C).C

orre

spon

ding

phas

econ

tras

tim

ages

ares

hown

inB

and

D.C

ells

inje

cted

with

anti

-Pol

mR

NA

were

iden

tifi

edba

sed

onth

eirl

ocat

ion

ona

grid

dedc

over

slip

.Bar

,20

Am

.

on June 15, 2006 www.jcb.orgDownloaded from

tai n i ng Tr i ton X- 100 and l ow sal t to so l ub i l i ze membranesand then were f i xed for i mmunof l uorescence mi croscopy tol ocal i ze the extract i on - res i stant ant i body. Under these con -d i t i ons , most so l ub l e protei ns are removed f rom the ER l u -men and f rom the cytop l asmi c space, wh i l e protei ns boundto i nso l ub l e structures are not extracted . Prev i ous workdemonstrated that gp210 remai ns associ ated wi th the deter -gent - stab l e pore comp l ex under these cond i t i ons (Gerace etal . , 1982 ; Greber et al . , 1990) . Fo l l owi ng detergent extrac-t i on of cel l s i n j ected wi th RL27 mRNA, mouse IgG wasl ocal i zed to the nucl ear envel ope i n a per i nucl ear "r i m" pat -tern by d i rect i mmunof l uorescence, wh i l e al most no mouseant i body remai ned i n more per i pheral cytop l asmi c reg i ons(Fi g . 3 A ; compare to Fi g . 2 A) . In contrast , cel l s i n j ectedwi th contro l ant i - DNA po l ymerase mRNA had no detect -ab l e IgG at the nucl ear per i phery and on l y very l ow l evel sof d i f fuse cytop l asmi c stai n i ng (Fi g . 3 C) . These data i nd i catethat ant i - gp2 l 0 IgG synthes i zed by mi cro i n j ected mRNAwas funct i onal i n the l umen of the nucl ear envel ope andbound to gp210.

Lumenal Ant i gp210 Ant i body Inh i b i ts Med i atedNucl ear Protei n ImportWe next exami ned whether b i nd i ng of ant i - gp210 ant i bodyto the pore comp l ex had an ef fect on med i ated nucl ear pro -tei n i mport i n v i vo . NRK cel l s were mi cro i n j ected wi thRL27 mRNA or wi th ant i - DNA po l ymerasemRNA togetherwi th an FITC- con j ugated l arge dextran to i dent i fy the RNAi n j ected cel l s (Fi g. 4 , A and D) . Af ter i ncubat i on for 2 1 / 2 hat 37 °C, cel l s were ch i l l ed to 0°C ( to arrest ongo i ng nucl eari mport dur i ng the mi cro i n j ect i on sess i on) , and i n j ected asecond t i me wi th rhodami ne- con j ugated nucl eop l asmi n , al arge nucl ear protei n of Xenopus oocytes that i s usefu l as aprobe for study i ng med i ated nucl ear i mport i n mammal i ancel l s (e . g. , Robb i ns et al . , 1991) . Cel l s then were i ncubatedfor 30 mi n at 37°C before bei ng f i xed and v i sual i zed byf l uorescence mi croscopy (Fi g . 4 , B and E , arrows) . As a fur -ther contro l , rhodami ne- con j ugated nucl eop l asmi n al so wasi n j ected i nto nearby non - mRNA- i n j ected cel l s .

The nucl ei of cel l s i n j ected wi th RL27 mRNA contai nedl i tt l e nucl eop l asmi n , al though some f l uorescence cou l d bedetected i n the nucl eo l i where the i ntranucl ear nucl eop l as -mi n tended to concentrate (Fi g. 4 B) . In contrast , nucl eo -p l asmi n - i n j ected cel l s wh i ch had not been prei n j ected wi thmRNA (Fi g . 4 A) showed strong nucl ear rhodami ne f l uores -cence and on l y a fai nt cytop l asmi c s i gnal (Fi g . 4 B) . Cel l swh i ch were i n j ected wi th contro l ant i - DNA po l ymerasemRNA accumu l ated nucl eop l asmi n i n the nucl eus to a s i mi -l ar extent as d i d non - RNA- i n j ected contro l s (Fi g . 4 E) .Therefore, express i on of RL27 ant i body had a speci f i c i n -h i b i tory ef fect on i mport of nucl eop l asmi n , al though i t d i dnot total l y b l ock nucl eop l asmi n uptake wi th our i n j ect i oncond i t i ons (see Mater i al s and Methods) . I t was not poss i b l eto determi ne whether express i on of h i gher l evel s of RL27ant i body wou l d comp l etel y i nh i b i t nucl eop l asmi n i mport ,s i nce h i gher concentrat i ons of mRNA cou l d not be i n j ectedi nto cel l s due to the v i scos i ty of these RNA so l ut i ons .

We used EM to quant i fy the ef fects of mi cro i n j ected RL27mRNA on med i ated nucl ear i mport . NRK cel l s were i n -j ected wi th RL27 mRNA or ant i - DNA po l ymerase mRNAand af ter a 2 1 / 2 - h i ncubat i on at 37°C, they were ch i l l ed to0°C and i n j ected wi th 10 - nm co l l o i dal go l d part i cl es coated

Greber and Gerace Inh i b i t i on of Nucl ear Import by Ant i bod i es ro 8p210

wi th nucl eop l asmi n . Cel l s were then i ncubated for 10 or 30mi n at 37°C to permi t med i ated transport , or were kept oni ce for 10 mi n to i nh i b i t med i ated i mport . Af ter f i xat i on andprocess i ng for th i n sect i on EM, the number of go l d part i cl esi n the nucl eus vs. cytop l asm was quant i f i ed i n cel l sect i onshav i ng a s i mi l ar rat i o of cytop l asmi c to nucl ear area . Th i sdata was expressed as the rat i o of nucl ear to cytop l asmi c par -t i cl es (N / C rat i o) .

