AUTOPHAGIC VACUOLES PRODUCED IN VITRO II. Studies ...AUTOPHAGIC VACUOLES PRODUCED IN VITRO II....

AUTOPHAGIC VACUOLES PRODUCED IN VITRO

II. Studies on the Mechanism of Formation

of Autophagic Vacuoles Produced by Chloroquine

MARTHA E. FEDORKO, JAMES G. H[RSCH, and ZANVIL A. COHN

From The Rockefeller University, New York 10021

ABSTRACT

Continuous phase-contrast observations have been made on macrophages following ex-posure to chloroquine. The initial abnormality is the appearance in the Golgi region ofsmall vacuoles with an intermediate density between that of pinosomes and granules. Overthe course of 1-2 hr these vacuoles grow larger and accumulate amorphous material orlipid. Pinosomes or granules frequently fuse with the toxic vacuoles. Chloroquine derivativescan be seen by fluorescence microscopy; the drug is rapidly taken up by macrophages andlocalized in small foci in the Golgi region. Chloroquine continues to produce vacuoleswhen pinocytosis is suppressed. Electron microscopic studies of chloroquine effects onmacrophages preincubated with colloidal gold to label predominately pinosomes or granulessuggest that toxic vacuoles can arise from unlabeled organelles. Later vacuoles regularlyacquire gold label, apparently by fusion, from both granules and pinosomes. L cells alsodevelop autophagic vacuoles after exposure to chloroquine. Smooth endoplasmic reticulumapparently is involved early in the autophagic process in these cells. Information nowavailable suggests an initial action of chloroquine on Golgi or smooth endoplasmic reticulumvesicles, and on granules, with alterations in their membranes leading to fusion with oneanother and with pinosomes.

INTRODUCTION

As reported in the preceding paper (1), chloro-quine induces formation of autophagic vacuolesin vitro in mouse macrophages. We present hereresults of experiments designed to shed light onthe primary intracellular site of chloroquinetoxicity and on the mechanism by which thistoxicity leads to abnormal vacuole formation.These experiments include study of the effects ofchloroquine on another type of tissue culture cell,the L-strain fibroblast.

MATERIAL AND METHODS

Cell Preparations

Mouse peritoneal macrophages were maintainedin a tissue-culture system as described previously

(2). Unless otherwise stated, culture was in medium199 with 20% newborn calf serum and 200 U/mlpenicillin.

Strain-L (Earle) cells were also studied. Thesecells were grown in 10% fetal calf serum in Eagle'sminimum essential medium, with subculture every4 days. Cells grown as a monolayer for 3 days inLeighton tubes with flying cover slips or in T flaskswere placed in fresh medium containing chloro-quine for observations of toxicity.

Cinemicrophotographic Observations

Cell cultures on cover slips in Leighton tubes werewashed and inverted over a thin rim of paraffinon glass slides so that chambers were made approxi-mately the depth of a No. cover slip. The chamberswere filled with fresh medium with or without

392

Dow

nloaded from http://rupress.org/jcb/article-pdf/38/2/392/1384606/392.pdf by guest on 12 August 2021

chloroquine and sealed with paraffin. Care was takenso that the specimens were not allowed to dry duringmounting. The slides were incubated at 38°C andobserved by oil-immersion, phase-contrast microscopyfor up to 2 hr. Motion pictures of control and drugexposed cells were also recorded by using techniquesdescribed previously (3). The movies were taken atfive frames per second. Projection of prints at 24frames per second thus allowed repeated study ofnormal and chloroquine-treated macrophages withaction speeded up approximately five times.

Fluorescence Microscopy

Chloroquine is detectable by fluorescence in cellsor tissues (4), so that it was possible to observedirectly its localization in macrophages. Concen-trations of chloroquine of 100 pzg/ml were requiredto produce fluorescence in treated cells. Theseobservations were made in slide chambers as de-scribed above by employing a dark-field fluorescentsystem with BG 12 exciter and OG-5 absorptionfilters.

