- Email: [email protected]
Nonimmune lymphocyte-macrophage interaction
CELLULAR
IMMUNOLOGY
92, 211-289 (1985)
Nonimmune
Lymphocyte-Macrophage
II. Evidence That the Interaction
Interaction
Involves Sulfated Polysaccharide
Recognition
ANITA S.-F. CHONGAND CHRISTOPHERR. PARISH Department of Microbiology, John Curtin School of Medical Research, Australian National University, Canberra, ACT 2601, Australia Received November 23, 1984; accepted December 31, 1984 This paper describesattempts to determine the molecular basis of the nonimmune interaction between lymphocytes and macrophages. Initial studies revealed that the interaction could be inhibited by simple sugars, six out of the thirty-five tested being inhibitory. Furthermore, the majority of the inhibitory sugars were charged and subsequent studies revealed that some sulfated polysaccharides, notably kappa-carrageenan, were potent inhibitors of the interaction. Further experiments revealed that the lymphocyte-macrophage interaction was indeed mediated nan-specific receptorson lymphocytes The results supporting such a conclusion by kmwere as follows: (a) When the interacting cells were preincubated with kappa-mrmgeenan, it was found that kappa-carrageenan exerted its inhibitory effect at the lymphocyte rather than the macrophage level. (b) separation of splenocytes into kappa-carrageenan-binding and -nonbinding subpopulations resulted in a corresponding enrichment and depletion of lymphocytes that reacted with macrophages. (c) Lymphocytes were found to express kappa-aumgeenanreactive molecules, these molecules being detected on the surface of lymphocytes by resetting and in detergent lysates as hemagglutinins. Furthermore, the polyanion specificity of these kappaeamipmnwfic receptors/hemagglutininsclosely resembledthe specScity of inhibition of the lymphocyte-macrophage interaction. (d) Pronase-msistant meterial in macrophage, but not lymphocyte lysates, effectively inhibited both the lymphocyte-macrophage interaction and the recognition of kappaveenan by lymphocytes, suggestingthat a kappa-carrageenan-like structure is expressed by macrophages. 0 1985 Academic PIGS, IIIC.
INTRODUCTION Despite voluminous experimental data and numerous models of celI adhesion (l-5), cell biologists as well as cellular immunologists are still waiting for a general theory of cell behavior to emerge that will arrange all these facts in a logical sequence and elucidate the molecular mechanisms that govern cell-cell interactions. The nonimmune interaction between lymphocytes and macrophages represents an interesting model of intercellular adhesion that could be used to analyze the molecular basis of cell adhesion. In particular, the description in the preceding paper of an automated calorimetric assay for measuring the interaction makes the system very amenable to study. Thus, this paper describes our initial attempts to determine the molecular basis of this interaction and presents evidence that the system is dependent upon sulfated polysaccharide recognition. 217 000%8749185$3.00 Copyright 0 1985 by Academic PIUS, Inc. All righk of repmdudon in any form werwd.
278
CHONG AND PARISH
MATERIALS
AND METHODS
Animals. All mice and sheep used were bred at the John Curtin School of Medical Research. Mice were used as donors of lymphocytes from 5 to 12 weeks of age. Chemicals. The monosaccharides and oligosaccharides used in the inhibition studies and their suppliers are listed in Table 1. Hyaluronic acid (Grade III-S from human umbilical cord), chondroitin-4-sulfate (chondroitin sulfate type A from whale cartilage), chondroitin-6-sulfate (chondroitin sulfate type C from shark cartilage), dermatan sulfate (chondroitin sulfate type B from porcine skin), keratan sulfate (from bovine cornea), fucoidan (from Fucus vesiculosus),pentosan polysulfate, kappa-carrageenan (type III from Eucheuma cottonii), lambda-carrageenan (type IV from Gigartina aciculaire and Gigartina pistillata), and iota-cairageenan (type V from Eucheuma spinosa) were purchased from Sigma Chemical Company (St. Louis, MO.). Heparin (bovine mucous) was supplied by Evans Medical Ltd. (Liverpool, U.K.) and Arteparon (Luitpold Werk, Munich, W. Germany) was a generous gift of Dr. P. Ghosh, Royal North Shore Hospital, St. Leonards, Sydney, Australia. Dextran sulfate-2.3 (2.3 sulfates/monosaccharide, MW 500,000) was purchased from Pharmacia Fine Chemicals (Uppsala, Sweden). All monosaccharides and oligosaccharides were stored at 80 mg/ml in phosphate-buffered saline (PBS), pH 7.0, and polyanions at 20 mg/ml in PBS or 0.15 M NaCl with the exception of TABLE 1 Monosaccharides and Oligosaccharides Used for Sugar Inhibition Studies Monosaccharides ~Fucose L-Fucose L-Rhamnose DRibosed l+Xyloseb N-Acetyl-neuraminic acid Glucose-6-phosphate Galactose-6-phosphate Mannosed-phosphate Glucose-6-sulfate Methyl-cY-r@ucopyranoside’ Methyl-c+Dgalactopyranoside’ Methyl$%r+galactopyranoside Methyl-a-D-mannopyranoside
L&lucose” N-Acetyk-glucosamine D-Glucosamine’ DGlucuronic acid’ DGalactose’ N-Acetyl-D-galactosamine D-Galactosamine D-Galacturonic acid’ DMannose t-Mannose N-Acetyl-Dmannosamine DMannosamine DMannitol Oligosaccharides Melihiose’ Maltose’ CellobioseC Sucrose
LaCt0S.C
Thiodigalactoside Stachyose Raffinose’
Note. Except as noted, sugars were supplied by Sigma Chemical Co., St. Louis, MO. a Ajax Chemicals Ltd., Sydney, Australia. b BDH, Poole, England. c Calbiochem, San Diego, Calif. d Fluka, London, England.
LYMPHOCYTE-MACROPHAGE
INTERACTION
279
hyaluronic acid (10 mg/ml) and the carrageenans (2 rng/ml). All sugars and polyanion solutions were stored frozen at -20°C. The pH of the acidic sugars was adjusted to 7.0 by the addition of 5 N NaOH. Preparation of cell suspensions.Thioglycolate-induced macrophageswere prepared as previously described (6). Lymphocytes for the lymphocyte-macrophage binding assay were prepared in Eagle’s minimum essential medium (F15; GIBCO, Grand Island, N.Y.) in 1% fetal calf serum (FCS), whereas lymphocytes for extraction of lectin were prepared in PBS containing 0.1% (w/v) bovine serum albumin (BSA; fraction V, Armour Pharmaceuticals, Eastbourne, England) as previously described (7). All lymphocytes were centrifuged on Isopaque-Ficoll to remove dead cells and erythrocytes before use (8). Blood was collected from sheep in Alsever’s solution and the red cells (SRBC) were washed four times in 20 vol of 0.15 M NaCl by centrifugation at 600g for 5 min at 20°C just before use. Preparation of lymphocyte and macrophage lysates. The preparation of lymphocyte and macrophage lysates has been previously described (6). Briefly, lymphocytes and macrophages were suspended in cold PBS with 5 X 10e4M phenyl methyl sulfonyl fluoride (PMSF) and 25 mM sodium cholate. After incubating the cells for 30 min on ice, the cell nuclei and remaining cellular debris were pelleted by centrifugation and the supernatant lysate was used for hemagglutination assays and/or pronase treatment. Preparation of pronase-treated lysates. Macrophage and thymocyte lysates (10’ cells/ml equivalents) were prepared by the procedure described above. Lysates were dialyzed for 20 hr against 0.1 mM sodium cholate/PBS and the insoluble material was pelleted by centrifugation at 27,000g for 20 min at 4°C. Supematants were collected and treated with pronase (Calbiochem-Behring, La Jolla, Calif.) at 1 mg/ml for 24 hr at 37°C. A control solution of PBS/O.1 mM sodium cholate was similarly treated. The lysates and control were then incubated in a boiling water bath for 20 min to inactivate the pronase. The resulting solutions were cooled and tested for inhibitory activity. Assay for quantljjing lymphocyte-macrophage interaction. Binding of lymphocytes to macrophages was quantified with an automated, calorimetric, rose bengal assay as described in detail in the accompanying paper (9). Inhibition of lymphocyte-macrophage interaction with sugars and pronase-treated cell lysates. In the sugar inhibition experiments 10 ~1 of lo-fold concentrated sugar in PBS was added to each well of preformed macrophage monolayers containing 100 ~1 of F15/1% FCS/Hepes-buffered medium. The sugars were left to incubate with the macrophages for 30 min at 37°C before addition of 3 X lo5 lymphocytes/ well in 100 ~1 of medium and a further 10 ~1 of lO-fold concentrated sugar. In the inhibition of the lymphocyte-macrophage interaction with pronase-treated lysates, lysates were serially diluted in PBS/O.1% BSA and coincubated with lymphocytes (3 X 106/ml) at room temperature for 30 min before addition to the macrophage monolayers. The binding assay was then carried out at 37°C as described previously (9). In some experiments, lymphocytes and macrophages were preincubated with kappa-carrageenan before mixing. Lymphocytes at 3 X lo6/ml and preformed macrophage monolayers ( 105/well) were preincubated with 50 &ml of kappa~nan in F15/1% FCS/Hepes-buffered medium for 30 min at 37’C. To remove excess sugars lymphocytes were washed twice by centrifugation in medium (F15/1% FCS),
