The subapical actin cytoskeleton regulates secretion and

81Journal of Cell Science 112, 81-96 (1999)Printed in Great Britain © The Company of Biologists Limited 1998JCS9862

The subapical actin cytoskeleton regulates secretion and membrane retrieval

in pancreatic acinar cells

Karine M. Valentijn, Francine D. Gumkowski and James D. Jamieson

Department of Cell Biology, Yale University School of Medicine, New Haven, CT, USAAuthor for correspondence (e-mail: [email protected])

Accepted 29 October; published on WWW 8 December 1998

We examined the effects of disruption of the actincytoskeleton by cytochalasin D (cytoD) on basal andcarbamylcholine-stimulated exocytosis and oncompensatory membrane retrieval in pancreatic acinarcells. Although the involvement of actin in exocytosis isreasonably well established, its role in these coupledprocesses is not understood. Our findings suggested thatcytoD inhibited stimulated secretion of amylase. However,morphometry revealed that exocytosis had occurred: thenumber of zymogen granules decreased, the size of thelumen increased, and large vacuolar structures continuouswith the lumen formed into which amylase accumulated.Large amounts of amylase were released to the medium onremoval of secretagogue and cytoD, suggesting that thesubapical actin network provides contractile forces thatexpel the lumenal contents. Strikingly, we observed that atthe apical pole of the cells where exocytosis occurred, cytoD

induced an accumulation of membrane invaginations intoa vastly enlarged apical membrane. These pits were oftensurrounded by a clathrin-like coat. Concomitantly, AP-2-,clathrin-, dynamin- and caveolin-like immunoreactivityconcentrated around the enlarged lumina, suggesting thatincorporation of zymogen granule membrane into theapical plasma membrane triggered the recruitment of theseproteins. After wash out of cytoD and carbamylcholine andreformation of the subapical actin cytoskeleton, the coatedinvaginations largely disappeared in association with areduction in lumenal size, and relocation of clathrin, AP-2,dynamin and caveolin into the cell. We suggest that theactin terminal web also controls compensatory membraneretrieval following exocytosis.

Key words: Actin, Exocytosis, Membrane retrieval, Acinar cell,Cytochalasin D, Dynamin, Clathrin, Caveolin, AP-2

SUMMARY

l., ofedreers,5;

wns,hently

p,on

ceryaThisicalas

gli

INTRODUCTION

The plasma membrane of eukaryotic cells is continuouremodeled due to incorporation of membrane duriexocytosis and compensatory removal of excess membranendocytosis (Arvan and Castle, 1998). The cytoskeleton mrepresent a regulator of the equilibrium between these functions since evidence indicates its involvement in bomembrane trafficking events (Trifaró and Vitale, 199Gottlieb et al., 1993; Jackman et al., 1994; Durrbach et 1996; Lamaze et al., 1997).

The participation of the cytoskeleton in exocytosis has bestudied in various models (Burgoyne and Cheek, 1987; Auand Bader, 1988; Trifaró and Vitale, 1993) where evidensuggests that under resting conditions, the actin cytoskelewhich is localized under the plasma membrane, prevesecretory granules from reaching their exocytic destinationagreement with this hypothesis, it has recently been shownphalloidin, which prevents actin disassembly, blocstimulated secretion in permeabilized pancreatic acinar c(Muallem et al., 1995). In most models, the actin cytoskeleis disassembled upon stimulation and rearranged therallowing secretory granules to reach the site of exocyto

slynge byay

twoth

3;al.,

ennisceton,nts. In thatksellstonebysis

(Cheek and Burgoyne, 1986; Vitale et al., 1991; Perrin et a1992). However, in the pancreatic acinar cell, disassemblycortical actin during stimulation of secretion has been reportonly in conditions where doses of secretagogue wesupramaximal and which have been shown to induce a lowsecretory response and pancreatitis (Burnham and William1982; O’Konski and Pandol, 1990; Jungermann et al., 199Muallem et al., 1995). Interestingly, several studies have shothat disruption of the actin cytoskeleton in exocrine glandinhibits stimulated secretion (Williams, 1977; Tojyo et al.1989; Muallem et al., 1995), which appears to contradict tbarrier hypothesis. Based on these observations, it has recebeen proposed that in addition to a role as a regulated clamthe actin cytoskeleton may also play a positive role in secretifrom exocrine glands (Muallem et al., 1995).

When exocytosis from glandular epithelial cells takes plathere is a significant amount of membrane from secretogranules which is incorporated into the apical plasmmembrane (Jamieson and Palade, 1971; Jena et al., 1994). excess membrane must be retrieved to prevent the apmembrane surface area from enlarging dramatically. Wherethe mechanisms of membrane retrieval followinneurotransmitter stimulation is well understood (De Camil

82

tehnthere

total

sin et0

de

in10

A,30M

atps.ry

gm

,das.bitr

tedotr30

C-ce1-

titykDa

mat

suepy

sY,ad

K. M. Valentijn, F. D. Gumkowski and J. D. Jamieson

and Takei, 1996), the compensatory membrane retrievalepithelial cells is less well understood. Morphological studihave shown that endocytic fluid phase tracers are taken upclathrin-coated pits and are transported to the Golgi alysosomes during regulated secretion from glandular epithecells (Herzog and Farquhar, 1977; Herzog and Reggio, 198While there is now growing evidence in constitutively secretinpolarized epithelial cells for the involvement of thcytoskeleton in endocytosis, especially in receptor-mediaendocytosis (Gottlieb et al., 1993; Jackman et al., 199Durrbach et al., 1996; Shurety et al., 1996), less is knownglandular epithelial cells.

The pancreatic acinar cell represents an excellent modegain insight into the role of the actin cytoskeleton in thregulation of exocytosis and membrane retrieval followinexocytosis. Using cytochalasin D (cytoD), a specific agewhich disrupts the actin cytoskeleton, we found that tsubapical actin network of pancreatic acinar cells is prerequisite for secretion by providing contractile forces thexpulse secretory proteins from the acinar lumen in respoto secretagogue stimulation without impeding exocytosis pse. Significantly, we found that concomitant treatment pancreatic lobules with cytoD and secretagogue stimulatof exocytosis leads to the accumulation of granule membrain the apical plasmalemma which was reversibly expandThis provided us with the opportunity to examine thmechanism of membrane retrieval during recovery frocytoD treatment. Our results suggest that expansion of apical membrane following exocytosis is a trigger for threcruitment of AP-2, clathrin, dynamin, and caveolin, leadinto compensatory membrane retrieval. The actin cytoskeleis mandatory in the process by regulating an early step in formation of retrieval vesicles and/or movement of membraback into the cell.

MATERIALS AND METHODS

AnimalsSprague-Dawley male rats (about 100 g, Camm Research InstitWayne, NJ, USA), were used throughout this study. Between 12 15 hours before experiments, they were fasted to ensure accumulaof zymogen granules in pancreatic acinar cells. Animals wesacrificed by cervical dislocation.

Pancreatic lobule preparationPancreatic lobules were prepared as previously described (Sch1983) and incubated in a Krebs-Ringer medium. The lobules wpre-incubated for 30 minutes with cytoD (5µg/ml, final DMSOconcentration lower than 0.1%) (Sigma, St Louis, MO, USA), athen stimulated for 60 minutes with carbamylcholine (Sigma, Louis, MO, USA) at a maximal effective dose of 10 µM (data notshown). At the end of the experiment, the lobules were either plainto fixative or used to measure amylase according to the methoBernfeld (1955) in both the incubation medium and the tissue. Reswere expressed as the percentage of total amylase released intincubation medium over 90 minutes.

