Proc. Nat. Acad. Sci. USAVol. 73, No. 3, pp. 837-841, March 1976Cell Biology

Effect of concanavalin A on expression of cell surfacesialyltransferase activity of mouse thymocytes

(glycosyltransferases/mitogens/plasma membrane)

RICHARD G. PAINTER AND ABRAHAM WHITEDepartment of Cell Biology, Institute of Biological Sciences, Syntex Research, Palo Alto, California 94304

Contributed by Abraham White, December 12, 1975

ABSTRACT Incubation of mouse thymocytes with mito-genic concentrations of concanavalin A causes a 2-fold in-crease in cell-surface-associated (but not total cell) sialyl-transferase activity (ectosialyltransferase, CMP-N-acetylneu-raminate:D-galactosyl-glycoprotein N-acetylneuraminyltrans-ferase, EC 2.4.99.1) as judged by incorporation of ["4C]sialicacid into endogenous cell acceptors and into added desialy-lated fetuin acceptor. The concanavalin-A-induced enhance-ment of enzymic activity is essentially complete within 1 hrafter addition of mitogen and remains at elevated levels for12 hr, declining rapidly thereafter. Intact cells labeled pre-viously with ["Clsialic acid and then incubated briefly withhydrolytic enzymes, including neuraminidase and insolubletrypsin, released 43-66% of total cell-associated radioactivi-ty without appreciably changing cell viability. Alterations insialyltransferase activity due to concanavalin A treatmentcould not be explained b a mitogen-mediated (a) uptake ofradioactive precursors, ( ) cell death, (c) increased productcatabolism, or (d) activation of sialyltransferase by mitogenbinding to the enzyme. Furthermore, the process does not re-quire active protein synthesis. The results are consistent witha rapid concanavalin-A-induced exposure of potential enzy-mic activity that was previously inaccessible to substrate.

Glycoprotein and glycolipid components of cell plasmamembranes that are localized primarily at the cell surfacecan function as receptors for plant mitogens such as con-canavalin A (Con A) and are, in part, responsible for expres-sion of a large variety of surface antigens (for a review seeref. 1).The binding of mitogens, such as Con A and phytohem-

agglutinin, to lymphocytes results in a variety of early bio-chemical changes, many of which occur at the level of theplasma membrane. For example, human peripheral lym-phocytes, within hours after exposure to mitogenic doses ofphytohemagglutinin, have increased rates of biosyntheticturnover of cholesterol (2), total phosphatidylcholine (3),phosphatidylinositol (4), and glycoproteins (5). Increasedturnover of membrane protein has been reported (6) in rab-bit thymocytes exposed to Con A.

Interpretation of data of this type is complicated by thefact that the appearance of isotope in cell membrane constit-uents may be sequential in nature, appearing in plasmamembrane fractions at relatively long periods after additionof radioactive precursor (7-9), depending upon the labeledprecursor used. Even in rapidly growing cell lines, transferof labeled galactose from the Golgi region into the plasmamembrane glycoprotein does not occur until at least 90 minor more after addition of sugar to the medium (8).A more direct approach to mitogen-induced alterations of

macromolecular components of cell plasma membranes in-volves assessing the appearance of cell surface markers sub-

Abbreviations: Con A, concanavalin A; CMP-NANA, cytidine-5'-monophosphate N-acetylneuraminic acid; a-MeMan, a-methyl-D-mannopyranoside.

837

sequent to exposure to mitogens. For example, mitogen-in-duced increases in insulin receptors (10), Na+,K+-dependentATPase (11), ouabain-binding sites (12), and alterations insurface antigens (13, 14) have been reported. A number ofadditional techniques are now available, including methodsapplicable to the radioactive labeling of both galactose (15)and N-acetylneuraminic acid (NANA) (16) localized specifi-cally at the cell surface. One of these (16) involves specificlabeling of surface glycoproteins (or glycolipids) by utilizingan endogenous sialyltransferase (ectosialyltransferase,CMP-N-acetylneuraminate:D-galactosyl-glycoprotein N-acetylneuraminyltransferase EC 2.4.99.1),* apparently lo-calized at the cell surface (16-18), to catalyze the transfer of[14C]NANA from CMP-[14C]NANA, which does not entercells, into membrane acceptors or, alternatively, into exo-genously added macromolecular acceptors.

