Interaction of human IgG chimeric antibodies with the human FcRI and FcRII receptors: Requirements for antibody-mediated host cell-target cell interaction

Molecular Immluroio~)‘, Vol. 26, No. 4. pp. 403-41

I,

1989

Printed in Great Britain.

OF HUMAN IgG CHIMERIC ANTIBODrES WITH THE HUMAN FcRI AND FcRII RECEPTORS: REQUIREMENTS FOR ANTIBODY-MEDIATED HOST CELL-TARGET CELL INTERACTION

~~TE~ACT~~~

M. R. WALKER,*? J. 7x4. WOOF,$ MM.BR~~GGEMANN,$ R. JEFFERIS’ and r). R. BURTON$ *Department of Immunology, The Medical Schoal, Edgbaston, Birmingham 315 2TJ, U.K.; SDepartment of Biochemistry, University of Sheffield, Sheffield SlO 2TN, U.K.; and $Division of Immunology, Department of Pathology, Addenbrooke’s Hospital, Cambridge CB2 2QQ, U.K. (Received

17 Jut?e 19SS; uccepted

22 Noliember

198s)

~~~racr~hirn~r~c monocfonai antibodies (McAb), specific for the hapten S-~odo-4-h~drox~~3-n~trophenacetyl (NIP), expressing human IgGl, IgG2, IgG3 and IgG4 subclass constant domains, have been examined for their ability to interact with the human FcRII receptor. Human red blood cells (RBC) sensitized by each of these McAbs have been assayed for their ability to form rosettes with the human histiocytic lymphoma U937 cell line, human B cell line Daudi and erythroblastoid K562 cell line. IgGl and IgG3 sensitized RBC formed significant rosettes with the FcR- and FcRII+ Daudi and K562 cell lines, the percentage ofcellsforming rosettes being directly proportional to the degree of sensitization of the REC. Bromehn treating Daudi celfs did not alter this pattern of reactivity, whereas brotnehn treated FcRI+ and FcRII+ U937 cells formed significant resettes with IgGl, lgG3 and IgG4 sensitized RBC, demonstrating a difference in the IgG subclass specificity between human FcRI and FcRII. Murine IgG2b anti-NIP sensitized RBC did not form rosettes with any cell line tested; however, RBC sensitized by some members of a panel of murine IgGl McAb, specific for the glycophorin A molecule, were able to form rosettes with Daudi, U937 and K562 cells. This interaction was enhanced by bromeIin treating the Daudi or U937 cells and can be correlated to the disposition of the epitopes recognized, relative to the target cell membrane, those McAbs recognizing epitopes furthest fforr the RBC surface being most efh%ve in interacting with FcRII. The data are interpreted in terms of a simple model for anti~dy-mediated ~ell*eIl interaction.

INTRODUCTION Three distinct FE: receptors for IgG (FcRf have been defined by a number of criteria on human leukocytes (Anderson and Looney, 1986) and termed FcRI, FcRII and FcRIII (previously FcR,,). Interaction of IgG with FcR mediates several biological processes, including phagocytosis of immune complexes (Silverstein et al., 19771, antibody dependent celt-mediated cytotoxicity (Shen et nt., 1987; Craziano and Fanger, 1987; Ortaldo et al., 1987) and stimulation of the release of superoxide (Yamamoto and Johnson, 1984; Anderson et al., 1986) and inflammatory mediators (Rouzer et n2., 1980). Understanding the mechanisms by whioh FcR mediate these processes, in terms of molecular spwifkity and FcR recognition sites on IgG, is critical in determining the clinical implications of restricted antibody responses to given antigens, e.g. rheumatoid arthritis (Lobatto et al., 1987), haemolytic desease of the newborn (Walker et al., 1988) and systemic lupus erythematosus (Salmon ef al., 1984, 1986). Indeed

-_tAuthor to whom correspondence should be addressed at: Department of Clinical Chemistry, Wolfson Research Laboratories, Queen Elizabeth Medical Centre, Birmingham BfS 2TH, U.K.