Af ter 30 mi n at 37°C, contro l cel l s i n j ected wi th ant i -DNA po l ymerase mRNA had a N / C rat i o of approx i matel y 2 ,whereas cel l s i n j ected wi th RL27 mRNA had an averageN / C rat i o of about 0. 5 (Tab l e I ) . Af ter on l y 10 mi n , the N / Crat i o i n ant i - DNA po l ymerase- i n j ected cel l s had reached as i mi l ar l evel as seen af ter 30 mi n wi th RL27 - i n j ected cel l s .As expected f rom prev i ous stud i es (R i chardson et al . , 1988) ,cel l s i n j ected wi th ant i - DNA po l ymerase mRNA and i n -cubated at 0°C showed al most no i mport of nucl eop l asmi n -go l d , wi th a N / C rat i o of 0 . 04 . Cons i dered together , thesedata show that express i on of RL27 IgG i nh i b i ts nucl ear i m-port of nucl eop l asmi n - go l d part i cl es approx i matel y four fo l dat 30 mi n , cons i stent wi th the qual i tat i ve resu l ts of f l uores -cence mi croscopy (Fi g . 4) .

E l ectron mi crographs of cel l s i n j ected wi th nucl eop l as -mi n - go l d con j ugates are shown i n Fi g . 5. There was no de-tectab l e d i f ference i n morpho l ogy of pore comp l exes be-tween cel l s i n j ected wi th RL27 mRNA and cel l s i n j ectedwi th ant i - DNA po l ymerase mRNA, al though on l y l ow reso -l ut i on i nformat i on on the arch i tecture of the pore comp l excan be deduced f rom th i n sect i on EM . In cel l s i n j ected wi thRL27 mRNA as i n cel l s i n j ected wi th ant i - DNA po l ymerasemRNA, pore comp l ex - associ ated go l d was f requent l y foundnear the cytop l asmi c sur face of pore comp l exes (Fi g . 5 A) ,s i mi l ar to prev i ous l y pub l i shed v i ews of cel l s i n j ected wi thnucl eop l asmi n - go l d con j ugates (Fel dherr et al . , 1984 ;R i chardson et al . , 1988) .

Pass i ve D if fus i on f rom the Cytop l asm to the Nucl eusIs Reduced by Ant i gp210To exami ne whether the RL27 ant i body af fects the funct i onald i f fus i on channel of the pore comp l ex , NRK cel l s were i n -j ected wi th RL27 mRNA or contro l ant i - DNA po l ymerasemRNA, together wi th rhodami ne- l abel ed IgG as a markerfor the i n j ected cel l s . Af ter i ncubat i on for 2 1 / 2 h to al l owexpress i on of IgG , cel l s were ch i l l ed on i ce . A 10 - kD f l uores -cent dextran was then i n j ected i nto the cytop l asm of mRNAi n j ected cel l s and cel l s were further i ncubated for 30 mi n i nthe co l d before f i xat i on and exami nat i on by f l uorescence mi -croscopy (Fi g . 6) . Vi rtual l y al l of the cel l s that were i n j ectedwi th RL27 mRNA had s i gn i f i cant l y l ower l evel s of nucl eardextran f l uorescence than cel l s i n j ected wi th ant i - DNA po l y -merase mRNA or cel l s that had not been i n j ected wi th mRNA(Fi g . 6 , B and E) . In three i ndependent exper i ments com-

pr i s i ng - 50 cel l s i n j ected wi th RL27 mRNA and 10 - kD dex -tran , - 40 % of the cel l s excl uded most dextran f rom the nu -cl eus (s i mi l ar to the two cel l s i nd i cated by arrows i n Fi g . 6B) . The other 60% of the cel l s had some nucl ear dextranf l uorescence (s i mi l ar to the cel l i nd i cated by an arrowheadi n Fi g . 6 B) , al though l ess than most contro l cel l s.

In al l contro l cel l s , the 10 - kD dextran was more concen -trated i n the nucl eus than i n the cytop l asm af ter the 30 - mi ni ncubat i on i n the co l d (Fi g . 6 B, i nternal contro l s and E , ar -rows) . A s i mi l ar phenomenon was al so observed i n prev i ous

21

on June 15, 2006 www.jcb.org

Downloaded from

on June 15, 2006 www.jcb.org

Downloaded from

Table I. Nuclear Protein Import Is Speci f ical ly Inhibitedby Anti -gp210 Antibody

aGp210, 30 ' , 37°CaPol , 10 ' , 37°CaPol , 30 ' , 37°CaPol , 10 ' , chi l led

Exponential ly growing NRK cel ls were microinjected with hybridoma mRNAcoding for anti -gp210 IgG or anti -DNA polymerase IgG in def ined areas ofpetri dishes . Cel ls were incubated for 2 1 /2 h in growth medium, chi l led onice, and injected a second time with nucleoplasmincoated 10-tun gold parti -

cles . After nuclear import reactions as indicated, cel ls were f ixed, and pro-cessed for thin section EM . Gold grains in the nucleus (N) and cytoplasm (C)were quanti f ied as described in Materials and Methods . In the f irst experimentwith anti -gp210 IgG and a 30 min at 37°C uptake incubation, sections acrossfour cel ls from two independent experiments were analyzed . In the second andfourth experiments with anti -DNA polymerase IgG import reactions for 10 minat 37°C and 10 min at 0°C, respectively, two cel ls from one experiment wereanalyzed in each case . In the third experiment with anti -DNA polymerase IgGand an import reaction for 30 min at 37°C, three cel ls from two independentexperiments were analyzed for gold particles . Values of the N /C particle ratioare shown with the standard deviation of this ratio within each group of di ffer-ent cel ls analyzed .