280
CHONG AND PARISH
whereas for the macrophage monolayer-s, the sugar solutions were thrown off, and the monolayers washed thrice with 200 &well of F15/1% FCS added with a multichannel pipet. All cells were placed in FI 5/l% FCS/Hepes medium for use in the rose bengal binding assay. Separation of lymphocyte subpopulations by a panning procedure. Polysaccharide (kappa-carrageenan or hyaluronic acid; 0.5 mg in 10 ml of PBS) was added to g-cm-diameter plastic petri dishes (Disposable Products Pty. Ltd., Adelaide, S. Australia) and left overnight at room temperature. Polysaccharide solutions were then thrown off and the dishes were washed thrice by dunking in a PBS bath at room temperature. Ten milliliters of PBS/O.1% BSA containing 4 X 10’ lymphocytes was added to each dish and the cells were left to settle and adhere for 90 min at 4°C. After gently swirling the plates, nonadherent cells were collected from both the kappa-carrageenan- and hyaluronic acid-treated plates. The adherent cells were harvested by vigorously pipetting the cells off the plastic with a short Pasteur pipet. Only kappa-carrageenan-adherentlymphocytes were collected as the yield of adherent lymphocytes from hyaluronic acid-coated plates was too low to allow testing in the macrophage-binding assay. The cells were washed once in F15/1% FCS/Hepesbuffered medium and resuspended at the required concentration for assessmentof binding to macrophage monolayers. Rosette and rosette inhibition assays. Thymocytes were rosetted with kappacarrageenan-coupled SRBC as previously described by Parish et al. (6). The polysaccharide was coupled onto the SRBC by CrQ ; the stock solution of CrClr used was prepared as previously reported (10). To 0.9 ml of 0.15 M NaCl were added 50 ~1 of packed SRBC and 0.4 mg of kappa-carrageenan dissolved in 0.2 ml of 0.15 M NaCl. After mixing, 60 ~1 of 0.1% CrQ in 0. I5 M NaCl was added. The mixture was left to react at room temperature for exactly 5 min and the reaction was then stopped by the addition of 4.0 ml of PBS. The coupled SRBC were pelleted by centrifugation (4OOg,5 min, 20°C) and then washed in 5.0 ml of PBS. The final SRBC pellet was resuspended in 5.0 ml of PBS/O.1% BSA/O.1% NaN3 and stored at 4°C for up to 7 days after coupling. The rosette-inhibition assay consisted of preincubating 25 ~1 of thymocytes (4 X 106/ml) with 25 ~1 of either polyanion- or pronase-treated lysate for 30 min on ice before the addition of 25 ~1 of 1% kappa-carrageenan-coupled SRBC, all TABLE 2 Sugars that Inhibited the Lymphocyte-Macrophage Interaction
Inhibitory sugar=
n
Glucosamine Glucuronic acid Galactosamine Galacturonic acid Galactose6-phosphate Thiodigalactoside
4 4 5 3 3 4
Thymocytes boundb (70control) 57.9 + 59.2 f 56.3 k 47.3 + 72.3 + 66.2 f
8.3 7.1 6.5 6.5 7.1 4.6
PC
o All sugarswere tested at 8 mg/ml in 3-5 experiments (n). b Results are the means of 3-5 experiments + standard error. cP values represent probability levels by Student’s t test that the differences are statistically significant.
LYMPHOCYTE-MACROPHAGE
281
INTERACTION
reactants being in PBS/O.1% BSA. Cells were mixed and then pelleted by centrifugation and left on ice for 30 min before being gently resuspended. Methyl violet (50 ~1 of 0.05% (w/v) in PBS) was added to each well and each sample was then transferred to a hemocytometer chamber and the percentage of rosette-forming cells assessed. Hemagglutination and hemagglutination-inhibition assays. The hemagglutination (HA) and HA-inhibition assayswere performed as described previously (6). Briefly, in the case of the HA-inhibition assays,serial dilutions of lymphocyte lysates were made in PBS/O.1% BSA in the wells of microplates and a constant concentration of a particular polyanion- or pronase-treated lysate was added to each well. After incubation for 60 min on ice, kappa-carrageenan-coupled SRBC were added and HA-inhibition titers determined. RESULTS Inhibition of Lymphocyte-Macrophage Interaction by Sugars and Polyanions Carbohydrate recognition has been implicated in many forms of intercellular adhesion (reviewed in (1 l-l 3)). To determine whether carbohydrate recognition is involved in the lymphocyte-macrophage interaction, we tested the ability of thirtyfive simple sugars (Table 1) at a concentration of 8 mg/ml to inhibit the interaction. Only six of the sugars tested showed significant inhibitory activity (Table 2). The most striking feature of this analysis was that, with the exception of thiodigalactoside, only charged sugars were inhibitory. It should be noted, however, that some charged TABLE 3 Effect of Different Polyanions on BALB/c Lymphocyte-macrophage Interaction Concentration (pgml) that reduces lymphocyte-macrophage interaction by 30% Substance tested Phosphated Mannan (yeast) Carboxylated Hyaluronic acid Lipopolysaccharide Poly-L-glutamic acid Sulfated Pentosan polysulfate Lambda-catrageenan Fucoidan Iota-carrageenan Rappa-carrageenan Sulfated and carboxylated Chondroitin4wlfate Chondroitin-6-sulfate Arteparon Heparin
Thymus
Spleen
>I000
>looo
>looo >looo >looo
z 1000 >I000 11000
Z-100 >I00 1000 100 10
>I00 >I00 1000 100 1
>looo >looo >loocl 1000
>looo 1looo >lOOO loo0
282
CHONG AND PARISH
sugars were not inhibitory, namely N-acetyl-neuraminic acid, D-mannosamine, glucosed-phosphate, and mannose-6-phosphate. On the other hand, although thiodigalactoside is usually uncharged, it carries a sulfide group which can be oxidized to sulfate, thus providing a charged group under certain conditions. Sugar inhibition of another lymphocyte recognition system, namely autorosetting, showed similar charged monosaccharides to be inhibitory (6, 14). Further work in that system showed that the lymphocytes expressedcell surface lectins that recognized sulfated polysaccharides on erythrocytes (6). Based on these findings the inhibitory effects of a range of polyanions were examined (Table 3). Only two of the polysaccharides tested showed strong inhibition of the lymphocyte-macrophage interaction, namely iota-carrageenan and kappa-carrageenan,the latter polysaccharide being the most potent inhibitor. Fucoidan and heparin exhibited weak inhibition at high concentrations, i.e., 1 mg/ml. It should be noted, however, that two of the polysaccharides, pentosan polysufate and lambda-carrageenan, could not be used at concentrations >O.l mg/ml due to their inhibition of rose bengal staining of cells. Figure 1 presentsthe complete inhibition curves for the effective polyanion inhibitors, again highlighting the potency of kappa-carrageenan as an inhibitor of the BALB/c thymocyte-macrophage interaction. Similar inhibition results were obtained for both the thymocyte-macrophage and splenocyte-macrophage interactions (Table 3). In fact, subsequent experiments revealed no significant difference in the specificity of inhibition of the two lymphocyte populations. Furthermore, although BALB/c cell populations were used for all the experiments presented in Table 3, a similar 120-
20 -
07
I 100
1000 INWBITCU
CCNCENll?ATKW
1 1
I 10 (u~Im0
FIG. 1. Ability of different polyanions to inhibit BALB/c thymocyte-macrophage interaction. Inhibitors (0), fucoidan (A), and heparin (A). are kappaxamgeman (O), iota canageenan (M), lambdameenan Results are expressed as percentages of control thymocy-te binding. The vertical lines represent the standard errors of means of three determinations.
