Proc. Natl. Acad. Sci. USAVol. 81, pp. 6383-6387, October 1984Cell Biology

Uptake of lectins by pulmonary alveolar type II cells: Subsequentdeposition into lamellar bodies

(Maclura pomifera lectin/binding/multivesicular bodies)

MARY C. WILLIAMSCardiovascular Research Institute and Department of Anatomy, University of California, San Francisco, CA 94143

Communicated by John A. Clements, June 22, 1984

ABSTRACT Type II cells of the alveolar epithelium ofadult rats have been shown to internalize the ferritin-labeledlectin from Maclurapomifera (MPA-F). This a-galactose-bind-ing molecule binds specifically to the apical plasma membraneof the cells. Once within the cell the lectin is cycled frompinocytic vesicles, to multivesicular bodies of two types, andfinally to lamellar bodies, the storage granules of surfactant.Those multivesicular bodies that first contain MPA-F lackdetectable lysosomal enzymes, while those labeled later arereactive. For 30-60 min, uptake of MPA-F is blocked byadding methyl a-D-galactoside to the instillate. Lectins thathave no or limited binding to type II cells are taken up inamounts similar to fluid phase markers. These observationsindicate that type II cells can take up substances from alveoliby the process of adsorptive endocytosis and deposit the in-gested material into lamellar bodies.

Based on the recent concept that type II cells of the alveolarepithelium recycle surfactant phospholipids and do so in arelatively efficient manner, there is renewed interest in theway in which type II cells interiorize materials from alveoli,and in the intracellular sites into which substances are de-posited after uptake. As shown in an earlier study (1), al-veolar type II cells internalize cationic ferritin and subse-quently transport it either into organelles of the secretorypathway-multivesicular bodies (MVBs) and lamellar bodies(LBs)-or into the interstitial space underlying the cells. Thefact that CF labels more than one intracellular pathway intype II cells is likely to be a consequence of its ability to bindto several populations of plasmalemmal molecules that candirect the ingested tracer to different intracellular sites. Toprovide a more specific probe by which this pathway can bestudied further, studies have been carried out with ferritin-labeled lectins. These molecules can be readily visualized byelectron microscopy with high resolution. The choice oflectins to be studied was based on earlier studies (2-4). Thetwo types of alveolar epithelial cells, type I lining cells andtype II secretory cells, display on their apical plasma mem-branes different affinities for lectins. Type II cells bindMaclura pomifera lectin (MPA), an a-galactose-specific lec-tin (5), whereas type I cells are labeled with Ricinus com-munis lectin I (RCA), which binds B-galactose. Neither cellbinds the lectins specific for the other cell. Type II cells bindlittle or no concanavalin A (Con A) (4). Uptake of ferritin-labeled Wisteria floribunda lectin (WFA-F), an acetylgalac-tosamine-binding lectin, was also studied because this mol-ecule does not bind to either cell type to any extent. The useof lectin tracers, with or without addition of the haptensugars, makes it possible to determine whether internaliza-tion is dependent on the ability of the tracer to bind to the cellmembrane.

On the basis of the observations reported here, such aninterdependence appears to be the case for type II cells as itis for many other cells. The organelles into which the lectinsare discharged after uptake by vesicles are constituents ofthe secretory pathway, namely MVBs (6) and LBs (7). Al-though it has not yet been tested by morphologic methodssuch as autoradiography, this pathway may account for thetransfer of surfactant from alveoli to LB fractions observedearlier by Hallman and co-workers (8). Some of the obser-vations reported here have been published previously in anabstract (3).

METHODSAnimals and Administration of Tracers. Experiments were

carried out on male Long-Evans rats, 275-350 g, anesthe-tized with sodium methohexital (7 mg/100 g of body weight).Animals were intubated intratracheally with a 16-gauge Tef-lon catheter inserted orally. Polyethylene tubing (PE-10, 0.61mm outside diameter) was inserted into the catheter andpushed gently several centimeters. Tracers were then in-stilled into the lung through the smaller tubing. These pro-cedures took 30-60 sec, after which both catheters wereremoved. After 10 min to 6 hr, animals were reanesthetizedand the lungs were fixed by vascular perfusion or inflation asdescribed elsewhere (9).