Metabolic Inhibitors

The following compounds were used (5): 2,4-dinitrophenol (1 X 10- 4 M, Eastman OrganicChemicals, Rochester, N. Y.), iodoacetate (5 X10- 4 M, Amend Drug and Chemical Co., New York),parafluorophenylalanine (1.0 mg/ml, NutritionalBiochemical Corporation, Cleveland, O.); sodiumfluoride (I X 10- 3 M, Merck & Co., Inc., New York);puromycin 1 ug/ml (Nutritional Biochemical Corpo-ration). Fresh solutions of inhibitors in culturemedium were adjusted, when necessary, to pH 7.3with NaOH or HCI. The inhibitors were added to24-hr-old macrophage cultures 1 hr prior to additionof chloroquine 30 g/ml; evaluation was made 2 hrlater.

Pinosome Counts

Cover slip cultures fixed in glutaraldehyde andmounted in water were observed under oil-immersionphase contrast. Numbers of clear pinocytic vesiclesin the pseudopods of 100 cells were counted and ex-pressed as pinosomes per 50 cells.

Variation in Tissue Culture Conditions

Some macrophages were cultured for 48 hr inmedium containing low (1-5%) concentration ofnewborn calf serum for production of cells of un-usually low dense granule content (6). Addition ofchloroquine then allowed assessment of the rela-tionship between granule content and toxic actionof this drug.

Labeling of Macrophage Organelles with

Colloidal Gold

Colloidal gold (2.5 mg/ml, Abbott Laboratories,Chicago, Ill.) was dialyzed against phosphate-buffered saline pH 7.4 and was added 1:5 or 1:10to the macrophage cultures. As established in previ-ous studies (7), gold particles enter the cell in pino-cytic vesicles and are rapidly concentrated withincentrally located pinocytic vacuoles. After incubationwith gold in the medium for several hours, goldparticles are seen primarily in dense bodies of macro-phages. Addition of colloidal gold simultaneouslywith chloroquine thus allowed observations onmorphology of early toxic events in cells with labeledpinosomes. Preincubation of macrophages with goldfor 6 hr followed by washing and addition of freshmedium containing chloroquine allowed similarstudy of cells with gold label primarily in densebodies.

Fixation and Processing for

Electron Microscopy

Fixation and processing for electron microscopywere according to methods described in the accom-panying report (1).

RESULTS

Phase-Contrast Microscopy of Macrophages

Exposed to Chloroquine

Macrophages growing on cover slips wereobserved by phase-contrast microscopy and photo-

graphed under "fast speed" motion picture con-ditions in a thin slide chamber. In well-spread cells

it was possible to observe clearly various activities(surface membrane ruffling, pinosome streaming,

and the like) and organelles (perinuclear Golgi

region and pinocytic vacuoles, transitional zone

dense granules, lipid droplets, peripheral mito-

chondria). Control cells not exposed to drug didnot change in their activities or morphology during

a 1 hr period of observation in the slide chambers.

Macrophages exposed to 50 ug/ml of chloroquine

under these conditions exhibited initially increasedsurface-membrane ruffling and some retraction of

their pseudopods. 10-15 min after addition of drugnumerous tiny (initially of a size at the limit of

resolution by light microscopy, i.e. about 200 mp)

vesicles appeared in the Golgi region of the cyto-plasm. These tiny vesicles appeared to be some-what more dense than did nearby pinosomes, butthey were much more lucent than the cytoplasmic

M. E. FEDORKO, J. G. HIRSCII, AND Z. A. ConN Autophagic Vacuoles II. 393

Dow


granules. In the course of the next 10 min or so,the toxic vesicles gradually increased in size and innumber, often to occupy most or all of the nuclearhof region. The motion-picture technique wasmost valuable for studying this very early phaseas it was possible to view the changes somewhatspeeded up, to rerun the action, or to reverse it atwill. Many instances were recorded of apparentfusion between toxic vacuoles and pinosomes orgranules. Markedly vacuolated cells continued toexhibit motility and slow or occasionally rapidtranslocation of the large vacuoles was associatedwith the motility. Some of the toxic vacuolesappeared to be partially collapsed or "floppy" andirregular in outline. An hour after chloroquineexposure most cells were severely vacuolated, andamorphous, dense material or refractile lipid drop-lets were often seen in Brownian movement withinvacuoles.

