Lactosamine type asialooligosaccharide recognition in NK cytotoxicity

Immunology Letters, 12 (1986) 83-90

Elsevier lmlet 713

LACTOSAMINE TYPE ASIALOOLIGOSACCHARIDE RECOGNITION IN NK CYTOTOXICITY* M. P O S P f ~ I L l, J. KUBRYCHT 1, K. B E Z O U ~ K A 2, O. T/kBORSKY 2, M. NOV,~K 3 and J. K O C O U R E K 2 ~Department of Immunology, Institute of Microbiology, Czechoslovak Academy of Sciences, 142 20 Prague 4, 2Department of Biochemistry, Faculty of Science, Charles University, Albertov 2030, 128 40 Prague 2, and 3Institute of Virology, Slovak Academy of Sciences, 809 39 Bratislava, Czechoslovakia

(Received 27 May 1985) (Modified version received 10 October 1985) (Accepted 26 October 1985)

I. Summary

2. Introduction

Inhibition o f pig NK cell activity by asialooligosaccharides (aOS) isolated from human serum glycoproteins was investigated. Triantennary aOS (aOSIII) of ceruloplasmin was found to be the most potent inhibitor up to the concentration 0.1 ~tg/ml, which is in agreement with its highly specific binding to NK-activityenriched pig lymphocytes (with a m o r p h o l o g y similar to h u m a n large granular lymphocytes (LGL)). Only lectins with the specificity to Gal(131--4)GlcNAc or Gal(/31--3)GalNAc structures exhibited inhibition of NK cytotoxicity. F(ab)2 fragments of rabbit antibodies against pig spleen membrane lectin cross-reacting with the pig liver membrane lectin completely inhibited NK activity when preincubated with the effectors or present in the incubation mixture during the assay. These data suggest that lectin receptors on cells of pig NK-activity-enriched fraction specific for aOSIII and antigenically related to membrane lectins isolated from pig spleen and liver, are involved in the NK recognition of several xenogeneic targets.

Natural killer (NK) cells play an important role in the immune response to cell-surface displayed antigens, such as those found on virus infected cells, t u m o r cells, etc. Many authors studied the inhibitory activity of mono- and disaccharides in NK cytotoxicity [1-3] but only few data pertaining to the question of the target cell recognition pattern were published. Till now the question of the nature of "lectin-like" structure involved in the cellular mechanism of NK cytotoxicity remains unclear [4, 5]. The increased level of NK activity when desialylated targets are used [ 6 - 8 ] , dependence on EDTA treatment [9] and hierarchy of inhibition by some sugars [1], led us to the assumption that binding sites of lectin type with specificity to aOS isolated from blood serum glycoproteins [10, 11] might be involved in the NK recognition. In our present work we confirmed and extended our preliminary findings by binding and inhibition studies.

3. Materials and Methods Key words: NK recognition - asialooligosaccharides - lec-

tins Note: All of the sugar residues have a D-configuration except

fucose which has an L-configuration. * Paper No. LXI of the series "Studies on lectins".

3.1. M a t e r i a l s Ficoll-Verografin was prepared from Ficoll 400 (Pharmacia Fine Chemicals, Uppsala, Sweden) and 76o-/0 Verografin (IA~iva, Prague, Czechoslovakia). aOS of the N-acetyllactosamine