In experiments using the fluid-phase marker FITC-dextr(10,000 MW, anionic, lysine fixable, Molecular Probes, Eugene, OUSA), the lobules were placed in the fluorescent probe solution (mg/ml) at the onset of the experiment.

In recovery experiments, lobules were incubated for 4 hours in presence of atropine (1 µM) (Sigma, St Louis, MO, USA) to block

ines viandlial0).g

eted4; in

l toegnt

hea

atnseerofionne

ed.emtheeg

tonthene

ute,andtionre

eele,ere

ndSt

cedd ofultso the

anR,1.5

the

the stimulatory effect of carbamylcholine. Thereafter a 60 minustimulation was given to lobules with 1 nM bombesin (ResearcBiochemicals Incorporated, Natick, MA, USA) to check for secretioresponsiveness after recovery. Amylase was measured in incubation medium from the initial 90 minute incubation, the 4 hourecovery period, the 60 minute bombesin stimulation and in thlobules, and the results were expressed as the percentage of amylase released during each incubation period.

ATP measurementsLobules were homogenized in ice-cold 2% TCA and ATP wadetermined with a firefly luciferase assay (Sigma). Proteconcentrations were determined according to the method of Lowryal. (1951). To deplete intracellular ATP, lobules were incubated for 9minutes with 5 µM antimycin A1 (Sigma, St Louis, MO, USA) andin the absence of glucose (replaced by 14 mM 2-deoxy-glucose).

Confocal microscopyLobules were fixed at 4°C for 24 hours in either 3% paraformaldehy(for immunofluorescence) or 3% paraformaldehyde, 0.25%glutaraldehyde (phalloidin staining and fluorescent dextran studies)5 mM phosphate buffer, pH 7.4, containing 0.3 M sucrose. 8 to µm frozen sections were prepared for fluorescence microscopy.

For immunofluorescence, sections were blocked with solution (20 mM phosphate buffer containing 0.5 M NaCl, pH 7.4supplemented with 0.1% saponin and 5% normal goat serum) for minutes. When glutaraldehyde was used, a washing step with 100 msodium borohydride in Solution A (without saponin and normal goserum) for 10 minutes was added to quench free aldehyde grouSections were incubated for 2-3 hours with primary and secondaantibodies diluted in solution A. Commercially available primaryantibodies used in this study were obtained from the followinsources: mouse anti-clathrin heavy chain antibody, clone 23, froTransduction Laboratories (Lexington KY, USA), mouse anti-α-adaptin (AP-2) antibody, clone 100/2, from Sigma (St Louis, MOUSA), mouse anti-caveolin-I antibody, clone Z034, from ZymeLaboratories (San Francisco, CA, USA). Rabbit anti-dynamin-2 wkindly provided by Dr R. Vallee (Worcester, Shrewsbury, MA)Secondary antibodies were goat anti-mouse or goat anti-rabcoupled to FITC, Texas Red, Alexa 488 or Alexa 594 (MoleculaProbes, Eugene, OR, USA). For controls, sections were incubawith secondary antibody alone which produced no staining (data nshown). Texas Red phalloidin (20 units/ml in solution A, MoleculaProbes, Eugene, OR, USA) was incubated with sections for minutes.

Fluorescence was visualized under either an MRC-600 or an MR1024 confocal laser scanning microscope (Bio-Rad MicroscienDivision, Cambridge, MA, USA) and optical sections ranging from to 2 µm were recorded at a resolution of 768 by 512 pixels (MRC600) or 512 by 512 pixels (MRC-1024) and 256 gray levels.

Antibody productionA polyclonal antiserum was raised in rabbits using purified rapancreatic amylase (Schramm and Loyter, 1966). Antibody specificwas assessed by western blots which detected an expected 55 band.

Electron microscopyLobules were fixed overnight in 2% glutaraldehyde in 0.1 M sodiucacodylate buffer, pH 7.4, at 4°C for 24 hours, treated for 1 hour room temperature with 1% OsO4 in cacodylate buffer, and furtherstained with 1% tannic acid (Simionescu and Simionescu, 1976). Tiswas dehydrated and embedded in Epon 812 (Electron MicroscoSciences, Fort Washington, PA, USA). Thick sections (0.5 µm) werestained with Toluidine Blue and examined by light microscopy (ZeisD7082 Axiophot Photomicroscope, Carl Zeiss, Inc., Thornwood, NUSA). Thin sections were counterstained with uranyl acetate and le

83Cytochalasin D inhibits membrane retrieval

was

hede-

ers,

ed. 1.

citrate and visualized under a Philips EM410 transmission electmicroscope (Philips Inc. Eindhoven, The Netherlands).

For immunoelectron microscopy, lobules were fixed for 24 houat 4°C in 3% paraformaldehyde and 0.25% glutaraldehyde in 5 mphosphate buffer, pH 7.4, containing 0.3 M sucrose and the tissueprocessed as previously described (Valentijn et al., 1997) localization on LR-White thin sections.

Morphometric analysisPrints from electron micrographs were scanned and the digital imawere analyzed with NIH image program (National Institutes Health, USA). The lumen perimeter of 15 acini from non-treated acytoD-treated lobules, 23 acini from carbamylcholine-treated lobuand 21 acini from cytoD and carbamylcholine-treated lobules wchosen randomly from 1 to 2 blocks from 4 independent experimeand the results were expressed as the mean ± s.e.m of the luperimeter in µm. The zymogen granules from all the acinar celsurrounding each analyzed lumina were counted and the results expressed as mean ± s.e.m of the total number of zymogen granper acinar lumen.

Fig. 1.CytoD in the presence of carbamylcholine induces dramaticfixed pancreatic lobules were stained with Toluidine Blue. LobulescytoD (C), or treated with both cytoD and carbamylcholine (D). Cystimulation with carbamylcholine, vacuolar structures (v) which ar

ron

rsM

wasfor

gesofndleserentsmenlswereules

Statistical analysisResults were expressed as mean ± s.e.m. Statistical significancedetermined using a two-tailed unpaired Student’s t-test.

RESULTS

The purpose of this study was to examine the role of tsubapical actin network in regulated exocytosis ancompensatory membrane retrieval following secretagogustimulated exocytosis. To this end, we used cytoD which entcell and binds to actin to inhibit its polymerization (Schliwa1982; Cooper, 1987; Ohmori and Toyama, 1992).

CytoD induces dramatic morphological changes inpancreatic lobulesThe effects of cytoD (5 µg/ml, an optimal dose; see below), onpancreatic lobules were examined on Toluidine Blue stain0.5µm sections. Representative examples are shown in Fig

changes in the morphology of pancreatic acinar cells. 0.5 µm sections from were non-treated (A), stimulated with carbamylcholine (B), treated withtoD induces an enlargement of the lumina (*) and, during concomitante often in contact with the lumen. Bar, 10 µm.