This communication reports the successful application ofthese techniques to measure the kinetics of early (<3 hr) ef-fects of mitogenic concentrations of Con A on (a) surfacesialic-acid-containing components and (b) the expression ofsurface sialyltransferase activity.

MATERIALS AND METHODS

Preparation of Mouse Thymocyte Suspensions. AdultC57B1/6J male mice 6-25 weeks of age (Jackson Laborato-ries, Bar Harbor, Me.) were sacrificed by cervical dislocationand thymus glands were excised under aseptic conditions.Glands from a number of animals were pooled, minced inHanks' balanced salt solution and the thymocytes were freedfrom tissue pieces by gently pressing the mince throughnylon mesh. After standing at room temperature for 10 min,the cell-containing supernatant was centrifuged at 325 X gfor 10 min and the cell pellets were twice washed at the cen-trifuge with 10 volumes of Hanks' solution. The cells werethen suspended in Dulbecco's minimum essential mediumsupplemented with 5% fetal calf serum (vol/vol), previouslyheated for 30 min at 560, 2 mM L-glutamine, 100 units/mlof penicillin, and 100 tg/ml of streptomycin at a final celldensity of 107 cells per ml. All tissue culture reagents werepurchased from Gibco (Berkeley, Calif.).

Cell suspensions in loosely capped polystyrene tubes (17 X100 mm, Falcon Plastics tube no. 2051) containing 1.0 ml al-iquots of 107 thymocytes in complete media were culturedin a 370 CO2 incubator equilibrated with 10% C02-90% air.Unless otherwise indicated, the final concentration of Con Awas 5 ,ug/ml.Measurement of Sialyltransferase. Unless otherwise

specified, cultures (in duplicate) were terminated after 1 hr

* The surface-localized enzyme, together with its acceptor mole-cules, defines the ectosialyltransferase system (18).

Dow

nloa

ded

by g

uest

on

Nov

embe

r 30

, 202

0

838 Cell Biology: Painter and White

and transferred quantitatively to capped conical centrifugetubes (Falcon Plastics tube no. 2095), which were centri-fuged at 325 X g for 10 min. The cell pellets were washedtwice at the centrifuge with cold isotonic assay buffer com-posed of 0.12 M NaCl-1 mM MgCl2-0.01 M potassiumphosphate buffer (pH 6.5). When measurement of incorpo-ration of [14C]NANA into endogenous cell acceptors wasdesired, the cell pellets were assayed by suspension in 60 ,lof assay buffer containing 8.56 ,M CMP-[14C]NANA (217mCi/mmol, lot no. 745-197, New England Nuclear, Boston,Mass.) labeled uniformly at carbons 4, 5, 6, 7, 8, and 9 ofNANA. For measurement of incorporation of [14C]NANA inthe presence of an exogenous acceptor, the assay was car-ried out as above, except that 250 itg of desialylated fetuinwas included in the assay medium. When total enzyme lev-els were measured, the assay medium contained, in additionto substrate and desialylated fetuin, 0.1% (vol/vol) TritonX-100. In a few experiments cells were assayed after briefsonication. In all experiments assays were carried out in a370 water bath with occasional shaking. The reaction wasterminated after 2 hr, unless otherwise indicated, by addi-tion of 1 ml of 1% phosphotungstic acid (wt/vol)-0.5 MHC1. The acid-insoluble material was collected by centrifu-gation and processed as described by McLean and Bosmann(18). After the final residue was dissolved by addition of 0.2ml of 0.5 M KOH, aliquots were taken for liquid scintillationcounting (0.15 ml) and cell protein (15 4ul) determination(19). Recovery of cell protein for control and Con-A-treatedcultures was generally identical. When desialylated fetuinwas present, the cell protein values were assumed to be iden-tical to those assay tubes in which desialylated fetuin was notpresent. Sialyltransferase activities were expressed as cpmincorporated/mg of cell protein per hr after correction forcpm incorporated in the absence of cells, which was essen-tially identical to the background counting rate. Unless spe-cifically indicated, values obtained in the presence of fetuinacceptor were not corrected for endogenous cell acceptors.