studies involving microinjected 10-kD dextran (Paine et al . ,1975 ; Peters, 1986) , possibly because of a higher level ofsolute exclusion from the cytoplasm compared to the nucleus(Horowitz and Moore, 1974) , or to higher nonspeci f ic inter-action with intranuclear components compared to cytoplas-mic structures .

In cel ls injected with RL27 mRNA, we also have observeda clearly decreased rate of nuclear entry at 0°C of microin-jected HRP, another smal l macromolecule that can freelydi ffuse across the nuclear envelope (data not shown) . Con-sidered together, our results indicate that the RL27 antibodysigni f icantly reduces the rate of di ffusion of smal l macro-molecules into the nucleus in addit ion to inhibit ing mediatednuclear import .

Expression ofAnti -Gp210 IgG Does NotRemove RLIAntigens from the Nuclear Pbre ComplexRecently, it has been demonstrated by an in vitro reconstitu-t ion assay that nuclear pore complexes lacking O- l inked gly-coproteins that bind wheat germ agglutinin are not functionalfor mediated import of nuclear proteins (Finlay and Forbes,1990) . We therefore tested whether inhibit ion of nuclear pro-tein import in our system was a result of the removal ofO- l inked glycoproteins of the pore complex, which are rec-ognized by the mAb RLl (Snow et al . , 1987) .

NRK cel ls were microinjected with RL27 mRNA and withcontrol anti -DNA polymerase hybridoma mRNA and incu-bated for 2 1 /2 h for expression of antibodies . After f ixation,cel ls were examined for RLl antigens by indirect immunof lu-orescence (Fig. 7) . mRNA- injected cel ls were identi f ied by

Greber and Gerace Inhibit ion of Nuclear Import by Antibodies to gp210

23

coinjection of f luorescently labeled BSA (Fig . 7, A and D) .Compared to a population of control cel ls, the population ofmicroinjected cel ls displayed a qual itatively indistinguish-able pattern of punctate staining in the nuclear envelope witha comparable signal intensity (Fig . 7, B and E) . Each f luo-rescent point probably represents either a single pore com-plex or a smal l cluster of pore complexes (Maul , 1977) . Thisimpl ies that expression of RL27 antibody did not lead to de-tectable removal or rearrangement of RLl antigens of the nu-clear envelope .

Direct quanti f ication of the frequency of pore complexesin di fferent populations of cel ls was performed by thin sec-t ion EM (see Materials and Methods) . At least 100 /Lm ofnuclear envelope were examined in NRK cel ls for each ex-perimental condit ion involving injection with RL27 mRNA,control mRNA, or noninjected cel ls. In al l three cases, anaverage of about one pore complex per 2 wm of nuclear enve-lope was obtained in two to four independent experiments .Considered together with the immunof luorescence stainingdiscussed above, these data indicate that the observed nu-clear import inhibit ion of nucleoplasmin caused by the RL27antibody was not simply a result of removal of pore com-plexes, but rather was because of constraints on pore com-plex functions .

Cel ls Are Arrested in G2 Phase by Antigp210 IgG

Since the expression of the RL27 antibody in interphase cel lsdrastical ly restricts the protein transport functions of thepore complex, we examined whether expression of this anti -body inf luences the progression of G2 phase NRK cel ls intomitosis . Such an effect would be expected i f nuclear importof proteins (or nuclear export of RNAs) were required for ac-t ivation of MPF, a primary init iator of M phase progression(Pines and Hunter, 1990) . Cel ls grown on coversl ips weresynchronized in S phase by a double thymidine block. Fivehours after thymidine release when most cel ls were in G2phase, cel ls were injected with RL27 mRNA or with controlanti -DNA polymerase mRNA together with a large FITC-conjugated dextran as a tracer. Cel ls were subsequently in-cubated at 37°C and monitored for progression through mi -tosis into Gl phase, as indicated by the appearance of cel ldoublets with roughly mirror image symmetry arising fromcel l division (e.g . , Fig . 8 C) . Normal ly, 60 to 80% of themRNA- injected cel ls were recovered on the dish for thisanalysis . 4 h postinjection, only 10% of the cel ls injectedwith RL27 mRNA had divided, whereas 78 % of the controlcel ls had separated into two daughter cel ls (Table II) . Typi -cal ly, G2-arrested cel ls that had been injected with RL27mRNA appeared larger than G2 control cel ls and often con-tained a nucleus almost twice the diameter of control nuclei(Fig. 8) . We have not directly determined whether the nu-

Figure 4. Microoinjection of hybridoma mRNA encoding RL27 IgG inhibits mediated nuclear import of nucleoplasmin. NRK cel ls in adef ined area on a gridded coversl ip were microinjected with mRNAencoding RL27 IgG (anti -gp210, A-C) or control anti -DNA polymeraseIgG (anti -Pol , D-F) and incubated for 2 1 /2 h for IgG expression . Petri dishes containing the injected coversl ips were chi l led on ice for30 min, and rhodamine-conjugated nucleoplasmin was then injected into the cytoplasm of cel ls in the previously def ined areas of the cover-sl ip (B and E) . The cel ls were subsequently moved to prewarmed medium, incubated for 30 min at 37°C, f ixed, and examined by f luores-cence microscopy. mRNA injected cel ls were identi f ied with coinjected 150-kD FIX-dextran (A and D) , and nucleoplasmin- injected cel lswere identi f ied by rhodamine f luorescence (B and E) . Arrows point to cel ls double injected with mRNA and nucleoplasmin (B and E) .The corresponding phase contrast images are shown in C and F. Bar, 20 pm .