altered Fc receptor expression and function has been d~monst~ted in systemic lupus erythematosus (Frank et nf., 1979) and rheumatoid arthritis (Katayama, 1981; Fries et al., 1984). Recent studies have shown that three FcR are expressed on distinctive and overlapping populations of cells (Anderson and Looney, 1986) and their mean mol. wts determined as 72,000, 40,000 and 50,00070,000 for FcRI, FcRII and FcRIII, respectively. The high affinity of FcRI for IgC monamer (rC, z 5 x I@/M) has enabled its affinity for the human IgG subclasses to be determined as IgGI = IgG3 > IgG4, whilst IgG2 is not bound (reviewed by Burton, 1985), and the interaction site of FcRI on IgGl to be identified (Woof et al., 1986; Duncan et& 1988). The specificity of FcRII on platelets has been determined to include all four subclasses, using chemically or heat induced oligomers of human paraproteins to mediate aggregation or release reactions (Pfueller and Luscher, 1972; IIenson and Spiegelberg, 1973; Martin et al., 1978). Inhibition experiments, however, suggest that FcRiI on platelets bind IgGl and IgG3 with equai affinity and IgG2 and IgG4 less well (Karas et al., 1982). FcRI and FcRII also appear to have differences in specificity for murine IgG subclasses, FcRI demonstrating a preferential atlinity for murine IgGfa (Woofet nf., 1986), whilst FcRII has a reported prefer-

403

M. R.

404

WALKEK et ui.

ence for the IgG2b and IgGl subclasses (Looney et al., 1986; Jones et al., 1985). The availability of chimeric IgG subclass monoclonal antibodies (McAb) recognizing the hapten 5-iodo-4-hydroxy-3~nitrophenacetyl (NIP) (Briiggemann et al., 1987) offers the potential to delineate the IgG subclass specificity of human FcRII. Within this study we have examined this specificity using these McAbs specifically bound to their ligand (NIP) on the surface of human RBC in a rosette assay. We have also carried out a series of experiments which allow us to characterize those factors important in rosetting or more generally in the antibody-mediated interaction of host and target cell. METHODS

Monoclonaf

AND MATERIALS

Red cell sensitizations

antibodies

The production of chimeric IgG subclass monoclonal anti-5-iodo-4-hydroxy-3-nitrophenacetyl (NIP) antibodies has been previously described (Briiggemann et al., 1987). Purified anti-NIP antibody was produced by affinity chromatography using NIP-caproate Sepharose column. Genetically engineered anti-NIP antibodies were a kind gift from Dr M. Neuberger (Cambridge). Ascitic fluid containing murine IgGl anti-glycophorin A McAb BRIG 93,89, 117, I 19 and 127 were kind gifts from Dr D. J. Anstee, South Western Regional Transfusion Service, Bristol. McAb LICRj LON R10 was also a kind gift from Dr P. Edwards. Ludwig Institute for Cancer Research, Surrey, whilst McAb CLB-ery-1 was a kind gift from Dr A. von dem Borne, Central Laboratory of The Netherlands Red Cross Blood Transfusion Service, Amsterdam. Monoclonal human IgG3 anti-Rhesus (D) antibody was purified from culture supernatant as previously described (Walker ef al., 1988). Paraproteins Human IgG subclass paraproteins from serum by DEAE-ceiluIose chromatography.

mouse serum for 30 min at 37 C. Non-adherent cells were removed by vigorous washing with RPMIIIO% FCS and the adherent cell fraction recovered by incubating with RPMI-10% FCS containing 5 miM EDTA (Sigma) for 5 min at 37°C. The adherent cell fraction was washed twice and resuspended in sterile saline prior to use in the assays. In all cases the percentage viability of the cells was assessed by their ability to exclude the dye Trypan Blue and was consistently > 95?&. In some* experiments U937 were cultured in the presence of 1000 IU/ml of recombinant human interferon gamma (IFN-7; Janssen Life Sciences Ltd) for 48 hr, or 1 mM dibutyryladenosine 3’,5’-cyclicmonophosphate (Bt,cAMP; Sigma) for 72 hr at 37 C. FolIowed by three washes in sterile PBS prior to use.

were isolated ion exchange

