Rossana Ese

8/14/2019 Rossana Ese

1/13

The Rockefeller University Press, 0021-9525/99/04/69/13 $2.00The Journal of Cell Biology, Volume 145, Number 1, April 5, 1999 6981

http://www.jcb.org 69

Golgi Structure Correlates with Transitional Endoplasmic ReticulumOrganization in Pichia pastori s and Sacchar omyces cerevisiae Olivia W. Rossanese, Jon Soderholm, Brooke J. Bevis, Irina B. Sears, James OConnor,Edward K . Williamson, and Benjamin S. GlickD epartment o f Molecular G enetics and Cell B iology, The U niversity of C hicago, Chicago, Illinois 60637

Abstract. G olgi stacks are often located nea r sites of transitional ER (tE R ), where COP II tra nsport vesi-cles are produced. This juxtapo sition may indicate t hatG olgi cisterna e form a t tE R sites. To explore this idea,we examined tw o budding yeasts: Pichia pastori s

, whichhas coherent G olgi stacks, and Saccharo myces cerevi - siae

, which has a dispersed G olgi. tER structures in thetwo yeasts were visualized using fusions between greenuorescent protein and CO PI I coa t proteins. We alsodetermined the localization of Sec12p, an ER mem-brane protein that initiates the CO PI I vesicle assemblypathw ay. In P

.

pastoris

, G olgi stacks are ad jacent to dis-crete tE R sites that contain COP II coat proteins aswell as Sec12p. This arra ngement of t he tE R -G olgi

system is independent o f microt ubules. In S

.

cerevisiae

,CO PI I vesicles appear to b e present throughout thecytopla sm and Sec12p is distributed t hroughoutthe ER , indicating that C OP II vesicles bud from theentire ER network. We propose that P

.

pastoris

ha sdiscrete tE R sites and therefore generates coherentG olgi stacks, whereas S

.

cerevisiae

has a delocalizedtE R and therefore generates a d ispersed G olgi. Thesendings open the wa y for a molecular genetic analysisof tE R sites.

Key words: apparatus, G olgi endoplasmic reticulum Pichia

Saccharomyces

microtubules

T

H E

transitional ER (tER )

1

is a specialized ER sub-doma in at which proteins destined for the G olgiapparatus are packaged into transport vesicles (Pa-

lade, 1975). tER sites are defined by the presence ofCO PI I vesicles, which carry secretory cargo out of the E R(Kuehn and Schekman, 1997). Vertebrate cells containmultiple tER sites, and the COPII components Sec23p,Sec13p, and Sar1p have b een localized to these sites (Or ciet al., 1991; Kuge et al., 1994; Shaywitz et al., 1995; Pac-caud et al., 1996; Tang et al., 1997). Although tER sites of-ten display an elabora te architecture (Ba nnykh and B alch,1997), the mechanisms that generate these structures arestill mysterious.

In many cell types, ranging from algae to pancreatic aci-nar cells, some or all of the tER sites are directly apposedto t he cis-face o f G olgi stacks (Whaley, 1975; Farq uharand Palade, 1981). Such morphological data led early in-

vestigators to propose the cisternal progression model forG olgi function (B eams a nd K essel, 1968; Morr, 1987). Inthis view, a new cis-G olgi cisterna f orms by the fusion ofmembranes derived from the tER. The cisterna thenprogresses through the stack to the trans face, where itfragments into secretory vesicles. However, this modelfailed to explain how resident G olgi proteins maintain apolarized d istribution across the stack, or w hy vesicles ap-pear to bud from a ll G olgi cisternae. As a n alternative, itwa s proposed tha t the G olgi consists of a series of stablecompartments, and that transport vesicles carry the sec-retory cargo from one compartment to the next (Dun-phy and Rothman, 1985; Farquhar, 1985; Rothman andWieland, 1996). According to the stable compartmentsmodel, the juxtaposition of tE R and G olgi structuresmight facilitate vesicular trafficking, but this juxtapositionwould not reflect a continual formation of new G olgi cis-ternae.

Recently, cisternal progression has been the subject ofrenewed interest (Bonfanti et al., 1998). An updatedmodel postulates that cisternal progression is coupled toretrogra de vesicular transport of resident G olgi proteins(Schnepf, 1993; Mironov et a l., 1997; Ba nnykh a nd B alch,1997; Love et a l., 1998). This cisterna l matura tionmodel can account for many experimental findings, in-cluding the polarity of the G olgi stack and the existence

Address correspondence to B enjamin S. G lick, D epartment of Mo lecularG enetics and Cell B iology, The U niversity of Chicago, 920 Ea st 58th

Street, Chicago, IL 60637. Tel.: (773) 702-5315. Fax: (773) 702-3172.E -mail: bsglick@midwa y.uchicago.edu

1. A bbreviations used in thi s paper:

D IC, differential interference contrast;G FP, green fluorescent protein; HA , hemagglutinin epitope; tE R, transi-tional ER .


2/13

The Journal of Cell Biology, Volume 145, 1999 70

of G olgi-derived tra nsport vesicles (G lick and Malhotra,1998; P elha m, 1998).

Can the cisternal maturation model help explain whyG olgi structure varies so dra matica lly between cell types(Mollenhauer a nd Morr, 1991)? If the G olgi is indeed anoutgrowth of the tE R , then some of the variation in G olgistructure may be due to differences in tER organization.To test this idea, we exa mined two b udding yeasts that d is-play different G olgi morphologies. Pichia pastoris

ha sstacked G olgi orga nelles (G ould et al., 1992; G lick, 1996).By contrast, the Saccharomyces cerevisiae

G olgi rarelyshows a stacked structure, but exists primarily as individ-ual cisternae dispersed throughout the cytoplasm (Preusset al., 1992). Here we demonstrate that tE R organizationis also fundamentally different in the two yeasts. P

.

pastoris

contains discrete tER sites. In S

.

cerevisiae

, the entireE R network apparently functions as tER . We suggest thatfixed tER sites in P

.

pastoris

produce coherent G olgistacks, whereas the delocalized tER in S

.

cerevisiae

pro-duces a dispersed G olgi. Thus, tE R organization may be akey determinant of G olgi structure in budding yeasts.

M ateri als and Methods

Str ains and Plasmids

Yeast strains and plasmids are listed in Table I. Experiments with S

.

cerevisiae

were carried out using strain D B Y1034 and d erivatives thereof.D B Y1034 was tra nsformed with plasmid pOH to obta in expression ofOch1p-HA . Strain D B Y1034-S13G , in which the endogeno us SEC13

genehas been replaced with SEC13-GFP

, was constructed as follows. The

U R A 3

cassette fro m pU C1318-U R A3 (B ndetti et a l., 1994) was excisedwith H indIII, blunted, and inserted into the SspI site of pU C19 (Ya nisch-Perr on et a l., 1985) to create pU C19-U R A3. An 1,166-bp HincII-HindI IIfragment spanning the 3

portion of SEC13

was then amplified by PCRfrom S

.

cerevisiae

genomic DNA a nd inserted into the corresponding sitesin pUC 19-UR A3. The resulting plasmid wa s mutagenized using theQuikChange kit (Strata gene Inc.) to replace the SEC13

stop codon with aSnaB I site. The E G F P

gene was excised from pEG FP-1 (Clontech) withB amH I a nd NotI, blunted, and inserted into this SnaBI site. The resulting

construct was linearized at the unique BstE II site and integrated into thechromosomal SEC13

gene. This po p-in strain wa s then plated on 5-fluo-rooro tic acid (Rot hstein, 1991) to select for the po p-out recombinan tstrain D B Y1034-S13G .

Stra in D B Y1034-S23G , in which the endogenous

SEC23

gene has been replaced with SEC23-GFP

, wa s constructed in thesame manner, beginning with the insertion into pUC 19-U R A3 of a1,120-bp AflII-HindIII fragment spanning the 3

portion of SEC23

. Simi-lar strategies were used to replace the endogenous SEC24

and SEC31

genes with EGFP

fusion genes. To construct str ain D B Y1034-S12m, inwhich the endogenous SEC12

gene has been replaced with SEC12-myc

, a1,685-bp HincII-XbaI fragment spanning the 3

portion of SEC12

was in-serted into pU C19-U R A3; a c-myc epitope sequence was then insertedbefore the stop codon using the QuikChange kit, and the resultingconstruct was linearized with SspI for pop-in integration into the

SEC12

locus.Experiments with P

.

pastoris

were carried out using the prototrophicwild-type strain PPY1 or the isogenic his4 arg4

auxotroph PPY12 and de-

rivatives thereof (Table I). G eneral methods for growth a nd transforma-tion of P

.

pastoris

have been described elsewhere (Sears et al., 1998).Strain PPY12-OH, which expresses Och1p-HA, was constructed as fol-lows. A modified gene encoding tagged Och1p (Chapman and Munro,1994) was excised from plasmid pOCHFT (a gift of Sean Munro, MedicalResearch Council, Cambridge, UK) by digesting at an upstream HindIIIsite (sequence including the start codon: AA G CTTAG AG ATCATG ),blunting, and digesting at a n Xb aI site immediately after the stop codon.This fragment w as subcloned into pI B 2 (Sears et al., 1998) that had b eendigested with SmaI and SpeI. The resulting plasmid was digested withBstEII and PstI, and the corresponding BstEII-PstI fragment from pOHwa s inserted to creat e pIB 2-OH , in which a gene encoding triple-hemag-glutinin epitope (HA)tagged Och1p is downstream of the strong consti-

tutive G A P

promoter. pIB2-OH was linearized with SalI and integratedinto the his4

locus of PPY 12. Strain P PY 12-S13G , in which the endoge-nous SEC13

gene has been replaced with SEC13-GFP

, was constructedas follows. pSG 464 (Go uld et al., 1992) was digested with Sna B I a nd NdeI,blunted, a nd religated to yield pU C19-AR G 4. This plasmid lacks the

PARS2

sequence present in pSG 464, and th erefore can only tran sform P

.

pastoris

by integration. A 461-bp HindIII-NsiI fragment spanning the 3

end of SEC13

was then amplified by PCR from P

.

pastoris

genomicD NA and inserted into pUC 19-AR G 4 that had b een cut with HindIII a ndPstI. This plasmid was mutagenized to replace the SEC13

stop codonwith a SnaBI site, and the E G F P

gene was inserted into this site as de-scribed above. The resulting construct was linearized at the unique MscI

site and integrated into the chromosomal SEC13

gene, yielding an intact

SEC13-GFP

fusion gene plus a 3

fragment of authentic SEC13

. A sim-ilar strat egy wa s used to create stra in PP Y12-S12m: a 1,328-bp B amH Ifragment spanning the 3

end of SEC12

was inserted into pUC 19-AR G 4;a c-myc epitope sequence was then inserted just upstream of the H D ELsequence, and the resulting construct was linearized with XhoI for integra-tion into the SEC12

locus. To express myc-tagged S

.

cerevisiae

Mnt1p(Chapman and Munro, 1994) in P

.

pastoris

, the gene encoding taggedMnt1p was excised from pY MT1B S (a gift of Sean Munro) and subclonedinto pOW3, a P

.

pastoris

episomal vector that contains the G A P

pro-moter (G ould et a l., 1992; Waterh am et a l., 1997; O.W. Rossanese, unpub-lished observations).