The Intracellular Localization of Chloroquine

as Observed by Fluorescence Microscopy

Since chloroquine or a derivative compoundfluoresces under ultraviolet illumination (4, 19), itwas possible to study its localization in macro-phages by observing exposed cells in slide culturechambers by fluorescence microscopy. A dose of100 yg/ml or more of chloroquine was required foradequate visualization of fluorescent drug or itsderivative within cells. As rapidly as slide chamberscould be examined under the microscope (3-5min), a fluorescent deposit in the form of fine dustor droplets was seen distributed in the Golgi region.After 10-15 min these droplets had changed inappearance into fluorescent vesicles or vacuoles,which, in size and location, were identical with thetoxic vacuoles described above. In this early phase,fluorescent matecial was never seen associatedwith lipid or other organelles in the peripheralcytoplasm.

Relationship Between Pinocytosis and

Autophagic Vacuoles Induced by Chloroquine

When pinocytosis is stimulated in macrophagesby increasing the concentration of serum in themedium or by adding certain polyanions (6, 8, 9),the cells develop over the course of a few hourslarge, clear vacuoles in the perinuclear region. Thegeneral appearance of macrophages vacuolated asa result of increased pinocytosis is similar to thatseen after exposure to chloroquine, and thus weinvestigated the possibility that chloroquine was

a,

06

oo oorn

'I

c

, 0> 73

Q ,

U Qo

-oO' C

240-

200-

160-

120-

80-

40-

I 2 3 4

Time in hours

FIGURE 1 Effect of chloroquine on pinocytosis inmacrophages. Arrows show period of chloroquine ex-posure.

acting, at least in part, by stimulating pinocyticactivity of the cells. As is shown in Fig. 1, macro-phage pseudopods contained somewhat increasednumbers of pinocytic vesicles 15 min after exposureto chloroquine, but by 30 min the pinosome countwas no greater than that of control cells; thereafterpinocytosis was in fact somewhat suppressed in cellsexposed to chloroquine. Pinocytosis remainedsuppressed to approximately 50% of control valuesfor 24 hr after the cells were washed and placed indrug free medium. Quantitation of pinocyticactivity by these methods is not precise, so that nofirm conclusions were possible; it seemed quiteunlikely, however, that the brief increase inpinocytic activity could be responsible for theextensive vacuole formation in macrophagesexposed to chloroquine.

Several inhibitors of metabolism or of proteinsynthesis have been shown to suppress or blockpinocytosis in macrophages (5). Addition to themacrophage cultures of 2,4-dinitrophenol (1 X10- 4 M), iodoacetate (5 X 10- 4 M), parafluoro-phenylalanine (1.0 mg/ml), sodium fluoride(1 X 10-3 M), or puromycin (1 I g/ml) I hr priorto addition of 30 ug/ml of chloroquine did notblock the development of vacuoles in the cells.Pinosome counts established that these agents didsuppress pinocytosis in parallel cells not treatedwith chloroquine.

One additional experimental approach was usedfor the study of the relationship between pino-cytosis and the toxic vacuoles produced by chloro-quine. Colloidal gold added to the medium servesas a convenient marker for pinocytic vesicles and

394 THE JOURNAL OF CELL BIOLOGY VOLUME 38, 1968

j,

Dow


FIGURE 2 The perinuclear region of a macrophage fixed 30 min after simultaneous exposure to chloro-quine and colloidal gold (see text). Pinocytic vacuoles (PV) exhibit characteristic irregular shape, inter-nal vesicles, and generally electron-transparent content and a high concentration of gold particles. Smallvesicles and amorphous material as present in these vacuoles may also be seen in pinocytic vacuoles ofmacrophages not exposed to chloroquine (see reference 23 for a discussion of the mechanism of trans-port of cytoplasmic residues into pinosomes). A large autoplhagic vacuole (AV) shows typical tubularextensions and a heterogeneous content of moderately electron-opaque amorphous material, mem-branous elements and lipid. This autophagic vacuole contains very few gold particles; this indicatesthat it did not arise from pinocytic vesicles or vacuoles. X 30,000.