84

n-ndngith inDig.ateslyndntht

sohat

dce

edIne).lls.f

ed


Lobules not treated with cytoD displayed numerous zymoggranules located at the apical pole of the acinar cells (Fig. 1When pancreatic lobules were stimulated wicarbamylcholine (10 µM, maximal effective dose), a reductioin the number of zymogen granules was noticed (Fig. 1Treatment with cytoD alone induced an enlargement of lumen; occasionally, vacuolar structures were observed incytoplasm of cytoD-treated acinar cells (Fig. 1C). Whlobules pre-treated with cytoD were stimulated wicarbamylcholine, the lumina were even further enlarged (F1D) and the number of zymogen granules was reducedaddition, the vacuolar structures already seen in tissues trewith cytoD alone occurred more frequently and were alarger. Interestingly, these vacuolar structures appeared tin contact with the lumen. Experiments in which FITC-dextraas a fluid phase tracer, was administered to the lobules duthe various treatments, showed that not only were the lumfilled with the fluorescent probe but also the vacuolar structu(Fig. 2). These experiments also showed that under conditions, intracellular labeled vesicles, probably endocycompartments whose nature is yet to be determined, wlocated in the supranuclear region of the cells.

In order to determine whether the morphological changobserved by light microscopy, were due to an effect of cyton the subapical actin cytoskeleton, we stained fixed fro

Fig. 2.CytoD treatment induces an accumulation of the fluid-phasacinar cells. Lobules were treated with cytoD and carbamylcholinconfocal microscope. Lobules were non-treated (A), stimulated wand carbamylcholine (D). Lumen (*), nucleus (N). Bar, 10 µm.

enA).

thnB).the theenthig.

. Inatedlsoo ben,ringinaresallticere

esoDzen

sections of lobules with Texas Red phalloidin (Fig. 3). In notreated lobules, strong phalloidin-staining was observed arouthe acinar lumen (Fig. 3A). Fainter staining was present alothe lateral membranes. Stimulation of pancreatic lobules wcarbamylcholine alone did not induce appreciable changesthe actin distribution (Fig. 3B). On the other hand, cytoalmost completely dispersed the subapical actin network (F3C), and induced the appearance of an intracellular punctstaining. Formation of actin aggregates has been previoureported to occur upon cytochalasin treatment (Edmonds aKoenig, 1989). Strong staining remained between adjacecells which may represent actin filaments emanating from tigand adhering junctions. A similar staining pattern was alobserved when the lobules were stimulated witcarbamylcholine in the presence of cytoD (Fig. 3D), except ththe lumina were expanded.

To examine more precisely the extent to which cytoD ancarbamylcholine caused changes in the lumen circumferenand the number of zymogen granules, we performquantitative electron microscopy (Table 1, Figs 4 and 5). non-treated tissue, lumina were small and microvilli werpresent at the apical membrane (Table 1; Fig. 4A and BZymogen granules were grouped at the apical poles of the ceIn carbamylcholine-stimulated lobules, the number ozymogen granules was significantly lower than in non-treat

e marker FITC-dextran into lumina and vacuolar structures of pancreatice in the presence of FITC-dextran, and cryosections were visualized under aith carbamylcholine (B), treated with cytoD (C) or treated with both cytoD


genntirh

esnnenes

Fig. 3.CytoD induces a marked disruption of the actin cytoskeleton in pancreatic acinar cells. Cryosections of fixed pancreatic lobules werestained with Texas Red phalloidin. Lobules were non-treated (A), stimulated with carbamylcholine (B), treated with cytoD (C) or treated withboth cytoD and carbamylcholine (D). Intracellular punctate staining appears during cytoD treatment. Lumen (*), nucleus (N). Bar, 10 µm.

tissue (Table 1 and Fig. 4C). Exocytotic figures were oftobserved (Fig. 4D). When compared to non-treated lobulumina were enlarged (by twofold) and mostly devoid microvilli (Table 1 and Fig. 4D). In cytoD-treated lobules, bounstimulated and stimulated, the lumina were significanexpanded by 1.5-fold (Table 1; Fig. 5A and B) and 3.5-fo(Table 1; Fig. 5C and D), respectively, compared to non-treatissue (Table 1; Fig. 4A and B).

The dramatic enlargement of the lumina i

mest,

blein theedicallar

fndrelyasere

Table 1. Effect of cytoD and carbamylcholine on thenumber of zymogen granules and the size of the acinar

lumenParameter

Number of zymogen Lumen perimeterTreatment per acinar lumen (µm) per acinus

Control 166.2±17.8 13.7±2.7Carbamylcholine 114.5±9.8** 31.9±4.4**CytoD 132.1±17.3 23.5±3.4*CytoD + Carbamylcholine 112.0±10.5** 49.8±6.9***

Lobules were pre-treated with cytoD (5 µg/ml) for 30 minutes andstimulated with carbamylcholine (10µM) for 60 minutes. Results areexpressed as the mean ± s.e.m (n=13 to 25). *P<0.05, **P<0.01,*** P<0.001, P values compared to non-treated tissue.

enles,ofthtlyldted

n

cytoD/carbamylcholine-treated lobules was in part due to larvacuolar structures whose continuity with the acinar lume(Fig. 5C and D) was confirmed by accessibility to fluorescedextran (Fig. 2C,D; see later) and by the similarity of theflocculent electron dense content within the lumen. Althougthe overall number of microvilli remained unchanged, stretchof plasma membrane were devoid of microvilli, especially ithe vacuolar structures (Fig. 5C and D). Quantificatioindicated a reduction of nearly 30% in the number of zymoggranules per lumen in cytoD/carbamylcholine-treated lobulcompared to non-treated tissue (Table 1). This was the samagnitude as in carbamylcholine-treated lobules. In contracytoD alone did not induce any significant changes (P>0.1) inthe number of zymogen granules compared to controls (Ta1). Actin microfilaments were observed around the lumen non-treated tissue, whereas most were absent beneathplasma membrane after cytoD treatment. In cytoD-treattissue, many zymogen granules were docked at the applasma membrane (Fig. 5) and the membranes of vacuostructures (Fig. 5C and D).

Strikingly, cytoD induced a dramatic accumulation omembrane invaginations in the apical membrane (Fig. 5B aD), whereas in non-treated lobules these structures were raobserved. No accumulation of membrane invaginations wobserved at the basolateral plasma membrane. These pits w

86

nttt

re

n

dnerin


Fig. 4.Carbamylcholineinduces a decrease in thenumber of zymogen granulesand an enlargement of theacinar lumen. Thin sectionsof fixed lobules werevisualized by electronmicroscopy at low (A,C) andhigh (B,D) magnification.Lobules were non-treated(A,B) or stimulated withcarbamylcholine (C,D). Non-treated tissue showsnumerous zymogen granulesand microvilli on the apicalplasma membrane (A,B). Thelumina of carbamylcholine-treated lobules are filled withflocculent electron densematerial and largely devoid ofmicrovilli (C,D). Bar, 2 µm(A,C); 1µm (B,D).

frequently covered with a clathrin-like coat (Fig. 5B, insePreliminary quantification revealed that about 70% of tmembrane invaginations were coated whereas 30% appenon-coated (data not shown). Invaginations with short newere occasionally observed continuous with the apiplasmamembrane. Vesicles coated with a clathrin-like coathe vicinity of the apical membrane were rarely detected.

CytoD inhibits carbamylcholine-stimulated amylaserelease from pancreatic lobulesGraded concentrations of cytoD alone (0.001 to 10 µg/ml) hadno effect on basal amylase release (Fig. 6). However, wpancreatic lobules were stimulated with a maximal dose

t).hearedckscalt in

hen(10

µM) of carbamylcholine, cytoD induced a dose-dependeinhibition of amylase release (Fig. 6). The maximal effec(ranging from 60 to 80% inhibition) of cytoD was observed a0.3 µg/ml.