Pla smid pAFB 584, which encodes glutathione S

-transferase fused tothe NH

2

-terminal acidic domain of S

.

cerevisiae

Sec7p (up to the NaeIsite in the coding sequence), wa s provided by Alex Franzusoff (U niversityof Colorado, D enver, CO).

N ocodazole Tr eatment

P

.

pastoris

cultures were grown overnight at 30

C to an OD

600

of

0.15 in1% yea st extra ct, 2% peptone, 2% glucose, 20 mg/liter a denin e sulfa te, 20mg/liter uracil, 50 mM sod ium maleat e, pH 5.5. H alf o f a culture then re-ceived nocodazole at a final concentration of 15

g/ml, dilut ed f rom a 10mg/ml stock solution in w ater-free dimethyl sulfoxide. A s a cont rol, theother half of the culture received only dimethyl sulfoxide. Incubation wascontinued fo r up to 2.5 h, and the cells were fixed for microscopy.

A nt ibodies for I mmunof luorescence

AntiHA monoclonal antibody (16B12; Berkeley Antibody Co.), antimyc monoclonal antibod y (9E10; B oehringer Mannheim B iochemicals)and antigreen fluorescent protein (GFP ) monoclonal antibody (a mix-

ture from clones 7.1 and 13.1; Boehringer M annh eim Biochemicals) wereused at 5

g/ml; in cells not expressing a ta gged protein, ea ch ant ibodygave only a fa int background signal. The antiG FP a ntibody was used tosupplement the endogenous G FP signal, which is diminished by the immu-nofluorescence procedure. Anti

-tubulin monoclonal antibody (KMX-1;B oehringer Ma nnheim Biochemicals) was used at 1

g/ml. The fo llow ingrabbit polyclonal antibodies were used. AntiPdi1p serum (a gift of PeterWalter, U niversity of C alifornia, San Francisco, CA ; originally producedby Victoria H ines, Chiron C orp., Emeryville, CA), raised against SD S-PAG Epurified S

.

cerevisiae

Pdi1p, was used at a dilution of 1:350. AlexFranzusoff generously provided a n a ntiserum raised aga inst a fusion be-tween

-galactosidase and the NH

2

-terminal portion of Sec7p (Franzusoffet al., 1991); this antibody was used at a dilution of 1:500. Oregon green488conjugated goa t antimouse IgG and Texas red-Xconjugated goa tantirabbit IgG (Molecular Probes, Inc.) were used at 20

g/ml.Controls were performed (not shown) to check the specificity of the

polyclonal antibodies. In the case of antiPdi1p, the immunofluorescence

signal could be quenched by preincubating with a protein fra gment (a giftof Robert Freedman, University of Kent at Canterbury, UK) comprisingthe COOH-terminal 308 residues of S

.

cerevisiae

Pdi1p. The antiSec7pantibody specifically recognized a glutathione S

-transferaseSec7p fusionprotein on an immunoblot of an extract from Escherichia

coli

cells ex-pressing this protein.

Many ra bbit antisera contain tra ces of reactivity against

-1,6-mannoselinkages, and because S

.

cerevisiae

and P

.

pastoris

both add

-1,6-mannoseresidues to glycoproteins transiting through t he G olgi (Orlean, 1997;Trimble et a l., 1991), these conta minating a ntibod ies give a spurious label-ing of G olgi structures (not show n). With S

.

cerevisiae

, anti

-1,6-man-nose antibo dies label G olgi cisternae in immunoelectron microscopy ex-periments (Preuss et al., 1992); and when used at very high dilutions for


3/13

Rossanese et al. Golgi and Transitional Endoplasmic Reticulum in Yeast

71

immunofluorescence, such ant ibodies primarily label early G olgi ele-ments. Similarly, with P

.

pastoris

, anti

-1,6-mannose antibodies give astrong and relatively specific G olgi labeling. In several cases, we observedpunctate staining with antisera raised aga inst various CO PI I proteins, butthis staining wa s actually due to conta minating anti

-1,6-mannose anti-bodies. This problem can be avoided by preincubating a diluted antibodyeither with fixed S

. cerevisiae cells or with 0.5 mg/ml purified ye ast ma n-nan (M-7504; Sigma C hemical Co .).

Electr on M icroscopy Thin-section electron microscopy was performed essentially as described(Ka iser and Schekman, 1990; G ould et al., 1992). In b rief, a 50-ml cultureof yeast cells in rich glucose medium was grown to an OD 600 of 0.5. Theculture was concentrated to a volume of 5 ml with a bottle-top vacuumfilter, and 40 ml of ice-cold 50 mM KP i, pH 6.8, 1 mM MgCl 2, 2% glutara l-dehyde was added rapidly with swirling. After fixation for 1 h on ice, thecells were washed repeatedly, and then resuspended in 0.75 ml 4%KMnO 4 and mixed for 1 h at room temperature. The cells were washed,and then resuspended in 0.75 ml 2% uranyl acetate and mixed for 1 h atroom t emperature. F inally, the cells were embedded in Spurrs resin; 50 mlof yea st culture yielded enough cells for three B EE M capsules. The resinwa s polymerized for 2 d at 68 C. Sections were stained with uranyl acetateand lead citrate, and viewed on an electron microscope (100 CX II; J EO LU.S.A. Inc.).

For immunoelectron microscopy, we used a modification of previouslypublished method s (K rgel et a l., 1996; R ieder et al., 1996). A 50-ml log-phase culture was concentrated to a volume of 5 ml (see above) andfixed by adding 40 ml of 50 mM KP i, pH 6.8, 1 mM MgCl 2, 4% formalde-hyde, 0.1% glutaraldehyde (aldehydes were EM grade from Ted Pella,Redding, CA ). After 1 h a t room temperature, the cells were washed twicewith 20 ml PB S, pH 7.4, conta ining 0.5% 2-mercapto etha nol, 0.5 mMo -phenanthroline, 0.25 mM phenylmethylsulfonyl fluoride, 1 M pepsta-tin, and then once more with 1 ml of the same solution. Treatment withperiodate was o mitted because this compound may destroy certain anti-gens by reacting with amino acid side chains (Clamp and Hough, 1965).The washed cells were embedded in 100 l of low-meltingtemperatureagarose, which was cut into cubic-millimeter blocks and infiltrated for 2 hat room temperature in 1.7 M sucrose, 25% polyvinylpyrrolidone K15(Fluka Ch emical Corp.) in PB S, prepared a ccording to Tokuyasu (1989).Individual agarose blocks were then mounted on copper pins, frozen inliquid nitrogen, and cryosectioned at 80 C. Thawed cryosections wereprepared a nd lab eled as described (G riffiths, 1993; K rgel et al., 1996), us-ing a blocking buff er consisting of P B S, 20 mM glycine, 1% fish gelatin(Sigma Chemical Co.), and 10% fetal calf serum; the serum had beenheated at 60 C for 1 h, and then centrifuged to remove complement. Dilu-tions were 1:10 for an affinity-purified polyclonal antiHA antibody (a giftof J an B urkhardt, U niversity of Chicago, Chicago, IL), 1:25 for an aff inity-purified polyclonal antiGFP antibody (a gift of C harles Zuker, Univer-sity of Ca lifornia, Sa n D iego, San D iego, CA), and 1:50 for protein A-gold(10 nm; G oldma rk B iologicals). After the protein A -gold step, the sec-tions were postfixed for 30 min in 1% glutara ldehyde in PB S, and th enwashed with distilled water and stained with 1.5% silicotungstic acid in2.5% polyvinyl alcohol (MW 15,000; ICN B iomedicals Inc.). The best re-sults were obtained using relatively thick sections and a thin layer of stain.

Electron micrographs were digitized and imported into Photoshop(Adobe Systems Inc.) to adjust brightness and contrast and to create com-posite images. For Figs. 3 and 7, the gold particles were darkened to im-prove visualization. Images were printed on an NP-1600 dye-sublimationprinter (Codonics).

I mmunof luorescence Mi croscopy

We modified previously described methods (Pringle et al., 1991) to en-hance both the preservation and the visualization of intracellular struc-tures. Improved preservation was achieved by rapidly digesting the cellwall with protease-free lyticase in sorbitol-free wash buffer, and by post-fixing the cells in acetone. Improved visualization was achieved by adher-ing the cells directly to coverslips (Weiss et al., 1989; Hell and Stelzer,1995) rather than to wells on a slide. A detailed protocol follows.