Tracers. Both MPA-F and WFA-F were purchased fromboth Polyscience (Warrington, PA) and E-Y Chemicals (SanMateo, CA). Con A-F was purchased from E-Y Chemicals.Biological grade carbon was obtained from Gunther-Wagner(Hanover, Germany) and was dialyzed against Gey's balianced salt solution buffered to pH 7.4 with 10 mM Hepesprior to use. Twenty microliters of carbon per animal wasadded to all tracers to provide a visible labeling of the smallarea of lung tissue that contained the lectins. Lectins wereadministered in volumes of 100 Al per animal at a proteinconcentration of 1 mg/ml. Hapten sugar control animalsreceived a combination of 100 1.l of MPA-F and 100 pul of 0.1M methyl a-D-galactoside (Sigma) or a combination of 100 A.lof Con A-F and 100 p.l of 0.1 M methyl a-D-mannoside.Sugars were dissolved in phosphate-buffered isotonic saline.

Fixatiohs and Processing for Electron Microscopy. Lectin-labeled lungs were fixed by inflation with 2% (wt/vol)glutaraldehyde/1% freshly prepared paraformaldehyde in 0.1M sodium phosphate buffer, pH 7.4, 370C, with or without4% (9) polyvinylpyrrolidone. Tissues to be used for thelocalization of acid phosphatase or aryl sulfatase were fixedby vascular perfusion with the above fixatives in 0.1 Msodium cacodylate buffer with 4% polyvinylpyrrolidone, pH7.4, 40C. Tissues were postfixed in osmium tetroxide, blockstained with uranium salts, dehydrated in cold graded ac-etone solutions, and embedded in Epon. Unstained thin

Abbreviations: MVB, multivesicular body; LB, lamellar body; MPA,Maclura pomifera lectin; WFA, Wisteria floribunda lectin; F, fer-ritin.

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

6383

Dow

nloa

ded

by g

uest

on

June

12,

202

0

Proc. Natl. Acad. Sci. USA 81 (1984)

F-'@FIFa o , r s,4. - 7

FIG. 1. (A) Scattered molecules of MPA-F are bound to the api-cal plasma membrane of a type II cell. Small vesicles (arrows), anMVB, and LB also contain MPA-F by 60 min after instillation.(x 77,500.) (B) After uptake into small vesicles (arrow), MPA-Fnext appears within large, irregularly shaped MVBs that lack anelectron-dense matrix. ( x 112,500.)

sections were examined in a Zeiss EM-10 electron micro-scope.

Cytochemistry. Aryl sulfatase or acid phosphatase wasdemonstrated in tissues from rats labeled with MPA-F for30-180 min. After fixation and washing, 125-pim-thick sec-tions were cut on a Sorvall TC-2 tissue sectioner (DuPontInstruments, Sorvall Division). Aryl sulfatase was demon-strated by the method of Goldfischer (10) as described byWall et al. (11). Acid phosphatase was demonstrated in amodified Gomori medium using p-glycerophosphate as asubstrate at pH 5.0 (12). Incubations were at 220C or 370C for2-3 hr. ;The number of animals studied for each tracer and the

range of times at which tissues were collected were MPA-F(nine animals, 10 min to 6 hrs), MPA-F plus methyl a-D-galactoside (three animals, 30 min to 6 hr), WFA-F (twoanimals, 60 or 75 min), Con A-F (three animals, 30 min to 2hr), and Con A-F plus methyl a-D-mannoside (one animal, 30mi).

RESULTSAs in the previous study (i), alveolar macrophages ingestedall tracers, including colloidal carbon, at all time periods.The presence of labeled vesicles in these cells indicated thatthe tracers had reached a specific alveolus and that theycould be readily detected by electron microscopy. Epithelialcells were studied only if labeled macrophages were nearby.None of the tracers produced damage to cellular structure,but at early time periods a few small distentions of the

FIG. 3. After 30-60 min, MPA-F is found in the nonlamellar ma-terial of LBs (arrows). The tracer is also present in an MVB that isapposed or attached to an LB. ( x 57,500.)