Because other cytochalasins, in particular A and B, aknown to deplete intracellular ATP by inhibitingmonosaccharide transport in addition to altering actipolymerization (Kletzien et al., 1972), we determined ATPlevels in pancreatic lobules incubated with cytoD. Non-treatelobules had basal ATP levels of 8.6±0.9 nmol/mg protei(n=3), treatment with antimycin A1 in the absence of glucosreduced ATP levels by 80%. Treatment with cytoD ocarbamylcholine alone did not cause significant changes


Fig. 5.CytoD induces an accumulation of membrane invaginations. Thin sections of fixed lobules were visualized by electron microscopy atlow (A,C) and high (B,D) magnification. Lobules were treated with cytoD alone (A,B) or in combination with carbamylcholine (C,D).Membrane invaginations (arrowheads) accumulate at the apical membrane in the presence of cytoD and cytoD plus carbamylcholine.Membrane invaginations are often covered with electron dense material resembling clathrin (short arrows in B, inset). Stretches of membranedevoid of microvilli are also observed (open arrows). Zymogen granules are found abutting the plasma membrane (long arrows). Organizedmicrofilaments in the terminal web are generally absent in cytoD treated-tissue. Bars: 2 µm (A,C); 1µm (B and inset, D).

88

ithhadend,indyinaere

s inolarthe

on theomied

ume

stillctsn

ofheingngt,

hent,

r toseasehehe

intheig.einaad

onlarhee.

theblytsd.

ledte


Fig. 6.CytoD inhibits carbamylcholine-stimulated secretion ofamylase from pancreatic acinar cells. Lobules were incubated in tpresence of graded concentrations of cytoD in the absence (opencircles) or presence (filled circles) of carbamylcholine (10 µM).Results are expressed as the mean ± s.e.m of the percentage ofamylase secreted in the incubation medium over 90 minutes (n=3).

Table 2. Effect of withdrawal of cytoD andcarbamylcholine on amylase secretion

Treatment

CytoD +Carbamyl- Carbamyl-

Condition Control choline CytoD choline

Initial 90 minutes 4.0±0.4 30.6±1.6 3.2±0.4 8.2±0.8Recovery 4 hours 6.2±0.5 7.3±0.6 5.8±0.5 21.3±1.3Bombesin 60 minutes 15.7±1.1 14.4±0.1 11.7±2.2 11.2±1.2*

Lobules were pre-treated with cytoD (5 µg/ml) for 30 minutes andstimulated with carbamylcholine (10µM) for 60 minutes still in the presenceof cytoD (initial 90 minutes). The lobules were rinsed and further incubatedfor 4 hours in the presence of atropine (1 µM) (recovery 4 hours). Then,bombesin (1 nM) was administered for 60 minutes.

*CytoD and carbamylcholine treated lobules that were not stimulated withbombesin after the recovery period released less than 3% amylase during the60 minute period. Results are expressed as the mean ± s.e.m. of thepercentage of total amylase released into the incubation medium (n=3 to 35).

basal ATP levels (9.1±0.5 and 9.4±0.4 nmol/mg protein (n=3),respectively) nor did the combination of the two drug(10.1±0.4 nmol/mg protein; n=3), indicating that structural andfunctional changes associated with cytoD treatment were trivial consequences of ATP depletion.

Amylase is secreted into the lumina and vacuolarstructures of cytoD and carbamylcholine-treatedpancreatic lobulesOur observations that the number of zymogen granules reduced when lobules were stimulated with carbamylcholin the presence of cytoD indicates that exocytosis itsproceeded in the presence of cytoD even though the aterminal web was depolymerized. Because the vacuostructures and lumina appeared to contain the same conand zymogen granules were found to be in close contact wtheir membranes, we determined if zymogen granules released secretory proteins into these structures. To this we performed immunogold electron microscopy on thsections from LR-White embedded lobules with an antiboagainst amylase (Fig. 7). As shown in Fig. 7C and D, lumand vacuolar structures as well as zymogen granules wstrongly labeled for amylase.

Recovery of amylase in the incubation medium afterwash out of cytoD and carbamylcholineThe above results led us to speculate that secretory proteinzymogen granules are released into the lumen and vacustructures but are not extruded from the acinar lumen into incubation medium.

One explanation could be that cytoD inhibits the alkalizatiand subsequent solubilization of the secretory products inductal system by blocking chloride/bicarbonate exchange frduct cells (Freedman et al., 1994). However, when we carr

s

not

wasineelfctinlartent,

out the experiments in a sodium bicarbonate buffered mediat pH 8.3, which should compensate for any inhibition of thchloride/bicarbonate exchanger, amylase release was inhibited by cytoD (data not shown), suggesting that the effeof cytoD are not likely due to an indirect effect of the drug oelectrolyte secretion.

Alternatively, by disrupting the actin network, cytoD maycounteract the contractile forces necessary for the extrusionamylase from the acinar lumina. Therefore, we studied teffect of removal of cytoD on amylase release measured dura 4 hour wash out of the drugs. As shown in Table 2, durithe wash out period from cytoD/carbamylcholine treatmen~3-fold higher amounts of amylase were recovered in tmedium when compared to tissue recovering from treatmewith either carbamylcholine or cytoD alone. After recoverytreated lobules responded to bombesin in a manner similanon-treated tissue (Table 2), indicating that the amyladetected into the medium was not due to leakage of amylfrom damaged tissue and was specifically due to tstimulatory effect of the peptide, since in the absence of tsecretagogue less than 3% of amylase was released.

Examination of the tissue stained with Texas Red phalloidindicated that the subapical actin network and the size of lumina were reestablished after wash out of the drugs (F8A). Incubation of the lobules with FITC-dextran during threcovery period also demonstrated that the size of the lumwas restored (Fig. 8B) and that most vacuolar structures hdisappeared in comparison to Fig. 2A. Thin section electrmicroscopy of lobules after wash out of the drugs was simito non-treated lobules (Fig. 1A) as shown in Fig. 9, and tsize of the lumina was similar to that of the non-treated tissuA filamentous electron dense material was observed aroundlumen beneath the plasma membrane, which presumarepresents actin microfilaments. Interestingly, the pisurrounded by a clathrin-like coat were now rarely observe

CytoD and carbamylcholine treatment recruitsclathrin, AP-2, dynamin and caveolin at the apicalmembraneThe detection of electron dense material which resembclathrin around forming endocytic vesicles led us to investiga

he


eretheei,sody

salate

neein,for

whether molecules, that are responsible for the formationcoated vesicles in other cell types, are also implicatedmembrane retrieval in the exocrine pancreaImmunofluorescence studies using an antibody directed agathe heavy chain of clathrin, revealed that the general stainpattern of clathrin in resting pancreatic acinar cells wpunctate and primarily localized to the supranuclear regionan area that may correspond to the Golgi complex (Fig. 10Large vesicular structures that might be lysosomes wstained. No staining was observed at the basal membrWhen membrane retrieval was arrested by cytoD acarbamylcholine treatment, the clathrin staining concentraaround the enlarged lumina and vacuolar structures (Fig. 10After wash out of the drugs, the staining pattern returnedthat observed in control tissue (Fig. 10C).