Incorporation of acid-soluble radioactivity was measuredby the method of Cunningham and Pardee (20) as modifiedby Datta (16).

Incorporation of [3H]thymidine (20 Ci/mmol, New En-gland Nuclear) was assessed by addition of 1.5 ,Ci dissolvedin 0.1 ml of complete medium to cultures during the last 24hr of a 48-hr cubation period. Cells were collected on glassfiber filters (Whatman) and washed successively with 10 mleach of 0.15 M NaCl, 10% (vol/vol) trichloroacetic acid andabsolute methanol. Filters were counted for 3H in 10 ml ofAquasol.

Fetuin acceptor was prepared from fetuin (Spiro type,Gibco, Inc.) as described by Kim et al. (21). Con A, obtainedfrom Miles-Yeda (Kankakee, Ill.), was repurified by affinitychromatography (22). Vibrio cholerae neuraminidase anda-methyl-D-mannopyranoside (a-MeMan) were purchasedfrom Calbiochem (La Jolla, Calif.). Cycloheximide, N-acetylneuraminic acid, and proteolytic enzymes were pur-chased from Sigma (St. Louis, Mo.). Trypsin coupled toSepharose beads was obtained from Worthington (Freehold,N.J.).

RESULTSViable (>90% as judged by trypan blue exclusion) thymo-cytes were preincubated at 370 with and without mitogenicconcentrations of Con A. Cultures were harvested after 1 hrby centrifugation, and washed twice with cold 0.12 M NaCl-0.01 M potassium phosphate (pH 6.5)-i mM MgCl2.

aW 200 -

0

0O /ZE 100

00 60 120 180

MINUTES

FIG. 1. Incorporation of [14CJNANA into trichloroacetic acid-insoluble (circles) and acid-soluble (squares) fractions of intactthymocytes incubated with 8.56 gM CMP-_[4C]NANA after treat-ment with (closed symbols) and withoat (open symbols) 5 ,g/ml ofCon A for 1 hr prior to assay. At various times duplicate aliquotscontaining 1.3 X 107 cells (0.138 mg of cell protein) were with-drawn and measurements were made of radioactivity incorporatedinto acid-soluble and -insoluble fractions.

Washed cell pellets were assayed for ability to incorporate[14C]NANA into endogenous cell acceptors as a function oftime at 37'.

Fig. 1 indicates that control thymocytes incorporate radio-activity into trichloroacetic acid-insoluble cell acceptors at arelatively constant rate for a period of 3 hr. The rate of in-corporation is equivalent to 0.53 pmol of [14C]NANA/mg ofcell protein per hr. Preincubation of thymocytes with mito-genic concentrations of Con A results in a doubling in rate of['4C]NANA incorporation to a value of 1.28 pmol/mg of cellprotein per hr. As in the control, the rate of incorporation isapproximately linear with time. On the other hand, incorpo-ration of radioactivity into acid-soluble fractions is quitelow, as might be expected from the known impermeabilityof cells to the nucleotide sugar. In any case, there is no de-tectable response of this parameter to the presence of Con A.When desialylated fetuin acceptor was included in the

assay medium, a similar response was observed, resulting inrates of incorporation equivalent to 1.0 and 2.0 pmol/mg ofcell protein per hr for control and Con-A-pretreated cells,respectively (data not shown).

In the above experiments, there was no significant loss ofcell viability during the incubation period. The relationshipbetween the Con A concentration used in the 1 hr preincu-bation period and sialyltransferase levels is depicted in Fig.2. The rates of incorporation of radioactivity into both cellu-lar acceptors and fetuin acceptor increase with increasingCon A concentration, reaching a maximum response at 5-10,ug/ml. The dose response of enzyme after 1 hr of cultureparallels that of the mitogenic response as assessed by cellu-lar incorporation of [3H]thymidine into DNA after 48 hr ofculture.