interstitial spaces indicated the presence of excess tissuefluid.MPA-F. By 10 min after its instillation, MPA-F was pres-

ent at low concentrations in small vesicles, some of whichwere clathrin coated (Figs. 1 and 2), and in a few electron-lucent MVBs. After 30 min the tracer was present in thesecompartments as well as in electron-dense MVBs (Fig. 3).By 30 min, fewer than 50% of the LBs in any particularlabeled cell contained small amounts of tracer at this time.Images suggestive of fusion of small labeled vesicles withMVBs, especially of the electron-lucent type, were frequent(Fig. 1B). No MPA-F was observed in the interior of thesmall vesicles of MVBs. By 60 min after instillation mostLBs of labeled cells contained MPA-F, and the amount oftracer within these organelles increased with longer expo-sures (Fig. 3). At 6 hr after instillation of the tracer, MPA-Fwas still adherent to the apical plasma membrane, and pi-nocytic vesicles, MVBs, and LBs were labeled as earlier. Noother organelles of type II cells, such as Golgi complex,endoplasmic reticulum, etc., contained MPA-F at any time,and none was observed in the interstitial compartment un-derlying either type II or type I cells. The addition of a-galactose to the MPA-F instillate blocked uptake by type II

FIG. 2. Some lectin-labeled vesicles have clathrin coats (arrows,CV), while others are smooth surfaced. (A, x216,000; B, x 172,000.)

FIG. 4. (A) When a-galactose is added to the instillate, MPA-Fis present within vacuoles and vesicles of alveolar macrophages(right) but not within adjacent type II cells. ( x 7,500.) (B) The typeII cell apex is devoid of MPA-F labeling, and none is found withinthe cell. ( x 62,500.)

6384 Cell Bidlog' : Williams

Dow

nloa

ded

by g

uest

on

June

12,

202

0

Proc. Natl. Acad. Sci. USA 81 (1984) 6385

FIG. 5. A few molecules of Con A-F are present in an MVB(left) after a 2-hr exposure, but most other MVBs and LBs containnone of this tracer. (x 90,000.)

cells, but not neighboring alveolar macrophages, for 30-60min (Fig. 4). By 6 hr, however, the inhibition was no longerevident and MVBs and LBs were labeled as before. This ismost likely due to a rapid loss of the low molecular weightsugar into adjacent alveoli, the interstitium, or both; thelarger MPA-F cannot penetrate tight junctions, may there-fore be trapped in alveoli, and eventually is taken up by typeII cells.Con A-F (Fig. 5) and WFA-F were found in type II cells in

much smaller amounts than MPA-F, although when presentthey were in the same organelles. By 2 hr after the instillationof Con A, most type II cells contained no tracer; in some

cells a few molecules were present in the nonlamellar areas

of LBs. Con A-F was also observed in type II cells of animalsgiven both the tracer and its hapten sugar, a-mannose.

Lysosomal enzymes were localized cytochemically in tis-sues exposed to MPA-F for 1-3 hr to determine whether theearliest MPA-F labeled MVBs Contain acid phosphatase or

aryl sulfatase. As shown by others (13), both enzymes couldbe demonstrated within most LBs, Golgi elements, and a fewsmall vesicles. A subpopulation of MVBs, mainly thosesmall and electron dense, was also enzymatically reactive(Fig. 6). MPA-F was present in many of the electron-lucentMVBs, but no enzymatic activity could be demonstrated inthese structures (Fig. 6). Often the negative MVBs were

located immediately adjacent to a highly reactive structuresuch as an LB. Similar results were obtained for the two

enzymes.In contrast to type II cells taking up MPA-F, type I cells

internalized significantly fewer molecules of all of the tracersstudied, even though lectins were present in those pinocyticvesicles that were open to the alveolar lumen. In type I cellsoccasional tracer molecules were observed in MVBs, butthese organelles are seen infrequently in this cell type. Trac-ers were not transported to the interstitial space underlyingtype I cells within the range of times studied.