0165-2478 / 86 / $ 3.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

83

type were prepared from human serum glycoproteins, desialylated, labeled with NaB[3H]H3 and their homogeneity evaluated as described [12]. Lectins from peanut (PNA), wheat germ (WGA), soya beans (SBA) and edible snail (HPA) were kindly provided by Dr. Filka, Laboratory for Production and Control of Lectin Preparations, Faculty of Science, Prague, Czechoslovakia. All other chemicals were of the highest purity availble. Porcine lymphocyte lectin LL I was isolated from pig spleen as previously described [13]. The protein was dialyzed against 1.0070 NHnHCO 3 and after lyophilization used for the i.m. immunization of chinchilla rabbits. Immunoglobulin G was prepared from rabbit antiserum by fractionation with (NH4)2SO 4 and subsequent DEAE cellulose chromatography. The solution of antibodies was adjusted to the concentration of protein 1 mg/ml and divided into two equal volumes. The first volume was passed through a column of Sepharose 4B with immobilized porcine lymphocyte lectin LL I (0.5 mg/ml settled gel), the second through a column of the same size containing unmodified Sepharose 4B. 0.5-ml fractions were pooled and checked for the specific antibodies against porcine lymphocyte lectin LL I [14], then IgG was cleaved with pepsin for 18 h at 37 °C and the reaction mixture chromatographed on a Sephadex G-150 column (1 x 100 cm) to prepare F(ab)2 fragments [15]. 3.2. Target cells The culture medium RPMI 1640 (Grand Island Biological Co., Grand Island, NY, U.S.A.) supplemented with 50 U/ml penicillin, 50/zg/ml streptomycin and 10% heat-inactivated calf serum (Bioveta, Ivanovice, Czechoslovakia) was used for lines K-562 [16], HL-60 [17] and for cytotoxicity testing. OTF 9 teratocarcinoma stem cells were kindly provided by Dr. Dr~iber (Institute of Molecular Genetics, Czechoslovakian Academy of Sciences, Prague, Czechoslovakia). These cells were culture in RPMI 1640 medium plus 10% FCS and subcultured every 2 days with minimal exposure to 0.02% EDTA (Institute of Molecular Genetics, Prague, Czechoslovakia). MSB-1 cell line was a gift from Dr. Max Cooper 84

(Tumor Institute, University of Alabama, Birmingham, AL, U.S.A.). The MSB-1 line is a lymphoblastoid cell line developed by in vitro propagation of cells of a spleen lymphoma from a chicken with Marek's disease [18]. These cells were adapted to RPMI 1640 medium plus 10070 chicken serum plus 8O7oFCS, supplemented with tryptose phosphate buffer (2.95 g/100 ml and 5 ml per 100 ml of medium). 3.3. Preparation of effector cells All experiments were performed with effector cells from heparinized pig peripheral blood (25 IU/ml) from Czech White improved breed not older than 1 yr. Mononuclear cells nonadherent to plastic were separated on Ficoll-Verografin density gradient and washed three times with Hanks' balanced salt solution (Institute of Molecular Genetics, Prague, Czechoslovakia). The isolated cells were resuspended in RPMI 1640 medium in the indicated concentrations and used for cytotoxic assays. In some experiments cells separated on Percoll discontinuous density gradient were used [19]. Seven different concentrations of Percoll (Pharmacia Fine Chemicals, Uppsala, Sweden) in isotonic medium were prepared ranging from 35 to 57.5%. After centrifugation at 550×g for 20 min at room temperature, cells from 7 layers were collected and washed with RPMI 1640 medium containing 5% FCS. The recovery of the cells was about 85% and their viability greater than 95% as judged by Trypan blue exclusion. Glass adherent cells were squeezed out by vigorous washing with BSS and separated into mononuclear and polymorphonuclear cells by centrifugation on Percoll discontinuous density gradients (50 and 70%, respectively). Binding studies were performed with mononuclear cells. Air-dried cell preparations were stained with M a y - Grunwald- Giemsa- Romanowski and morphologically differentiated on the slides by oil immersion microscopy. Cells similar to human LGL with basophilic cytoplasm and several azurophilic granules were counted as pig LGL cells.

3.4. Binding studies Individual [3H]aOS (see Fig. 2 for structures) were incubated with 106 of cells in the concentration range 10 - 5 - 1 0 -1° mol/1. After 1 h at 4°C, cells were separated by centrifugation, washed and cell-bound radioactivity measured by liquid scintillation counting. For the determination of binding capacity of cell fractions, [3H]aOS in the concentration 10 -6 mol/1 were used. Results were analyzed according to Scatchard [20].