LYMPHOCYTE-MACROPHAGE
283
INTERACTION
hierarchy of inhibition was obtained when CBA/H lymphocytes and macrophages were tested (data not shown). In order to analyze the inhibition phenomenon further, preincubation experiments with kappa-carrageenan were performed. This involved preincubating each cell type separately with kappa-carrageenan, washing off excesspolysaccharide, and using the cells in the standard binding assay. It was found that lymphocytes preincubated with kappa-carrageenan could not bind to macrophages (7.1% of control), whereas macrophages preincubated with kappa-carrageenan very effectively bound lymphocytes, and in fact exhibited slightly augmented binding (116.2% of control) (Table 4). Thus, the results of the preincubation experiment indicate that kappa-carrageenan exerts its inhibitory effect at the lymphocyte rather than the macrophage level. Involvement of Lymphocyte Receptorsfor Kappa-Carrageenan in the LymphocyteMacrophage Interaction Recent studies in this laboratory have demonstrated that lymphocytes carry receptors for kappa-carrageenan (15). Thus, the kappa-carrageenan inhibition (Table 3) and preincubation (Table 4) experiments suggestthat these lymphocyte receptors may either mediate the lymphocyte-macrophage interaction or be closely associated with the recognition molecules involved in the interaction. If this were the case, since only approximately 60% of mouse spleen cells carry receptors for kappacarrageenan (15), separation of spleen cells into kappa-carrageenan-binding and -nonbinding subpopulations should result in a corresponding enrichment and depletion of lymphocytes that interact with macrophages. Mouse spleen cells were separatedinto kappa-carrageenan-binding and -nonbinding populations by “panning” cells on kappa-carrageenan-coatedpetri dishes. Hyaluronic acid-coated petri dishes were used as a control as it has been shown that lymphocytes lack receptors for hyaluronic acid (15). Using this approach, approximately 60 and 30% of the added lymphocytes did not bind to the hyaluronic acid- and kappacarrageenan-coated dishes, respectively, and 30% of added lymphocytes were subsequently eluted from the kappa-carrageenan-coated dishes. It was found that the kappa-carrageenan-nonbinding splenocytes exhibited a greatly reduced ability to bind to macrophages, whereas the kappa-carrageenan-binding cells (compared with TABLE 4 Effect of Preincubation of Thymocytes and Macrophages with Kappa-Carrageenan on Thymocyte-Macrophage Interaction Cell type
Kappa-carrageenan (50 ah-4
Macrophage Thymocyte Macrophage + thymocyte Macrophage + thymcqte
Preincubationb Preincubation Preincubation Continuously’
Thymocyte binding (% control) 116.2 + 5.4 7.1 + 0.5 16.9 f 1.7 2.9 +- 0.3
’ Results are the means of triplicate determinations f standard errors. * Cells were preincubated with kappa-carrageenan (50 &ml) at 37°C for 30 min, then washed free of polysacchatide before addition of appropriate cells. ’ Kappawnan present throughout experiment.
284
CHONG AND PARISH
the hyaluronic acid control) showed a slightly augmented ability to bind to macrophages (Fig. 2). Previous studies have shown that kappa-carrageenan receptors can be readily detected on the surface of lymphocytes by rosetting with kappa-carrageenan-coupled SRBC and, in fact, potent hemagglutinating activity for kappa-carrageenan-SRBC can be detected in detergent lysates of lymphocytes ( 15). If these kappa-carrageenanspecific receptors are actually involved in macrophage recognition, it would be anticipated that they would be inhibited by a similar array of polyanions as the lymphocyte-macrophage interaction. Table 5 presents the results of such a comparison. It was found that, with one exception, all phosphated, carboxylated, and sulfatedcarboxylated polyanions that failed to inhibit the lymphocyte-macrophage interaction also were noninhibitory in the rosetting and hemagglutination assays.The exception was arteparon that was inhibitory at the highest concentration (200 p/ml) tested in the HA assay. Similarly, with one exception, the four polyanions that inhibited the lymphocyte-macrophage system showed a similar hierarchy of inhibition in the rosetting and HA assays,namely kappa-carrageenan > iota-carrageenan > fucoidan 3 heparin (compare Tables 3 and 5). The exception was again in the HA assay where fucoidan was a stronger inhibitor than iota-carrageenan. Finally, another important similarity was that kappa-carrageenan was the most potent inhibitor in all three systems although the rosetting and HA assayswere inhibited, respectively, by 50 and 500 times lower concentrations of kappa-canageenanthan the lymphocyte80
r
I 0.625
I
I
I
1.25
2.5
5.0
CELLS/WELL
X 1ti5
FIG. 2. Ability of hyaluronic acid-depleted (m), kappacarrageenandepleted (0), and kappa-carrageenanenriched (0) splenocytes to bind to syngeneic macrophages. Results are expressed as percentages of splenocytes bound and vertical lines represent the standard errors of means of triplicate determinations.
LYMPHOCYTE-MACROPHAGE
285
INTERACTION
TABLE 5 Ability of Different Polyanions to Inhibit Kappa-Carrageenan-Specific Receptors Expressed by Thymocytes Concentration (&ml)
Substance tested Phosphated Mannan Carboxylated Hyaluronic acid Lipopolysaccharide Poly-L-glutamic acid Sulfated Pentosan polysulfate Lambda-canageenan Fucoidan Iota-carrageenan Kappa-carrageenan Sulfated and wboxylated Chondroitin4sulfate Chondroitin-6-sulfate Arteparon Heparin
Thymocyte resetting by 30%
that reduces Hemagglutination titer by 50% >200
>200 >200 2200 200
0.2 (P) 20
2 (P) 0.2 >200 >200 >200 20
>200 >200 2200
0.2 (P) 0.02 (P) 0.2 2 0.02 2200 >200 200
20
(P)
Note. p = partial inhibition at the highest concentration tested (200 &ml).
macrophage interaction. Figure 3 depicts the complete inhibition curves for kappacarrageenan, comparing the inhibitory activity of this polysaccharide with the other two carrageenans-iota and lambda. On the other hand, although pentosan polysulfate and lambda-carrageenan did not inhibit the interaction between macrophages and lymphocytes at the highest concentration tested ( 100 &ml, Table 3), both polysaccharides inhibited the rosetting and HA assays.This discrepancy could be due to the greater sensitivity to inhibition of the rosetting and HA assays(see above), combined with, in most cases, the partial nature of the inhibitions observed, e.g., note the partial inhibition by lambda-carrageenan in Fig. 3. Collectively these data suggestthat kappa-carrageenan-specific receptors on lymphocytes are involved in the lymphocyte-macrophage interaction. It could be argued, however, that these receptors do not mediate the interaction but are closely associatedwith the relevant recognition molecules, the binding of kappa-carrageenan merely sterically masking these recognition molecules. In order to overcome this criticism, attempts were made to demonstrate a kappa-carrageenan-like molecule on macrophages. This was achieved by exhaustively digesting cholate lysates of macrophages with pronase, boiling the lysates to inactivate the enzyme and testing the resulting material for its ability to inhibit the lymphocyte-macrophage interaction and the corresponding rosetting and HA assays. Controls consisted of thymocyte lysates or pronase alone treated in a similar manner. It was found that only the pronase-treated macrophage lysate was inhibitory, the thymocyte lysate and pronase
286
CHONG AND PARISH
O/ 20
O-0 5
1.25
0.32
0.06
CARRAGEENAN
100 CONCENTRATION
, 6.25
,
, 0.38
,
, 0.025
, 0.0015
&~/ml)
FIG. 3. Ability of different carrageenans to inhibit the rosetting by BALB/c thymocytes (A) and the hemagglutinating activity of detergent lysates of BALB/c thymocytes (B) for kappa-carrageenan-coupled SRBC. Carrageenan inhibitors are kappa (O), lambda (0), iota (m). The broken lines represent the percentage rosetting-forming cells and the HA titers of lymphocyte lysates in the absence of inhibitors. The thymocyte lysate for the HA assay was used at an initial dilution of 1:10.
control lacking inhibitory activity (Fig. 4). The macrophage lysate inhibited the thymocyte-macrophage interaction and the thymocyte rosetting of kappa-carrageenan-coupled SRBC by 30% at dilutions of 1:8 of the original material and reduced the HA titer of the thymocyte lysate by 50% at a dilution of 1:64. Similar specificities of inhibition were obtained in the spleen cell-macrophage interaction and in the HA of kappa-carrageenan-coupled SRBC by spleen cell lysates (data not shown). DISCUSSION This paper describes attempts to determine the molecular mechanisms involved in the nonimmune interaction between lymphocytes and macrophages. Since carbohydrate recognition has been implicated in many forms of intercellular adhesion (1 l-l 3, 16), we tested the ability of a range of simple sugars to inhibit this interaction. Initial studies revealed that only six of the thirty-five sugars tested exhibited significant inhibitory activity and five of the inhibitory sugarswere charged (Table 2). Subsequent studies demonstrated that some sulfated polysaccharidesnotably kappa-carrageenan-were potent inhibitors of the interaction (Table 3). This inhibitory activity could not be attributed to a nonspecific electrostatic effect of the negatively charged kappa-carrageenan since other similarly charged (e.g., hyaluronic acid) or more highly charged (e.g., chondroitin sulfates) polyanions were noninhibitory. Further evidence for the highly specific nature of the inhibition was obtained when the potency of kappa-carrageenan as an inhibitor was compared to
LYMPHOCYTE-MACROPHAGE
INTERACTION
287
FIG. 4. Ability of pronase-treated lysates to inhibit the thymocyte-macrophage interaction (A), resetting by BALB/c thymocytes (B), and HA activity of detergent lysates of BALB/c thymocytes (C) for kappacarmgeenan-coupled SRBC. The pronase-treated lysates were from macrophages (0) and thymocytes (m). Heat-inactivated pronase (0) was used as a control. Results of the thymocyte-macrophage interaction are expressed as mean percentages of the controls, and the vertical lines represent the standard errors of the means of triplicate determinations. The broken lines represent the percentage rosette-forming cells and the HA titers of lymphocyte lysates in the absence of inhibitors. The thymocyte lysate for the HA assay was used at an initial dilution of 1:lO.