During the course of the formation of coated vesicle

Fig. 7.Amylase is present in lumina and vacuolar structures of cytand embedded in LR-White and sections were stained with an anNon-treated tissue (A) shows strong staining in the content of zymStaining with the secondary antibody alone of tissue treated with low power electron micrograph (C) of an acinus treated with cytoDin direct contact with the lumen (*). At higher magnification (D), nustructure. Bar, 1 µm.

of ins.instingas, inA).ereane.ndtedB). to

s,

clathrin molecules are recruited to the target membrane whthey associate with adaptor proteins (AP-2 in the case of plasma membrane) (Robinson, 1994; De Camilli and Tak1996). Therefore we investigated whether AP-2 was alrecruited at the apical plasma membrane. Using an antibodirected against the α subunit of the AP-2 complex, we foundthat AP-2 relocated from a Golgi-like area in control lobule(Fig. 11A) to the apical membrane when membrane retrievwas blocked by cytoD and carbamylcholine (Fig. 11B) in pattern similar to that observed with clathrin. After wash ouof the drugs, AP-2 like-immunoreactivity was restored to thGolgi area (Fig. 11C).

Since we had observed that the majority of the membrainvaginations were arrested before pinching off of thmembrane, it was of interest to investigate whether dynamwhich has been postulated to play a role as a ‘pinchase’

oD and carbamylcholine-treated pancreatic acinar cells. Lobules were fixedtibody directed against amylase and a 10 nm-gold labeled secondary antibody.ogen granules, whereas mitochondria are devoid of staining (top right).

cytoD and carbamylcholine (B) shows very little non-specific background. A and carbamylcholine shows a large vacuole (V) which does not appear to bemerous gold particles are found both in the lumen and the vacuolar

90

s thednelaene

lin-hend).

on

er

alndd

calofneP-nn-allsoeportas

ileto

,anein%)peing tobens

toityr

ngcleies,

dd

e


Fig. 8.Withdrawal of cytoD and carbamylcholine restores the normmorphology of pancreatic acinar cells. Lobules were incubatedinitially with cytoD and carbamylcholine and subsequently rinsed 4 hours in the presence of atropine. An intense staining around thlumen is observed with Texas Red phalloidin (A). Experiments witFITC-dextran administered during the wash out, show that the sizthe lumen is restored and that vacuolar structures have largelydisappeared (B). Bars, 10 µm.

endocytic pits in neurons (De Camilli and Takei, 1996), wparticipating in membrane retrieval. Because it has previoubeen demonstrated that only dynamin-2 is expressed in exocrine pancreas (Cook et al., 1996), we used an antibwhich specifically recognizes dynamin-2 (Okamoto et a1997), and found that although some staining was alrepresent at the apical plasma membrane in resting cells intracellularly as small dots (Fig. 12A), there was a pronouncaccumulation of the punctate staining around enlarged lum(Fig. 12B). No staining was found at the basal membrane. Aremoval of the drugs, the staining pattern of dynamin-2 wnot different from that of control tissue (Fig. 12C).

As mentioned before, not all the membrane invaginatio

asslytheodyl.,

adyandedinafteras

ns

appeared to be coated with clathrin (Fig. 5). Although it ipossible that a coat was present on these invaginations butpreparation did not allow for its detection, we questionewhether a mechanism other than clathrin-mediated membraretrieval was also involved. In the exocrine pancreas, caveoare observed, however sporadically, at the basal membra(Roettger et al., 1995). We tested an antibody against caveoI and observed that caveolin was distributed throughout tsupranuclear region of the acinar cell in a punctate manner, ano staining was found at the basal membrane (Fig. 13AInterestingly, caveolin relocated to the apical membrane upcytoD and carbamylcholine treatment (Fig. 13B). Thisrelocalization was transient since the caveolinimmunoreactivity returned to the supranuclear region aftnormal conditions were restored.

DISCUSSION

In this study, we have shown that the disruption of the apicactin cytoskeleton with cytoD induces striking changes ipancreatic acinar morphology, especially under stimulateconditions where the lumina are dramatically enlarged annumerous membrane invaginations accumulate at the apimembrane. Our results suggest that the incorporation secretory granules membrane into the plasma membraduring exocytosis serves as a trigger for the recruitment of A2, clathrin, and dynamin, which are known to participate iendocytosis. Strikingly, another mechanism, namely caveolimediated endocytosis, might also be involved in apicmembrane retrieval in the exocrine pancreas. Our results asuggest that the actin cytoskeleton is involved in an early stof the formation of coated vesicles from coated pits and ftheir subsequent movement into the cell. Finally, our daindicate that the subapical actin cytoskeleton facilitatesecretion from the exocrine pancreas by providing contractforces that expel secretory products from the acinar lumen inthe extracellular space.

Involvement of the actin cytoskeleton in membraneretrievalWe found that pancreatic acinar cells treated with cytoDpossessed a dramatic increase in the number of membrinvaginations in the apical plasma membrane only (and not the basolateral plasma membrane), many of which (about 70were surrounded by a clathrin-like coat. Because of the shaof the invaginations and the presence of short necks connectthe plasma membrane and these structures, it is temptingspeculate that they are forming endocytic vesicles. It could argued that the accumulation of the coated invaginatiooriginates from a stimulatory effect of cytoD on theirformation. However, if that were the case, one would expect see an increase in the number of coated vesicles in the vicinof the plasma membrane but this did not occur. Ouobservations suggest that in the presence of cytoD, formiendocytic vesicles are blocked in a stage preceding vesibudding. These results are in agreement with two recent studon the effect of cytoD on another endocytic mechanismnamely receptor-mediated endocytosis, in MDCK cells anCaco-2 cells (Gottlieb et al., 1993; Shurety et al., 1996), ancontribute to the growing evidence for the involvement of th

al

forehe of


Fig. 9.Withdrawal of cytoD and carbamylcholine restores the size of the lumen of pancreatic acini. Lobules were incubated for 90 minutes withcytoD and carbamylcholine and then rinsed for 4 hours in the presence of atropine. Thin sections were visualized by electron microscopy.(A) Non-treated lobule; (B) cytoD and carbamylcholine treated lobule. The size of the lumen in cytoD and carbamylcholine-treated tissue iscomparable to that of non-treated tissue. Membrane invaginations coated with clathrin-like material observed in Fig. 5 have disappeared and thesubapical actin cytoskeleton has reformed. Bar, 1 µm.

Fig. 10.CytoD and carbamylcholine treatment induces an accumulation of clathrin-like immunoreactivity around the acinar lumen andvacuolar structures. Cryosections of fixed pancreatic lobules were stained with an antibody against the heavy chain of clathrin (A-C).(A′,B′,C′) Phase contrast micrographs of the fields shown in A,B,C. Lobules were non-treated (A,A′), treated with cytoD and carbamylcholine(B,B′) or treated with cytoD and carbamylcholine and then rinsed for 4 hours (C,C′). Clathrin staining relocated from the supranuclear region tothe apical membrane when membrane retrieval was arrested by cytoD and carbamylcholine treatment (B) and moved back into the cell uponreformation of the subapical actin cytoskeleton (C). The arrows point to the lumina; N, nucleus. Bar, 10 µm.