Nucleus Cytoplasm N /C

2,549 4,867 0 .52 f 0.051,050 1,784 0 .59 t 0.093,675 1,827 2 .01 t 0.55

90 2,174 ` 0 .04 t 0 .01

on June 15, 2006 www.jcb.org

Downloaded from

Figure S. Visual ization of nucleoplasmin-coated gold particles at the nuclearenvelope of cel ls injected with RL27 mRNA . NRK cel ls grownin a def ined area of a plastic dish were injected with mRNA encoding anti -gp210 IgG (A) and control anti -DNA polymerase IgG (B) ,incubated for 2 1 /2 h at 37°C and chi l led on ice for 30 min. Nucleoplasmin-coated gold (10 nm) was injected into the cytoplasm of cel lsin the def ined areas of the dish, cold medium was removed and prewarmed medium was added, and the cel ls were incubated at 37°C for30 min (A) or for 10 min (B) . The cel ls were then processed for thin section EM . Arrowheads indicate examples of nuclear pore complexes,most of which have associated nucleoplasmin-gold particles . A 10-min timepoint is shown for cel ls preinjected with anti -DNA polymerasemRNA and incubated at 37°C, since by 30 min in these cel ls most gold had been imported in the nucleus (Table I) and l ittle gold couldbe seen at the nuclear envelope (data not shown) . Bar, 200 nm.

cleus of G2-arrested cel ls actual ly increased in volume assuggested by this morphological change, or whether theysimply became more f lattened . 24 h postinjection, only 44of the cel ls injected with RL27 mRNA had divided, whi le94% of the control cel ls had undergone division (Table II) .This indicates that the arrest in G2 phase is reversible . Thiscould be due either to turnover of the RL27 IgG and /or ofthe mRNA encoding RL27, or to the low levels of mediatedimport and passive di ffusion encountered in RL27- injectedcel ls .

A simi lar G2 arrest phenotype was obtained by microin-jecting a 1 mg /ml solution of wheat germ agglutinin into thecytoplasm of late G2 cel ls (data not shown) . Wheat germ ag-glutinin is known to inhibit mediated nuclear protein importbut not passive di ffusion in vivo (Dabauval le et al . , 19ß8b ;Yoneda et al . , 1987) and in vitro (Finlay et al . , 1987 ; Adam

The Journal of Cel l Biology, Volume 116, 1992

24

et al . , 1990) . This result supports the notion that the ob-served G2 arrest caused by expression of anti -gp210 anti -body could be due to inhibit ion of mediated protein transportacross the nuclear envelope .

DiscussionIn this study we have obtained direct evidence that gp210,an integral membrane protein of the nuclear pore complex,plays an important role in organization of the pore complexassociated with mediated transport and passive di ffusion .We introduced a mAb (RL27) against gp210 into the lumenof the ER and nuclear envelope of cultured rat cel ls bymicroinjection of polyA+ mRNA from a hybridoma cel l l inesecreting RL27. The RL27 antibody, which reacts with thelumenal domain of gp210, bound to the nuclear envelope

Figure 6. The rate of di ffusion of a 1041) dextran across the nuclear envelope is reduced in cel ls microinjected with RL27 mRNA . NRKcel ls in a def ined area on a gridded coversl ip were injected with hybridoma mRNA encoding RL27 IgG (anti -gp210, A-C) or anti -DNApolymerase IgG (anti -Pol , D-F) . After an incubation of 2 1 /2 h, the cel ls were chi l led on ice for 30 min and subsequently injected with104D FTPC-dextran (B and E) . After a further incubation for 30 min on ice, cel ls were f ixed and observed in the f luorescence microscope.RNAinjected cel ls were visual ized with coinjected rhodamine- labeled goat IgG (A and D) . Arrows point to cel ls that were double injectedwith mRNA and dextran (B and E) . Corresponding phase contrast images are shown in C and F. Bar, 20 gym .

on June 15, 2006 www.jcb.org

Downloaded from

Greber and ûerace Inhibit ion of Nuclear Import by Antibodes to gp21Q

25

on June 15, 2006 www.jcb.org

Downloaded from

Figure 7. Pore complex antigens recognized by the RLl antibody persist in the nuclear envelope of cel ls microinjected with RL27 mRNA .NRK cel ls were microinjected with hybridoma mRNAs encoding anti -gp210 IgG (A-C) and anti -DNA polymerase IgG (D-F) and in-cubated for 2112 h at 37°C before f ixation and further processing for immunof luorescence microscopy using the RLl monoclonal antibody(B and E) . mRNA injected cel ls were identi f ied with coinjected FITC-BSA (A and D) and are indicated by arrows in B and E. The corre-

sponding phase contrast images are shown in C and F. Bar, 20 um .