the other two carrageenans, iota and lambda, which are less inhibitory. Kappacarrageenan is composed of carbohydrate chains of a-( 1,3)-linked units of carrabiose, forming a linear alternating sequence of Dgalactose4sulfates and 3,6-anhydro+ galactose units (17). Iota-carrageenan differs from kappa-carrageenan only in the extra sulfate group per disaccharide it carries, while lambda-carrageenan carries three sulfate groups per dissaccharide and exhibits a differing linkage of the carbohydrate residues. Further experiments revealed that the lymphocyte-macrophage interaction was indeed mediated by kappa-carrageenan-specific receptors on lymphocytes that recognized a kappa-carrageenan-like structure on the surface of macrophages. The experimental data favoring this conclusion were as follows: (i) Preincubation of the interacting cells with kappa-carrageenan indicated that kappa-carrageenan exerted its inhibitory effect on the lymphocyte rather than the macrophage. Furthermore, this experiment ruled out the possibiity that kappa-carrageenan was inhibitory due to its well known selective cytotoxicity for macrophages (18). (ii) Separation by panning of splenocytes into kappa-carrageenan-binding and -nonbinding subpopulations resulted in a corresponding enrichment and depleteion of lymphocytes that reacted with macrophages. (iii) Lymphocytes were found to express kappa-carrageenan-reactive molecules, these molecules being detected on the surface of lymphocytes by resetting and in detergent lysates of lymphocytes as potent hemagglutinins of kappa-carrageenan-coupled SRBC. More importantly, the polyanion specificity of these kappa-carrageenan-specific receptors/hemagglutinins closely resembled the specificity of inhibition of the lymphocyte-macrophage interaction (compare Tables
288
CHONG AND PARISH
3 and 5). (iv) Pronase-resistant material in macrophage but not lymphocyte lysates effectively inhibited both the lymphocyte-macrophage interaction and the recognition of kappa-caxrageenan by lymphocytes (Fig. 4). This experiment suggests that macrophagesexpressa kappa-carrageenan-like structure and overcomes the argument that inhibition observed with kappa-carrageenan is the result of steric hindrance caused by the binding of kappa-carrageenan to molecules closely associated with the relevant receptors. Previous work by Parish and Snowden (15) has demonstrated the presence of a diverse array of lymphocyte surface lectins specific for sulfated polysaccharides. They proposed that these lymphocyte receptors are involved in cell-cell communication and lymphocyte homing. The kappa-carrageenan receptor on lymphocytes could thus represent one of these lectins whose function is to recognize a complementary molecule on macrophages and thus mediate the lymphocyte-macrophage interaction. Earlier studies by Parish et al. (6) also identified a unique plasma protein termed autorosette inhibition factor (AIF) that had a high affinity for certain sulfated polysaccharides and was a potent inhibitor of autorosetting via blocking sulfated carbohydrate structures on erythrocytes. This finding implied that the AIF molecule could regulate cell-cell interactions involving sulfated polysaccharide recognition. However, when tested in the rose bengal assay,AIF did not inhibit the lymphocyte-macrophage interaction (A. S.-F. Chong, unpublished data), suggesting that AIF has a much more specific regulatory role. It must be emphasized at this point that kappa-carrageenan is a polysaccharide that is derived from the cell wall of the red-brown alga E. cottonii, and therefore a possible murine correlate of this molecule on the macrophage membrane could be a glycosaminoglycan. The most likely glycosaminoglycans to participate in cell-cell interactions are the heparan sulfates, principally due to their cell surface expression (19) and huge potential for structural diversity (20). However, until further chemical characterization of the macrophage-derived molecule is performed, other sulfated macromolecules on the cell surface, namely sulfated glycoproteins and glycolipids, cannot be excluded as possible candidates. In contrast to the proposition that a lymphocyte lectin specific for sulfated sugars mediates the lymphocyte-macrophage interaction, Agrwal and Thomas (30) have postulated that the interaction involves recognition of macrophage Ia antigens. However, it is interesting to note the report by Sant et al. (3 1) that sulfated polysaccharides are tightly associated with Ia antigens. Sulfated polysaccharides have been implicated as specific acceptors in other systems of intercellular recognition. These include sperm-egg fusion in brown algae (21), sea urchins (22), hamsters (23), and guinea pigs (24); embryogenesis in Volvox carterii (25), and sea urchins (26); and specific aggregation of teratocarcinoma stem cell lines (27). Finally, in the case of lymphocytes there are reports showing that sulfated polysaccharides can inhibit lymphocyte recirculation in vivo (28), the binding of lymphocytes to high endothelial venules in vitro (29) and lymphocyte autorosetting (6). Thus, evidence is mounting to suggestthat sulfated polysaccharide recognition may represent the molecular basis of intercellular adhesion in many systems. REFERENCES 1. Dolowy, K., In “Cell Adhesion and Motility” (A. S. G. Curtis and J. D. Pit@.,F!ds.),pp. 39-65. Cambridge Univ. Press,Cambridge, 1980.
LYMPHOCYTE-MACROPHAGE
INTERACTION
289
2. Steinberg, M. S., J. Exp. Zool. 173, 395, 1970.
3. Monroy, A., and Moscona, A. A., “Introductory Concepts in Developmental Biology,” Univ. Chicago Press,Chicago/London, 1979. 4. Hood, L., and Prahl, J., Adv. Immunol. 5, 291, 1973. 5. Shur, B. D., and Roth, S., Biochim. Biophys. Acta 415,473, 1975. 6. Parish, C. R., Rylatt, D. B., and Snowden, J. M., J. Cell Sci. 67, 145, 1984. 7. Parish, C. R., Kirov, S. M., Bowem, N., and Blanden, R. V., Eur. .I. Immunol. 4, 808, 1974. 8. Davidson, W. F., and Parish, C. R., J. Immunol. Methods 7, 291, 1975. 9. Chong, A. S.-F., and Parish, C. R., Cell. Immunol. 92, 265, 1985. 10. Parish, C. R., and McKenzie, I. F. C., J. Immunol. Methods 20, 173, 1978. 11. Monsigny, M., Kieda, C., and Roche, A.-C., Biol. Cell 47, 95, 1983. 12. Rauvala, H., Trends Biochem. Sci. 8, 323, 1983. 13. Roth, S., In “Cell Interactions and Development: Molecular Mechanisms” (K. M. Yamada, Ed.), pp. 77-97. Wiley, New York, 1983. 14. Sia, D. Y., and Parish, C. R., Immunogenetics 12, 587, 1981. 15. Parish, C. R., and Snowden, J. M., Cell. Immunol. 91, 1985. 16. Barondes, S. H., Ann. Rev. Biochem. 58,207, 1981. 17. Thomsom, A. W., and Fowler, E. F., Agents Actions 11, 3, 198 1. 18. Catanzaro, P. J., Schwartz, H. J., and Graham, R. C., Jr., Amer. J. Pathol. 64, 387, 1971. 19. Kraemer, P. M., Biochemistry 10, 1445, 197 1. 20. Dietrich, C. P., Nader, H. B., and Straus, A. H., B&hem. Biophys. Res. Commun. 111, 865, 1983. 21. Bolwell, G. P., Callow, J. A., and Evans, L. V., J. CeZZSci. 43, 209, 1980. 22. Glabe, C. G., Ycdnock, T., and Rosen, S. D., J. Cell Sci. 61, 475, 1983. 23. Ahuja, K. K., Exp. Cell Rex. 140, 353, 1982. 24. Huang, T. T. F., and Yanagimachi, R., Exp. Cell Res. 153, 363, 1984. 25. Wetzel, G. G., and Kettman, J. R., J. Immunol. 126, 723, 1981. 26. Akasaka, K., and Terayama, H., Exp. Cell Res. 150,226, 1984. 27. Grabel, L., Glabe, C., Singer, M., Martin, G., and Rosen, S., Biochem. Biophys. Res. Commun. 102, 1165, 1981. 28. Bradfield, J. W. B., and Born, G. V. R., Cell. Immunol. 14, 22, 1974. 29. Stoolman, L. M., and Rosen, S. D., J. Cell Biol. 96, 72, 1983. 30. Agnval, N., and Thomas, D. W., CeZZ.Immunol. 84, 352, 1984. 31. Sam, A. J., Schwartz, B. D., and Cullen, S. E., J. Exp. Med. 158, 1979, 1983.