Assay of cells for sialyltransferase activity as a function ofpreincubation time following the addition of mitogen (Fig.3) revealed that the magnitude of ,he response to Con A,when compared to that in control cells, reaches a maximumwithin the first few hours of culture with regard to both in-corporation into endogenous acceptors (Fig. 3A) and in thepresence of exogenous acceptors (Fig. 3B). The respectiveenzymic levels remain relatively constant for 12 hr in Con-A-treated cultures, declining rather abruptly between 12

Proc. Nat. Acad. Sci. USA 73 (1976)

Dow

nloa

ded

by g

uest

on

Nov

embe

r 30

, 202

0

Proc. Nat. Acad. Sci. USA 73 (1976) - 839

I

w0_

LU-

ZT_0 IC a:

0 Ij

0 oUJ- "

- x

3 30 W

-4iIm

_<-

20m

ZC)( C

100:Dr

M--I10 0 mmI mqm0

00 5 10

FINAL CON A CONCENTRATION jig/mlFIG. 2. Effect of Con A concentration on surface sialyltrans-

ferase levels (-) and mitogenic activity (--- -) of thymocytes. Cellswere preincubated for 1 hr with the indicated concentration of ConA, washed, and assayed for surface sialyltransferase either in thepresence (0) or absence (0) of desialylated fetuin acceptor (seeMaterials and Methods). Mitogenic activity was assessed by mea-surement of incorporation of [3H]thymidine. Vertical lines indicateaverage deviation of the mean when greater than size of symbol.

and 24 hr. However, Con-A-stimulated cultures did not re-turn to corresponding control levels for as long as 72 hr afterinitiation of the culture. In contrast, control cultures exhibita rather gradual decrease in enzymic specific activity.

Since the data were consistent with a possible increasedexpression of sialyltransferase rather than cell surface ac-ceptors, experiments were designed to determine whetherde novo protein synthesis was an obligatory requirement forthe observed effect (Table 1). Cycloheximide, even at a rath-er high concentration, had only a marginal inhibitory effecton the response of intact cells to Con A. In addition, whenthe total sialyltransferase levels of control (Con A + a-MeMan, a specific inhibitor of Con A binding to cells) andCon-A-treated cells are compared, no appreciable effect ofCon A on total cellular enzyme is apparent. Note, however,that prevention of Con A binding by a-MeMan blocks com-pletely the response of enzyme when measured in intactcells.One possible mechanism by which Con A might affect si-

alyltransferase activity would involve binding of the mito-gen directly to the enzyme. To explore this possibility, soni-cates of untreated thymocytes were prepared and the enzy-mic activity was titrated against increasing concentrations ofCon A. The results (not shown) indicated no appreciable ef-fect of Con A on the sialyltransferase activity of cell homog-enates at Con A concentrations as high as 50 ,g/ml, despitethe fact that Con-A-induced agglutination had taken place.

Additional evidence for incorporation of [14C]NANA intosurface receptors was sought, even though this phenomenonhas been well documented in other instances (16-18, 23).Con-A-pretreated thymocytes were labeled for 3 hr, usingconditions similar to those described in Fig. 1. After label-ing, the cells were washed four times with NaN3-containingRPMI 1640. The cells (1.6 X 107/ml) were suspended inRPMI 1640 containing either trypsin, Pronase, Vibrio chol-erae neuraminidase, insoluble trypsin, or no enzyme and in-cubated at 370 for 15 min. After the reaction was quenchedby addition of 0.25 ml of fetal calf serum, the cells were col-lected by centrifugation. The radioactivity released into the

I

LLJ0-

LLJ

0~

0--J-J

(9

a-U;

IE 1500Ir

cLzu0o 10000c--J-J

O 500-i

0~

U

0 24 48 72HOURS OF CU LTUR E

0 24 48 72HOURS OF CU LTUR E

FIG. 3. Effect of preincubation time on surface sialyltransfer-ase activities of thymocytes cultured with (0) and without (0)Con A (5 gg/ml). Cultures were harvested at the indicated periodsafter initiation of incubation and assayed for an additional 2 hr forsialyltransferase activity in the absence (A) and presence (B) ofdesialylated fetuin as described in Fig. 2. Vertical lines indicate av-erage deviation of the mean when greater than size of symbol.

supernatant fraction was measured and calculation wasmade of the percent of the total recovered radioactivity re-leased.

As shown in Table 2, approximately 66% of the total ra-dioactivity was released by neuraminidase, while trypsin lib-erated as much as 55% of the total radioactivity and Pronase46%. In the absence of any added enzyme, 23% of the totalradioactivity was released.