The Journal of Cel l Biology, Volume 116, 1992

26

on June 15, 2006 www.jcb.org

Downloaded from

Fi gure 8 . Mi cro i n j ect i on of mRNA encod i ng RL27 IgG i nh i b i ts progress i on of G2 cel l s i nto M phase. NRK cel l s were synchron i zed i nS phase by a doub l e thymi d i ne b l ock and subsequent l y grown for 5 h to accumu l ate the popu l at i ons i n G2 phase . They then were mi cro i n -j ected wi th hybr i doma mRNAs encod i ng RL27 IgG (ant i - gp210 , A and B) and ant i - DNA po l ymerase IgG (ant i - Po l , C and D) . mRNA-i n j ected cel l s were i dent i f i ed wi th co i n j ected 150 - kD FITC- dextran (A and C) . Correspond i ng phase contrast i mages are shown i n B andD. Arrows po i nt to cel l s i n j ected wi th mRNA . Bar , 20 um .

dur i ng the course of i n v i vo express i on , as i nd i cated by i tsspeci f i c retent i on i n a detergent - i nso l ub l e structure of the nu -cl ear envel ope . I t was poss i b l e to subsequent l y measure sev -eral funct i onal propert i es of the pore comp l ex , i ncl ud i ngmed i ated transport and pass i ve d i f fus i on .

Greber and Gerace Inh i b i t i on of Nucl ear Import by Ant i bod i es to gp110

Wh i l e the ep i tope for RL27 was strong l y detectab l e i ngp210 of al l cu l tured rat cel l s tested , i t was absent f rommouse gp210 based on i mmunof l uorescence mi croscopy andi mmunopreci p i tat i on stud i es per formed on mouse 3T3 cel l s(data not shown) . Th i s l ack of cross - react i v i ty wi th gp210

2 7

on June 15, 2006 www.jcb.org

Downloaded from

Tab l e 17. Express i on ofAnt i - gp210 Ant i body i n G2 Cel l sBl ocks Progress i on i nto Mi tos i s

The exper i ment was per formed as descr i bed i n Fi g . 8 . The data are f rom threei ndependent exper i ments (4 h af ter i n j ect i on) and two i ndependent exper i ments(20 h af ter i n j ect i on) .

of mouse, the source of the hybr i doma cel l s , i s understanda-b l e cons i der i ng the i nh i b i tory propert i es of RL27 on cel l cy -cl e progress i on .

Inh i b i t i on of Med i ated Protei n Import and Pass i veD i f fus i on i nto the Nucl eus by RL27

Transport Inh i b i t i on and Organ i zat i on of thePbre Comp l ex

We found that i n v i vo express i on of the ant i - gp210 ant i bodydramat i cal l y i nh i b i ted the ab i l i ty of the cel l to i mport f l uo -rescent l y l abel ed nucl eop l asmi n i nto the nucl eus . Quant i ta-t i ve anal ys i s showed that nucl ear i mport of nucl eop l asmi n -go l d con j ugates was reduced approx i matel y four fo l d at 30mi n af ter i n j ect i on . Li ght mi croscope i mmunocytochemi s -try and EM demonstrated that th i s ef fect was not becauseof removal or l arge- scal e structural al terat i on of pore com-p l exes . Cons i dered together , the l i ght and el ectron mi cro -scop i c anal yses i nd i cate that the ant i - gp210 ant i body actedby d i mi n i sh i ng the rate of nucl ear protei n uptake . Whetheral l pore comp l exes are i nh i b i ted to a s i mi l ar extent by th i sant i body i s unknown .

In add i t i on to i nf l uenci ng med i ated protei n i mport , theRL27 ant i body al so apparent l y reduced the s i ze of the func-t i onal d i f fus i onal channel of the pore comp l ex , s i nce the rateof nucl ear entry of a 10 - kD dextran was s i gn i f i cant l y reducedby the RL27 ant i body i n essent i al l y al l cel l s . Cons i dered to -gether , the resu l ts of the nucl eop l asmi n and dextran transportstud i es i nd i cate that gp210 i s ei ther d i rect l y or i nd i rect l yl i nked to structures i mportant for med i ated protei n transportand pass i ve d i f fus i on across the pore comp l ex .

Cel l s that were i n j ected wi th RL27 IgG were found to beal most comp l etel y i nh i b i ted i n thei r ab i l i ty to progress f romG2 phase to mi tos i s . Th i s b l ock i n cel l cycl e progress i on per -s i sts for a l ong per i od of t i me, s i nce even 20 h af ter mRNAi n j ect i on ,

of the RL27 ant i body rai se the i nterest i ng poss i b i l i ty thatpore comp l ex permeab i l i ty al so may be regu l ated f rom thel umenal s i de of the ER / nucl ear envel ope system .

Wh i l e i t i s poss i b l e that RL27 "nonphys i o l og i cal l y" per -turbs pore comp l ex funct i ons v i a al terat i on of gp210 organ i -zat i on , i t al so i s concei vab l e that i nh i b i t i on med i ated byb i nd i ng of RL27 part i al l y mi mi cs some phys i o l og i cal path -way . Accord i ng to th i s scenar i o , gp210 (or associ ated l u -menal protei ns) cou l d serve as a condu i t for transmi tt i ng s i g -nal s f rom the l umen of the ER to the cytop l asmi cal l yl ocal i zed transport apparatus . Wh i l e b i nd i ng of RL27 togp210 d i mi n i shes the permeab i l i ty of the pore comp l ex , l u -menal s i gnal s act i ng through a transmembrane pathway atthe pore comp l ex i n pr i nci p l e cou l d ei ther i nh i b i t or st i mu -l ate nucl ear transport .