1A _Q

1B

eo

8O

6O

60

~o

t,O

6

.~2o

20

10

30

IO0

3O0

i tt 1000 ~'

J 1

3

10

30

100

E:T

3.5. Cytolytic assays Effector cells obtained after Ficoll-Verografin or Percoll separation ( 0 . 5 - 3 . 0 × 106 in 100 #1) were mixed in U-bottom microtiter wells in triplicate with target cells at the appropriate effector-target (E:T) ratios. 100/~1 of the suspension of target ceils (104 cells) prelabeled with Na 5lCr (specific activity 1.21 GBq/ml, Amersham, U.K.) were added. 5~Cr release into 100/~1 supernatant was measured in each well in a gamma counter. The percentage of specific cytolysis was calculated according to the formula: [(exp- SR):(max- SR)] × 100. The release of label from target cells without effector cells was defined as spontaneous release (SR), label released from targets in the presence of effectors as experimental (exp) and label released after the addition of 50 #1 of 0.5°70 Triton solution as maximal (max). Spontaneous relase of the label did not exceed 5°-/0 of the total release. The values are presented as the mean +_ SD of triplicate samples. NK activity of separated cell fractions was expressed in lytic units necessary to give 30°7o specific cytotoxicity as proposed by Cerottini and Brunner (LU30) [21]. To examine the effects of aOS and other substances the ceils were either pretreated with a solution of the substance and then washed or the substance was added in 50/zl to the incubation mixture during the cytotoxic assay.

4. Results

The relation between the percentage of specific cytotoxicity and effector:target cells ratio with K-562 as targets is given in Fig. 1A (pig

Fig. 1. Percentageof specific cytotoxicityas a function of effector:target cell ratio (E:T) for pig mononuclear leucocytes (1A) and NK-activity-enrichedcell fraction (1B, see Table 2) as effectors.

mononuclear leucocytes as effectors) and B (NKactivity-enriched cell fraction as effectors). The linear part of this curve occurs with the ratio effectors:targets from 50:1 to 300:1 (mononuclear cells) and 5:1 to 30:1 (enriched fraction). Significant inhibition of NK cytotoxicity to four xenogeneic cell lines is achieved with aOS of the lactosamine type the structures of which are given in Fig. 2. Triantennary aOS are especially powerful inhibitors even in the concentration 0.1 #g/ml (aOSIII4) or 1 ttg/ml (aOSIII6, Table 1). However, the inhibition is not absolutely restricted to this type of structure as in the concentration 10/~g/ml (Table 1) or 100 izg/ml (Fig. 3) inhibition is observed also with other aOS. The importance of lactosamine type aOS was also shown by the experiments concerning the inhibition of NK cytotoxicity with lectins. Only those lectins that can bind lactosamine type aOS are efficient as inhibitors (Fig. 4). To assess the localization of the NK receptors the binding of aOS to effector as well as target cells was measured (Table 2). Target and glassadherent cells show only a weak, unspecific binding, while pig lymphocytes bind all compounds examined specifically with the association constants in the range 106-1010 1/mol -]. Remarkable agreement between high inhibitory activity in the cytotoxic test and the highly specific binding towards NK-activity-enriched cell fraction (in pigs, cells with morphology similar 85

Oat

Oat Gel al,4 ~1,4 OlcN~ O L ~ ~1,2 B1,2 Man Man ,,1,6 ,tl,3 Man

Oat

Got

Oat

al,4 B14 al 4 OlcNAc OlcNAc OlcNAc e16 B12 ~12

~Iv~n

Iv~n

Man

~I,6 #,1,3 Man t~1,4 Oc N/~,c

~1~ OlcNAc 814

ach~

o 0 S II

Got

Man ~

Got

Got

814 /~1,4 ~14 C4cNAc OtcNAc OlcN/~ ~1,2 e12 e14,

'

at4 Oldq~ ~,16 ' ~

,tl,3 Man t,l(. GIcNAc

t~1,4

£J.cNAc

0tcNAc

o0S III 6

Got

Oat

~14

~1,~

OtcNAc Otr.lqAc OtcNAc B12 ~12 B14

Man ,tl.6

,.,!,6

1~14

0al

at4

M'an~ ,d,3 Ivtan

OLoNAc

C-t.cNAc

o0S IV

a0S II14

Fig. 2. Structures of aOS used in the inhibition studies.