DISCUSSIONThe data presented demonstrate that exposure of mouse thy-mocytes to a mitogenic concentration of Con A produces anapparent 2-fold enhancement of activity of plasma-mem-brane-bound sialyltransferase. The effect is expressed re-gardless of whether enzymic assays are based upon incorpo-ration of radioactivity into endogenous cellular acceptors orinto exogenously added fetuin acceptor. Although the datado not rigorously exclude an increased availability of endog-enous acceptors, they are more compatible with a Con-A-induced increase in enzyme availability (or activity) towardits substrates at the cell surface. De novo protein synthesis isapparently not a requirement for expression of the effect(Table 1).

Assay of sialyltransferase activity as a function of the time

Cell Biology: Painter and White

Dow

nloa

ded

by g

uest

on

Nov

embe

r 30

, 202

0

840 Cell Biology: Painter and White

Table 1. Effect of inhibitors on exogenous sialyltransferase activity of intact and detergent-disrupted thymocytes*

Sialyltransferase activity (pmol of [ "4C ] NANAincorporated into fetuin acceptor/hr per mg of

cell protein)t

Cell treatment Intact cells Disrupted cells

Preincubation 60 min with medium aloneprior to assay 0.5 ± 0.1 Not done

Preincubation 60 min with 5 jig/ml ofCon A prior to assay 1.4 ± 0.1 6.4 ± 0.3

Preincubation 60 min with 5 ,;g/ml ofCon A and 0.05 M a-MeMan priorto assay 0.6 ± 0.2 7.3 ± 0.5

Preincubation with 5 mig/ml of Con A and100jg/ml of cycloheximide 60 minprior to assay 1.2 ± 0.1 Not done

Preincubation with 5 ,4g/ml of Con A and0.05 M a-MeMan and 100 mg/ml ofcycloheximide 60 min prior to assay 0.6 ± 0.03 Not done

* Duplicate cultures treated as indicated above prior to assay were assayed for surface (intact cells) and total (disrupted cells) sialyltrans-ferase activity (see Materials and Methods for details). Inhibitors, when used, were present throughout the course of the experiment.

t Incorporation into fetuin acceptor was calculated by subtracting values obtained in its absence from those obtained in its presence. Varia-tion is expressed as average deviation of the mean.

of preincubation with Con A, utilizing either endogenous or

exogenous acceptors, reveals elevated levels of enzymic ac-

tivity after 1 hr of contact with mitogen (Fig. 3). The datastrongly indicate that the activation process is essentiallycompleted by the first hour. Support for this is seen by thelack of further increases in precursor incorporation at laterperiods of culture (Fig. 3) and by the linearity of the kineticsof incorporation of label into acid-insoluble cell acceptorsover a 3-hr period after 1 hr of Con A pretreatment (Fig. 1).Indeed, if the activation process were proceeding during theassay period, a nonlinear acceleration of isotope incorpora-tion might be expected.

Several trivial explanations for the observed stimulation ofsialyltransferase may be eliminated. The Con A effect can-

not be due to an increased rate of transport of substrates or

their breakdown products, since (a) no effect of Con A couldbe demonstrated on incorporation of radioactivity into acid-soluble cell components beyond that observed in controlcells (Fig. 1); (b) no detectable degradation (as judged bypaper chromatography) of CMP-['4C]NANA, a cell-imper-meable compound (16, 17), could be detected under theconditions of assay either with intact or sonicated cell sus-

pensions. Furthermore, free NANA (1 mM), the most proba-ble product of substrate degradation, did not inhibit uptakeof label into cell acceptors (unpublished results). In addition,possible Con-A-induced cell damage with release of intracel-lular enzyme cannot be the basis for the increased enzymicactivity, since measurements of cell viability at the end ofthe assay showed that Con-A-treated thymocytes were essen-

tially as viable (88% viable) as control cells (90% viable).Furthermore, sialyltransferase-mediated incorporation of ra-

dioactivity actually decreases as a function of culture time,as does cell viability (Fig. 3). Finally, the lack of effect ofCon A on sialyltransferase of disrupted cell preparations (seebelow) strongly indicates that maintenance of cell integrityis required.