vacuoles (7). Addition of colloidal gold and chloro-

quine at the same time thus allowed a comparison

to be made at intervals thereafter of the gold

content of pinocytic and of toxic vacuoles. After30 min (see Fig. 2) the pinocytic vacuoles were

heavily labeled with gold particles, whereas the

toxic vacuoles (typified by their heterogenous

content, invaginations, and protrusions) werelabeled lightly or not at all. 3 hr after simultaneous

exposure to gold and chloroquine, the autophagic

vacuoles commonly showed moderately intense

gold labeling. The results thus suggested that the

autophagic vacuoles did not arise primarily frompinocytic structures but that fusion or some mecha-

nism of interchange between these organelles came

into play with passage of time.

Relationship Between Granules andAutophagic Vacuoles Induced by Chloroquine

An obvious possible source of membrane andof hydrolases for the autophagic vacuoles was thepopulation of cytoplasmic dense bodies or granulesin macrophages maintained under these con-ditions. Perhaps chloroquine was acting on theselysosomal structures to bring about fusion with oneanother or with other organelles leading to vacuoleformation.

The granule content of macrophages varieswidely depending on age of the culture and com-position of the medium (6). Macrophages main-tained in a medium of low serum concentrationhad only a few small granules, but on exposure tochloroquine they formed toxic vacuoles in a

M. E. FEDORKO, J. G. HIRSCH, AND Z. A. COHN Autophagic Vacuoles II. 395

Dow


FIGURE 8 A portion of a macrophage which had been preincubated with colloidal gold for 6 hr to labelthe granules, and then transferred to medium containing chloroquine but no gold. After 0 min theearly autophagic vacuoles (AV), identified by their shape and moderately electron-transparent content,contain only a few gold particles, whereas granules (G) nearby and in other parts of the cell are heavilylabeled with gold and exhibit a dense matrix. A structure of indeterminate morphology (arrow) hasfeatures between those usually seen in granules and toxic vacuoles and is heavily labeled with gold; itmay thus have been formed by fusion of a granule with a toxic vacuole. X 30,000.

manner not significantly different from that seenwith heavily granulated macrophages which hadbeen cultured in a high serum medium. Freshmouse peritoneal macrophages or human bloodmonocytes are relatively immature cells with onlysmall numbers of dense bodies. Chloroquine alsoinduced extensive vacuole formation in thesecells; this finding eliminated cultivation in vitroand preexisting large granule content as prereq-uisites for the toxic effect.

In macrophages cultivated in a suitable me-dium, the pinocyticvacuoles acquire hydrolases andin the course of several hours mature into densebodies (6, 7, 10). If colloidal gold is present duringthis process, the cells show heavy deposits of goldparticles in their granules. Macrophages thuslabeled were washed and placed in medium con-taining chloroquine. The toxic vacuoles appearing15-30 min later contained gold (Fig. 3), but theconcentration of gold in the irregularly shaped,large autophagic vacuoles was generally lower thanthat in typical dense bodies.

The results thus suggest that preformed granulesare not the primary site of chloroquine action andautophagic vacuole genesis, but they participate

in the process early and deliver their contents to thevacuole, probably by fusing with it.

Effects of Chloroquine on

L-Strain Fibroblasts

Studies were made to establish the toxic effects,if any, of chloroquine on a cell type other than themacrophage. The Earle strain of L cell wasselected for this purpose, as L cells differ in struc-ture and function from macrophages in manyregards. L cells multiply rapidly in culture, forinstance, whereas macrophages do not undergocell division in vitro. L cells display much lessphagocytic and pinocytic activity, and they con-tain fewer dense bodies than do macrophages.