IMMUNOLOGY
92, 211-289 (1985)
Nonimmune
Lymphocyte-Macrophage
II. Evidence That the Interaction
Interaction
Involves Sulfated Polysaccharide
Recognition
ANITA S.-F. CHONGAND CHRISTOPHERR. PARISH Department of Microbiology, John Curtin School of Medical Research, Australian National University, Canberra, ACT 2601, Australia Received November 23, 1984; accepted December 31, 1984 This paper describesattempts to determine the molecular basis of the nonimmune interaction between lymphocytes and macrophages. Initial studies revealed that the interaction could be inhibited by simple sugars, six out of the thirty-five tested being inhibitory. Furthermore, the majority of the inhibitory sugars were charged and subsequent studies revealed that some sulfated polysaccharides, notably kappa-carrageenan, were potent inhibitors of the interaction. Further experiments revealed that the lymphocyte-macrophage interaction was indeed mediated nan-specific receptorson lymphocytes The results supporting such a conclusion by kmwere as follows: (a) When the interacting cells were preincubated with kappa-mrmgeenan, it was found that kappa-carrageenan exerted its inhibitory effect at the lymphocyte rather than the macrophage level. (b) separation of splenocytes into kappa-carrageenan-binding and -nonbinding subpopulations resulted in a corresponding enrichment and depletion of lymphocytes that reacted with macrophages. (c) Lymphocytes were found to express kappa-aumgeenanreactive molecules, these molecules being detected on the surface of lymphocytes by resetting and in detergent lysates as hemagglutinins. Furthermore, the polyanion specificity of these kappaeamipmnwfic receptors/hemagglutininsclosely resembledthe specScity of inhibition of the lymphocyte-macrophage interaction. (d) Pronase-msistant meterial in macrophage, but not lymphocyte lysates, effectively inhibited both the lymphocyte-macrophage interaction and the recognition of kappaveenan by lymphocytes, suggestingthat a kappa-carrageenan-like structure is expressed by macrophages. 0 1985 Academic PIGS, IIIC.
INTRODUCTION Despite voluminous experimental data and numerous models of celI adhesion (l-5), cell biologists as well as cellular immunologists are still waiting for a general theory of cell behavior to emerge that will arrange all these facts in a logical sequence and elucidate the molecular mechanisms that govern cell-cell interactions. The nonimmune interaction between lymphocytes and macrophages represents an interesting model of intercellular adhesion that could be used to analyze the molecular basis of cell adhesion. In particular, the description in the preceding paper of an automated calorimetric assay for measuring the interaction makes the system very amenable to study. Thus, this paper describes our initial attempts to determine the molecular basis of this interaction and presents evidence that the system is dependent upon sulfated polysaccharide recognition. 217 000%8749185$3.00 Copyright 0 1985 by Academic PIUS, Inc. All righk of repmdudon in any form werwd.
278
CHONG AND PARISH
MATERIALS
AND METHODS
Animals. All mice and sheep used were bred at the John Curtin School of Medical Research. Mice were used as donors of lymphocytes from 5 to 12 weeks of age. Chemicals. The monosaccharides and oligosaccharides used in the inhibition studies and their suppliers are listed in Table 1. Hyaluronic acid (Grade III-S from human umbilical cord), chondroitin-4-sulfate (chondroitin sulfate type A from whale cartilage), chondroitin-6-sulfate (chondroitin sulfate type C from shark cartilage), dermatan sulfate (chondroitin sulfate type B from porcine skin), keratan sulfate (from bovine cornea), fucoidan (from Fucus vesiculosus),pentosan polysulfate, kappa-carrageenan (type III from Eucheuma cottonii), lambda-carrageenan (type IV from Gigartina aciculaire and Gigartina pistillata), and iota-cairageenan (type V from Eucheuma spinosa) were purchased from Sigma Chemical Company (St. Louis, MO.). Heparin (bovine mucous) was supplied by Evans Medical Ltd. (Liverpool, U.K.) and Arteparon (Luitpold Werk, Munich, W. Germany) was a generous gift of Dr. P. Ghosh, Royal North Shore Hospital, St. Leonards, Sydney, Australia. Dextran sulfate-2.3 (2.3 sulfates/monosaccharide, MW 500,000) was purchased from Pharmacia Fine Chemicals (Uppsala, Sweden). All monosaccharides and oligosaccharides were stored at 80 mg/ml in phosphate-buffered saline (PBS), pH 7.0, and polyanions at 20 mg/ml in PBS or 0.15 M NaCl with the exception of TABLE 1 Monosaccharides and Oligosaccharides Used for Sugar Inhibition Studies Monosaccharides ~Fucose L-Fucose L-Rhamnose DRibosed l+Xyloseb N-Acetyl-neuraminic acid Glucose-6-phosphate Galactose-6-phosphate Mannosed-phosphate Glucose-6-sulfate Methyl-cY-r@ucopyranoside’ Methyl-c+Dgalactopyranoside’ Methyl$%r+galactopyranoside Methyl-a-D-mannopyranoside
L&lucose” N-Acetyk-glucosamine D-Glucosamine’ DGlucuronic acid’ DGalactose’ N-Acetyl-D-galactosamine D-Galactosamine D-Galacturonic acid’ DMannose t-Mannose N-Acetyl-Dmannosamine DMannosamine DMannitol Oligosaccharides Melihiose’ Maltose’ CellobioseC Sucrose
LaCt0S.C
Thiodigalactoside Stachyose Raffinose’
Note. Except as noted, sugars were supplied by Sigma Chemical Co., St. Louis, MO. a Ajax Chemicals Ltd., Sydney, Australia. b BDH, Poole, England. c Calbiochem, San Diego, Calif. d Fluka, London, England.
LYMPHOCYTE-MACROPHAGE
INTERACTION
279
hyaluronic acid (10 mg/ml) and the carrageenans (2 rng/ml). All sugars and polyanion solutions were stored frozen at -20°C. The pH of the acidic sugars was adjusted to 7.0 by the addition of 5 N NaOH. Preparation of cell suspensions.Thioglycolate-induced macrophageswere prepared as previously described (6). Lymphocytes for the lymphocyte-macrophage binding assay were prepared in Eagle’s minimum essential medium (F15; GIBCO, Grand Island, N.Y.) in 1% fetal calf serum (FCS), whereas lymphocytes for extraction of lectin were prepared in PBS containing 0.1% (w/v) bovine serum albumin (BSA; fraction V, Armour Pharmaceuticals, Eastbourne, England) as previously described (7). All lymphocytes were centrifuged on Isopaque-Ficoll to remove dead cells and erythrocytes before use (8). Blood was collected from sheep in Alsever’s solution and the red cells (SRBC) were washed four times in 20 vol of 0.15 M NaCl by centrifugation at 600g for 5 min at 20°C just before use. Preparation of lymphocyte and macrophage lysates. The preparation of lymphocyte and macrophage lysates has been previously described (6). Briefly, lymphocytes and macrophages were suspended in cold PBS with 5 X 10e4M phenyl methyl sulfonyl fluoride (PMSF) and 25 mM sodium cholate. After incubating the cells for 30 min on ice, the cell nuclei and remaining cellular debris were pelleted by centrifugation and the supernatant lysate was used for hemagglutination assays and/or pronase treatment. Preparation of pronase-treated lysates. Macrophage and thymocyte lysates (10’ cells/ml equivalents) were prepared by the procedure described above. Lysates were dialyzed for 20 hr against 0.1 mM sodium cholate/PBS and the insoluble material was pelleted by centrifugation at 27,000g for 20 min at 4°C. Supematants were collected and treated with pronase (Calbiochem-Behring, La Jolla, Calif.) at 1 mg/ml for 24 hr at 37°C. A control solution of PBS/O.1 mM sodium cholate was similarly treated. The lysates and control were then incubated in a boiling water bath for 20 min to inactivate the pronase. The resulting solutions were cooled and tested for inhibitory activity. Assay for quantljjing lymphocyte-macrophage interaction. Binding of lymphocytes to macrophages was quantified with an automated, calorimetric, rose bengal assay as described in detail in the accompanying paper (9). Inhibition of lymphocyte-macrophage interaction with sugars and pronase-treated cell lysates. In the sugar inhibition experiments 10 ~1 of lo-fold concentrated sugar in PBS was added to each well of preformed macrophage monolayers containing 100 ~1 of F15/1% FCS/Hepes-buffered medium. The sugars were left to incubate with the macrophages for 30 min at 37°C before addition of 3 X lo5 lymphocytes/ well in 100 ~1 of medium and a further 10 ~1 of lO-fold concentrated sugar. In the inhibition of the lymphocyte-macrophage interaction with pronase-treated lysates, lysates were serially diluted in PBS/O.1% BSA and coincubated with lymphocytes (3 X 106/ml) at room temperature for 30 min before addition to the macrophage monolayers. The binding assay was then carried out at 37°C as described previously (9). In some experiments, lymphocytes and macrophages were preincubated with kappa-carrageenan before mixing. Lymphocytes at 3 X lo6/ml and preformed macrophage monolayers ( 105/well) were preincubated with 50 &ml of kappa~nan in F15/1% FCS/Hepes-buffered medium for 30 min at 37’C. To remove excess sugars lymphocytes were washed twice by centrifugation in medium (F15/1% FCS),