Fix ati on and Spheroplastin g. A yeast culture is grown overnight in richor selective medium with good aeration to an OD 600 of 0.251.0. 8 ml ofthis culture is pipetted onto a 150-ml bottle-top f ilter, and the liquid is re-moved by vacuum. The cells are immediately resuspended in 5 ml offreshly prepared 50 mM KP i, pH 6.5, 1 mM MgCl 2, 4% formaldehyde (EM

grade; Ted Pella). After fixation for 2 h at room temperature in a 15-mltube (Falcon Labware), the cells were centrifuged for 3 min at 1,000 g (2,000 rpm) in a tabletop centrifuge, and the supernatant is aspirated com-pletely with a Pasteur pipet. The cells are resuspended in 5 ml of freshlyprepared wash buffer (100 mM KP i, pH 7.5, 1 mM MgCl 2, 0.25 mM phe-nylmethylsulfonyl fluoride, 1 M pepstatin), and then centrifuged a gain asabove. Finally, the cells are resuspended in wash buffer to an OD 600 of 10.To 100 l of concentrated cell suspension is added 0.6 l of 2-mercapto-ethanol followed by 20 l of recombinan t y east lytic enzyme (20,000 U/ml;IC N B iomedicals Inc.). Aft er mixing end-over-end for 1530 min at roomtemperature, the spheroplasts are centrifuged for 2 min at 400 g (2,000rpm) in a microfuge, gently resuspended in 100 l wash buffer, and then

centrifuged and resuspended once again.Adh esion and Permeabili zation. To create wells o n a coverslip, a suit-

able piece is cut from the adhesive backing of an express courier docu-ment pouch, holes are mad e with a rotating-head hole punch, and t he per-forated plastic is attached to the coverslip. 10 l of 0.1% polylysine(P-1524; Sigma Chemical Co.) is added to each well, and then removed byvacuum aspiration. Each well is washed three times with wa ter, and 10 lof spheropla st suspension is add ed. Af ter 3 min, excess liquid is blotted o ffwith a cotton swab, and the dried coverslip is immersed in 40 ml of ace-tone precooled to 20 C in a 50-ml tube (Falcon Labware). (This organicsolvent fixation step is important for preserving tE R and G olgi structure.For some antigens methanol works better, but in most cases acetone is su-perior.) After 5 min, the coverslip is removed, inverted onto a paper tow elto blot off excess solvent, and allowed to d ry.

Ant ibody Incubations. Each well is precoated for 30 min with a drop ofPB S-Block (PB S, pH 7.4, 1% dried milk, 0.1% bovine serum albumin,0.1% octyl glucoside). 10 l of primary antibody mixture in PB S block is

then added, and the coverslip is incubated in a humid chamber for 1 h.Ea ch well is washed eight times with PB S block, and 10 l of secondaryantibody mixture in PB S block is added. 2 g/ml 33258 (H oechst AG ) isincluded in the secondary a ntibody solution to sta in DNA . After a 30-minincubation in the dark, each well is washed eight times with PBS block.The liquid is aspirated completely after the final wash. Each well then re-ceives 5 l of phenylenediamine-containing mounting medium (Pringle etal., 1991). The coverslip is inverted onto a slide a nd sea led with nail polish.

M icr oscopy. Samples were viewed on an Axioplan microscope equippedwith a 100 Plan-Apo 1.4 NA objective (Carl Zeiss, Inc.) and with band-pass filters for visualizing H oechst dye, fluorescein/Orego n green, a ndTexas red. Images were captured with a Photometrics CCD camera usingthe Openlab software package (Improvision). For each field of cells, mul-tiple images were recorded from focal planes spaced 0.2 M apart; threeto four of these images were deblurred using a simple deconvolution algo-rithm, and then combined to form a single projected image. Where indi-cated, the same cells were also photographed in differential interference

contrast (DIC ) mode at a single focal plane. Photoshop was used for colori-zation, adjustments of brightness and contrast, and creation of compositeand merged images. Manipulations with the O penlab and P hotoshop pro-grams improved the quality of the images, but did not significantly altertheir information content. Fluorescence and D IC images were printed on aStylus Photo color inkjet printer (Epson E lectronics America Inc.).

To quantify the overlap between Och1p-HA and Sec7p, digital imageswere examined with Phot oshop. Cells that showed clear staining with bothantibodies were used for q uantitation. A spot of one color was scored asoverlapping with a spot of the second color if the center of the first spotwas conta ined within the second spot. To estimate the percentage of G olgistructures that are perinuclear in P . pastoris , the perinuclear and periph-eral Och1p-HA spots were counted in 50 cells that showed faint nuclearenvelope staining (see Fig. 2 C). To estimate the average percentage ofthe cell area occupied by fluorescent spots of a given marker protein, 25cells stained to reveal Och1p-HA and Sec7p were analyzed using NIH Im-ag e (http: //rsb.inf o.nih .gov/nih-image/): the tot al a rea s of the O ch1p-H A

and Sec7p spots were summed and divided by tw ice the total areas of thecells.

Visualization of GFP Fusion Proteins in I nt act Cells

Yea st cells expressing a G FP-labeled CO PII protein were grown to logphase with good aeration. A 400- l aliquot of the culture was then mixedrapidly with 400 l of 100 mM KP i, pH 6.5, 2 mM MgCl 2, 8% formalde-hyde, 0.5% glutaraldehyde. After fixation for 1 h in the dark, the cellswere wa shed twice with 1 ml PB S, and t hen resuspended in 50 l PBS. Forviewing by fluorescence microscopy, 12 l of the resuspended cells werespotted on a slide and spread with a coverslip.


4/13


Results

Golgi Stacks in P. pastor is Ar e Closely A ssociated withtER Sites

Sta cks of G olgi cisternae can be rea dily visualized in P .pastoris cells, with ea ch stack containing abo ut four cister-nae (Fig. 1) (G ould et al., 1992; G lick, 1996). P reviouslypublished images of P . pastori s showed G olgi stacks near

the nucleus (G ould et al., 1992; R amb ourg et al., 1995;G lick, 1996). G olgi structures might form b y grow ing outof the nuclear envelope, which constitutes a large fractionof the ER in budding yeasts (Preuss et al., 1991). Alterna-tively, P . pastori s G olgi stacks might be positioned nearthe nucleus by microtubule-dependent transport towardthe centrosome, as occurs in vertebrate cells (Thybergand Moskalewski, 1985; Kreis, 1990; Lippincott-Schwartz,

Figure 1. Thin-section elec-tron microscopy of G olgi andER structures in P . p astori s.(A) A representative P . pastoris cell. G olgi stacks arefound adjacent to ER mem-branes, including both thenuclear envelope and pe-ripheral ER elements. (B) Arepresentative P . pastori s cellafter treatment with nocoda-zole for 2.5 h. The nucleushas failed to divide. How-ever, G olgi morphology isunaffected by the drug treat-ment. This experiment wasperformed with strain PPY1.B ars, 0.5 m. G , G olgi stack;N, nucleus; ER, peripheralER membranes; M, mito-chondrion; V, va cuole.


5/13

Rossanese et al. Golgi and Transitional Endoplasmic Reticulum in Yeast 73

1998). Such a role fo r microtubules is excluded by the f ol-lowing two observations.

First, thin-section electron microscopy indicates that atypical P . pastoris cell contains several distinct G olgistacks, only some of which are located near the nucleus(Fig. 1 A). O ther G olgi stacks are fo und next to the pe-ripheral E R elements that underlie the plasma membrane.Second, treatment of P . pastori s cells with nocodazoledoes not visibly alter the structure or positioning of G olgistacks (Fig. 1 B). A s with S . cerevi siae (Jacobs et al., 1988),nocodazole treatment of P . pastori s d isrupts microtubulesand inhibits nuclear division (see Fig. 4), but G olgi stacksare still observed next to the nuclear envelope and periph-eral ER. In both untreated and nocodazole-treated cells,vesicular profiles are frequently seen in ER regions adja-cent to the G olgi cisternae (Fig. 1) (G ould et a l., 1992;G lick, 1996). These vesiculating E R regions resemble thetER sites seen in vertebrate cells (Palade, 1975; Bannykhand Balch, 1997). Hence, the morphological data suggestthat G olgi stacks in P . pastori s are a ssociated not with thecentrosome, but rather with tE R sites.

Comparison of ER and Golgi Str uctu res in t he Two

Yeasts by I mmunostaining Electron microscopy indicates that general ER structure issimilar in P . pastor is and S . cerevisiae , but that G olgiorganelles in P . pastori s are more coherent (see above).We confirmed this interpretation by visualizing variousmarker proteins (Table I) using immunofluorescence mi-croscopy. For these experiments, we developed a modi-fied immunofluorescence protocol that consistently yields

high-qua lity images (see Mat erials and Method s). B ecauseP . pastori s is closely related to S . cerevisiae (Higgins andCregg, 1998), polyclonal antibodies raised against S . cerevisiae ant igens often cross-react w ith the P . pastori s homo-logues. Alternatively, marker proteins were modified withepitope ta gs and visualized using specific monoclonal ant i-bodies.

To visualize the general ER in P . pastori s , fixed cellswere labeled with an antibody against protein disulfideisomerase (Pdi1p), a marker for the general E R (Fig. 2 D ,red) (Sitia and Meldolesi, 1992; Nishikawa et al., 1994).The same cells were a lso incubat ed with H oechst dye (Fig.2 D , blue) to lab el D NA. The antiPd i1p antibod y high-lights the nuclear envelope as well as peripheral ER ele-ments. This pattern resembles the E R distribution seen inS . cerevisiae (R ose et al., 1989; Preuss et a l., 1991).

Fig. 2 A show s the S . cerevisiae G olgi. This organelle ap-pears in immunofluorescence images as a set of punctatespots, with different G olgi markers often displaying only apartia lly overlapping localization (Segev et a l., 1988; Fran-zusoff et al., 1991; Antebi and Fink, 1992; Lussier et al.,1995). In S . cerevi siae cells expressing an HA-tagged ver-sion of the ea rly G olgi marker Och1p (Nakaya ma eta l., 1992; G ay nor et al., 1994; Ha rris and Waters, 1996;G aynor a nd E mr, 1997), an antiHA monoclonal antibody(Fig. 2 A, green) reveals multiple spots of Och1p-HA la-beling per cell. The same cells were also labeled with apolyclonal antibody (red) against Sec7p, a protein that isconcentrated in lat e G olgi elements (Franzusoff et a l.,1991). The tw o ma rkers show no significant overlap (Fig. 2A, Merged; Tab le II).

D o early a nd late G olgi elements ever colocalize in

Table I. Marker Proteins, Yeast Strains, and Yeast Plasmids Used in This Study

Marker proteinsPdi1p Lumenal ER protein. General ER marker.