The ultrastructure of L cells cultured on glassfor 72 hr and processed by mixed fixation isillustrated in Fig. 4. Nuclei were very large, show-ing dispersed chromatin and prominent nucleoli.Cytoplasm in the nuclear hof area showednumerous small vesicles, well-developed complexesof Golgi cisternae, and a few large clear vacuoles.Peripheral cytoplasm of L cells contained large,irregularly shaped mitochondria, numerous poly-ribosomes, a few multivesicular bodies, and a few

396 ThE JOURNAL OF CELL BIOLOGY VOLUME 8, 1968

Dow


FIGURE 4 An L-strain mouse fibroblast cultured on glass for 3 days and processed by mixed fixation. Thenucleus (N) exhibits a dispersed chromatin pattern. Golgi (GO) vesicles and cisternae, and a few largeclear vacuoles (CV) are seen in the nuclear hof area. Note the large number of very small vesicles inthe centrosomal region. Large mitochondria (M), many with irregular shapes, are widely distributed in thecytoplasm. In some places (arrow) rough ER is seen in continuity with dilated smooth ER; the dilatedsmooth ER often contains a small, round, very electron-opaque body. In cross-sections (double-stemmedarrows) of transitional ER these dense bodies may appear to be in vesicles. The L-cell cytoplasm ex-hibits many polyribosomes and amorphous moderately electron-opaque material. A few multivesicularbodies (MVB) may be seen; dense bodies or granules are rare or absent. X 24,000.

397

Dow


FIGURE 5 A portion of an L cell 30 min after exposure to 30 g/ml of chloroquine. Numerous toxicvacuoles (V) are seen; many of these vacuoles contain small dense bodies resembling those seen in di-lated terminal smooth ER of control cells (see Fig 4). X 21,000.

strips of partly rough, partly smooth endoplasmicreticulum (ER). The smooth ER was dilated, andfrequently contained a small, round (600 my), veryelectron-opaque body, possibly of viral nature.These dense bodies appeared to be in small vesiclesor vacuoles in transverse sections of the dilatedsmooth ER; they were distributed throughout thecytoplasm but were especially numerous in thetransitional zone between the Golgi region and thegeneral cytoplasm.

L cells exposed to 30 g/ml chloroquinedeveloped large cytoplasmic vacuoles generallyresembling those seen in macrophages. IndividualL cells seemed to vary markedly in their suscepti-bility to chloroquine toxicity, in contrast to thefairly uniform response of macrophages. A smallpercentage of exposed L cells showed extensiveautophagic vacuole formation within 15 min, andafter 3 hr of incubation with chloroquine, approxi-mately 75% of the cells showed changes rangingfrom mild to severe, whereas the remaining 25%still appeared normal.

Early toxic changes in L cells exposed to chloro-quine appeared under the phase-contrast micro-scope as clear, round vacuoles about the Golgi

region. Electron microscopy of early toxic vacuoles(Fig. 5) showed a distinct limiting membrane and acontent of amorphous material, tiny vesicles, andsmall dense bodies resembling those seen in dilatedterminal smooth ER of the normal cell (see above).The vacuoles in other affected cells were larger andhighly irregular in shape (Fig. 6). More severechloroquine toxicity in L cells took the form oflarge autophagic vacuoles which varied widely intheir content (Fig. 7). In one vacuole a well-pre-served mitochondrion was demonstrated; othervacuoles contained electron-opaque amorphousdeposits, fine granular material, or stacks of mem-branes in an onionskin pattern. A double limitingmembrane with intervening clear space was seenabout some of the toxic vacuoles. The cytoplasm ofthese severely vacuolated L cells showedapparently normal polyribosomes, short strips ofER, and normal elongated mitochondria. Golgicisternae were sometimes dilated.

DISCUSSION

Macrophages exposed to 30 Mg/ml of chloroquinedevelop, within an hour, many large vacuoles inthe perinuclear region. Obviously the total surface


Dow


FIGURE 6 Illustrates a slightly more advanced state of toxic vacuolization of L cells exposed to chloro-quine. Large vacuoles are grouped about the Golgi region in a manner quite similar to that seen in affectedmacrophages (compare to Fig. 10 in accompanying paper, reference 1). The vacuoles are generally electrontransparent and contain amorphous material and small vesicles. Nucleus, Golgi elements, and peripheralcytoplasmic elements are essentially normal. X 20,000.