DISCUSSIONAlveolar type II cells synthesize, store, and secrete a com-

plex phospholipid-protein material called pulmonary surfaceactive material or surfactant, a mixture of neutral and anionicphospholipids and proteins. Prior to its secretion this ma-

terial is stored in LBs, which contain, in addition to surfac-tant, lysosomal enzymes (13, 14). As indicated later in thisdiscussion, these secretory cells are now believed to retrievefrom alveoli some or part of their own secretory product andrepackage it for storage in LBs.Much recent experimental evidence indicates that various

types of cells ingest materials by both nonspecific and highlyspecific mechanisms. These studies on uptake of lectins by

FIG. 6. Those MVBs that first contain MPA-F (left) contain nodetectable aryl sulfatase (shown here) or acid phosphatase, whilethe population labeled later (right) contains these enzymes.(x72,500.)

type II cells were undertaken to seek evidence for or againstthe occurrence of a specific uptake process, adsorptive en-docytosis, in these secretory cells. The general structuraland metabolic characteristics of adsorptive endocytosis havebeen reviewed recently by Steinman et al. (15), by Brown etal. (16), and by others. As determined mainly from studies ofcultured cells or isolated perfused organs, these character-istics include the high-affinity binding of a ligand to plasma-lemmal ldi, followed by a very rapid uptake of the ligahdinto pinocytic vesicles. Within this time period the ligand issubsequently deposited into an irregularly shaped, smooth-surfaced "endosome" (15, 17) or "receptosome" (18). Thisnewly identified organelle appears not to contain lysosomhalenzymes (11, 19) and may be the site of uncoupling ofreceptor-ligand complexes (20) from which the receptorsmay return to the cell surface. The concentration of ligand inenddsomes exceeds that of the pinocytic vesicles labeledearlier. In most cases fusion with lysosomes follows. In somecells certain molecules such as cationic ferritin, which bindelectrostatically to cell membranes, proceed to the Golgiapparatus with or without an intermediate stop at a lysosome(21, 22). These characteristics contrast with those of fluidphase pinocytosis, which requires no binding. Different trac-ers are taken up at approximately the same rate, perhaps asa constitutive cellular process. The results of the presentstudies indicate that type II cells take up MPA-F by adsorp-tive endocytosis and deposit it into organelles of the secre-tory pathway.

Several observations provide evidence for the specificityof MPA-F binding and uptake by type II cells. First, bindingof MPA-F to type II cells is inhibited by the addition ofexcess concentrations of the hapten sugar in vivo, as in thisstudy, in lung slices (unpublished observations) and in typeII cells cultured for 24 hr (23). Second, MPA-F binds only totype II cells and not to adjacent type I cells even though theyare likely to be exposed simultaneously to equal concentra-tions of tracer. Like binding, endocytosis of MPA-F is alsocell specific. Type I cells do not endocytose any of thetracers in significant amounts, whereas alveolar macro-phages rapidly ingest large amounts of all tracers. In con-trast, type II cells discriminate between the various tracersstudied and subsequently ingest in significant quantities onlythose two molecules for which there is evidence of binding,namely cationic ferritin (1) and MPA-F.

After uptake, MPA-F and some ingested cationic ferritinare transported to certain organelles of the secretory path-way. This pathway resembles in certain respects the adsorp-tive endocytic pathway in hepatocytes, fibroblasts, and othercells where ligands are deposited in lysosomes. This simi-larity is a consequence of the peculiar nature of the organ-elles (LBs and at least some MVBs) that participate insurfactant transfer and storage; these organelles containlysosomal enzymes as well as the phospholipid lamellae ofsurfactant. It is therefore unclear whether the role played bythis pathway in normal alveolar function is one related toproduction, storage, and recycling of surfactant compo-

Cell Biology: Williams

Dow

nloa

ded

by g

uest

on

June

12,

202

0

Proc. Natl. Acad. Sci. USA 81 (1984)

AcPF

FIG. 7. Schematic of pathways taken by tracers and MVBs.Molecules that bind to the cell membrane (X) are taken into vesicles(1), which become or empty into acid phosphatase-negative (AcP-)MVBs (2). These receive lysosomal enzymes (3) by fusion with pri-mary lysosomes or MVBs derived from the Golgi area (RER, roughendoplasmic reticulum) (4). Some of the tracer-laden enzyme-posi-tive MVBs fuse with forming LBs (5), while others may carry outtypical hydrolytic functions of traditional lysosomes (6). The returnof membrane to the cell surface and the formation of containedvesicles of MVBs by invagination are uncertain (?) but feasible.