Sec12p ER membrane protein. Initiates the COPII assembly pathway.Sec13p, Sec23p, Sec24p, Sec31p COPII coat proteins. Incorporated at a late stage of coat formation.Och1p Golgi membrane protein. Early Golgi marker.Sec7p Peripherally associated Golgi protein. Late Golgi marker.*

S. cerevisiae strains and plasmidsStrains

DBY1034 MATa his4 lys2 ura3 Segev and Botstein, 1987DBY1034-S13G DBY1034 SEC13-GFP This studyDBY1034-S23G DBY1034 SEC23-GFP This studyDBY1034-S24G DBY1034 SEC24-GFP This studyDBY1034-S31G DBY1034 SEC31-GFP This studyDBY1034-S12m DBY1034 SEC12 -myc This studyCTY214 MATa ura3 ade2 leu2 his4 sec14-1 ts Vytas Bankaitis

PlasmidspOH OCH1-HA URA3 CEN6 Harris and Waters, 1996

P. pastoris strains and plasmidsStrains

PPY1 Prototrophic Gould et al., 1992PPY12 his4 arg4 Gould et al., 1992PPY12-OH PPY12 OCH1-HA HIS4 This studyPPY12-S13G PPY12 SEC13-GFP ARG4 This studyPPY12-S12m PPY12 SEC12 -myc ARG4 This study

PlasmidspOW3-FLMNT1 MNT1- myc ARG4 PARS2 This study

*Sec7p is thought to function in both intra-Golgi and ER-to-Golgi transport (Franzusoff et al., 1991; Lupashin et al., 1996; Wolf et al., 1998). However, the immunofluorescencestaining pattern of Sec7p apparently represents late Golgi elements and shows little overlap with the staining pattern of early Golgi markers (Franzusoff et al., 1991; Antebi andFink, 1992). Vytas Bankaitis (University of Alabama, Birmingham, AL).


6/13


S . cerevi siae ? I n tempera ture-sensitive sec14 mutants of S .cerevisiae , multilamellar G olgi structures accumulat e atthe nonpermissive temperature (No vick et al., 1980; R am-bourg et a l., 1996); these structures are reminiscent o f theG olgi stacks seen in higher eukaryo tes. Therefore, wetested whether Och1p-HA and Sec7p colocalize in a sec14 mutant after incubation at the nonpermissive temperatureof 37 C. The 37 C t reatment causes redistribution of Sec7pinto la rge clusters (Fig. 2 B , red) (Fra nzusoff et a l., 1991).However, the staining pattern of Och1p-HA is largely un-changed in the sec14 mutant (Fig. 2 B, green), and there isstill no significant overlap bet ween O ch1p-H A a nd Sec7p.It seems that S . cerevi siae is incapable of genera ting coher-ent G olgi stacks that contain both early and late G olgiproteins.

Fig. 2 C shows the P . pastori s G olgi. U sing an integrat-ing expression vector (Sears et al., 1998), we generated aP . pastori s strain that stably expresses HA-tagged S . cerevisiae Och1p. P . pastori s contains an -1,6-mannosyltrans-ferase activity like that ascribed to Och1p (Trimble et al.,1991; Nakanishi-Shindo et al., 1993), and expression ofO ch1p-H A in P . pastori s has no effect on growth or G olgimorphology (not shown). Och1p-HA localizes to two tosix spots per P . pastori s cell (Fig. 2 C , green). As predictedfrom electron microscopy, a subset ( 45%) of the O ch1p-H A spots clearly a djoin the nucleus (see Fig. 2 C , legend).When P . pastoris cells are labeled with the antibodyagainst S . cerevi siae Sec7p, two to six spots are again de-tected (Fig. 2 C, red), presumably because this antibodyreacts with the P . pastori s homologue of Sec7p. By con-

Figure 2. Immunofluorescencestaining of G olgi and E R struc-tures. (A) G olgi labeling in S .cerevisiae . Fixed cells of strainD B Y 1034/pOH were incubatedwith a monoclonal antiHA an-tibody to label Och1p-HA, fol-lowed by Oregon greencon-jugated antimouse antibody(green). The same cells werealso incubated with a polyclonalantibody against Sec7p, fol-lowed by Texas redconjugatedantirabbit antibody (red). Themerged image shows very littleoverlap between the two G olgimarkers. (B) G olgi labeling in asec14 mutant of S . cerevisiae .Strain CTY214 was grown a t thepermissive temperature of 23 C,and then shifted to the nonper-missive temperature of 37 C for1 h before fixation. Antibody la-beling was as in A. After thetemperature shift, Sec7p-con-taining structures reorganize

into large clusters, but the stain-ing pattern of Och1p-HA is notsignificantly altered. Surpris-ingly, as illustrated by the cell onthe right, budded sec14 cells of-ten partition the Sec7p clustersinto the bud. (C) G olgi labelingin P . pastoris . Fixed cells ofstrain PPY12-OH were incu-bated with the same primaryand secondary antibodies as inA. Och1p-HA staining overlapsstrongly with Sec7p staining, asshown in the merged image. Asin some strains of S . cerevisiae

(O.W. Rossanese, unpublishedobservations), Och1p-HA oftengives a f aint staining of the gen-eral ER in P . pastori s ; for exam-

ple, the cell on the lower left shows weak nuclear envelope staining, which reveals that tw o o f the three labeled G olgi spots adjoin thenucleus. (D ) G eneral ER labeling in P . pastori s . Fixed cells of strain P PY 12 were incubated with a polyclonal a ntibody a gainst S . cerevi - siae P di1p, followed by Texas redconjugated a ntirabbit a ntibody (red). The same cells were also incubated with H oechst dye to stainD NA (blue). As shown in the merged image, Pdi1p is present in the nuclear envelope and peripheral E R structures. Ba r, 2 m.


7/13


trast to the situation in S . cerevi siae , O ch1p-H A a nd Sec7pexhibit nearly quantitative overlap in P . pastori s (Fig. 2 C,Merged; Table II). Similarly, when a tagged version of theS . cerevisiae media l-G olgi protein K re2p/Mnt1p (Cha p-

man and Munro, 1994; Lussier et al., 1995) is expressed inP . pastori s , it colocalizes with Sec7p (not shown). Co local-ization at the light microscopy level indicates that twomarkers are either in the same compartment or in closelyapposed structures. These dat a support t he conclusion tha tG olgi organelles are more coherent in P . pastori s than inS . cerevi siae .

To confirm that the structures visualized in P . pastori s by immunofluorescence are indeed G olgi membranes, weperformed immunoelectron microscopy on thawed cryo-sections. Using a polyclonal antiHA antibody to detectOch1p-H A, we see a specific lab eling of G olgi stacks (Fig.3; Tab le II I).

Golgi Or ganizati on in P. pastor is Is U naff ected byM icrotubule Di srupt ion or C ell Cycle Progression

P . pastori s resembles vertebrate cells in having stackedG olgi cisternae. We tested whether the tw o cell types arealso similar w ith regard to G olgi dynamics. Pre-G olgi ele-ments in vertebrate cells are transported along microtu-bules (Lippincott-Schwartz, 1998), but our electron mi-croscopy data indicate that in P . pastori s , microtubules donot inf luence G olgi structure (see Fig. 1). This conclusionwas confirmed at the immunofluorescence level. In un-treated P . pastori s cells, microtubules are present a nd nu-clei partition into d aughter cells during mitosis (Fig. 4 A).Microtubule distribution shows no detectable relationshipto G olgi distribution (Fig. 4 A). Nocodazole treatment de-polymerizes microtubules and blocks nuclear migration,but the G olgi staining pattern is unaffected (Fig. 4 B).Moreo ver, the G olgi markers Sec7p and O ch1p-H A stillcolocalize in nocodazole-treated P . pastoris cells (notshown).

To determine whether cell cycle progression altersG olgi organization in P . pastori s , as it does in vertebratecells (Rabouille and Warren, 1997), we analyzed P . pastoris in the G 1, S/G 2, and M pha ses of the cell cycle. Im-munofluorescence was used to quantify the average num-

ber o f G olgi structures per cell. This number is similar ineach phase of the cell cycle (Table IV). Thus, when ob-serving P . pastori s under a var iety of conditions, we consis-tently find that each cell contains a small number of dis-

tinct G olgi stacks.COPII Coat Proteins Show D if ferent D istri butions in S. cerevisiae and P. pastoris

The electron microscopy dat a suggest that P . pastori s con-

Table II. Overlap between Golgi Markers as Visualized by Immunofluorescence

Marker protein Second markerSpots thatcolocalize

Spots that failto colocalize Percent overlap

Saccharomyces cerevisiae (DBY1034/pOH)Och1p-HA Sec7p 50 611 7.6Sec7p Och1p-HA 50 437 10.3

Pichia pastoris (PPY12-OH)Och1p-HA Sec7p 215 12 94.7Sec7p Och1p-HA 215 18 92.3

Yeast cells expressing Och1p-HA were processed for double-label immunofluores-cence, and overlap between Och1p-HA and Sec7p was quantified. Colocalizationmeans that a spot labeled for the marker protein was also labeled for the secondmarker. 50 cells of each species were examined, with budded cells being counted asone. Based on the fraction of the cell area occupied by fluorescent spots, the probabil-ity that a given spot would overlap by chance with a spot representing the secondmarker was estimated at 8.9% for S. cerevisiae and 8.3% for P. pastoris . With P.

pastoris , a small percentage of the fluorescent spots are labeled for only one of the twoGolgi markers; it is unclear whether this phenomenon is meaningful or whether it re-flects a limitation of the immunofluorescence method.

Figur e 3. Immunoelectron microscopy of a G olgi marker in P .pastori s. P . pastori s cells of strain P PY 12-OH were fixed and cryo-sectioned, and then incubated with a polyclonal antiHA anti-body fo llowed by protein A-gold to detect Och1p-HA . G old par-ticles are consistently observed over G olgi stacks. In cells of theparental PPY12 strain, which does not express Och1p-HA, nospecific labeling is seen with the a ntiHA antibody (not shown).B ar, 0.5 m. N, nuclei.

Table III. Quantitation of Immunoelectron Microscopy Data for P. pastoris

CompartmentTotal number

of gold particlesAverage numberof gold particles

per m2

Och1p-HA labeling

General ER 0 0.0Transitional ER 2 1.6Golgi 28 16.3

Sec13p-GFP labelingGeneral ER 6 1.3Transitional ER 63 47.0Golgi 3 1.4

Data are from the experiments described in Figs. 3 (for Och1p-HA) and 7 (forSec13p-GFP). In each case, 20 cells that showed good membrane preservation werephotographed without regard to the labeling pattern. Stereology was then performedusing the method of Griffiths (1993). The antiHA and antiGFP antibodies both gavenegligible labeling of nuclei, vacuoles, and mitochondria.