Table 1 Inhibition of NK activity by aOS of the N-acetyllactosamine type

E:T = 50:1 aOSll aOSllI6 aOSIII4 aOSIV Control

0.1 /~g/ml

1 ~g/ml

10 # g / m l

22.5 _+ 1.2 16.8 +_0.7 2.1 _+3.0 25.4 + 0.2

21.4 6.7 2.8 24.6 24.7

20.6 + 1.2 1 . 1 -+ 1 . 0

70 qTROL

>I----

0.8_+0.2

20.6+2.3

ox I-->~J

50

E:T = 100:1 aOSI1 aOSIlI6 aOSIII4 aOSIV Control

48.0_+ 1.7 25.0 _+ 1.2 2.5 _+0.1 47.0 _+ 1.9

25.8 + 0.7 3.7 + 0.6 0.7 + 0.5 23.3 _+0.6 45.7_+2.3

20.0 _+2.4 0.4_+0.4

i,

0.1_+0.1 25.0+ 1.8

~t~ CO 30

E:T = 300:1 aOSll aOSI I l6 aOSIII4 aOSIV Control

63.1 24.7 2.4 69.3

46.6+ 1.0 3.3 + 1.5 0.7 _ 0.7 45.9_+0.9 75.0 + 3.6

23.0_+ 1.4 2.1+1.8 0.4+0.3 25.3 _+ 1.0

10

+0.8 + 0.4 _ 0.8 + 3.4

Effector cells were pig mononuclear cells after F i c o l l - V e r o grafin separation; K-562 cells were used as targets in the effector:target cell ratios indicated, aOS (see Fig. 2 for structures) were present in the incubation mixture during the 5~Cr release assay in the given final concentrations; in control experiments they were replaced with medium. Results are expressed as the mean +_ SD.

86

HL-60

K-562 -+ 0.5 + 1.5 _+0.6 _+2.7 _+0.6

II II16 1114 IV

II

1116 1114 IV

Fig. 3. Inhibition of NK activity of pig mononuclear leucocytes (Ficoll-Verografin separation) against K-562 and HL-60 target cells by aOS of the lactosamine type. aOSIl, aOSIll6, aOSII14 and aOSIV (see Fig. 2 for structures) were added to the 5~Cr release assay mixture (final concentration 100/~g/ml). E:T was 300:1. Bars represent mean +_ SD.

Table 2 Morphology, NK activity and [3H]OS binding for cell fractions of pig leucocytes Cell fractions ~

% LGL

Mononuclear leucocytes Glass adherent cells Percoll fractions: 35.0070 40.007o 42.5°70 45.0070 47.5070 50.0% K-562 target cells a

NK activity against K-562

Association constants Ka~ at 4 °C aOSII

aOSIll6

aOSlll4

I4.3 ND

t.7 x 109 8.5 x 104

1.3 x 10s 5.0 x 104

0.9 x 107 4.7 x 104

2.3 x 106 3.2 x 104

0 190 250 200 40 0 ND

ND ND 2 . 0 x 10r ND ND 2 . 3 x 109 8.3 x 103

ND ND 1.8 × 109 ND ND 1.7X108 6 . 4 x 103

ND ND 2.1 × t0 m ND ND 2.2X 107 6.1 x 103

ND ND 2.3 x 107 ND ND 2.6X 10~ 10.6x 103

LUg0

LU30/I07 cells

5.0 ND

172 ND

0.0 20.0 50.0 50.0 1.0 0.0 ND

0 38 62 66 6 0 ND

aOSIV

For the preparation of cell fractions see experimental section. ND, not determined. Binding of aOS to B lymphocytes, T lymphocytes and polymorphonuclear leucocytes was described in detail in [13].