Cell lines The human monocytic cell line U937, human erythroblastic line K562, Burkitt lymphoma line Daudi and myelocytic cell line HL60 were maintained in RPMI 1640 medium (Gibco U.K. Ltd) supplemented with 10% foetal calf serum (FCS) at 37’C in a humidified, 5% CO, gassed incubator. Cells were washed three times in sterile phosphate buffered saline (PBS) prior to suspension to the appropriate density in PBS for use in the assays. Murine macrophages were obtained by washing the peritoneal cavity of BALB/c mice with sterile PBS 6 days after they had received 0.1 ml of pristane (Aldrich) i.p. The cells obtained were washed twice with RPM1 medium containing 10% FCS and incubated for 1 hr at 37°C in glass bottles pretreated with

Human 0’ (RlR2) red blood cells (RBC), kindly provided by Dr D. McDonald (Blood Transfusion Service, Edgbaston, Birmingham), were washed five times in sterile PBS and packed by centrifugation at 7508 for 5 min. Washed RBC were sensitized with human monoclonal anti-Rhesus (D) antibody, as previously described (Walker c-‘tal., l988), or with mouse antiglycophorin A McAb by incubating 60 ~1(2 drops) of packed RBC with 900~1 of McAb ascitic fluid at a 1 in IO0 dilution in PBS for I hr at 37 C. Sensitized RBC were then washed three times in sterile PBS and resuspended to a 1% suspension in PBS. For sensitization with anti-NIP McAb. washed RBC were washed twice with isotonic borate buffer, pH 8.3, and packed in the same buffer. Packed RBC (I ml) were incubated for I hr at room temperature on a rotator with 9 ml of borate buffer, pH 8.3, containing 100 pg of NIP-caproateO-succinimide (Cambridge Research Biochemicals, Harston, Cambridge) predissolved in dimethylformamide (Sigma). The NIP derivatized RBC were then washed four times with PBS and 200,ul of a 10% suspension of RBC were incubated with 100 btl of purified anti-NIP McAb at 200, 100, 50, 25, IO and 5i~g/ml in PBS for I hr at 37°C. Sensitized red cells were further washed three times with PBS and resuspended to I % suspension. Rosette assays Rosetting of sensitized RBC to effector cells was performed according to the protocol described by Anderson et al. (1986), using a ratio of 100 RBC:l effector cell, as previously described (Walker ct al., 1988). Inhibition of rosette formation was performed by incubating the reaction mixture in the presence of 50 ItI of purified IgG diluted in PBS to O.ll9.7 mgiml as indicated in the text. Direct binding ussq Polyclonal rabbit anti-mouse lambda light chain (RAM-,?.) antiserum, purified by DEAE-cellulose ionexchange chromatography, was a kind gift from Dr

Interaction

of human

S. Oldfield. This preparation was radiolabelled with 12’iodine using the chloramine T method (Hunter and Greenwood, 1962), as previously described (Walker et al., 1987). Direct binding of ‘251-labelled antiserum to sensitized RBC was performed according to the method of Woof et al. (1984, 1986). Essentially, 2 x 10’ RBC in a balanced salt solution containing 0.2% BSA and 0.05% NaN, (BBN) were incubated with ‘251-labelled RAM-I (sp. act. of 0.25 pCi/ml) in a total reaction volume of 40 ~1 for 1 hr at 37°C. The sp. act. of the cell bound label was measured using an LKB Wallac 1275 Minigamma, following separation by centrifugation through water-immiscible oil (Versilube F50; a kind gift from Dr J. Brock, Glasgow). In each experiment duplicate tubes were used and the non-specific binding, determined using non-sensitized NIP derivatized RBC, subtracted. Haemagglutination

assay

Haemagglutination was performed in U-shaped microtitre trays (Titertek), as previously described (Lowe et al.. 1982). Ascitic fluid containing murine McAb specific for epitopes on human IgG (Lowe et al., 1982; Nik Jaafar et al., 1983) or polyclonal RAM antiserum were doubly diluted from a 1 in 10 external dilution in HEPES buffer containing 2% FCS. One drop (30 ~1) of a 0.2% suspension of sensitized RBC were added to each well and the agglutination pattern assessed, relative to control cells, in the absence of antibody after 30min1 hr. Reverse passive haemagglutination was similarly performed using RBC sensitized with murine McAbs specific for the human IgG subclasses, as previously described (Walker et al., 1986). RESULTS Rosetting