As pointed out by Hakomori and coworkers (24), an in-crease in a surface sialidase could result in an apparent in-crease in sialyltransferase activity by increasing the avail-ability of acceptors for the transferase. In order to assess this

possibility, [14C]NANA labeled fetuin was isolated from cell-free supernatents by Sephadex G-25 chromatography. Morethan 90% of the acid-insoluble radioactivity in these super-natants is associated with the fetuin-containing fraction.When this fraction was incubated for 3 hr with intact or son-icated thymocytes under the standard conditions of assay, nosignificant release of radioactive material into the acid-solu-ble fraction was observed. On the other hand, V. choleraeneuraminidase caused release of 95% of the original fetuin-associated radioactivity (unpublished results).Con A apparently does not affect enzymic activity by

binding directly with the enzyme, since the mitogen had nomeasurable effect on the sialyltransferase activity of cell son-icates, even at concentrations which clearly caused aggluti-nation (50,g/ml).The localization of enzyme and its acceptors at the cell

surface is a central assumption of this paper. While there is

Table 2. Release of cell-associated radioactivity by variousenzymes from Con-A-treated thymocytes prelabeled with

CMP_[14CINANA*

Total radioactivityreleased into cell-free supernatant

Enzyme Concentration cpm %t

None 130 23Vibrio cholerae 20 units/ml 277 49neuraminidase 100 units/ml 373 66

Crystalline 4 p g/ml 190 33trypsin 40,jg/ml 246 43

Insoluble trypsin 3 units/ml 149 266 units/ml 314 55

Pronase 4 mg/ml 209 3740,ug/ml 263 46

* See text for details.t The total radioactivity was 569 cpm and was determined bycounting aliquots of the original suspension.

Proc. Nat. Acad. Sci. USA 73 (1976)

Dow

nloa

ded

by g

uest

on

Nov

embe

r 30

, 202

0

Proc. Nat. Acad. Sci. USA 73 (1976) 841

biochemical (16-18) and autoradiographic (23) evidence inother mammalian cells to support this view, further evi-dence was sought for thymocytes.When prelabeled cells are treated briefly with various hy-

drolytic enzymes under conditions in which the cells remainviable, release of considerable cell-associated radioactivity isobserved (Table 2). Neuraminidase released at least 50-70%of the total radioactivity into the cell-free supernatant,values which agree well with the known percentage of neu-raminidase-susceptible NANA residues found in murine thy-mocytes (25). Proteases, including trypsin covalently at-tached to Sepharose beads, also released a considerable frac-tion of the total radioactivity, suggesting that a significantfraction of the incorporated [14C]NANA is associated withsurface membrane glycoproteins.The results obtained are consistent with a Con-A-induced

process, not dependent upon de now protein synthesis,which involves the exposure or increased expression of newenzymic activity at the cell perimeter.

At least two mechanisms for the mediation by Con A ofthe increased expression of enzyme at the cell surface arecompatible with the present data. One would envision thebinding of Con A to a surface carbohydrate inhibitor of si-alyltransferase while the other postulates a Con-A-stimulat-ed transfer of pre-existing intracellular membrane present,e.g., in the Golgi apparatus, into the cell surface membrane.The latter possibility is at least consistent with a number ofexperimental facts. First, it is known that mitogens cause aconsiderable increase in phagocytotic and endocytotic up-take (26) which would be expected to lead to loss of cell sur-face membrane. Since mitogens lead to an ultimate increasein cell volume and therefore membrane surface area, con-siderable amounts of new membrane might be expected tobe incorporated into the existing plasma membrane. Second,shortly after addition of mitogens, there is increased expres-sion or turnover of a large variety of different membraneconstituents, including alterations of surface charge (27), in-creased transport rates of metabolites (28, 29), and increasesin NA',K+-dependent ATPase activity and ouabain-bindingsites (12), as well as of new Con-A-binding sites (30). In thelast case, existence of "preformed membrane glycan" wasproposed to explain the results. All of the above changes inmembrane properties occur within the first few hours afterexposure to mitogen and most of these proceed indepen-dently of de novo protein biosynthesis.The ability of Con A to increase the expression of glycosy-

lated components at the cell surfaces may be a general phe-nomenon. In another mammalian cell system, the hamsterfibroblast, low concentrations of Con A make certain gly-colipids more accessible to attack at the cell surface by ga-lactose oxidase (31).