280
CHONG AND PARISH
whereas for the macrophage monolayer-s, the sugar solutions were thrown off, and the monolayers washed thrice with 200 &well of F15/1% FCS added with a multichannel pipet. All cells were placed in FI 5/l% FCS/Hepes medium for use in the rose bengal binding assay. Separation of lymphocyte subpopulations by a panning procedure. Polysaccharide (kappa-carrageenan or hyaluronic acid; 0.5 mg in 10 ml of PBS) was added to g-cm-diameter plastic petri dishes (Disposable Products Pty. Ltd., Adelaide, S. Australia) and left overnight at room temperature. Polysaccharide solutions were then thrown off and the dishes were washed thrice by dunking in a PBS bath at room temperature. Ten milliliters of PBS/O.1% BSA containing 4 X 10’ lymphocytes was added to each dish and the cells were left to settle and adhere for 90 min at 4°C. After gently swirling the plates, nonadherent cells were collected from both the kappa-carrageenan- and hyaluronic acid-treated plates. The adherent cells were harvested by vigorously pipetting the cells off the plastic with a short Pasteur pipet. Only kappa-carrageenan-adherentlymphocytes were collected as the yield of adherent lymphocytes from hyaluronic acid-coated plates was too low to allow testing in the macrophage-binding assay. The cells were washed once in F15/1% FCS/Hepesbuffered medium and resuspended at the required concentration for assessmentof binding to macrophage monolayers. Rosette and rosette inhibition assays. Thymocytes were rosetted with kappacarrageenan-coupled SRBC as previously described by Parish et al. (6). The polysaccharide was coupled onto the SRBC by CrQ ; the stock solution of CrClr used was prepared as previously reported (10). To 0.9 ml of 0.15 M NaCl were added 50 ~1 of packed SRBC and 0.4 mg of kappa-carrageenan dissolved in 0.2 ml of 0.15 M NaCl. After mixing, 60 ~1 of 0.1% CrQ in 0. I5 M NaCl was added. The mixture was left to react at room temperature for exactly 5 min and the reaction was then stopped by the addition of 4.0 ml of PBS. The coupled SRBC were pelleted by centrifugation (4OOg,5 min, 20°C) and then washed in 5.0 ml of PBS. The final SRBC pellet was resuspended in 5.0 ml of PBS/O.1% BSA/O.1% NaN3 and stored at 4°C for up to 7 days after coupling. The rosette-inhibition assay consisted of preincubating 25 ~1 of thymocytes (4 X 106/ml) with 25 ~1 of either polyanion- or pronase-treated lysate for 30 min on ice before the addition of 25 ~1 of 1% kappa-carrageenan-coupled SRBC, all TABLE 2 Sugars that Inhibited the Lymphocyte-Macrophage Interaction
Inhibitory sugar=
n
Glucosamine Glucuronic acid Galactosamine Galacturonic acid Galactose6-phosphate Thiodigalactoside
4 4 5 3 3 4
Thymocytes boundb (70control) 57.9 + 59.2 f 56.3 k 47.3 + 72.3 + 66.2 f
8.3 7.1 6.5 6.5 7.1 4.6
PC
o All sugarswere tested at 8 mg/ml in 3-5 experiments (n). b Results are the means of 3-5 experiments + standard error. cP values represent probability levels by Student’s t test that the differences are statistically significant.
LYMPHOCYTE-MACROPHAGE
281
INTERACTION
reactants being in PBS/O.1% BSA. Cells were mixed and then pelleted by centrifugation and left on ice for 30 min before being gently resuspended. Methyl violet (50 ~1 of 0.05% (w/v) in PBS) was added to each well and each sample was then transferred to a hemocytometer chamber and the percentage of rosette-forming cells assessed. Hemagglutination and hemagglutination-inhibition assays. The hemagglutination (HA) and HA-inhibition assayswere performed as described previously (6). Briefly, in the case of the HA-inhibition assays,serial dilutions of lymphocyte lysates were made in PBS/O.1% BSA in the wells of microplates and a constant concentration of a particular polyanion- or pronase-treated lysate was added to each well. After incubation for 60 min on ice, kappa-carrageenan-coupled SRBC were added and HA-inhibition titers determined. RESULTS Inhibition of Lymphocyte-Macrophage Interaction by Sugars and Polyanions Carbohydrate recognition has been implicated in many forms of intercellular adhesion (reviewed in (1 l-l 3)). To determine whether carbohydrate recognition is involved in the lymphocyte-macrophage interaction, we tested the ability of thirtyfive simple sugars (Table 1) at a concentration of 8 mg/ml to inhibit the interaction. Only six of the sugars tested showed significant inhibitory activity (Table 2). The most striking feature of this analysis was that, with the exception of thiodigalactoside, only charged sugars were inhibitory. It should be noted, however, that some charged TABLE 3 Effect of Different Polyanions on BALB/c Lymphocyte-macrophage Interaction Concentration (pgml) that reduces lymphocyte-macrophage interaction by 30% Substance tested Phosphated Mannan (yeast) Carboxylated Hyaluronic acid Lipopolysaccharide Poly-L-glutamic acid Sulfated Pentosan polysulfate Lambda-catrageenan Fucoidan Iota-carrageenan Rappa-carrageenan Sulfated and carboxylated Chondroitin4wlfate Chondroitin-6-sulfate Arteparon Heparin
Thymus
Spleen
>I000
>looo
>looo >looo >looo
z 1000 >I000 11000
Z-100 >I00 1000 100 10
>I00 >I00 1000 100 1
>looo >looo >loocl 1000
>looo 1looo >lOOO loo0
282
CHONG AND PARISH
sugars were not inhibitory, namely N-acetyl-neuraminic acid, D-mannosamine, glucosed-phosphate, and mannose-6-phosphate. On the other hand, although thiodigalactoside is usually uncharged, it carries a sulfide group which can be oxidized to sulfate, thus providing a charged group under certain conditions. Sugar inhibition of another lymphocyte recognition system, namely autorosetting, showed similar charged monosaccharides to be inhibitory (6, 14). Further work in that system showed that the lymphocytes expressedcell surface lectins that recognized sulfated polysaccharides on erythrocytes (6). Based on these findings the inhibitory effects of a range of polyanions were examined (Table 3). Only two of the polysaccharides tested showed strong inhibition of the lymphocyte-macrophage interaction, namely iota-carrageenan and kappa-carrageenan,the latter polysaccharide being the most potent inhibitor. Fucoidan and heparin exhibited weak inhibition at high concentrations, i.e., 1 mg/ml. It should be noted, however, that two of the polysaccharides, pentosan polysufate and lambda-carrageenan, could not be used at concentrations >O.l mg/ml due to their inhibition of rose bengal staining of cells. Figure 1 presentsthe complete inhibition curves for the effective polyanion inhibitors, again highlighting the potency of kappa-carrageenan as an inhibitor of the BALB/c thymocyte-macrophage interaction. Similar inhibition results were obtained for both the thymocyte-macrophage and splenocyte-macrophage interactions (Table 3). In fact, subsequent experiments revealed no significant difference in the specificity of inhibition of the two lymphocyte populations. Furthermore, although BALB/c cell populations were used for all the experiments presented in Table 3, a similar 120-
20 -
07
I 100
1000 INWBITCU
CCNCENll?ATKW
1 1
I 10 (u~Im0
FIG. 1. Ability of different polyanions to inhibit BALB/c thymocyte-macrophage interaction. Inhibitors (0), fucoidan (A), and heparin (A). are kappaxamgeman (O), iota canageenan (M), lambdameenan Results are expressed as percentages of control thymocy-te binding. The vertical lines represent the standard errors of means of three determinations.