of human IgG subclass

sensitized

RBC

Using the FcRII+ cell lines Daudi (Fig. la) and K.562 (Fig. lb) significant rosette formation was evident with RBC sensitized with IgGl, IgG3g and IgG3b anti-NIP McAb, the percentage of cells forming rosettes being proportional to the degree of sensitization of the RBC. Neither IgG2 nor IgG4 sensitized RBC mediated rosette formation at any level of sensitization. Significant rosette formation was also mediated by IgG3 sensitized RBC using the FcRI+ and FcRII+ cell lines U937 (Fig. lc) and HL60 (Fig. Id), independent of the IgG3 allotype used here. However, increasing the density of FcRI on U937 cells by stimulation with 1000 U/ml IFN-y enabled IgGl, as well as IgG3 sensitized RBC, to form rosettes (Fig. le). In contrast, the concomitant up-regulation of FcRII and down-regulation of FcRI on U937 cells, on stimulation with Bt,cAMP (Sheth et al., 1988), resulted in abolition of IgGl-mediated rosette formation and diminution of IgG3-mediated rosette formation, such that a significant percentage of rosettes were only formed at the highest sensitization levels (Fig. If).

IgG chimeric

antibodies

405

Direct binding of ‘251-labelled rabbit anti-mouse lambda light chain (RAM-I) antiserum to sensitized RBC revealed that all the IgG subclasses sensitized RBC to equivalent degrees (Fig. 2). Similarly, haemagglutination, using murine McAb specific for epitopes on human IgG Fc, or RAM polyclonal antiserum, gave equivalent titres for each IgG subclass sensitization (data not shown). Inhibition

of rosette formation

by human IgG subclass

paraproteins

IgG3 anti-Rhesus (D) CB6 sensitized RBC have been demonstrated to form rosettes via FcRI but not FcRII (Walker et al., 1988). This rosette formation was significantly ( > 90%) inhibited by human IgGl, IgG3 and IgG4 paraproteins, but not by IgG2 paraproteins, at concns of the order of 10m6 M (Table 1). IgG3 anti-NIP sensitized RBC rosetting via FcRII expressed on Daudi cells could also be significantly inhibited by IgGl and IgG3 paraproteins (60-90%) at 1 mg/ml (7 x 10m6M), but not by IgG2 or IgG4 paraproteins at the same concn. Increasing the concn of the IgG4 used as inhibitor to 6.7 and 9.7 mg/ml could, however, produce some inhibition (2540%), consistent with contamination of these preparations with polyclonal IgGl and/or IgG3. Reverse passive haemagglutination, using sheep RBC sensitized McAbs specific for IgG3 (ZG4) and IgGl (JL512), revealed that, at the concentrations used, IgG3 plus IgGl accounted for 0.7 mg, sufficient to produce the observed inhibition of rosette formation (data not shown). Removal of IgGl and IgG3 contaminants by positive affinity chromatography, employing an IgG4 specific monoclonal antibody (RJ4) linked to Sepharose 4B (Pharmacia), abolished the rosette inhibition previously observed for the IgG4 paraproteins tested (data not shown). Murine

IgGl

anti-glycophorin

A

sensitized

RBC

rosetting

RBC sensitized with murine IgGl anti-glycophorin A McAb varied widely in their ability to mediate rosette formation with FcRI and/or FcRII expressing cell lines (Table 2). No significant rosette formation was evident with any McAb using U937 or HL60 cells. Stimulation of U937 cells with IFN-y, thereby upregulating FcRI, did not appreciably alter the degree of rosette formation with any McAb sensitization. However, stimulation of U937 cells with Bt,cAMP, thereby up-regulating FcRII, did produce a significant increase in the level of rosette mediated by CLB-ery-I (from 5 to > 15%). BRIC 93 and BRIC 89 did not mediate rosette formation with the FcRIand FcRII+ cell lines Daudi and K562, whilst BRIC 117, 119, 127, LICR/ LON RIO and CLB-ery-I sensitized RBC mediated rosette formation to different degrees (from 10 to 80%, respectively).

M. R.

406

Murine IgG2b anti-NIP

sensitized

WALKER

al.

Eflect

RBC rosetting