area of membrane around these toxic vacuoles isconsiderable. Many of the studies reported herewere done to shed light on origin of this vacuolarmembrane. The possibility that the membrane isnewly synthesized in response to chloroquineexposure seems highly unlikely, in view of therapidity with which vacuoles appear and in viewof the fact that various inhibitors of metabolismand of protein synthesis do not block vacuoleformation. If it is assumed then that the vacuolarmembrane arises from preexisting cytomembranes,the structures which are possible contributorsinclude the following: (a) Golgi cisternae, (b)

perinuclear vesicles or vacuoles deriving from theGolgi complex or smooth ER, (c) pinocyticvacuoles, (d) dense bodies or granules, (e) tubulesor cisternae of ER, and (f) cell surface membranevia pinocytic activity. The observations made hereseem to eliminate all of these possibilities except (b)the smooth vesicles in the Golgi region andpossibly (d) the dense granules. As is pointed out inthe accompanying paper (1), Golgi cisternaeappear to be unchanged during the early stages ofchloroquine toxicity; the Golgi abnormalities seenin some macrophages after several hours ofexposure to chloroquine involve dilitation of the

M. E. FEDoRKo, J. G. HIRasH, AND Z. A. CouN Autophagic Vacuoles 11. 399

Dow


FIGURE 7 Shows various other forms of autophagic vacuoles in an L cell exposed to chloroquine for 2hr. Some of the vacuoles appear to have a distinct double membrane (double-stemmed arrows). Onevacuole contains a well-preserved mitochondrion (arrow) in addition to amorphous material and mem-branous matter. The content of other vacuoles varies widely, including membranous lamellae in anonionskin pattern, very electron-opaque structureless masses, and areas which appear identical withcytoplasmic matrix. X 20,000.

cisternae, but there is no significant disappearanceof the Golgi stacks. The experiments in whichpinocytic vacuoles were labeled with colloidal goldshow that the early chloroquine-toxic vacuoles donot arise primarily from these structures, for thetoxic vacuoles in this circumstance contained littleor no gold. Origin of the vacuoles entirely fromcytoplasmic dense bodies also seems unlikely fromthe gold-labeling observations and from the factthat extensive vacuolization appears in cells havingbut few dense bodies, e.g. macrophages culturedin low serum medium, monocytes, L cells. Themacrophages studied exhibit only a few short stripsof ER, so that origin of the toxic vacuoles from thispotential membrane source is quite unlikely.Finally, pinocytic activity of the cells can be variedby varying the serum concentration in the mediumor can be suppressed or blocked by inhibitors(puromycin, amino acid analogues, dinitrophenol,etc), without markedly affecting the vacuolizationwhich appears following chloroquine exposure.

Although the evidence thus indicates that theearly toxic vacuoles can and probably usually doarise from structures other than pinosomes, gran-ules, and ER, both phase microscopy and ultra-structural observations show that these organellesdo contribute their content to the vacuoles withpassage of time, probably by fusion.

Incrimination of vesicles of the Golgi region asthe likely source of vacuolar membrane fits withfluorescence microscopy showing that the drug israpidly taken up by macrophages and localizedin the Golgi area in tiny foci, and with light andelectron microscopic observations showing that thevacuoles make their appearance in this area. Bothphase and electron microscopy show the density ofthe content of the early toxic vacuoles to be some-what greater than that of pioncytic vacuoles orvesicles, and considerably less than that of thedense granules. We think it likely, then, that chloro-quine or its derivative is somehow concentrated inthe Golgi vesicles and alters their membranes, with


Dow


a resulting tendency of these vesicles to fuse withone another, and with other membrane limitedcytoplasmic organelles. Perhaps granules are sim-ilarly directly affected.

Previous studies have established the pathway ofhydrolase synthesis in stimulated macrophages(7). The enzymes are made in the rough ER, thenare transported to the Golgi region, and finallyappear in dense bodies. Almost certainly the ve-sicular elements in and around the Golgi regioncontain hydrolases, and if we are correct in con-cluding that these vesicles fuse with one anotherto form the early vacuoles, the vacuoles would beendowed with hydrolases from the start, and theywould likely acquire additional digestive enzymesby fusing with dense bodies. Fusion of the toxicvacuoles with pinocytic vacuoles would then con-vert them into heterophagic vacuoles (containingextracellular material and digestive enzymes).Incorporation of cytoplasmic components into thevacuoles by invagination or trapping (see Dis-cussion in accompanying paper, reference 1)would in turn convert them into autophagicvacuoles as well.