to human LGL) is evident in the case of triantennary aOS (aOSIII4 and aOSIII6). Finally, we tested the relation of the proposed lectin receptors on the surface of effector cells to

7O

nTl::-q

MSB-1

>-

the previously described membrane lectins found on the surface of pig lymphocytes. Fig. 5 presents data about various effects of F(ab)2 fragments of rabbit antibodies raised against one of these lectins. Strong inhibition can be produced upon preincubation with the effectors or when they are present in the incubation mixture during the cytotoxicity test. The specificity

CONTROL

I 1(-562

HL-60 50

50-

.-L

IX

>O

3o

3,0

30-

g U.

0

o

~ 0

10

I0-

C 12

1I 11161114IV

U 11161114IV

Fig. 4. Inhibition of NK activity of pig mononuclear leucocytes (Ficoll-Verografin separation) against OTF-9 and MSB-I target cells by aOS of the lactosamine type. aOSII, aOSlll6, aOSIlI4 and aOSIV (see Fig. 2 for structures) were added to the ~lCr release assay mixture (final concentration 100/~g/ml). E:T was 300:1. Bars represent mean + SD.

3 4 5

C 1 2 345

LECTINS

Fig. 5. Inhibition of NK activity of pig mononuclear leucocytes (Ficoll-Verografin separation) against K-562 and HL-60 target cells with lectins. P N A (1), W G A (2) and HPA (3) were added to the ~Cr assay mixture (final concentration 5 #g/ml), and lectins from pig liver (4) and pig spleen lectin LL 1 (5) (final concentration 1/~g/ml). EfT was 300:1. Bars represent mean + SD.

87

Table 3 Inhibition of NK cytotoxicity by IgG antibodies against porcine lymphocyte lectin LL 1 E :T

50:1

100:1

300:1

lgG + Sepharose 1:1 1:10 1:100

0.4_+0.2 18.6_+0.2 22.7_+0.8

0.5_+0.3 36.9_+1.0 49.3_+0.8

0.7_+0.6 56.7_+2.7 65.7_+1.2

IgG + LL l-Sepharose 1:1 1:10 1:100

25.3-+0.1 24.1_+0.6 25.1_+1.3

52.4+1.1 53.9_+0.7 48.5_+0.5

74.4_+0.2 68.4-+0.2 68.0-+0.9

Control

25.3___0.3

51.0_+0.4

68.5+_1.0

Effector cells were pig mononuclear leucocytes (Ficoll-Verografin separation); K-562 cells were used as target cells using the effector:target cells ratios indicated. The solution of the specific antibody in the incubation medium (lgG + Sepharose 1:1) was adjusted to the final concentration of protein in the incubation mixture 1 m g / m l (which corresponds to approx. 0.1 m g / m l specific antibodies) and then diluted 10 times (lgG + Sepharose 1:10) or 100 times (IgG + Sepharose 1:100). The solution of IgG from which specific antibodies were depleted by adsorption on LL lSepharose (IgG + LL I-Sepharose 1:1) was adjusted to the same final concentration of protein (1 m g / m l of the incubation mixture) and diluted in the same manner. In control experiments the mixture contained the medium instead of the solution of antibody. Results are expressed as the mean +_ SD.

and dose dependence of this inhibition were evaluated using rabbit antiserum depleted of the specific antibody as a control (Table 3).