92

r ofalin-77;tis,sen,e

red

elesen

eb

heingal


Fig. 11.CytoD and carbamylcholine treatment induces an accumulation of AP-2-like immunoreactivity around enlarged acinar lumina.Cryosections of fixed pancreatic lobules were stained with an antibody against the α-subunit of the AP-2 complex (A-C). (A′,B′,C′) Phasecontrast micrographs of the fields shown in A,B,C. Lobules were non-treated (A,A′), treated with cytoD and carbamylcholine (B,B′) or treatedwith cytoD and carbamylcholine and then rinsed for 4 hours (C,C′). AP-2 staining was concentrated around the enlarged acinar lumina whenmembrane retrieval was blocked by cytoD and carbamylcholine treatment (B) and was restored to an intracellular location upon wash out ofcytoD and carbamylcholine (C). The arrows point to the lumina; N, nucleus. Bar, 10 µm.

actin cytoskeleton in the general process of endocytosisvarious cell types (Jackman et al., 1994; Durrbach et al., 19Lamaze et al., 1997).

It is well established that compensatory membrane retriefollows immediately after exocytosis in various cell type(Herzog and Farquhar, 1977; Geisow et al., 1985; Patzak Winkler, 1986; De Camilli and Takei, 1996). We have founthat the luminal circumference of the acinar pancreatic cincreases upon secretagogue stimulation, as repopreviously (Jamieson and Palade, 1971; Jena et al., 1994),that cytoD further extended the size of the lumen, in additto an accumulation of endocytic pits. Furthermore, wobserved that after removal of the drugs (cytoD acarbamylcholine), the size of the lumina was restored andendocytic pits disappeared. It is therefore reasonable to susthat the enlargement of the lumina and the simultaneaccumulation of membrane invaginations are interrelated.

Recruitment of clathrin, dynamin, AP-2 and caveolinduring membrane retrievalMembrane retrieval following exocytosis occurs via clathricoated pits in different cell types (De Camilli and Takei, 199

in96;

valsanddellrted andione

nd thepectous

n-6;

Geisow et al., 1985; Patzak and Winkler, 1986). Ouobservation of accumulation of coated pits in the presencecytoD indicates that at least part of the membrane retrievprocess in the exocrine pancreas is mediated via clathrcoated pits, as proposed before (Herzog and Farquhar, 19Herzog and Reggio, 1980). According to our currenunderstanding of the molecular mechanisms of endocytosAP-2 is first recruited to the plasma membrane which allowclathrin coats to assemble and the pinching off of thmembrane is thought to be mediated by dynamin (Robinso1994; De Camilli and Takei, 1996). Our immunofluorescencanalysis indicated that AP-2, clathrin, and dynamin werecruited at the plasma membrane upon cytoD ancarbamylcholine treatment, which coincides with thaccumulation of coated pits, suggesting that these molecuare involved in compensatory membrane retrieval in thpancreas. The pinching off of coated pits took place whecytoD was washed out suggesting that the actin terminal wis crucial for this and later steps in membrane retrieval.

Our results lead us to suggest that the incorporation of tzymogen granule membrane into the plasma membrane durexocytosis triggers the recruitment of AP-2 to the apic


1ory;

net

nheD

alsoas.ichthe4;

-us

ettedles

Fig. 12.CytoD and carbamylcholine treatment induces an accumulation of dynamin-2-like immunoreactivity around enlarged acinar lumina.Cryosections of fixed pancreatic lobules were stained with an antibody against dynamin-2 (A-C). (A′,B′,C′) Phase contrast micrographs of thefields shown in A,B,C. Lobules were non-treated (A,A′), treated with cytoD and carbamylcholine (B,B′) or treated with cytoD andcarbamylcholine and then rinsed for 4 hours (C,C′). Dynamin-2 staining accumulated at the apical membrane when membrane retrieval wasblocked by cytoD and carbamylcholine treatment. The arrows point to the lumina; N, nucleus. Bar, 10 µm.

membrane. Although the cytoplasmic tail motifs of sevemembrane proteins have been shown to interact with the 2 complex (Kirchhausen et al., 1997), it is not known whthese proteins are at the apical membrane of the acinar especially given the fact that classical receptor-mediaendocytosis occurs only at the basal membrane in pancreacinar cells (Cruz et al., 1984). Possible candidates inclsynaptotagmin which is known to be in secretory vesimembrane and to bind AP-2 (Zhang et al., 1994). Others mbe membrane proteins that are ineffectively sorted frompurposefully included in secretory granules during thformation but whose cytoplasmic tails remain masked aunavailable for AP-2 interaction until exocytosis. Unmaskinmight result from physical changes accompanying granmembrane fusion with the apical plasma membrane suchprotein aggregation or covalent modifications such phosphorylation which is known to accompany exocytosisacinar cells (Marlowe et al., 1998). Protein phosphorylationthe tail motif of furin is required for AP-1 interaction withimmature secretory granules in neuroendocrine cells (Dittiéal., 1997). It is of interest to note that the staining pattern AP-2 in acinar cells does not correspond to that expected

ralAP-atcell,tedatic

udecleight oreirndg

ule asas in of

etfor for

zymogen granules. This is similar to the distribution of AP-in other regulated secretory cells where mature secretgranules lack AP-1 immunoreactivity (Dittié et al., 1997Klumperman et al., 1998).

The absence of a distinct coat on some of the membrainvaginations led us to look for caveolin given its involvemenin insulin-stimulated exocytosis or endocytosis of GLUT4 iadipocytes (Fan et al., 1983; Scherer et al., 1994). Taccumulation of caveolin at the apical membrane during cytoand carbamylcholine treatment suggests that caveolae may be involved in membrane retrieval in the exocrine pancreThese observations are in agreement with previous data whhave shown that aggregation of caveolae is induced by disassembly of the actin cytoskeleton (Parton et al., 199Fujimoto et al., 1995; Deckert et al., 1996).

Involvement of the actin cytoskeleton in exocytosisAlthough we found that cytoD inhibited carbamylcholineinduced amylase release which is in agreement with previostudies on various exocrine glands (Williams, 1977; Muller al., 1985; Tojyo et al., 1989), several observations suggesthat exocytosis did occur: (i) the number of zymogen granu

94

onm

tiningtheand

is aarnm

erltsinceh

mai etl.,thes

theely


Fig. 13.CytoD and carbamylcholine treatment induces an accumulation of caveolin-like immunoreactivity around enlarged acinar lumina.Cryosections of fixed pancreatic lobules were stained with an antibody against caveolin-I (A-C). (A′,B′,C′) Phase contrast micrographs of thefields shown in A,B,C. Lobules were non-treated (A,A′), treated with cytoD and carbamylcholine (B,B′) or treated with cytoD andcarbamylcholine and then rinsed for 4 hours (C,C′). Caveolin staining relocated from an intracellular area to the apical membrane whenmembrane retrieval was arrested by cytoD and carbamylcholine treatment. The arrows point to the lumina; N, nucleus. Bar, 10 µm.

was decreased, (ii) lumina were significantly enlarged, (lumina and vacuolar structures were filled with amylase, a(iv) substantial amounts of amylase were released into incubation medium after wash out of the drugs. These resshow that despite an apparent inhibition of secretion, cytoD not inhibit exocytosis per se, which persisted in the preseof the drug, as previously shown in other cell types (Orci et 1972; Sontag et al., 1988; Matter et al., 1989).