LYMPHOCYTE-MACROPHAGE
283
INTERACTION
hierarchy of inhibition was obtained when CBA/H lymphocytes and macrophages were tested (data not shown). In order to analyze the inhibition phenomenon further, preincubation experiments with kappa-carrageenan were performed. This involved preincubating each cell type separately with kappa-carrageenan, washing off excesspolysaccharide, and using the cells in the standard binding assay. It was found that lymphocytes preincubated with kappa-carrageenan could not bind to macrophages (7.1% of control), whereas macrophages preincubated with kappa-carrageenan very effectively bound lymphocytes, and in fact exhibited slightly augmented binding (116.2% of control) (Table 4). Thus, the results of the preincubation experiment indicate that kappa-carrageenan exerts its inhibitory effect at the lymphocyte rather than the macrophage level. Involvement of Lymphocyte Receptorsfor Kappa-Carrageenan in the LymphocyteMacrophage Interaction Recent studies in this laboratory have demonstrated that lymphocytes carry receptors for kappa-carrageenan (15). Thus, the kappa-carrageenan inhibition (Table 3) and preincubation (Table 4) experiments suggestthat these lymphocyte receptors may either mediate the lymphocyte-macrophage interaction or be closely associated with the recognition molecules involved in the interaction. If this were the case, since only approximately 60% of mouse spleen cells carry receptors for kappacarrageenan (15), separation of spleen cells into kappa-carrageenan-binding and -nonbinding subpopulations should result in a corresponding enrichment and depletion of lymphocytes that interact with macrophages. Mouse spleen cells were separatedinto kappa-carrageenan-binding and -nonbinding populations by “panning” cells on kappa-carrageenan-coatedpetri dishes. Hyaluronic acid-coated petri dishes were used as a control as it has been shown that lymphocytes lack receptors for hyaluronic acid (15). Using this approach, approximately 60 and 30% of the added lymphocytes did not bind to the hyaluronic acid- and kappacarrageenan-coated dishes, respectively, and 30% of added lymphocytes were subsequently eluted from the kappa-carrageenan-coated dishes. It was found that the kappa-carrageenan-nonbinding splenocytes exhibited a greatly reduced ability to bind to macrophages, whereas the kappa-carrageenan-binding cells (compared with TABLE 4 Effect of Preincubation of Thymocytes and Macrophages with Kappa-Carrageenan on Thymocyte-Macrophage Interaction Cell type
Kappa-carrageenan (50 ah-4
Macrophage Thymocyte Macrophage + thymocyte Macrophage + thymcqte
Preincubationb Preincubation Preincubation Continuously’
Thymocyte binding (% control) 116.2 + 5.4 7.1 + 0.5 16.9 f 1.7 2.9 +- 0.3
’ Results are the means of triplicate determinations f standard errors. * Cells were preincubated with kappa-carrageenan (50 &ml) at 37°C for 30 min, then washed free of polysacchatide before addition of appropriate cells. ’ Kappawnan present throughout experiment.
284
CHONG AND PARISH
the hyaluronic acid control) showed a slightly augmented ability to bind to macrophages (Fig. 2). Previous studies have shown that kappa-carrageenan receptors can be readily detected on the surface of lymphocytes by rosetting with kappa-carrageenan-coupled SRBC and, in fact, potent hemagglutinating activity for kappa-carrageenan-SRBC can be detected in detergent lysates of lymphocytes ( 15). If these kappa-carrageenanspecific receptors are actually involved in macrophage recognition, it would be anticipated that they would be inhibited by a similar array of polyanions as the lymphocyte-macrophage interaction. Table 5 presents the results of such a comparison. It was found that, with one exception, all phosphated, carboxylated, and sulfatedcarboxylated polyanions that failed to inhibit the lymphocyte-macrophage interaction also were noninhibitory in the rosetting and hemagglutination assays.The exception was arteparon that was inhibitory at the highest concentration (200 p/ml) tested in the HA assay. Similarly, with one exception, the four polyanions that inhibited the lymphocyte-macrophage system showed a similar hierarchy of inhibition in the rosetting and HA assays,namely kappa-carrageenan > iota-carrageenan > fucoidan 3 heparin (compare Tables 3 and 5). The exception was again in the HA assay where fucoidan was a stronger inhibitor than iota-carrageenan. Finally, another important similarity was that kappa-carrageenan was the most potent inhibitor in all three systems although the rosetting and HA assayswere inhibited, respectively, by 50 and 500 times lower concentrations of kappa-canageenanthan the lymphocyte80
r
I 0.625
I
I
I
1.25
2.5
5.0
CELLS/WELL
X 1ti5
FIG. 2. Ability of hyaluronic acid-depleted (m), kappacarrageenandepleted (0), and kappa-carrageenanenriched (0) splenocytes to bind to syngeneic macrophages. Results are expressed as percentages of splenocytes bound and vertical lines represent the standard errors of means of triplicate determinations.
LYMPHOCYTE-MACROPHAGE
285
INTERACTION
TABLE 5 Ability of Different Polyanions to Inhibit Kappa-Carrageenan-Specific Receptors Expressed by Thymocytes Concentration (&ml)
Substance tested Phosphated Mannan Carboxylated Hyaluronic acid Lipopolysaccharide Poly-L-glutamic acid Sulfated Pentosan polysulfate Lambda-canageenan Fucoidan Iota-carrageenan Kappa-carrageenan Sulfated and wboxylated Chondroitin4sulfate Chondroitin-6-sulfate Arteparon Heparin
Thymocyte resetting by 30%
that reduces Hemagglutination titer by 50% >200
>200 >200 2200 200
0.2 (P) 20
2 (P) 0.2 >200 >200 >200 20
>200 >200 2200
0.2 (P) 0.02 (P) 0.2 2 0.02 2200 >200 200
20
(P)
Note. p = partial inhibition at the highest concentration tested (200 &ml).
macrophage interaction. Figure 3 depicts the complete inhibition curves for kappacarrageenan, comparing the inhibitory activity of this polysaccharide with the other two carrageenans-iota and lambda. On the other hand, although pentosan polysulfate and lambda-carrageenan did not inhibit the interaction between macrophages and lymphocytes at the highest concentration tested ( 100 &ml, Table 3), both polysaccharides inhibited the rosetting and HA assays.This discrepancy could be due to the greater sensitivity to inhibition of the rosetting and HA assays(see above), combined with, in most cases, the partial nature of the inhibitions observed, e.g., note the partial inhibition by lambda-carrageenan in Fig. 3. Collectively these data suggestthat kappa-carrageenan-specific receptors on lymphocytes are involved in the lymphocyte-macrophage interaction. It could be argued, however, that these receptors do not mediate the interaction but are closely associatedwith the relevant recognition molecules, the binding of kappa-carrageenan merely sterically masking these recognition molecules. In order to overcome this criticism, attempts were made to demonstrate a kappa-carrageenan-like molecule on macrophages. This was achieved by exhaustively digesting cholate lysates of macrophages with pronase, boiling the lysates to inactivate the enzyme and testing the resulting material for its ability to inhibit the lymphocyte-macrophage interaction and the corresponding rosetting and HA assays. Controls consisted of thymocyte lysates or pronase alone treated in a similar manner. It was found that only the pronase-treated macrophage lysate was inhibitory, the thymocyte lysate and pronase
286
CHONG AND PARISH
O/ 20
O-0 5
1.25
0.32
0.06
CARRAGEENAN
100 CONCENTRATION
, 6.25
,
, 0.38
,
, 0.025
, 0.0015
&~/ml)
FIG. 3. Ability of different carrageenans to inhibit the rosetting by BALB/c thymocytes (A) and the hemagglutinating activity of detergent lysates of BALB/c thymocytes (B) for kappa-carrageenan-coupled SRBC. Carrageenan inhibitors are kappa (O), lambda (0), iota (m). The broken lines represent the percentage rosetting-forming cells and the HA titers of lymphocyte lysates in the absence of inhibitors. The thymocyte lysate for the HA assay was used at an initial dilution of 1:10.
control lacking inhibitory activity (Fig. 4). The macrophage lysate inhibited the thymocyte-macrophage interaction and the thymocyte rosetting of kappa-carrageenan-coupled SRBC by 30% at dilutions of 1:8 of the original material and reduced the HA titer of the thymocyte lysate by 50% at a dilution of 1:64. Similar specificities of inhibition were obtained in the spleen cell-macrophage interaction and in the HA of kappa-carrageenan-coupled SRBC by spleen cell lysates (data not shown). DISCUSSION This paper describes attempts to determine the molecular mechanisms involved in the nonimmune interaction between lymphocytes and macrophages. Since carbohydrate recognition has been implicated in many forms of intercellular adhesion (1 l-l 3, 16), we tested the ability of a range of simple sugars to inhibit this interaction. Initial studies revealed that only six of the thirty-five sugars tested exhibited significant inhibitory activity and five of the inhibitory sugarswere charged (Table 2). Subsequent studies demonstrated that some sulfated polysaccharidesnotably kappa-carrageenan-were potent inhibitors of the interaction (Table 3). This inhibitory activity could not be attributed to a nonspecific electrostatic effect of the negatively charged kappa-carrageenan since other similarly charged (e.g., hyaluronic acid) or more highly charged (e.g., chondroitin sulfates) polyanions were noninhibitory. Further evidence for the highly specific nature of the inhibition was obtained when the potency of kappa-carrageenan as an inhibitor was compared to
LYMPHOCYTE-MACROPHAGE
INTERACTION
287
FIG. 4. Ability of pronase-treated lysates to inhibit the thymocyte-macrophage interaction (A), resetting by BALB/c thymocytes (B), and HA activity of detergent lysates of BALB/c thymocytes (C) for kappacarmgeenan-coupled SRBC. The pronase-treated lysates were from macrophages (0) and thymocytes (m). Heat-inactivated pronase (0) was used as a control. Results of the thymocyte-macrophage interaction are expressed as mean percentages of the controls, and the vertical lines represent the standard errors of the means of triplicate determinations. The broken lines represent the percentage rosette-forming cells and the HA titers of lymphocyte lysates in the absence of inhibitors. The thymocyte lysate for the HA assay was used at an initial dilution of 1:lO.