Our conclusions on origins of the limiting mem-branes of autophagic vacuoles agree reasonablywell with those drawn by others in different sys-tems (discussed in references 11-13). Novikoff andhis colleagues have proposed that autophagicvacuoles in nerve and liver cells arise from smoothER, in the Golgi zone (14, 15). The macrophagesand L cells studied here show a highly developedGolgi complex, but contain only a few ER ele-ments, so that it seems likely that most ofthe vesicles in the nuclear hof of these cells arederived from the Golgi rather than from smoothER. In any event, both Golgi and smooth ERvesicles represent protolysosomes or primary lyso-somes of these cells, and it is difficult or impossibleto distinguish between them by techniquescurrently available.

The mechanism of chloroquine uptake by cellsis unknown. Chloroquine or its fluorescent deriva-tive enters the cells within minutes after its addi-tion to the medium. That the rapid uptake andtoxic effects are not blocked by various metabolicinhibitors suggests that chloroquine is taken into

the cells by some means other than pinocytosis orenergy-dependent active transport. The only con-dition found to inhibit chloroquine uptake oraction is low temperature. Cells held at 40C in thepresence of 30 pg/ml chloroquine for 2 hr and thenwashed in the cold prior to incubation at 38°Cdo not develop toxic vacuoles. There are similar-ities between chloroquine and at least two otheragents, neutral red and acridine orange (16-19),each of which is approximately the same molecularweight, is rapidly taken up by certain cells, isapparently concentrated in lysosomes by mecha-nisms unknown, exhibits or develops fluorescenceor photosensitizing reactivity, and producesvacuolization or other cytotoxic effects when addedto cells in sufficiently high concentration.

The possible membrane-damaging action ofchloroquine is at present unexplained. In bacteriathis drug combines with deoxyribonucleic acidand blocks protein synthesis (20-22), but no evi-dence for a similar action in mammalian cells isavailable. The mechanism of the antimalarialeffects of chloroquine are unknown.

The studies of chloroquine toxicity on L cellsestablishes that the drug produces autophagicvacuoles in a cell type different from the macro-phage. The presence of a small dense body in thesmooth ER makes it possible to follow this markerin the L cells exposed to chloroquine, and thus toshow that the smooth ER vesicles and cisternaealso participate in formation of, or contribute theircontents to, the toxic vacuoles Finally, the cell-to-cell variability in time of toxic response in the Lcells, in contrast to the more or less uniform re-sponsiveness of the macrophages, leads us tospeculate that dividing cells may be susceptible tochloroquine intoxication only during certainphases of their growth cycle. Further study of thispossibility might shed additional light on themechanisms involved in the toxic response.

The authors wish to thank Mrs. Carol D. Stearnsfor her excellent technical assistance.

This work was supported by the United StatesPublic Health Service, grants No. Al 01831 and AI07012.

Received for publication 14 March 1968.

REFERENCES

1. FEDORKO, M. E., J. G. HIRSCH, and Z. A. COHN.

1968. Autophagic vacuoles produced in vitro.I. Studies on cultured macrophages exposed tochloroquine. J. Cell Biol. 38:377.

M. E. FEDORKO, J. G. HIRSCH, AND Z. A. COIIN Autophagic Vacuoles I. 401

Dow


2. COHN, Z. A., and B. BENSON. 1965. The differentia-tion of mononuclear phagocytes. J. Exptl. Med.121:153.

3. HIRSCH, J. G. 1962. Cinemicrophotographicobservations on granule lysis in polymor-phonuclear leucocytes during phagocytosis.J. Exptl. Med. 116:827.

4. BAUMER, A., H. PAU, and A. CONRADS. 1959.

Darstellung von Resochin bzw. Resochin-derivaten in Geweben. Z. Rheumaforsch. 18:433.