nents, or one related to catabolism of material ingested fromalveoli, or both.The sequence of structures that transport MPA-F is mor-

phologically similar to that of organelles known to participatein receptor-mediated endocytosis. Uptake occurs by meansof small pinocytic vesicles. Although some of these haveclathrin coats, this modification is no longer believed to bespecific for receptor-mediated internalization (24). The MPA-F is then deposited into a large-irregularly shaped MVB'thatlacks detectable lysosomal enzymes. At this point someMPA-F molecules appear to be bound to the MVB mem-brane; others are detached and free in the lumen or appear toadhere to the outer surface of the contained vesicles. SomeMPA-F may, therefore, be dissociated from its receptor atthis time. MPA-F is then concentrated in the smaller MVBs,which are in general positive for lysosomal enzymes. Themechanism by which this latter population ofMVBs acquireslysosomal enzymes is not certain, but is presumably byfusion with a primary lysosome or with a Golgi-derived,enzyme-positive MVB. The size difference between the smallearly labeling MVBs and those labeled later is suggestive ofa loss of membrane to some other compartment at this time.Much evidence indicates that the small MVI3s then fuse withlamellar bodies and, in doing so, contribute to them'the smallcontained vesicles, lysosomal enzymes, and in this caseMPA-F derived from the alveolar lumen.These observations have implications relating to the for-

mation of LBs, especially with regard to functions of MVBs(Fig. 7). MVBs are present in unusually large numbers intype IX cells (6, 25). On the basis of their autoradiographicstudies, Chevalier and Collet (6) suggested that MVBs trans-port newly synthesized protein, buit not phospholipid, fromtthe Golgi area to developing LBs. As shown here, MVBs canalso collect materials from the alveolus en route to LBs.Whether the two populations of MVBs described here are infact originally derived from two different sources, one theapical plasma membrane and adjacent structures and theother the Golgi area, is uncertain but seems likely. If this iscorrect, the two populations are likely to merge later-becauseMPA-F labels most MVBs with time. The fact that bothpopulations of MVBs contain internal vesicles, whether de-rived from the Golgi complex or elsewhere, argues that allMVBs transport phospholipid to LBs. These small vesicles

are often observed in LBs but they seem to disappear as theLBs enlarge, perhaps by fusion to adjacent lamellae. Thesource of the membrane of the contained vesicles is ingeneral unknown, as is the mechanism by which the vesiclesare freed into the MVB lumen. In certain cells the containedvesicles of MVBs may be derived from patches of plasmamembrane that contain receptors for specific hormoneligands (26). In other cases they react positively with antiseraagainst the Golgi apparatus (27) and have similar histochem-ical properties to Golgi membranes (28).These studies were initiated because of an interestiin the

cellular mechanisms that account for the recently describedphenomenon of recycling of surfactant phospholipids. Gei-ger reported in an early study that type II cells internalizedradiolabeled aerosolized dipalmitoyl lecithin, a major con-stituent of surfactant (29). Hallman et al. (8) showed thatlabels from surfactant phospholipids, prepared biosyntheti-cally and instilled into alveoli, appeared in LBs isolated fromlung within minutes after instillation of the tracer. A maximalspecific activity was reached in the LB fraction in about 60min. Using more theoretical approaches, Jacobs et al. (30),Glatz et al. (31), Jobe and co-workers (32), and Magoon et al.(33) calculated that about 85-95% of natural surfactantphospholipid may be recycled. As discussed by Hallman, ofthe several mechanisms possible for surfactant recycling,uptake in vesicles is consistent with data to the present. Thetime course, fate, and specificity of MPA-F uptake are com-patible with such a bulk uptake process being performed bytype II cells. The organelles that carry MPA-F are devoid ofidentifiable phospholipid, other than the contained vesicles.If, however, phospholipid is recycled in a nonbilayer form,the techniques used for this study may be unable to fix itwithin the tissue or to render the complex electron dense andresolvable. Clearly a variety of experimental approaches willbe needed to define factors that regulate surfactant recyclingand the cellular mechanisms by which it is accomplished.

Dr. Alan E. Brandt generously shared his observations on MPAlabeling of type II cells prior to publication. Ms. Lennell Allen andMs. Marcia Hansen provided excellent technical assistance, and Dr.Leland G. Dobbs assisted in the instillation of various tracers. Drs.John A. Clements and Susan R. Walker kindly read the manuscriptprior to publication. These studies were supported by ProgramProject Grant HL-24075 from the National Heart, Lung, and BloodInstitute.