Lumenal regu l at i on of nucl ear envel ope permeab i l i ty cou l dbe i mportant i n a number of d i f ferent phys i o l og i cal s i tua-t i ons . For examp l e, th i s cou l d be a component of short - termresponses to cel l s to growth factors (Berr i dge and I rv i ne,1989) . In numerous cases , growth factor s i gnal i ng i nvo l vesa trans i ent decrease i n the f ree poo l of ER cal ci um, wh i chi n turn cou l d d i rect l y or i nd i rect l y af fect the structure ofgp210. I t cou l d al so be a part of an i ntermed i ate- term re-sponse of cel l s to env i ronmental cond i t i ons that regu l atesecret i on , membrane traf f i c, and the l umenal structure of theER . Incl uded i n th i s l atter category wou l d be the heat shockresponse . Fi nal l y , d i f ferent types of l umenal regu l at i on of thepore comp l ex cou l d be associ ated wi th speci f i c d i f ferent i a-t i on states of the cel l .

In concl us i on , the resu l ts of the present study underscorethe i mportance of gp210 i n the funct i onal organ i zat i on of thepore comp l ex , and emphas i ze the need for h i gh reso l ut i onb i ochemi cal and structural anal ys i s of gp210 and associ atedprotei ns. Furthermore, th i s work suggests that a poss i b l ei nf l uence of the ER l umenal compartment on nucl eop l asmi ctransport shou l d be exp l ored i n more detai l .

We wou l d l i ke to thank Jon Bl ev i tt for prepar i ng u l trath i n sect i ons for EManal ys i s , Steve Adam for the g i f t of nucl eop l asmi n and U l r i ch Seydel forhel p wi th i l l ustrat i ons . We al so thank Steve Adam and Joanne Westendor ffor hel pfu l comments on the manuscr i pt .

Th i s work was supported by the Nat i onal Inst i tutes of Heal th and by theG . Haro l d and Lei l a Y . Mathers Foundat i on .

Recei ved for pub l i cat i on 30 June 1991 and i n rev i sed form 11 September1991 .

References

Adam, S . A . , and L . Gerace . 1991 . Cytoso l i c protei ns that speci f i cal l y b i nd nu -cl ear l ocat i on s i gnal s are receptors for nucl ear i mport . Cel l . 66 : 837 - 847 .

Adam, S . A. , R . E . Sterne Mart , and L . Gerace . 1990 . Nucl ear protei n i mporti n permeab i l i zed mammal i an cel l s requ i res so l ub l e cytop l asmi c factors . J.Cel l Bi o l . 111 : 807 - 816 .

Aeb i , U . , M . Jami k, R . Rei chel t , and A . Engel . 1990 . Structural anal ys i s ofthe nucl ear pore comp l ex by convent i onal and scann i ng transmi ss i on el ec-tron mi croscopy . EMSA Bu l l et i n . Key , C . W . 1989 . Interact i ons and struc-ture of the nucl ear pore comp l ex reveal ed by cryo - el ectron mi croscopy . J.Cel l Bi o l . 109 : 955 - 970 .

Akey , C . W . 1989 . Interact i ons and structure of the nucl ear pore comp l ex re-veal ed by cryo - el ectron mi croscopy . J. Cel l Bi o l . 109 : 955 - 970 .

Berr i dge, M . J . , and R . F . I rv i ne . 1989 . Inos i to l phosphates and cel l s i gnal l i ng .Nature (Lord. ) . 341 : 197 - 205 .

Berr i os , M . , A . J . Fi l son , G . Bl obel , and P. A . Fi sher . 1983 . A 174 kDa ATP-ase / dATPase po l ypept i de and a g l ycoprotei n of apparent l y i dent i cal mo l ecu -l ar wei ght are common but d i st i nct components of h i gher eucaryot i c nucl earstructural protei n subf ract i ons . J. Bi o l . Chem . 258 : 13384 - 13390 .

Bonner , W . M . 1975 . Protei n mi grat i on i nto nucl ei . I . Frog oocyte nucl ei i nv i vo accumu l ate mi cro i n j ected h i stones , al l ow entry of smal l protei ns , and

Greber and Gerace Inh i b i t i on of Nucl ear Import by Ant i bod i es to gp210

excl ude l arge protei ns . J. Cel l Bi o l . 64 : 421 - 430 .Burke, B . 1990 . The nucl ear envel ope and nucl ear transport . Curr . Op . Cel l

Bi o l . 2 : 514 - 520 .Burke, B . , and G . Warren . 1984 . Mi cro i n j ect i on of mRNA cod i ng for an ant i -

Go l g i ant i body i nh i b i ts i ntracel l u l ar transport of a v i ral membrane protei n .Cel l . 36 : 847 - 856 .

Dabauval l e, M . - C . , R . Benavente, and N . Chal y . 19ß8a . Monocl onal ant i bod -i es to a Mr 68 , 000 pore comp l ex g l ycoprotei n i nter fere wi th nucl ear protei nuptake i n Xenopus oocytes . Chromosoma. 97 : 193 - 197 .

Dabauval l e, M . - C . , B . Schu l z , U . Scheer , and R . Peters . 19ß8b. Inh i b i t i on ofnucl ear accumu l at i on of karyoph i l i c protei ns i n l i v i ng cel l s by mi cro i n j ect i onof the l ect i n wheat germ agg l ut i n i n . Exp . Cel l Res . 174 : 291 - 296 .