5. Discussion

Pig cells with NK activity were separated from peripheral blood lymphocytes on discontinuous Percoll density gradients. Cell fractions isolated by 40, 42.5 and 45°7o Percoll contained cells morphologically similar to human LGL with cytoplasmic granules. No typical monocytes were detected in these fractions. During the investigation of NK cell cytotoxicity, several papers appeared suggesting the role of carbohydrate structures and lectins in the recognition of target cells. More detailed analyses of 88

saccharide structures probably involved in the NK recognition were performed by Muchmore et al. [1] and Harris et al. [5]. Muchmore and coworkers have shown participation of Oglycosidically bound saccharides with terminal /31--3 galactosyl structure in the inhibition of NK activity. Our experiments with animal and plant lectins suggested a more profound role of carbohydrate structures with terminal/31--4 galactosyl residues (pig spleen and liver lectins as compared with peanut agglutinin with the preferential affinity to/31--3 galactosyl residues) [22] and thus we successfully employed this type of structure naturally occurring in mammals. From the structural point of view the fact that not only the presence of the terminal galactosyl residue, but also the type of branching of the oligosaccharide may play an essential role in the NK recognition appears to be important. This statement was supported in our inhibition studies in which triantennary aOS with/31--4 Nacetyllactosamine branch manifested their inhibitory effect in much lower concentrations than did similar aOS with different branchings. The relation between the target sugar structure described by Harris et al. [5] and that proposed by us is not fully understood. Stereochemical comparison of both structures cannot exclude the possibility that a fucosylated or partially fucosylated lactosamine branch can form a part of the target sugar structure. In our experiments monoclonal antibodies against SSEA-I antigen of teratocarcinoma cells inhibited also NK cytotoxicity (unpubl. results). The observation that some aOS may inhibit NK cytotoxicity raised the question about the role of desialylation in the modulation of NK activity in vivo. Elevated concentrations of desialylated serum glycoproteins during inflammation and after malignant transformation are of special importance in this context. In addition, desialylated forms of glycoproteins were found in some liver diseases, such as hepatic cirrhosis, hepatitis and hepatomas [23]. In order to explain the inhibitory effects observed we searched for the appropriate cell surface receptors capable of the specific recognition of the galactose-terminated carbohydrate struc-

HL- 60

K 562

CONTROL

80

40

60

30

40

20

20

10

g f.D

n c~

50

1013

300

50

100

300 E:T

Fig. 6. Inhibition of NK activity of pig mononuclear leucocytes (Ficoll-Verografin separation) against K-562 and HL-60 target cells by F(ab) 2 fragments of rabbit antibodies against pig spleen lectin LL I. Line h target cells preincubated with F(ab)2 (1 mg/ml protein) for 1 h at 4 ° C and washed before 5~Cr release assay. Line 2: F(ab): present during ~'Cr release assay in a final concentration of 0.1 mg/ml protein. Line 3: effector cells preincubated with F(ab)2 (1 mg/ml of protein) for 1 h at 4 ° C and washed before 5~Cr release assay. Effector:target cell ratio was from 50:1 to 300:1.

tures. We examined the binding o f tritiumlabeled lactosamine-type aOS towards target cell lines as well as pig lymphocytes used as effectors. While only weak, unspecific binding to the target cell line K-562 was observed, pig lymphocytes express at least two highly specific binding systems for [3H]aOS. Receptors specific for biantennary aOS were detected predominantly on T as well as on B lymphocytes [13] and a system with the high affinity for triantennary aOS is the major binding system of the NK-activity-enriched cell fraction (LGL). Our attention was focused on the relation of these proposed membrane receptors of NK cells to the previously described membrane lectins isolated from liver, spleen and lymph nodes [10, 11, 13]. As specific antibodies to the NK-activityenriched cell surface receptors were not available to us, we employed rabbit antibodies raised

against the pig lymphocyte lectin LL I, that have a strong cross-reactivity with similar lectins on hepatocytes and NK-activity-enriched cell fraction. The question that we now address concerns detailed molecular characteristics o f receptors reacting with these antibodies which exhibit a strong blocking effect on the NK cytotoxicity when either preincubated with the effector cells (but not with target cells) or present in the incubation mixture. These experimental data support the view that membrane lectins with the specificity for triantennary aOS of the N-acetyllactosamine type and antigenically related to membrane lectins isolated from the spleen and liver play an important role in the NK recognition of several xenogeneic targets by pig NK cells. These membrane receptors can exhibit some degree of cross-reactivity towards various cell surface glycoconjugates with 89

the terminal galactosyl residues and thus represent a non-specific, generally spread type of receptor, the existence of which was suggested in previous studies concerning the specificity of NK recognition (cold target inhibition studies, monolayer cell adsorption procedures) [24]. Experiments concerning the isolation and detailed molecular characterization of these lectin receptors from the pig NK cells are currently under way in our laboratories.