8/13


tains discrete tER sites, whereas S . cerevisiae lacks suchsites. To explore this idea further, we compared the local-izations of CO PI I proteins in the two yeasts. G FP wa sfused to the COOH terminus of S . cerevi siae Sec13p, acoat protein that is incorporated at a late stage of COPII

vesicle assembly (Pryer et al., 1993; Kuehn and Schek-man, 1997). R eplacement of the endo genous SEC13 genewith the SEC13-GFP gene has no detectable effect oncell growth, indicating tha t the Sec13p-G FP f usion canperform the essential function of Sec13p (Pryer et al.,1993). When S . cerevisiae cells expressing Sec13p-G FP a reviewed by fluorescence microscopy, each cell contains

3050 tiny spots (shown in Fig. 5 A, although this patternis difficult to photograph accurately). The spots are dis-tributed almost evenly throughout the cytoplasm, withsome cells showing an apparent concentration of spots onthe nuclear envelope. We surmise that these spots repre-sent individual COPII vesicles. A vesicle is smaller thanthe resolution limit of light microscopy (Lacey, 1989), sothe appa rent size of the spots is probab ly mislead ing, but asingle C OP II vesicle contains many copies of Sec13p-G FPand hence should produce a detectable fluorescence sig-nal. G FP w as also fused to three other S . cerevi siae COPI Icoat proteins (Kuehn and Schekman, 1997): Sec23p,Sec24p, and Sec31p. All of these fusions are functional,and they a ll give the same fluorescence pattern a s Sec13p-G FP (Fig. 5 B and da ta not shown), indicating that wehave visualized the normal distribution of COPII vesicles.These results strongly support the notion tha t C OP II vesi-cles bud from the entire E R in S . cerevi siae .

In parallel, the P . pastor is SE C13 gene was replacedwith a SEC13-GFP fusion gene. The resulting fluores-cence pattern is strikingly different from that seen in S .cerevisiae . Sec13p-G FP in P . pastori s is concentrated inonly tw o to six large spots per cell (Fig. 6 A). When exam-ined by immunofluorescence microscopy, the Sec13p-G FPspots are a djacent to, but not q uite overlapping, the G olgi

Figur e 4. Immunofluorescence ana lysis of nocoda zole-treated P .pastoris cells. (A) D NA, tub ulin, and Sec7p distributions in un-treated P . pastori s cells. Fixed cells of strain PPY12 were incu-bated w ith Hoechst dye to stain D NA. Microtubules were visual-ized with a monoclonal antitubulin antibody followed byOregon greenconjugated antimouse antibod y. The G olgimarker Sec7p wa s visualized as in Fig. 2. (B ) D NA, tubulin, andSec7p distributions in P . pastori s cells treated with nocodazolefor 2.5 h. Fixed cells were labeled as in A. As previously de-scribed for S . cerevisiae (Ja cobs et a l., 1988), microtub ules are vir-tually undetectable within 1 h after nocodazole addition (notshown). Ba r, 2 m.

Table IV. The P. pastoris Golgi Retains Its Organization

Throughout the Cell CyclePhase of cell cycle Average number of Och1p-HA spots per cell

G1 3.8 1.2S/G2 4.1 0.9M 3.9 1.3

An unsynchronized culture of P. pastoris strain PPY12-OH was processed for im-munofluorescence as in Fig. 2 C. Individual cells were assigned to a phase of the cellcycle based on morphology and nuclear distribution (Lew et al., 1997). 50 cells wereexamined for each phase of the cell cycle to determine the average number of Och1p-HA spots. The numbers listed refer to nucleated cells; buds lacking nuclei were ex-cluded from the analysis. Standard deviations are indicated.

Figur e 5. Visualization of G FP-labeled CO PI I coat proteins inintact S . cerevisiae cells. (A) Sec13p-G FP fluorescence in S . cerevisiae . Cells of strain D B Y1034-S13G were fixed and viewed di-rectly. The same cells were imaged in D IC and fluorescencemodes. (B ) Sec23p-G FP fluo rescence in S . cerevisiae strainD B Y1034-S23G . The experiment was performed a s in A. B ar, 2 m.


9/13


spots marked b y the a ntiSec7p antibo dy (Fig. 6 B ). Thisresult suggests tha t Sec13p-G FP is localized to tE R sites.Indeed, immunoelectron microscopy with an antiG FP a n-tibody revealed tha t Sec13p-G FP is present on t ubulovesic-ular structures at the interface between ER membranes

and G olgi stacks (Fig. 7; Table I II ). We conclude tha tCOPII vesicle budding is restricted to discrete tER sites inP . pastor is .

Sec12p Localizes to tER Sit es in P. pastori s

The earliest known player in the C OP II assembly pathwa yis Sec12p, a membrane-bound guanine nucleotide ex-change factor tha t recruits the small G TPa se Sar1p to theER membrane (Nakano et al., 1988; Nakano and Mura-matsu, 1989; Barlowe and Schekman, 1993; Kuehn andSchekman, 1997). In previous studies of S . cerevisiae ,Sec12p exhibited general ER staining (Nishikawa and Na-kano, 1991, 1993). Our results confirm those earlier find-ings. S . cerevi siae Sec12p has tra ditiona lly been visualizedin strains overexpressing this protein (Nishikawa and Na-kano, 1993; Nishikawa et al., 1994). To eliminate possibleambiguities resulting from overexpression, we replacedthe chromosomal SEC12 gene with a myc-tagged version.The resulting strain grow s like the wild-type (not shown),implying that Sec12p-myc can functionally repla ce the es-sential wild-type protein (Nakano et al., 1988). Sec12p-myc (Fig. 8 A, green) colocalizes with P di1p (Fig. 8 A, red)in the nuclear envelope and in peripheral ER membranes.Although Sec12p-myc sometimes shows a discontinuousstaining pattern, the f luorescence signal is relatively wea k,and w e have observed that the ER network often appearsdiscontinuous when it is weakly stained (not shown).Hence, the combined data suggest that Sec12p-myc ispresent throughout the E R in S . cerevi siae .

We also replaced P . pastori s SEC12 with a myc-taggedversion. Once again the fluorescence signal is weak, butthe pat tern is clearly visible. In this case, the ant imyc an-tibody does not give a general ER staining, but instead la-bels severa l spots per cell (Fig. 8 B , green). L ike Sec13p-G FP (see ab ove), Sec12p-myc localizes to sites tha t areimmediately a djacent to Sec7p-containing G olgi structures(Fig. 8 B ). Thus, in P . pastori s , components at bo th earlyand late stages of the COPII assembly pathway are con-centrated at tE R sites.

Figure 6. Visualization of Sec13p-G FPin P . pastori s. (A) Sec13p-G FP fluores-cence in intact cells. P . pastori s cells ofstrain PP Y12-S13G were fixed andviewed directly. D IC and f luorescenceimages were collected separately, andthen combined to generate the mergedimage. (B ) Immunofluorescence local-ization o f Sec13p-G FP and Sec7p. FixedP . pastoris cells of strain PPY12-S13Gwere incubated with a monoclonal antiG FP antibody followed by Oregongreenconjugated antimouse antibody(green). The same cells were also sta inedfor Sec7p (red) as in Fig. 2. The mergedimage shows that Sec13-G FP a ndSec7p-containing structures are closelyapposed, but distinct. Ba r, 2 m.

Figur e 7. Immunoelectron microscopy of a tER marker in P .pastori s. P . pastori s cells of stra in PP Y12-S13G were fixed a ndcryosectioned. The localization of Sec13p-G FP wa s determinedby incubating with a polyclonal antiG FP a ntibody followed byprotein A-gold. G old particles are consistently observed overtER regions, which are often associated with indentations of thenuclear envelope. In cells of the pa rental P PY 12 strain, which ex-presses wild-type Sec13p, no specific labeling is seen with theantiG FP ant ibody (not shown). Bar, 0.5 m. N, nuclei.


10/13


D iscussion Why is the G olgi appara tus more photogenic in P . pastoris than in S . cerevi siae ? These two yeasts are morpho-logically very similar; yet G olgi cisterna e in P . pastori s are orga nized into stacks, whereas G olgi cisternae in S .cerevisiae are scattered t hroughout the cytopla sm. We pro-pose the following hypothesis. In P . pastori s , COP II vesi-cles bud from fixed tE R sites, and then fuse with one an-other to create new G olgi cisternae, which mature toyield polarized sta cks (G lick and Ma lhotra , 1998). InS . cerevisiae , CO PII vesicles bud throughout the ER , andtherefore each G olgi cisterna forms at a different loca-tion (F ig. 9).

Our immunofluorescence data confirm that S . cerevisiae and P . pastori s have fundamentally different G olgistructures. A hallmark of the G olgi in S . cerevisiae isthat various marker proteins often show distinct punc-tate distributions (Antebi and Fink, 1992; Chapman andMunro, 1994; Lussier et al., 1995). For example, the earlyG olgi protein Och1p-H A exhibits virtually no overlapwith the late G olgi protein Sec7p. Although multilamel-

lar G olgi structures are seen in temperature-sensitivesec7 and sec14 mutants of S . cerevi siae (Novick et al.,1980; Svodoba and Necas, 1987; Rambourg et al., 1993,1996), we find tha t early a nd late G olgi markers still donot colocalize in sec14 mutant cells (Fig. 2), indicatingthat S . cerevisiae cannot ma ke coherent G olgi stacks.With P . pastoris , on the other hand, Och1p-HA andSec7p overlap almost completely (Fig. 2), as expected ifeach G olgi stack represents an ord ered set of early, mid-dle, and late cisternae.

D oes this difference in G olgi structure correlate with adifference in ER organization? The ER of both yeastscomprises the nuclear envelope plus peripheral elements.