Dav i s , L . I . , and G . Bl obel . 1987 . Nucl ear pore comp l ex contai ns a fami l y ofg l ycoprotei nsthat i ncl udes p62 : G l ycosy l at i on through a prev i ous l y un i den -t i f i ed pathway . Proc . Nod . Acad . Sci . USA. 84 : 7552 - 7556 .

Dav i s , L . I . , and G . R . Fi nk . 1990 . The NUPI gene encodes an essent i al com-ponent of the yeast nucl ear pore comp l ex . Cel l . 61 : 965 - 978 .

D i ngwal l , C . , and R . A. Laskey . 1986 . Protei n i mport i nto the cel l nucl eus .Annu. Rev. Cel l Bi o l . 2 : 367 - 390 .

D i ngwal l , C . , S . V . Sharn i ck, and R . A . Laskey . 1982 . A po l ypept i de domai nthat speci f i es mi grat i on of nucl eop l asmi n i nto the nucl eus . Cel l . 30 : 449 - 458 .

Dworetzky , S. I . , and C . M . Fel dherr . 1988 . Trans l ocat i on of RNA- coated go l dpart i cl es through the nucl ear pores of oocytes . J. Cel l Bi o l . 106 : 575 - 584 .

Farquhar , M . , and G . Pal ade . 1955 . Cel l j unct i ons i n amph i b i an ski n . J. Cel lBi o l . 26 : 261 - 291 .

Featherstone, C . M. , M . K . Darby , and L . Gerace . 1988 . A monocl onal ant i -body agai nst the nucl ear pore comp l ex i nh i b i ts nucl eocytop l asmi c transportof protei n and RNA i n v i vo . J. Cel l Bi o l . 107 : 1289 - 1297 .

Fel dherr , C . M . , and D . Aki n . 1990 . The permeab i l i ty of the nucl ear envel opei n d i v i d i ng and nond i v i d i ng cel l cu l tures. J. Cel l Bi o l . 111 : 1 - 8 .

Fel dherr , C . M . , R . J. Cohen , and J . A . Ogbum . 1983 . Ev i dence for med i atedprotei n uptake by amph i b i an oocyte nucl ei . J. Cel l Bi o l . 96 : 1486 - 1490 .

Fel dherr , C . M . , E . Kal l enbach , and N . Schu l tz . 1984 . Movement of a karyo -ph i l i c protei n through the nucl ear pore of oocytes . J. Cel l Bi o l . 99 : 2216 -2222 .

Fi n l ay , D . R . , and D . J . Forbes . 1990 . Reconst i tut i o n of b i ochemi cal l y al terednucl ear pores : transport can be el i mi nated and restored . Cel l . 60 : 17 - 29 .

Fi n l ay , D . R . , D . D . Newmeyer , T . M . Pr i ce, and D . J . Forbes . 1987 . Inh i b i -t i on of i n v i tro nucl ear transport by a l ect i n that b i nds to nucl ear pores . J.Cel l Bi o l . 104 : 189 - 200 .

Franke, W . W . 1974 . Structure, b i ochemi stry and funct i ons of the nucl ear enve-l ope . Int . Rev . Cyto l . Supp l . 4: 71 - 236 .

Franke, W . W . , U . Scheer , G . Krohne, and E. D . Jarasch . 1981 . The nucl earenvel ope and the arch i tecture of the nucl ear per i phery . J. Cel l Bi o l . 91 : 39s -50s .

Gerace, L . , and G . Bl obel . 1980 . The nucl ear envel ope l ami na i s revers i b l y de-po l ymer i zed dur i ng mi tos i s . Cel l . 19 : 277 - 287 .

Gerace, L . , and B . Burke . 1988 . Funct i onal organ i zat i on of the nucl ear enve-l ope . Annu . Rev. Cel l Bi o l . 4 : 335 - 374 .

Gerace, L. , Y. Ottav i ano , and C . Kondor - Koch . 1982 . Ident i f i cat i on of a maj orpo l ypept i de of the nucl ear pore comp l ex . J. Cel l Bi o l . 95 : 826 - 837 .

Go l dfarb , D . 1989 . Nucl ear transport . Curr . Op . Cel l Bi o l . 1 : 441 - 446 .Go l dfarb , D . S . , J . Gar i epy , G. Schoo l n i k, and R . D . Kornberg . 1986 . Syn -

thet i c pept i des as nucl ear l ocat i on s i gnal s . Nature (Land. ) . 326 : 641 - 644 .Greber , U . F . , A . Sen i or , and L . Gerace . 1990 . A maj or g l ycoprotei n of the

nucl ear pore comp l ex i s a membrane- spann i ng po l ypept i de wi th a l arge l u -menal domai n and a smal l cytop l asmi c tai l . EMBO (Eur . Mo l . Bi o l . Org . )J. 9 : 1495 - 1502 .

Horowi tz , S . B . , and L . C . Moore . 1974 . The nucl ear permeab i l i ty , i ntracel l u -l ar d i str i but i on , and d i f fus i on of i nu l i n i n the amph i b i an oocyte . J. Cel l Bi o l .60 : 405 - 415 .

Hurt l ey , S . M . , and A . Hel en i us . 1989 . Protei n o l i gomer i zat i on i n the endo -p l asmi c ret i cu l um . Annu . Rev . Cel l Bi o l . 52 : 277 - 307 .

Laemml i , U . 1970 . Cl eavage of structural protei ns dur i ng the assemb l y of thehead of bacter i ophage T4 . Nature (Load. ) . 227 : 680 - 685 .