We observed that when lobules were stimulated wcarbamylcholine, cytoD induced an enlargement of the lumin which amylase and FITC-dextran accumulated aconsequently cytoD inhibited the expulsion of amylase frothe acinar lumen. Expansion of the lumen was not duesimple back pressure from occluded ducts since the lumsurface area is also expanded following cytoD treatmentpancreatic acini where the lumen is in direct continuity withe medium (K. M. Valentijn and J. D. Jamiesonunpublished). It is reasonable to suggest that the aterminal web surrounding the acinar lumen generacontractile forces that expel luminal contents into the dusystem. Whether this occurs independently of or is a resulconcomitant membrane retrieval with resulting reduction the luminal volume is unknown. Under physiologica

iii)ndtheultsdidnceal.,

ithenndm toinal ofth,

ctintesct

t ofofl

conditions, it is also possible that the actin cytoskeletprovides contractile forces to extrude secretory products frofusing zymogen granules. In support of this idea, acmicrofilaments have been observed around dischargzymogen granules at the electron microscope level in exocrine pancreas and parotid (Palade, 1975; Segawa Yamashina, 1989).

Recently, it has been proposed that actin disassembly final trigger for exocytosis in permeabilized pancreatic acincells, since introduction of low concentrations of actimonomer-binding proteins induced amylase release fropermeabilized pancreatic acinar cells with no furthstimulatory factors required (Muallem et al., 1995). Our resuon intact acinar cells do not agree with these observations scytoD alone did not elicit amylase secretion, althougzymogen granules were docked at the apical plasmembrane. In agreement with other studies, however (Orcal., 1972; Williams, 1977; Sontag et al., 1988; Tojyo et a1989), we find that exocytosis occurs in acinar cells when actin cytoskeleton was disrupted by cytoD, only if a stimuluwas concurrently administered.

Taken together, the results presented here indicate thatsubapical actin cytoskeleton of pancreatic acinar cells clos


lls

lls.

tic

n

torys.

in

lls.

in

s.

byal

is.

ter

th

regulates the balance between exocytosis and compensamembrane retrieval.

We acknowledge the Center for Cell Imaging for help with thconfocal and electron microscopy, Dr Spyridon Artavanis-Tsakonfor the use of the MRC-1024 confocal scanning microscope, DonLacivita for technical help and Mr Henry Tan for his excellenphotographic work. We are grateful to Mr Aaron Granger for his hein the preparation of the amylase antibody. We also are indebteDr Richard Vallee for providing us with a dynamin-2 polyclonaantibody. This work was supported by United States Public HeaService Grant DK 17389 to James D. Jamieson.

REFERENCES

Arvan, P. and Castle, D. (1998). Sorting and storage during secretory granubiogenesis: looking backward and looking forward. Biochem. J.332, 593-610.

Aunis, D. and Bader, M.-F. (1988). The cytoskeleton as a barrier to exocytosin secretory cells. J. Exp. Biol.139, 253-266.

Bernfeld, P. (1955). Amylases alpha and beta. Meth. Enzymol.1, 149-158.Burgoyne, R. D. and Cheek, T. R. (1987). Reorganisation of peripheral actin

filaments as a prelude to exocytosis. Biosci. Rep.7, 281-288.Burnham, D. B. and Williams, J. A. (1982). Effects of high concentrations

of secretagogues on the morphology and secretory activity of the pancra role for microfilaments. Cell Tissue Res.222, 201-212.

Cheek, T. R. and Burgoyne, R. D. (1986). Nicotine-evoked disassembly ocortical actin filaments in adrenal chromaffin cells. FEBS Lett.207, 110-114.

Cook, T., Mesa, K. and Urrutia, R. (1996). Three dynamin-encoding geneare differentially expressed in developing rat brain. J. Neurochem.67, 927-931.

Cooper, J. A. (1987). Effects of cytochalasin and phalloidin on actin. J. CellBiol. 105, 1473-1478.

Cruz, J., Posner, B. I. and Bergeron, J. J. M. (1984). Receptor-mediatedendocytosis of [125I]insulin into pancreatic acinar cells in vivo.Endocrinology115, 1996-2008.

De Camilli, P. and Takei, K. (1996). Molecular mechanisms in synapticvesicle endocytosis and recycling. Neuron16, 481-486.

Deckert, M., Ticchioni, M. and Bernard, A. (1996). Endocytosis of GPI-anchored proteins in human lymphocytes:role of glycolipid-based domaactin cytoskeleton, and protein kinases. J. Cell Biol.133, 791-799.

Dittié, A. S., Thomas, L., Thomas, G. and Tooze, S. A. (1997). Interactionof furin in immature secretory granules from neuroendocrine cells with AP-1 adaptor complex is modulated by casein kinase II phosphorylatEMBO J.16, 4859-4870.

Durrbach, A., Louvard, D. and Coudrier, E. (1996). Actin filamentsfacilitate two steps of endocytosis. J. Cell Sci.109, 457-465.

Edmonds, B. T. and Koenig, E. (1989). ATP and calmodulin dependenactomyosin aggregates induced by cytochalasin D in goldfish retiganglion cell axons in vitro. J. Neurobiol.21, 555-566.

Fan, J. Y., Carpentier, J.-L., Van Obberghen, E., Grunfeld, C., Gorden, P.and Orci, L. (1983). Morphological changes of the 3T3-L1 fibroblasplasma membrane upon differentiation to the adipocyte form. J. Cell Sci.61, 219-230.

Freedman, S. D., Kern, H. F. and Scheele, G. A. (1994). Apical membranetrafficking during regulated pancreatic exocrine secretion--Role of alkalpH in the acinar lumen and enzymatic cleavage of GP2, a GPI-linkprotein. Eur. J. Cell Biol.65, 354-365.

Fujimoto, T., Miyawaki, A. and Mikoshiba, K. (1995). Inositol 1,4,5-triphosphate receptor-like protein in plasmalemmal caveolae is linkedactin filaments. J. Cell Sci.108, 7-15.

Geisow, M. J., Childs, J. and Burgoyne, R. D. (1985). Cholinergicstimulation of chromaffin cells induces rapid coating of the plasma mebraEur. J. Cell Biol.38, 51-56.

Gottlieb, T. A., Ivanov, I. E., Adesnik, M. and Sabatini, D. D. (1993). Actinmicrofilaments play a critical role in endocytosis at the apical but not basolateral surface of polarized epithelial cells. J. Cell Biol.120, 695-710.

Herzog, V. and Farquhar, M. G. (1977). Luminal membrane retrieved afte

tory

easnatlp

d tollth

le

is

eas:

f

s

ins,

theion.

tnal

t

ineed

to

ne.

the

r

exocytosis reaches most Golgi cisternae in secretory cells. Proc. Nat. Acad.Sci. USA74, 5073-5077.

Herzog, V. and Reggio, H. (1980). Pathways of endocytosis from luminalplasma in rat exocrine pancreas. Eur. J. Cell Biol.21, 141-150.

Jackman, M. R., Shurety, W., Ellis, J. A. and Luzio, J. P. (1994). Inhibitionof apical but not basolateral endocytosis of ricin and folate in Caco-2 ceby cytochalasin D. J. Cell Sci.107, 2547-2556.

Jamieson, J. D. and Palade, G. E. (1971). Synthesis, intracellular transport,and discharge of secretory proteins in stimulated pancreatic exocrine ceJ. Cell Biol.50, 135-158.

Jena, B. P., Gumkowski, F. D., Konieczko, E. M., Fischer von Mollard, G.,Jahn, R. and Jamieson, J. D. (1994). Redistribution of a rab-3 like GTPbinding protein from secretory granules to the Golgi complex in pancreaacinar cells during regulated exocytosis. J. Cell Biol.124, 43-53.