the other two carrageenans, iota and lambda, which are less inhibitory. Kappacarrageenan is composed of carbohydrate chains of a-( 1,3)-linked units of carrabiose, forming a linear alternating sequence of Dgalactose4sulfates and 3,6-anhydro+ galactose units (17). Iota-carrageenan differs from kappa-carrageenan only in the extra sulfate group per disaccharide it carries, while lambda-carrageenan carries three sulfate groups per dissaccharide and exhibits a differing linkage of the carbohydrate residues. Further experiments revealed that the lymphocyte-macrophage interaction was indeed mediated by kappa-carrageenan-specific receptors on lymphocytes that recognized a kappa-carrageenan-like structure on the surface of macrophages. The experimental data favoring this conclusion were as follows: (i) Preincubation of the interacting cells with kappa-carrageenan indicated that kappa-carrageenan exerted its inhibitory effect on the lymphocyte rather than the macrophage. Furthermore, this experiment ruled out the possibiity that kappa-carrageenan was inhibitory due to its well known selective cytotoxicity for macrophages (18). (ii) Separation by panning of splenocytes into kappa-carrageenan-binding and -nonbinding subpopulations resulted in a corresponding enrichment and depleteion of lymphocytes that reacted with macrophages. (iii) Lymphocytes were found to express kappa-carrageenan-reactive molecules, these molecules being detected on the surface of lymphocytes by resetting and in detergent lysates of lymphocytes as potent hemagglutinins of kappa-carrageenan-coupled SRBC. More importantly, the polyanion specificity of these kappa-carrageenan-specific receptors/hemagglutinins closely resembled the specificity of inhibition of the lymphocyte-macrophage interaction (compare Tables
288
CHONG AND PARISH
3 and 5). (iv) Pronase-resistant material in macrophage but not lymphocyte lysates effectively inhibited both the lymphocyte-macrophage interaction and the recognition of kappa-caxrageenan by lymphocytes (Fig. 4). This experiment suggests that macrophagesexpressa kappa-carrageenan-like structure and overcomes the argument that inhibition observed with kappa-carrageenan is the result of steric hindrance caused by the binding of kappa-carrageenan to molecules closely associated with the relevant receptors. Previous work by Parish and Snowden (15) has demonstrated the presence of a diverse array of lymphocyte surface lectins specific for sulfated polysaccharides. They proposed that these lymphocyte receptors are involved in cell-cell communication and lymphocyte homing. The kappa-carrageenan receptor on lymphocytes could thus represent one of these lectins whose function is to recognize a complementary molecule on macrophages and thus mediate the lymphocyte-macrophage interaction. Earlier studies by Parish et al. (6) also identified a unique plasma protein termed autorosette inhibition factor (AIF) that had a high affinity for certain sulfated polysaccharides and was a potent inhibitor of autorosetting via blocking sulfated carbohydrate structures on erythrocytes. This finding implied that the AIF molecule could regulate cell-cell interactions involving sulfated polysaccharide recognition. However, when tested in the rose bengal assay,AIF did not inhibit the lymphocyte-macrophage interaction (A. S.-F. Chong, unpublished data), suggesting that AIF has a much more specific regulatory role. It must be emphasized at this point that kappa-carrageenan is a polysaccharide that is derived from the cell wall of the red-brown alga E. cottonii, and therefore a possible murine correlate of this molecule on the macrophage membrane could be a glycosaminoglycan. The most likely glycosaminoglycans to participate in cell-cell interactions are the heparan sulfates, principally due to their cell surface expression (19) and huge potential for structural diversity (20). However, until further chemical characterization of the macrophage-derived molecule is performed, other sulfated macromolecules on the cell surface, namely sulfated glycoproteins and glycolipids, cannot be excluded as possible candidates. In contrast to the proposition that a lymphocyte lectin specific for sulfated sugars mediates the lymphocyte-macrophage interaction, Agrwal and Thomas (30) have postulated that the interaction involves recognition of macrophage Ia antigens. However, it is interesting to note the report by Sant et al. (3 1) that sulfated polysaccharides are tightly associated with Ia antigens. Sulfated polysaccharides have been implicated as specific acceptors in other systems of intercellular recognition. These include sperm-egg fusion in brown algae (21), sea urchins (22), hamsters (23), and guinea pigs (24); embryogenesis in Volvox carterii (25), and sea urchins (26); and specific aggregation of teratocarcinoma stem cell lines (27). Finally, in the case of lymphocytes there are reports showing that sulfated polysaccharides can inhibit lymphocyte recirculation in vivo (28), the binding of lymphocytes to high endothelial venules in vitro (29) and lymphocyte autorosetting (6). Thus, evidence is mounting to suggestthat sulfated polysaccharide recognition may represent the molecular basis of intercellular adhesion in many systems. REFERENCES 1. Dolowy, K., In “Cell Adhesion and Motility” (A. S. G. Curtis and J. D. Pit@.,F!ds.),pp. 39-65. Cambridge Univ. Press,Cambridge, 1980.
LYMPHOCYTE-MACROPHAGE
INTERACTION
289
2. Steinberg, M. S., J. Exp. Zool. 173, 395, 1970.
3. Monroy, A., and Moscona, A. A., “Introductory Concepts in Developmental Biology,” Univ. Chicago Press,Chicago/London, 1979. 4. Hood, L., and Prahl, J., Adv. Immunol. 5, 291, 1973. 5. Shur, B. D., and Roth, S., Biochim. Biophys. Acta 415,473, 1975. 6. Parish, C. R., Rylatt, D. B., and Snowden, J. M., J. Cell Sci. 67, 145, 1984. 7. Parish, C. R., Kirov, S. M., Bowem, N., and Blanden, R. V., Eur. .I. Immunol. 4, 808, 1974. 8. Davidson, W. F., and Parish, C. R., J. Immunol. Methods 7, 291, 1975. 9. Chong, A. S.-F., and Parish, C. R., Cell. Immunol. 92, 265, 1985. 10. Parish, C. R., and McKenzie, I. F. C., J. Immunol. Methods 20, 173, 1978. 11. Monsigny, M., Kieda, C., and Roche, A.-C., Biol. Cell 47, 95, 1983. 12. Rauvala, H., Trends Biochem. Sci. 8, 323, 1983. 13. Roth, S., In “Cell Interactions and Development: Molecular Mechanisms” (K. M. Yamada, Ed.), pp. 77-97. Wiley, New York, 1983. 14. Sia, D. Y., and Parish, C. R., Immunogenetics 12, 587, 1981. 15. Parish, C. R., and Snowden, J. M., Cell. Immunol. 91, 1985. 16. Barondes, S. H., Ann. Rev. Biochem. 58,207, 1981. 17. Thomsom, A. W., and Fowler, E. F., Agents Actions 11, 3, 198 1. 18. Catanzaro, P. J., Schwartz, H. J., and Graham, R. C., Jr., Amer. J. Pathol. 64, 387, 1971. 19. Kraemer, P. M., Biochemistry 10, 1445, 197 1. 20. Dietrich, C. P., Nader, H. B., and Straus, A. H., B&hem. Biophys. Res. Commun. 111, 865, 1983. 21. Bolwell, G. P., Callow, J. A., and Evans, L. V., J. CeZZSci. 43, 209, 1980. 22. Glabe, C. G., Ycdnock, T., and Rosen, S. D., J. Cell Sci. 61, 475, 1983. 23. Ahuja, K. K., Exp. Cell Rex. 140, 353, 1982. 24. Huang, T. T. F., and Yanagimachi, R., Exp. Cell Res. 153, 363, 1984. 25. Wetzel, G. G., and Kettman, J. R., J. Immunol. 126, 723, 1981. 26. Akasaka, K., and Terayama, H., Exp. Cell Res. 150,226, 1984. 27. Grabel, L., Glabe, C., Singer, M., Martin, G., and Rosen, S., Biochem. Biophys. Res. Commun. 102, 1165, 1981. 28. Bradfield, J. W. B., and Born, G. V. R., Cell. Immunol. 14, 22, 1974. 29. Stoolman, L. M., and Rosen, S. D., J. Cell Biol. 96, 72, 1983. 30. Agnval, N., and Thomas, D. W., CeZZ.Immunol. 84, 352, 1984. 31. Sam, A. J., Schwartz, B. D., and Cullen, S. E., J. Exp. Med. 158, 1979, 1983.