Acknowledgment is made of the valuable technical assistance of

Ms. Susan Haag.

1. Cook, G. M. W. & Stoddart, R. W. (1973) Surface Carbohy-drates of the Eukaryotic Cell (Academic Press, Inc., NewYork).

2. Chen, H. W., Heiniger, H. J. & Kandutsch, A. A. (1975) Proc.Nat. Acad. Sci. USA 72, 1950-1954.

3. Fisher, D. B. & Mueller, G. C. (1969) Biochim. Biophys. Acta176,316-323.

4. Fisher, D. B. & Mueller, G. C. (1968) Proc. Nat. Acad. Sci.USA 60, 1396-1402.

5. Hayden, G. A., Crowley, G. M. & Jamieson, G. A. (1970) J.Biol. Chem. 245,5827-5832.

6. Schmidt-Ullrich, R., Wallach, D. F. H. & Ferber, E. (1974)Biochim. Biophys. Acta 356,288-299.

7. Bennett, G. & Leblond, C. P. (1970) J. Cell Biol. 46, 409-416.8. Zagury, D., Uhr, J. W., Jamieson, J. D. & Palade, G. E. (1970)

J. Cell Biol. 46,52-63.9. Hirano, H., Parkhouse, B., Nicolson, G. L., Lennox, E. S. &

Singer, S. J. (1972) Proc. Nat. Acad. Sct. USA 69,2945-2949.10. Krug, U., Krug, F. & Cuatrecasas, P. (1972) Proc. Nat. Acad.

Sci. USA 69,2604-2608.11. Quastel, M. R. & Kaplan, J. G. (1970) Exp. Cell Res. 63, 230-

233.12. Wright, P., Quastel, M. R. & Kaplan, J. G. (1972) in Proc. 7th

Leukocyte Culture Conference, ed. Daguillard, F. (AcademicPress, Inc., New York), pp. 87-101.

13. Mackintosh, P., Hardy, D. A. & Aviet, T. (1971) Lancet i,1019.

14. Yamamoto, K., Natsu-ume, S. & Takahashi, M. (1975) Jpn. J.Exp. Med. 45, 49-51.

15. Gahmberg, C. G. & Hakomori, S. (1973) J. Biol. Chem. 248,4311-4317.

16. Datta, P. (1974) Biochemistry 13,3987-3991.17. Patt, L. M. & Grimes, W. J. (1974) J. Biol. Chem. 249, 4157-

4165.18. McLean, R. J. & Bosmann, H. B. (1975) Proc. Nat. Acad. Sci.

USA 72, 310-313.19. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.

(1951) J. Biol. Chem. 193,265-275.20. Cunningham, D. D. & Pardee, A. B. (1969) Proc. Nat. Acad.

Sci. USA 64,1049-1056.21. Kim, Y. S., Perdomo, J. & Norberg, J. (1971) J. Biol. Chem.

246,5466-5476.22. Agrawal, B. B. L. & Goldstein, I. J. (1972) in Methods in En-

zymology, ed. Ginsburg, V. (Academic Press, Inc., New York),vol. 23, pp. 313-318.

23. Porter, C. W. & Bernacki, R. J. (1975) Nature 256,648-650.24. Yogeeswaran, G. & Hakomori, S. (1975) Biochemistry 14,

2151-2156.25. Despont, J. P., Abel, C. A. & Grey, H. M. (1975) Cell. Immu-

nol. 17, 487-494.26. Hirschhorn, R., Brittinger, G., Weissman, G. & Hirschhorn, K.

(1968) J. Cell Biol. 37, 412-423.27. Currie, G. A. (1967) Nature 216,694-695.28. Peters, J. H. & Hausen, P. (1971) Eur. J. Biochem. 19, 509-

513.29. Van Den Berg, K. J. & Betel, I. (1974) Exp. Cell Res. 84,

412-418.30. Karsenti, E. & Avrameas, S. (1973) FEBS Lett. 32, 238-242.31. Gahmberg, G. C. & Hakomori, S. (1975) J. Biol. Chem. 250,

2447-2451.

Cell Biology: Painter and White

Dow

nloa

ded

by g

uest

on

Nov

embe

r 30

, 202

0