1. Williams, M. C. (1984) Proc. Natl. Acad. Sci. USA 81, 6054-6058.

2. Branidt, A. (1982) Fed. Proc. Fed. Am. Soc. Exp. Biol. 41, 755(abstr.).

3. Williams, M. C. (1983) Am. Rev. Respir. Dis. 127, 271 (abstr.).4. Dixon, M. T. & Jersild, R. A. (1983) Am. J. Pathol. 113,

389-395.5. Bausch, J. N. & Poretz, R. D. (1977) Biochemistry 16,

5790-5794.6. Chevalier, G. & Collet, A. J. (1975) Anat. Rec. 174, 289-310.7. Gil, J. & Reiss, 0. K. (1973) J. Cell Biol. 58, 152-171.8. Hallman, M., Epstein, B. L. & Gluck, L. (1981) J. Clin. Invest.

68, 742-751.9. Williams, M. C. & Benson, B. J. (1981) J. Histochem. Cyto-

chem. 29, 291-305.10. Goldfischer, S. (1%5) J. Histochem. Cytochem. 13, 520-523.11. Wall, D. A., Wilson, G. & Hubbard, A. L. (1980) Cell 21,

79-93.12. Nichols, B. A. (1976) J. Exp. Med. 144, 920-932.13. Goldfischer, S., Kikkawa, Y. & Hoffman, L. (1968) J. His-

tochem. Cytochem. 16, 102-109.14. Hook, G. E. R. & Gilmore, L. EN. (1982) J. Biol. Chem. 257,

9211-9220.15. Steinman, R. M., Mellman, I. S., Muller, W. A. & Cohn,

Z. A. (1983) J. Cell Biol. 96, 1-27.16. Brown, M. S., Anderson, R. G. W. & Goldstein, J. L. (1983)

Cell 32, 663-667.

6386 Cell Biology: Williams

Dow

nloa

ded

by g

uest

on

June

12,

202

0

Cell Biology: Williams

17. Simons, K., Garoff, H. & Helenius, A. (1982) Sci. Am. 246 (2),58-66.

18. Willingham, M. C. & Pastan, 1. H. (1980) Cell 21, 67-77.19. Merion, M. & Sly, W. S. (1983) J. Cell Biol. 96, 644-650.20. Geuze, H. J., Slot, J. W., Strous, G. J. A. M., Lodish, H. F.

& Schwartz, A. L. (1983) Cell 32, 277-287.21. Farquhar, M. G. (1978) J. Cell Biol. 77, R35-R42.22. Herzog, V. & Miller, F. (1979) Eur. J. Cell Biol. 19, 203-215.23. Dobbs, L. G., Williams, M. C. & Brandt, A. E. (1983) J. Cell

Biol. 97, 332 (abstr.).24. Aggeler, J. & Werb, Z. (1982) J. Cell Biol. 94, 613-623.25. Sorokin, S. P. (1967) J. Histochem. Cytochem. 14, 884-897.26. McKanna, J. A., Haigler, H. T. & Cohen, S. (1979) Proc.

Natl. Acad. Sci. USA 76, 5689-5693.

Proc. Nati. Acad. Sci. USA 81 (1984) 6387

27. Tougard, C., Louvard, D., Picart, R. & Tixier-Vidal, A. (1983)J. Cell Biol. 96, 1197-1207.

28. Friend, D. S. (1969) J. Cell Biol. 44, 269-279.29. Geiger, K., Gallagher, M. L. & Hedley-Whyte, J. (1975) J.

Appl. Physiol. 39, 759-766.30. Jacobs, H., Jobe, A., Ikegami, M. & Jones, S. (1982) J. Biol.

Chem. 257, 1805-1810.31. Glatz, T., Ikegami, M. & Jobe, A. (1982) Pediatr. Res. 16,

711-715.32. Jacobs, H., Jobe, A., Ikegami, M. & Conway, D. (1983) J. Biol.

Chem. 258, 4159-4165.33. Magoon, M. W., Wright, J. R., Baritussio, A., Williams,

M. C., Goerke, J., Benson, B. J., Hamilton, R. L. & Clem-ents, J. A. (1983) Biochim. Biophys. Acta 750, 18-31.

Dow

nloa

ded

by g

uest

on

June

12,

202

0