5. COHN, Z. A. 1966. The regulation of pinocytosisin mouse macrophages. I. Metabolic require-ments as defined by the use of inhibitors.J. Exptl. Med. 124:557.

6. COHN, Z. A., and B. BENSON. 1965. The in vitrodifferentiation of mononuclear phagocytes.II. The influence of serum on granule forma-tion, hydrolase production, and pinocytosis.J. Expil. Med. 121:835.

7. COHN, Z. A., M. E. Fedorko, and J. G. Hirsch.1966. The in vitro differentiation of mono-nuclear phagocytes. V. The formation ofmacrophage lysosomes. J. Exptl. Med. 123:757.

8. COIIN, Z. A., and E. PARKS. 1967. The regulationof pinocytosis in mouse macrophages. II.Factors inducing vesicle formation. J. Exptl.

Med. 125:213.9. COHN, Z. A., and E. PARKS. 1967. The regulation

of pinocytosis in mouse macrophages. III.The induction of vesicle formation by nucleo-tides and nucleosides. J. Exptl. AMed. 125:457.

10. COHN, Z. A., J. G. HIRSCH, and M. E. FEDORKO.1966. The in vitro differentiation of mono-nuclear phagocytes. IV. The ultrastructure ofmacrophage differentiation in the peritonealcavity and in culture. J. Exptl. Med. 123:747.

11. NOVIKOFF, A. B., E. ESSNER, and N. QUINTANA.1964. Golgi apparatus and lysosomes. Federa-

tion Proc. 23:1010.12. BRANDES, D., D. E. BUETOW, F. BERTINI, and D.

B. MALHOFF. 1964. Role of lysosomes in

cellular lytic processes. I. Effect of carbonstarvation in Euglena gracilis. Expil. Aol. Pathol.3:583.

13. DE DUVE, C., and R. WATTIAUX. 1966. Functions

of Lysosomes. Ann. Rev. Physiol. 28:435.14. NOVIKOFF, A. B., and W. SHIN. 1964. The endo-

plasmic reticulum in the Golgi zone and itsrelations to microbodies, Golgi apparatus andautophagic vacuoles in rat liver cells. J.Microscopie. 3:187.

15. HOLTZMAN, E., A. B. NOVIKOFF, and H. VILLA-VERDE. 1967. Lysosomes and GERL in normal

and chromatolytic neurons of the rat ganglionnodosum. J. Cell Biol. 33:419.

16. MORGAN, W. S., J. FERNANDO, and M. ALOUSI.

Studies of the biological activity of neutralred. Exptl. Mol. Pathol. 5:491.

17. ALOUSI, M. A., W. S. MORGAN, and R. J.STENGER. 1967. Evolution of autophagic

vacuoles in pancreatic acinar cells afterneutral red injections. Am. J. Pathol. 50:7 a.(Abstr.)

18. RoaBBINS, E., and P. I. MARCUS. 1963. Dynamicsof acridine orange-cell interactions. J. CellBiol. 18:237.

19. ALLISON, A. C., and M. R. YOUNG. 1964. Up-

take of dyes and drugs by living cells in culture.Life Sci. 3:1407.

20. ALLISON, J. L., R. L. O'BRIEN, and F. E. HAHN.1965. DNA: reaction with chloroquine. Science.

149:1111.21. COHEN, S. N., and K. L. YIELDING. 1965.

Inhibition of DNA and RNA polymerasereactions by chloroquine. Proc. Natl. Acad.Sci. U.S. 54:521.

22. CIAK, J., and F. E. HAHN. 1966. Chloroquine:mode of action. Science. 151:347.

23. HIRSCH, J. G., M. E. FEDORKO, and Z. A.COHN. 1968. Vesicle fusion and formation atthe surface of pinocytic vacuoles in macro-phages. J. Cell Biol. 38: in press.


Dow


AUTOPHAGIC VACUOLES PRODUCED IN VITRO II. Studies ...AUTOPHAGIC VACUOLES PRODUCED IN VITRO II....

Documents

Transcript of AUTOPHAGIC VACUOLES PRODUCED IN VITRO II. Studies ...AUTOPHAGIC VACUOLES PRODUCED IN VITRO II....