Jungermann, J., Lerch, M. M., Weidenbach, H., Lutz, M. P., Krüger, B.and Adler, G. (1995). Disassembly of rat pancreatic acinar cell cytoskeletoduring supramaximal secretagogue stimulation. Am. J. Physiol. Gastrointest.Liver Physiol.268, G328-G338.

Kirchhausen, T., Bonifacino, J. S. and Howard, R. (1997). Linking cargoto vesicle formation: receptor tail interactions with coat proteins. Curr. Opin.Cell Biol. 9, 488-495.

Kletzien, R. F., Perdue, J. F. and Springer, A. (1972). Cytochalasin A andB. Inhibition of sugar uptake in cultured cells. J. Biol. Chem.247, 2964-2966.

Klumperman, J., Kuliawat, R., Griffith, J. M., Geuze, H. J. and Arvan, P.(1998). Mannose 6-phosphate receptors are sorted from immature secregranules via adaptor protein AP-1, clathrin, and syntaxin 6-positive vesicleJ. Cell Biol.141, 359-371.

Lamaze, C., Fujimoto, L. M., Yin, H. L. and Schmid, S. L. (1997). Theactin cytoskeleton is required for receptor-mediated endocytosis mammalian cells. J. Biol. Chem.272, 20332-20335.

Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951).Protein measurement with the Folin phenol reagent. J. Biol. Chem.193, 265-275.

Marlowe, K. J., Farshori, P., Torgerson, R. R., Anderson, K. L., Miller, L.J. and McNiven, M. A. (1998). Changes in kinesin distribution andphosphorylation occur during regulated secretion in pancreatic acinar ceEur. J. Cell Biol.75, 140-152.

Matter, K., Dreyer, F. and Aktories, K. (1989). Actin involvement inexocytosis from PC12 cells: studies on the influence of botulinum C2 toxon stimulated noradrenaline release. J. Neurochem. 52, 370-376.

Muallem, S., Kwiatkowska, K., Xu, X. and Yin, H. L. (1995). Actin filamentdisassembly is a sufficient final trigger for exocytosis in nonexcitable cellJ. Cell Biol.128, 589-598.

Muller, P., Chambaut-Guerin, A.-M. and Rossignol, B. (1985).Comparative effects of cytochalasin D on the protein discharge induced alpha- and beta-adrenergic or cholinergic agonists in rat exorbital lacrimglands. Biochim. Biophys. Acta844, 158-166.

O’Konski, M. S. and Pandol, S. J. (1990). Effects of caerulein on the apicalcytoskeleton of the pancreatic acinar cell. J. Clin. Invest.86, 1649-1657.

Ohmori, S. and Toyama, S. (1992). Direct proof that the primary site of actionof cytochalasin on cell motility processes is actin. J. Cell Biol.116, 933-941.

Okamoto, P. M., Herskovits, J. S. and Vallee, R. B. (1997). Role of the basic,proline-rich region of dynamin in Src homology 3 domain binding andendocytosis. J. Biol. Chem.272, 11629-11635.

Orci, L., Gabbay, K. H. and Malaisse, W. J. (1972). Pancreatic beta-cellweb: its possible role in insulin secretion. Science175, 1128-1130.

Palade, G. E. (1975). Intracellular aspects of the process of protein synthesScience189, 347-356.

Parton, R. G., Joggerst, B. and Simons, K. (1994). Regulated internalizationof caveolae. J. Cell Biol.127, 1199-1215.

Patzak, A. and Winkler, H. (1986). Exocytotic exposure and recycling ofmembrane antigens of chromaffin granules: ultrastructural evaluation afimmunolabeling. J. Cell Biol.102, 510-515.

Perrin, D., Möller, K., Hanke, K. and Söling, H.-D. (1992). cAMP andCa2+-mediated secretion in parotid acinar cells is associated wireversible changes in the organization of the cytoskeleton. J. Cell Biol.116, 127-134.

Robinson, M. S. (1994). The role of clathrin, adaptors and dynamin inendocytosis. Curr. Opin. Cell Biol.6, 538-544.

Roettger, B. F., Rentsch, R. U., Hadac, E. M., Hellen, E. H., Burghardt, T.P. and Miller, L. J. (1995). Insulation of a G protein-coupled receptor on

96

-

ry

at

ged

l


the plasmalemmal surface of the pancreatic acinar cell. J. Cell Biol. 130,579-590.

Scheele, G. (1983). Pancreatic lobules in the in vitro study of pancreatic acincell function. Meth. Enzymol.98, 17-28.

Scherer, P. E., Lisanti, M. P., Baldini, G., Sargiacomo, M., Mastick, C. C.and Lodish, H. F. (1994). Induction of caveolin during adipogenesis anassociation of GLUT4 with caveolin-rich vesicles. J. Cell Biol.127, 1233-1243.

Schliwa, M. (1982). Action of cytochalasin D on cytoskeletal networks. J. CellBiol. 92, 79-91.

Schramm, M. and Loyter, A. (1966). Purification of alpha-amylases byprecipitation of amylase-glycogen complexes. Meth. Enzymol.8, 533-537.

Segawa, A. and Yamashina, S. (1989). Roles of microfilaments in exocytosisA new hypothesis. Cell Struct. Funct.14, 531-544.

Shurety, W., Bright, N. A. and Luzio, J. P. (1996). The effects of cytochalasinD and phorbol myristate acetate on the apical endocytosis of ricinpolarised Caco-2 cells. J. Cell Sci.109, 2927-2935.

Simionescu, N. and Simionescu, M. (1976). Galloylglucoses of lowmolecular weight as mordant in electron microscopy. I. Procedure, aevidence for mordanting effect. J. Cell Biol.70, 608-621.

ar

d

:

in

nd

Sontag, J.-M., Aunis, D. and Bader, M.-F. (1988). Peripheral actin filamentscontrol calcium-mediated catecholamine release from streptolysin-Opermabilized chromaffin cells. Eur. J. Cell Biol.46, 316-326.

Tojyo, Y., Okumura, K., Kanazawa, M. and Matsumoto, Y. (1989). Effectof cytochalasin D on acinar cell structure and secretion in rat parotid salivaglands in vitro. Arch. Oral Biol.34, 847-855.

Trifaró, J.-M. and Vitale, M. L. (1993). Cytoskeleton dynamics duringneurotransmitter release. Trends Neurosci.16, 466-472.

Valentijn, J. A., La Civita, D. Q., Gumkowski, F. D. and Jamieson, J. D.(1997). Rab4 associates with the actin cytoskeleton in developing rpancreatic acinar cells. Eur. J. Cell Biol.72, 1-8.

Vitale, M. L., Del Castillo, A. R., Tchakarov, L. and Trifaró, J.-M. (1991).Cortical filamentous actin disassembly and scinderin redistribution durinchromaffin cell stimulation precede exocytosis, a phenomenon not exhibitby gelsolin. J. Cell Biol.113, 1057-1068.

Williams, J. A. (1977). Effects of cytochalasin B on pancreatic acinar celstructure and secretion. Cell Tissue Res.179, 453-466.

Zhang, J. Z., Davletov, B. A., Südhof, T. C. and Anderson, R. G. W. (1994).Synaptotagmin I is a high affinity receptor for clathrin AP-2: Implicationsfor membrane recycling. Cell 78, 751-760.

The subapical actin cytoskeleton regulates secretion and

Documents

Transcript of The subapical actin cytoskeleton regulates secretion and