References [1] Muchmore, A. V., Decker, J. M. and Blaese, K. M. (1981) lmmunobiology 158, 191. [2] Brunda, J. J., Wiltrout, R. H., Holden, H. T. and Varesic, L. (1983) Int. J. Cancer 31, 373. [3] Stutman, O., Died, P., Wisun, R. E. and Lattime, E. C. (1980) Proc. Natl. Acad. Sci. (U.S.A.) 77, 2895. [4] Vodenlich, R. L., Sutherland, R., Schneider, C., Newman, R. and Greaves, M. (1983) Proc. Natl. Acad. Sci. (U.S.A.) 80, 835. [5] Harris, J. F., Chin, J., Jewet, M. A. S., Kennedy, M. and Gorczynski, M. (1984) J. Immunol. 132, 2502. [6] Pospi~il, M., Kov~i~fi, F., Trebichavsk~,, I., Hofman, J., Kov~i% H. and Lima, P. (1979) in: Immunology of Reproduction, Proc. 4th Int. Symp. (K. Bratanov, Ed.) pp. 640-645, Varna. [7] Yogeeswaren, G., Gronberg, A., Hanson, M., Daliana, T., Kiessling, R. and Welsh, R. M. (1981) Int. J. Cancer 28, 517.

90

[8] Rooney, C. M. and Munro, A. J. (1984) Immunology 51, 193. [9] Hiserodt, J. C., Britvan, L. J. and Targan, S. R. (1982) J. Immunol. 129, 1782. [101 Ashwell, G. and Harford, J. (1982) Ann. Rev. Biochem. 51, 531. [11] Kieda, C. M. T., Bowles, D. J., Ravid, A. and Sharon, N. (1978) FEBS Lett. 94, 391. [12] Bezou~ka, K., T~borsk~, O., Kubrycht, J., Pospi~il, M. and Kocourek, J. (1985) Biochem. J. 227, 345. [131 Bezou~ka, K., Kar~iskov~i, H., T~iborsk~, O., Kofrofaov~i, O., Vo~i~ek, J., Kubrycht, J. and Kocourek, J. (1985) Lectins Biol. Biochem. Clin. Biochem. 4, 353. [14] Ouchterlony, O. (1958) in: Progress in Allergy, (P. Kal16s, Ed.) Vol. 5, pp. 1-78, S. Karger, Basel. [15] Zik~in, J. (1980) Folia Microbiol. 25, 246. [16] Klein, E., Ben-Bassat, H., Neumann, H., Halph, P., Zeuthen, J., Polliack, A. and V~inky, F. (1976) Int. J. Cancer 18, 421. [17] Collins, S. J., Gallo, R. C. and Galagher, R. E. (1977) Nature (London) 270, 347. [18] Akiyama, Y. and Kato, S. (1974) Biken J. 17, 105. [19] Timonen, T., Ortaldo, J. R. and Herberman, R. B. (1981) J. Exp. Med. 153, 569. [20] Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51, 660. [21] Cerottini, J.-C. and Brunner, K. T. (1971) in: In Vitro Methods in Cell-Mediated Immunity (B. R. Bloom and P. R. Glade, Eds.) pp. 369-373, Academic Press, New York. [22] Sharon, N. (1980) in: Progress in Immunology, Vol. IV (M. Fougerau and J. Dausset, Eds.) p. 254, Academic Press, London. [23] Marshall, J. S. and Stanford, W. (1978) Biochim. Biophys. Acta 543, 41. [24] Ortaldo, J. R., Oldham, R. K., Cannon, G. C. and Herberman, R. B. (1977) J. Natl. Cancer Inst. 59, 77.