H owever, in P . pastori s , vesicles can o ften be seen buddingspecifically from regions of the ER adjacent to G olgistacks (Fig. 1) (G ould e t a l., 1992; G lick, 1996). Such vesic-ulating ER regions have not been seen in S . cerevi siae (Kuehn and Schekman, 1997). These observations sug-gested to us that P . pastori s contains discrete tER sites,whereas S . cerevi siae does not. To test this interpretation,we used COP II coa t proteins as markers for the tER (Orci

Figur e 8. Immunofluorescence local-ization of tagged Sec12p in the twoyeasts. (A) C olocalization of Sec12p-myc with Pdi1p in S . cerevisiae . Fixedcells of stra in D B Y 1034-S12m wereincubated with a monoclonal antimyc an t ibody fo l lowed by Ore -gon greenconjugated antimouseantibody to visualize Sec12p-myc

(green). The same cells were alsostained for Pdi1p (red) as in Fig. 2.As shown in the merged image, thestaining patterns for the two proteinslargely overlap. (B) Localization ofSec12p-myc in P . pastori s . Fixed cellsof strain PPY12-S12m were stainedas in A to visualize Sec12p-myc(green); Sec7p (red) was visualizedas in Fig. 2. The merged imageshows that the two staining patternsare closely apposed, but distinct.B ar, 2 m.

Figure 9. Summary and in-terpretation of the experi-mental data. P . pastori s con-tains discrete tER sites thatproduce coherent G olgistacks. In S . cerevisiae ,COPII vesicles bud through-out the ER network, result-ing in a dispersed G olgi. Theshaded portions of the ERare the regions that can func-tion as tER. See text for de-tails.


11/13


et al., 1991; Kuge et al., 1994; Shaywitz et al., 1995; Pac-caud et a l., 1996; Tan g et a l., 1997). With S . cerevisiae , fus-ing G FP to Sec13p or other CO PI I coat proteins revealedmany small fluorescent spots that prob ably represent in-dividual COPII vesicles (Fig. 5). These spots are foundthroughout the cytoplasm, consistent with the not ion thatCOPII vesicles bud at random from the entire ER. Bycontrast, a Sec13p-G FP fusion in P . pastori s localizes to asmall number of discrete regions (Fig. 6). These regionsare immediately adjacent to G olgi stacks, and theycontain tubulovesicular membranes (Fig. 7). Hence, wepropose that Sec13p-G FP marks a compart ment in P .pastoris that is analogous to the tER sites described pre-viously in vertebr at e cells (Pa lad e, 1975; B anny kh andB a lch, 1997).

H ow is Sec13p recruited to tE R sites in P . pastori s ? Theassembly of CO PI I vesicles is an o rdered process that b e-gins with the a ction of Sec12p (B arlow e and Schekman,1993; Kuehn and Schekman, 1997). We compared the lo-calization o f epitope-tagged Sec12p in the two yeasts. Co n-sistent with previous reports (Nishikawa and Nakano,1993; Nishikawa et al., 1994), Sec12p-myc in S . cerevi siae isdistributed throughout the E R . In P . pastori s , on the otherhand, Sec12p-myc is concentrated at tER sites (Fig. 8).This observation suggests a w orking model for t he orga ni-zation of tE R and G olgi compartments in the two yeasts.We postulate tha t in P . pastori s , Sec12p is anchored a t tE Rsites by unknown partner proteins that comprise a tE Rscaffold. B ecause Sec12p initiates the assembly of C OP IIvesicles, these vesicles bud exclusively fro m tE R sites, andsuccessive G olgi cisterna e form a t fixed locations to gener-ate polarized stacks. This model assumes that tER sitesare relatively stable entities, and, indeed, our studies ofSec13p-G FP dyna mics in P . pastori s confirm that tE R sitesare long lived and slow mo ving (B.J . B evis, unpublishedobservations). The situation is quite different in S . cerevi - siae this yeast apparently lacks a tER scaffold, so Sec12pis free to diffuse throughout the ER. As a result, COPIIvesicles bud from the entire E R , and successive G olgi cis-ternae form at different locations, yielding a dispersed or-ganelle (Fig. 9). In this view, G olgi structure and position-ing are strongly influenced by tE R organization.

An important test of our model will be to alter tER or-ganization in P . pastori s and ask whether G olgi structure iscorrespondingly affected. Such an experiment will revealwhether the existence of G olgi stacks in P . pastori s is duesolely to the presence of fixed tE R sites. The simplest view is that G olgi stacking is a kinetic phenomenon, w ith cister-nal formation and maturation occurring too quickly forsuccessive cisternae to diffuse away from one another. Al-ternat ively, G olgi cisternae in P . pastori s might be held to-

gether by a cytosolic matrix (Mollenhauer and Morr,1978; Cluet t and B row n, 1992; Staehelin and Mo ore , 1995;B ar r et a l., 1997). These possibilities can be distinguishedby generating P . pastori s mutants in which CO PI I proteinsare delocalized, and then asking whether the loss of tERsites leads to G olgi dispersal. One promising approach fo-cuses on determining whether P . pastori s Sec12p containsa tER localization signal that is recognized by specificpartner proteins.

Although S . cerevisiae lacks discrete tER sites, thisyeast can be used to explore the more general question of

whether the G olgi is an outgrowth of the ER . Early in S-phase of the S . cerevisiae cell cycle, the small b uds invari-ably contain both E R and G olgi structures (Segev et al.,1988; Redding et al., 1991; Preuss et al., 1991, 1992; O.W.Rossanese, unpublished observations). It is likely that theE R elements present in the emerging bud give rise to new G olgi cisternae. If so, mutants that fail to transport ERmembra nes into the bud should also lack G olgi cisternaein the bud. We are currently testing this prediction bycharacterizing S . cerevi siae mutants defective in G olgi in-heritance.

A comparison of budding yeasts with other eukaryotescan indicate which aspects of the tE R -G olgi system arecell typespecific. First, unlike ma ny euka ryotes, S . cerevi - siae contains neither stacked G olgi organelles nor discretetER sites. It is unclear whether the absence of these struc-tures in S . cerevi siae is adaptive, or whether it reflects aloss-of-function mutation during the evolution of thisyeast. Second, tubular connections betw een G olgi stacksare present in vertebrate cells (Mironov et al., 1997), buthave not been detected in P . pastori s , the fission yeastSchi zosaccharom yces pom be (Chappell and Warren, 1989)or certain insect cells (Stanley et al., 1997). Third, theG olgi breaks down during mitosis in vertebrat e cells(Rabouille and Warren, 1997; Acharya et al., 1998), butnot in higher plant s (D riouich and St aehelin, 1997) or b ud-ding yea sts (Tab le IV) (Maka row , 1988).

Which aspects of the tE R -G olgi system a re universal?We propose that G olgi cisternae a lways form by the coa-lescence of tER-derived membranes. This relationship isparticularly evident in cells that contain G olgi stacks im-mediately a djacent to tE R sites (Whaley, 1975; Farq uharand Pa lad e, 1981; B racker et al., 1996). We have no w d oc-umented such an a ssociation in P . pastori s . A close appo-sition between tE R sites and G olgi stacks is probab ly ageneral feature of cells that do not tra nsport G olgi ele-ments along cytoskeletal tracks. Consistent w ith this idea,microtubules have no influence on G olgi structure or po-sitioning in P . pastor is (Figs. 1 and 4). However, in somecell types, nascent G olgi elements are transported aw ayfrom tE R sites, thereby obscuring the tER -G olgi connec-tion. For example, in S . pombe , microtubules play a rolein cisternal stacking and possibly in G olgi movement(Ay scough et al., 1993). In higher pla nts, G olgi structuresare tra nsported along actin filaments (D riouich and Sta e-helin, 1997; B oev ink et a l., 1998). The best -studie d exa m-ple of cytoskeleton-mediated G olgi movement is pro-vided by vertebrate cells, which employ microtubules andassociated motor proteins to generate a juxtanuclearG olgi ribbon (Lippincott-Schwar tz, 1998; B urkhard t, 1998).As a consequence, tE R sites and G olgi elements nor-

mally do not colocalize in vertebrate cells. Yet recent evi-dence suggests that na scent G olgi structures in vertebratecells initially coalesce at tER sites (Presley et al., 1997;Scales et al., 1997; Rowe et al., 1998); in the presence ofmicrotubule-depolymerizing a gents, entire G olgi stacksare found next to tER sites (Cole et al., 1996; Storrie etal., 1998). Thus, when micro tubules ha ve been disrupted,the tE R -G olgi system in a vertebrate cell resembles thetE R -G olgi system in P . pastori s . It seems that in all eu-karyotes, the G olgi can be viewed as a dynamic out-growth of the tER .


12/13


Thanks for reagents and advice to Randy Schekman, Alex Franzusoff,Chris Kaiser, Akihiko Nakano, Vytas Bankatis, Suresh Subramani, JimCregg, Steve G ould, Sean Munro, P eter Walter, G erry Waters, TomStevens, Jan Burkhardt, Charles Zuker, and Robert Freedman, and forhelp with microscopy to B ob J osephs, Yimei Chen, Sharon P armet, JudithAustin, and J im McIlvain. We are grat eful to Ada m Linstedt, Lelio O rci,and Hewson Sw ift for help with initial explorations of the P . pastoris sys-tem, and to Nava Segev, Vivek Malhotra, and Adam Hammond for criti-cal reading of the manuscript.

O.W. Rossanese, J. Soderholm, and B .J. B evis were supported by Na-tional Institutes of Health training grant 5-20942. E.K. Williamson oper-ates the EM core facility funded by the U niversity of C hicago Cancer R e-search Center. B.S. G lick was supported by Nat ional Science Foundationgrant M CB -9604342, a Pilot a nd Feasibility Study Awa rd from the D iabe-tes Research Foundation, a Basil OConnor Starter Scholar ResearchAw ard (5-FY 96-1138) from the Ma rch of D imes Birth D efects Founda -tion, a Young Investigator A ward from the Ca ncer Research Foundation,an Institutional Research G rant from the American Ca ncer Society, and agrant from the P ew Cha ritable Trusts.

Received for publication 4 June 1998 and in revised form 23 February1999.

References

Acharya, U ., A. Mallabiabarrena, J .K. Acharya, and V. Malhotra. 1998. Signal-ing via mitogen-activated protein kinase (MEK1) is required for G olgi frag-mentation during mitosis. Cell . 92:183192.

Antebi, A., and G .R. Fink. 1992. The yeast Ca 2 -ATPase homologue, PMR1, isrequired for normal G olgi function a nd localizes in a novel G olgi-like distri-bution. M ol. Biol. Cell . 3:633654.