Lanford , R . E . , P . Kanda, and R . C . Kennedy . 1986 . Induct i on of nucl ear trans -port wi th a synthet i c pept i de homo l ogous to the SV40 T ant i gen transports i gnal . Cel l . 46 : 575 - 582 .

Leun i ssen , J . L. M . , and l . R . DeMer . 1989 . Preparat i on of go l d probes . InImmuno - Go l d Label i ng i n Cel l Bi o l ogy . Verkl ej i , A . J . , and J . L. M .Leun i ssen , ed i tors . CRL Press , Boca Raton , FL . 3 - 16 .

Mau l , G . G . 1977 . The nucl ear and cytop l asmi c pore comp l ex . Structure, dy -nami cs , d i str i but i on and evo l ut i on . Int . Rev . Cyto l . Supp l . 6 : 75 - 186 .

Nehrbass , U . , H . Kern , A . Mutvei , H . Horstman , B . Marshal l say , and E . C .Hurt . 1990 . NSPI : a yeast nucl ear envel ope protei n l ocal i zed at the nucl earpores exerts i ts essent i al funct i on by i ts carboxy - termi nal domai n . Cel l . 61 :979 - 989 .

Newmeyer , D . , and D . Forbes . 1988 . Nucl ear i mport can be separated i nto twod i st i nct steps i n v i tro : nucl ear pore b i nd i ng and trans l ocat i on . Cel l . 52 : 641 -653 .

Newmeyer , D . D . , D . R . Fi n l ay , and D . J . Forbes . 1986 . In v i tro transportof a f l uorescent nucl ear protei n and excl us i on of non - nucl ear protei ns . J. Cel lBi o l . 103 : 2091 - 2102 .

Pai ne, P . L . , L. C . Moore, and S . B . Horowi tz . 1975 . Nucl ear envel ope perme-ab i l i ty . Nature ( fond. ) . 254 : 109 - 114 .

29

on June 15, 2006 www.jcb.org

Downloaded from

Peters , R . 1986 . Fl uorescence mi crophoto l ys i s to measure nucl eocytop l asmi ctransport and i ntracel l u l ar mob i l i ty . Bi ochem . Bi ophys . Acta . 864 : 305 - 359 .

Pi nes , J . , and T . Hunter . 1990 . p34 / cdc2 : the S and M ki nase? New Bi o l og i st .2 : 389 - 401 .

Rei chel t , R . , A. Ho l zenburg , E . L. Buh l e, Jr . , M . Jarn i k, A. Engel , and U .Aeb i . 1990 . Correl at i on between structure and mass d i str i but i on of nucl earpore comp l ex components . J. Cel l Bi o l . 110 : 883 - 894 .

R i chardson , W. D . , A . D . Mi l l s , S . M . D i l worth , R . A . Laskey , and C . D i ng -wal l . 1988 . Nucl ear protei n mi grat i on i nvo l ves two steps : rap i d b i nd i ng atthe nucl ear envel ope fo l l owed by s l ower trans l ocat i on through the nucl earpores . Cel l . 52 : 655 - 664 .

Robb i ns , J . , S . M . D i l worth , R . A . Laskey , and C . D i ngwal l . 1991 . Two i nter -dependent bas i c domai ns i n nucl eop l asmi n nucl ear target i ng sequence :i dent i f i cat i on of a cl ass of b i part i te nucl ear target i ng sequence . Cel l . 64:615 - 623 .

Sambrook, J. , E . F . Fr i tsch , and T . Man i at i s . 1989 . 2nd ed i t i on. Mo l ecu l arCl on i ng : A Laboratory Manual . Co l d Spr i ng Harbor Laboratory Press , Co l d

The Journal of Cel l Bi o l ogy , Vo l ute 116 , 1992

Spr i ng Harbor , NY. 7 . 18 - 7 . 22 and 7 . 26 - - 7 . 29 .S i l ver , P . A . 1991 . How protei ns enter the nucl eus . Cel l . 64 : 489 - 497 .Snow, C . M . , A . Sen i or , and L. Gerace . 1987 . Monocl onal ant i bod i es i dent i fy

a group of nucl ear pore comp l ex g l ycoprotei ns . J. Cel l Bi o l . 104 : 1143 -1156 .

Unwi n , P . N . T . , and R . A . Mi l l i gan . 1982 , A l arge part i cl e associ ated wi ththe per i meter of the nucl ear pore comp l ex . J. Cel l Bi o l . 93 : 63 - 75 .

Wh i te, E . M . , C. D . Al l i s , D . S. Go l dfarb , A . Sr i vastva, J . W . Wei r , andM . A . Gorovsky . 1989 . Nucl eus - speci f i c and temporal l y restr i cted l ocal i za-t i on of protei ns i n Terrahymena mi cronucl ei and mi cronucl ei . J. Cel l Bi o l .109 : 1983 - 1992.

Wozn i ak, R . W . , E . Barmi k, and G . Bl obel . 1989 . Pr i mary structure anal ys i sof an i ntegral membrane g l ycoprotei n of the nucl ear pore . J. Cel l Bi o l . 108 :2083 - 2092 .

Yoneda, Y . , N . Imamoto - Sonobe, M . Yamai zumi , and T . Uch i da . 1987 . Re-vers i b l e i nh i b i t i on of protei n i mport i nto the nucl eus by wheat germ agg l ut i -n i n i n j ected i nto cu l ture cel l s . Exp. Cel l Res . 173 : 586 - 595 .

30

on June 15, 2006 www.jcb.org

Downloaded from