Ayscough, K., N.M.A. H ajibagheri, R . Watson, and G . Warren. 1993. Stackingof G olgi cisternae in Schizosaccharomy ces pom be requires intact microtu-bules. J. Cell Sci. 106:12271237.

Ba nnykh, S.I., and W.E. Ba lch. 1997. Membrane dynamics at the endoplasmicreticulum-G olgi interface. J. Cell B iol. 138:14.

B arlowe, C ., and R . Schekman. 1993. SEC12 encodes a guanine-nucleotide-exchange factor essential for transport vesicle budding from the ER. Nature .365:347349.

Ba rr, F.A., M. Puype, J. V andekerckhove, and G . Warren. 1997. GR ASP 65, aprotein involved in the stacking of G olgi cisternae. Cell . 91:253262.

Bea ms, H.W., and R .G . Kessel. 1968. The G olgi apparatus: structure and func-tion. In t. Rev. Cytol. 23:209276.

Bndetti, H ., S. R aths, F. Crusaz, and H. R iezman. 1994. The E N D 3 gene en-codes a protein that is required for the internalization step of endocytosisand for actin cytoskeleton organization in yeast. M ol. Biol. Cell . 5:10231037.

Bo evink, P., K. Oparka, S. Santa Cruz, B. Ma rtin, A. Betteridge, and C. H awes.1998. Stacks on tracks: the plant G olgi appara tus traffics on an actin/E R net -work. Plant J. 15:441447.

Bonfanti, L., A.A. Mironov, Jr., J. Martnez-Menrguez, O. Martella, A.Fusella, M. Ba ldassarre, R. B uccione, H.J. G euze, A.A. Mironov, and A.Luini. 1998. Procollagen tra verses the Go lgi stack without leaving the lumenof cisternae: evidence for cisternal maturation. Cell . 95:9931003.

Bra cker, C.E., D .J. Morr, and S.N. G rove. 1996. Structure, differentiation andmultiplication of G olgi apparatus in fungal hyphae. Protoplasma . 194:250274.

B urkhardt , J.K . 1998. The role of microtubule-based motor pro teins in main-taining the structure and function of the G olgi complex. Biochim. B iophys.A cta . 1404:113126.

Chapman, R .E., and S. Munro. 1994. The functioning of the yeast G olgi appara-tus requires an ER protein encoded by A N P 1 , a member of a new family ofgenes affecting the secretory pathway. EM BO (Eur. Mol. Biol. Organ.) J. 13:48964907.

Cha ppell, T.G ., and G . Warren. 1989. A galactosyltran sferase from the fissionyeast Schizosaccharomy ces pom be . J. Cell B iol. 109:26932702.

Clamp, J .R., a nd L . H ough. 1965. The periodate oxidation of amino acids withreference to studies of glycoproteins. Bi ochem. J. 94:1724.

Cluett, E.B ., and W.J. B rown. 1992. Adhesion of G olgi cisternae by protein-aceous interactions: intercisternal bridges as putative adhesive structures. J.Cell Sci. 103:773784.

Cole, N.B., N. Sciaky, A. Marotta, J. Song, and J. Lippincott-Schwartz. 1996.G olgi dispersal during microtubule disruption: regeneration of G olgi stacksat periphera l endoplasmic reticulum exit sites. M ol. Biol. Cell . 7:631650.

D riouich, A., and L.A . Staehelin. 1997. The plant G olgi apparat us: structuralorganization and functional properties. I n The G olgi Apparatus. E.G .B erger and J. Ro th, editors. Birkh user, B asel, Switzerland . 275301.

D unphy, W.G ., and J.E . Ro thman. 1985. Compartmental organization of theG olgi stack. Cell . 42:1321.

Farquhar, M.G . 1985. P rogress in unraveling pathways of G olgi traffic. Annu.

Rev. Cell Bi ol. 1:447488.Farquhar, M.G ., and G .E. P alade. 1981. The G olgi apparatus (complex)

(19541981)from a rtifact to center sta ge. J. Cell B iol. 91:77s103s.Franzusoff, A., K. R edding, J. Crosby, R .S. Fuller, and R . Schekman. 1991. Lo-

calization of components involved in protein transport and processingthrough the yeast G olgi apparatus. J. Cell B iol. 112:2737.

G aynor, E.C ., and S.D. E mr. 1997. COPI -independent anterograde transport:cargo-selective ER to G olgi protein transport in yeast CO PI mutants. J. Cell Biol. 136:789802.

G aynor, E.C ., S. te Heesen, T.R. G raham, M. Aebi, and S.D . Emr. 1994. Signal-mediated retrieval of a membrane protein from the G olgi to the ER in yeast.J. Cell B iol. 127:653665.

G lick, B.S. 1996. Cell biology: alterna tives to bakers yeast. Curr. Biol. 6:15701572.

G lick, B .S., and V . Malhotra. 1998. The curious status of the G olgi apparatus.Cell . 95:883889.

G ould, S.J., D. M cCollum, A.P. Spong, J.A. H eyman, and S. Subramani. 1992.D evelopment of the yeast Pichia pastoris as a model organism for a geneticand molecular analysis of peroxisome assembly. Y east . 8:613628.

G riffiths, G . 1993. Fine Structure Immunocytochemistry. Springer-Verlag, B er-lin. 459 pp.

Ha rris, S.L., and M.G . Waters. 1996. Localization of a yeast early G olgi manno-syltransferase, O ch1p, involves retrograde transport. J. Cell B iol. 132:985998.

Hell, S.W., and E.H.K. Stelzer. 1995. Lens aberrations in confocal fluorescencemicroscopy. I n Handbook of Biological Confocal Microscopy. J.B. Pawley,editor. P lenum Publishing C orp., New Y ork. 347354.

Higgins, D.R., and J.M. Cregg. 1998. Pichia P rotocols. Methods in M olecularBiology. Vol. 103. J.M. Walker, editor. Humana Press, Totowa, NJ. 270 pp.

Jacobs, C .W., A.E .M. Ada ms, P.J. Szaniszlo, and J .R. Pringle. 1988. Functionsof microtubules in the Saccharo myces cerevisiae cell cycle. J. Cell B iol. 107:14091426.

Ka iser, C.A., and R . Schekman. 1990. Distinct sets of SEC genes govern trans-port vesicle formation and fusion early in the secretory pathway. Cell . 61:723733.

Krgel, E ., R. Menzel, H. Ho neck, F. Vogel, A. Bhmer, and W.-H. Schunck.1996. Candida maltosa NAD PH -cytochrome P450 reductase: cloning of afull-length cD NA, hetero logous expression in Saccharom yces cerevisiae a ndfunction of the N-terminal region for membrane anchoring and proliferationof the endoplasmic reticulum. Y east . 12:333348.

Kreis, T.E. 1990. Role of microtubules in the organisation of the G olgi appara -tus. Cell M otil. Cytoskelet. 15:6770.

Kuehn, M.J., and R. Schekman. 1997. CO PII and secretory cargo capture intotransport vesicles. Curr. Opin. Cell B iol. 9:477483.

Kuge, O., C. D ascher, L. Orci, T. Row e, M. Amherdt, H. Plutner, M. Ra vaz-zola, G . Tanigawa , J.E . R othman, and W.E. Ba lch. 1994. Sar1 promotes ves-icle budding from the endopla smic reticulum but not G olgi compartments. J.Cell Biol . 125:5165.

Lacey, A.J. 1989. Light Microscopy in Biology. A Practical Approach. IRLPress, Oxford, U K. 329 pp.

Lew , D.J ., T. Weinert, and J .R . Pringle. 1997. Cell cycle control in Saccharomy- ces cerevisi ae . I n The Molecular and Cellular Biology of the Yeast Saccharo- myces . Vol. 3. J.R. Pringle, J.R. Broach, and E.W. Jones, editors. ColdSpring Harbor Labo ratory P ress, Cold Spring Harbor, NY. 607695.

Lippincott-Schwa rtz, J. 1998. Cytoskeleta l proteins and G olgi dynamics. Curr.Opin. Cell Biol . 10:5259.

Love, H.D ., C.-C. Lin, C.S. Short, and J. O stermann. 1998. Isolation of func-tional G olgi-derived vesicles with a po ssible role in retrograde tra nsport. J.Cell Biol . 140:541551.

Lupa shin, V.V., S. Ha mamoto , and R.W. Schekman. 1996. Biochemical re-quirements for the targeting and fusion of ER-derived transport vesicleswith purified yeast G olgi membranes. J. Cell B iol. 132:277289.

Lussier, M., A.M. Sdicu, T. Ketela, and H. Bussey. 1995. Localization and tar-geting of the Saccharom yces cerevisiae K re2p/Mnt 1p 1,2-mannosyltrans-ferase to a medial -G olgi compartment. J. Cell B iol. 131:913927.

Makarow, M. 1988. Secretion of invertase in mitotic yeast cells. E M B O ( E u r.M ol. Biol. Or gan.) J. 7:14751482.

Mironov, A.A., P. Weidman, and A. Luini. 1997. Variations on the intracellulartransport theme: maturing cisternae and trafficking tubules. J. Cell B iol. 138:481484.

Mollenhauer, H.H., a nd D .J. Morr. 1978. Structural compartmentation of thecytosol: zones of exclusion, zones of adhesion, cytoskeletal and intercisternalelements. Subcell. Bi ochem. 5:327359.

Mollenhauer, H.H., and D .J. Morr. 1991. Perspectives on G olgi apparatusform and function. J. Electron M icrosc. Tech. 17:214.

Morr, D .J. 1987. The G olgi appara tus. In t. Rev. Cytol. 17:211253.Nakanishi-Shindo, Y., K . Nakayama, A. Tanaka, Y. Toda, a nd Y . Jigami. 1993.

Structure of the N -linked oligosaccharides that show the complete loss of-1,6-polymannose outer chain from och1 , och1 mnn1 , and och1 mnn1 alg3

mutants of Saccharom yces cerevisiae . J. Biol. Chem. 268:2633826345.Nakano, A., D . Bra da, and R. Schekman. 1988. A membrane glycoprotein,

Sec12p, required for protein transport from the endoplasmic reticulum tothe G olgi apparatus in yeast. J. Cell B iol. 107:851863.

Naka no, A., and M. Mura matsu. 1989. A novel G TP-binding protein, Sar1p, isinvolved in transport from the endoplasmic reticulum to the G olgi appara-tus. J. Cell B iol. 109:26772691.


13/13

Nakayama, K., T. Nagasu, Y. Shimma, J. Kuromitsu, and Y. Jigami. 1992.O C H 1 encodes a novel membrane bound mannosyltransferase: outer chainelongation of asparagine-linked oligosaccharides. EM BO (Eur. Mol . B iol .Or gan.) J. 11:25112519.

Nishikawa , S., and A . Nakan o. 1991. The G TP-binding Sar1 protein is localizedto the early compartment of the yeast secretory pathway. Biochim. B iophys.Acta . 1093:135143.

Nishikawa, S., and A. Nakano. 1993. Identification of a gene required for mem-brane protein retention in the early secretory pathway. Proc. Natl. A cad. Sci.U SA . 90:81798183.

Nishikawa, S.-I., A. Hirata, and A. Nakano. 1994. Inhibition of endoplasmicreticulum (E R )-to-G olgi transport induces relocalizatio n of binding protein(BiP ) within the ER to form the BiP bodies. M ol. Biol. Cell . 5:11291143.

Novick, P., C. Field, and R. Schekman. 1980. Identification of 23 complementa-tion groups required for post-translational events in the yeast secretorypathway. Cell . 21:205215.

Orci, L., M. Ravazzola, P. Meda, C. Holcomb, H.-P. Moore, L. Hicke, and R.Schekman. 1991. Mammalian Sec23p homologue is restricted to the endo-plasmic reticulum transitional cytoplasm. Proc. N atl. A cad. Sci. USA . 88:86118615.

Orlean, P. 1997. Biogenesis of yeast wall and surface components. I n The Mo-lecular and C ellular B iology of the Yeast Saccharomyces . Vol. 3. J.R. Prin-gle, J.R. B roach, and E.W. Jones, editors. Cold Spring Harbor La boratoryPress, Co ld Spring H arbo r, NY. 229362.

Paccaud, J.-P., W. Reith, J.-L. Carpentier, M. Ravazzola, M. Amherdt, R.Schekman, and L. Orci. 1996. Cloning and functional characterization ofmammalian homologues of the COPII component Sec23. M ol. Biol. Cell .7:15351546.

Pa lade, G . 1975. Intra cellular aspects of the process of protein synthesis. Sci- ence . 189:347358.

Pelham, H.R .B. 1998. G etting through the G olgi complex. Trends Cell Biol .8:4549.

Presley, J.F., N.B. Cole, T.A. Schroer, K. H irschberg, K.J.M. Z aal, and J. L ip-pincott-Schwart z. 1997. ER -to-G olgi transport visualized in living cells. N a- ture . 389:8185.

Preuss, D ., J. Mulholland, A. Franzusoff, N. Segev, and D . B otstein. 1992.Chara cterization of the Saccharomyces G olgi complex through the cell cycleby immunoelectron microscopy. M ol. Biol. Cell . 3:789803.

Preuss, D ., J. Mulholland, C.A. Ka iser, P. Orlean, C. A lbright, M.D. R ose, P.W.R obbins, and D . Bot stein. 1991. Structure of the yeast endoplasmic reticu-lum: localization of ER proteins using immunofluorescence and immuno-electron microscopy. Y east . 7:891911.

Pringle, J.R., A.E.M. A dams, D.G . Drubin, and B .K. Ha arer. 1991. Immunoflu-orescence methods for yeast. M ethods Enzymol. 194:565602.

Pryer, N.K., N.R. Salama, R. Schekman, and C.A. Kaiser. 1993. CytosolicSec13p complex is required for vesicle formation from the endoplasmicreticulum in vitro. J. Cell B iol. 120:865875.

Ra bouille, C., a nd G . Warren. 1997. Changes in the architecture of the G olgiapparatus during mitosis. I n The G olgi Apparatus. E.G . Berger and J. Roth,editors. B irkhuser, B asel, Switzerla nd. 195217.

Ra mbourg, A., Y. C lermont, and F. Kps. 1993. Modulation of the G olgi ap-paratus in Saccharo my ces cerevisi ae sec7 mutants as seen by three-dimen-sional electron microscopy. A nat. Rec. 237:441452.

Ra mbourg, A., Y. C lermont, J.M. Nicaud, C. G aillardin, and F. Kepes. 1996.Transformations of membrane-bound organelles in sec14 mutants of theyeasts Saccharom yces cerevisiae and Yarrowia lipolytica . A nat. Rec. 245:447458.

Rambourg, A., Y. Clermont, L. Ovtracht, and F. Kps. 1995. Three-dimen-sional structure of tubular netw orks, presumably G olgi in nature, in variousyeast strains: a comparative study. A nat. Rec. 243:283293.

Redding, K ., C. Holcomb, and R.S. Fuller. 1991. Immunolocalization of K ex2protease identifies a putative late G olgi compartment in the yeast Saccharo- myces cerevisiae . J. Cell B iol. 113:527538.

Rieder, S.E., L.M. B anta, K . Khrer, J.M. McCaf fery, and S.D. E mr. 1996. Mul-tilamellar endosome-like compartment accumulates in the yeast vps28 vacu-olar protein sorting mutant. M ol. Biol. Cell . 7:985999.

Ro se, M.D., L .M. Misra, and J .P. Vo gel. 1989. K A R 2 , a karyogamy gene, is theyeast homolog of the mammalian B iP/G RP 78 gene. Cell . 57:12111221.

Rothman, J.E., and F.T. Wieland. 1996. Protein sorting by transport vesicles.Science . 272:227234.

Rothstein, R. 1991. Targeting, disruption, replacement, and allele rescue: inte-grative DNA transformation in yeast. M ethods Enzymol. 194:281301.

Ro we, T., C. D ascher, S. B annykh, H. P lutner, and W.E. B alch. 1998. Role ofvesicle-associated synta xin 5 in the assembly of pre-G olgi intermediates. Sci- ence . 279:696700.

Scales, S.J., R. P epperkok, and T.E. K reis. 1997. Visualization of E R -to-G olgitransport in living cells reveals a sequential mode of action for C OP II andCOPI. Cell . 90:11371148.

Schnepf, E . 1993. G olgi appara tus and slime secretion in plants: the early impli-cations and recent models of membrane traffic. Protoplasma . 172:311.

Sears, I.B., J . OC onnor, O.W. Rossanese, and B .S. G lick. 1998. A versatile setof vectors for constitutive and regulated gene expression in Pichia pastoris .Y east . 14:783790.

Segev, N., and D . B otstein. 1987. The ras -like yea st YPT1 gene is itself essentialfor growth, sporulation, and starvation response. M ol. Cell. Biol. 7:23672377.

Segev, N., J. Mulholland , and D . Bo tstein. 1988. The yeast G TP-binding Y PT1protein and a mammalian counterpart are associated with t he secretion ma-chinery. Cell . 52:915924.

Shaywitz, D .A., L. Orci, M. Ra vazzola, A. Swaroop, a nd C.A. K aiser. 1995. Hu-man SEC13Rp functions in yeast and is located on transport vesicles bud-ding from the endoplasmic reticulum. J. Cell B iol. 128:769777.

Sitia, R., and J. Meldolesi. 1992. Endoplasmic reticulum: a dynamic patchworkof specialized subregions. M ol. Biol. Cell . 3:10671072.

Staehelin, L .A., and I . Moore. 1995. The plant G olgi appara tus: structure, func-tional organization and trafficking mechanisms. A nnu. Rev. Plant Physiol.Plant M ol. Biol. 46:261288.

Stanley, H., J . B otas, and V. Malhotra. 1997. The mechanism of G olgi segrega-tion during mitosis is cell type-specific. Proc. Natl. A cad. Sci. USA . 94:1446714470.

Storrie, B., J . White, S. R ttger, E.H .K. Stelzer, T. Suganuma, and T. Nilsson.1998. Recycling of G olgi-resident glycosyltransfera ses through the ER re-veals a novel pathway and provides an explanation for nocodazole-inducedG olgi scattering. J. Cell B iol. 143:15051521.

Svodo ba, A., and O . Necas. 1987. Ultra structure of Saccharom yces cerevisiae cells accumulating G olgi organelles. J. Basic M icrobiol. 27:603612.

Tang, B.L ., F. Peter, J. Krijnse-Locker, S. Low, G . G riffiths, and W. Hong.1997. The mammalian ho molog of yeast Sec13p is enriched in the int ermedi-ate compartment and is essential for protein transport from the endoplasmicreticulum to the G olgi apparatus. M ol. Cell. Biol. 17:256266.

Thyberg, J., and S. Moskalewski. 1985. Microtubules and the organization ofthe G olgi complex. Exp . Cell Res. 159:116.

Tokuyasu, K.T. 1989. Use of poly(viny lpyrrolidone) a nd poly(vinyl a lcohol) forcryoultramicrotomy. H istochem. J. 21:163171.

Trimble, R.B., P.H. Atkinson, J.F. Tschopp, R.R. Townsend, and F. Maley.1991. Structure of oligosaccharides on Saccharom yces SUC2 invertase se-creted by the methylotrophic yeast Pichia pastoris . J. Biol. Chem. 266:2280722817.

Waterham, H.R., M.E . D igan, P.J. K outz, S.V. Lair, and J .M. Cregg. 1997. Iso-lation of the Pichia pastoris glyceraldehyde-3-phosphate dehydrogenasegene and characterization of its promoter. Gene . 186:3744.

Weiss, D.G ., W. Maile, and R .A. Wick. 1989. Video microscopy. I n Light Mi-croscopy in Biology. A Practical Approach. A.J. Lacey, editor. IRL Press,Oxford , U K. 221278.

Whaley, W.G . 1975. The G olgi Appa ratus. Springer-Verlag, Vienn a. 190 pp.Wolf, J., M. Nicks, S. D eitz, E. va n Tuinen, and A . Franzusoff . 1998. An N-end

rule destabilization muta nt reveals pre-G olgi requirements for Sec7p inyeast membrane traffic. Bi ochem. Bi ophys. Res. Comm un. 243:191198.

Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage clon-ing vectors and host strains: nucleotide sequences of the M13mp18 andpU C19 vectors. Gene . 33:103119.

Documents

Rossana Ese