Murine ‘wild type’ IgG2b anti-NIP sensitized RBC were unable to mediate rosette formation with U937, Daudi or K562 cells, but did mediate significant rosette formation ( > 30%) with mouse peritoneal macrophages (Table 3). However, RBC sensitized with murine IgG2b anti-NIP EL235 containing Leu instead of Glu at residue 235 formed significant rosettes with the FcRI+ U937 cells.

(a)

et

on rosetting

of pretreating

: IO pg

ICI u937

: 5

IgG

cells ulith

brome*in U937 cells treated with bromelin formed rosettes with IgGl (80%), IgG3 ( > 90%) and IgG4 (> 60%) anti-NIP sensitized RBC, but not with IgG2 anti-NIP sensitized RBC (Table 4), whereas only IgG3 sensitized RBC formed significant rosettes (50%) with control U937 cells. Significant rosettes were also formed with IgG3 anti-Rhesus (D) CB6 sensitized RBC (> 90%).

DAUDI

, 20

efector

i 25

per

IgG2 . IgG4 05

:

I

2u

2 x IO7 RBC

10 /1g

le 1 IFN

(d)

HL60

20

10

5

IgG

per

5

IgG

I

25

per

2

x IO7

- 8 stimulated

05 pg

RBC

(f

U937

)

IgG

CAMP

per

2

x IO’

stlmutated

RBC

U937

c pg

25 2 x IO7

I RBC

05

20

IO pg

IgG

5 per

25 2 x IO7

I RBC

05

20

IO pg

IgG

5 per

25 2 x IO7 RBC

Fig. 1. Rosette formation between RBC sensitized with varying amounts of human IgG anti-NIP McAb and (a) Daudi cells, (b) K562 cells, (c) U937 cells, (d) HL60 cells, (e) U937 cells stimulated with IFN-y and (f) U937 cells stimulated with Bt,cAMP. RBC sensitized with(D) IgGl, (A) IgG2, (0) IgG3b. (0) IgG3g and (0) IgG4 anti-NIP McAb.

05

Interaction of human IgG chimeric antibodies

407

n

IgGl

0

IgG2

3 IgG3g l

IgG3b

0 IgG4

Fig. 2. Direct binding of ‘251-labelledRAM light chain antiserum to RBC sensitized by varying amounts of human IgG anti-NIP McAb.

murine IgG2b anti-NIP EL235 sensitized RBC (> 80%) and RBC sensitized with murine IgGl antiglycophorin A McAb BRIC 117, 119, 127, LICR/ LON RlO and CLB-ery-1 (22-73%), but not with RBC sensitized by McAb BRIC 89 or 93. Using control U937 cells, rosettes were formed with CB6 sensitized RBC (75%), EL235 sensitized RBC (25%), but not with RBC sensitized by any of the murine IgGl anti-glycophorin A McAbs (Table 2). Greater than 80% of bromelin treated Daudi cells formed rosettes with human IgG3 and IgGl, but not IgG2 or IgG4 anti-NIP sensitized RBC, whilst control Daudi cells also formed rosettes with only IgG3 and IgGl RBC. No rosetting with bromelin treated Daudi cells

Table I. Inhibition of human IgG3g anti-NIP and IgG3 anti-D sensitized RBC rosette formation by human IgG paraproteins

% Inhibition of rosette formation IgG3g-RBC CBCRBC + U937”

Inhibitor

“Inhibitors ‘Inhibitors ‘Figure in dFigure in

0 85 93 II II 63 73 17 I3 (28)’ I5 (43)d

0 99 97 7 5 99 99 90 92 89

PBS alone Wid IgGl Cr IgGl Pe IgG2 Gi IgG2 Ga IgG3 GB6 IgG3 Rea IgG4 Am IgG4 AS IgG4 at 0.1 mg/ml. at I .O mg/ml. brackets obtained brackets obtained

using inhibitor using inhibitor

Table 2. Rosette

McAb BRIC 93 BRIC 89 BRIC I17 LICR/LON BRIC 119 BRIC 127 CLB-cry-I

RIO

between + Daudi”

at 9.7 mg/ml. at 6.7 mg/ml.

formation

of murine

IgGl

No. epitopes per RBC”

u937

% Rosettes formed with u937 u937 + IFN-yh + CAMP HL60

I ,ooo,ooo I ,ooo,ooo 200,000 500,000 200,000 200,000 ND

0 0 0 352 0 0 312

“Determined in laboratory of origin. ‘Stimulated with IO00 U/ml IFN-y. ‘Stimulated with ImM Bt,cAMP. dND = not determined.

anti-glycophorin

0 0 2+1 I+1 NDd ND 4*1

0 0 0 0 ND ND 15+3

A McAb sensitized

0 0 0 0 0 0 6?3

RBC

Daudi

K562

0 3*3 l3&5 27k II 35 *3 36 + I 71 i-7

0 0 IO& 20 f 25 f 28 f 53 *

1 I 3

I 2

408

M. R. WALKER

Table 3. Murk

IgGZb anti-NIP

sensitized

RBC rosette formation

% Rosettes formed with “lCJ”X

RBC sensitization”

u937

KS62

Daudl

macrophages

Wild type lgG2b EL235 IgG2bh

0 14k6

0 ND

0 ND

31+4 ND

“2 x IO’ RBC sensitized with 20 pg IgG. hEL235 = Glu to Leu interchange at residue 235 (Eu index). ‘ND = not determined. Table 4. Effect of bromelin RBC sensltizatlons IgGlh IgGZh IgG3’ IgG4h BRIC 89 BRIG 93 BRIG 117 BRIG 119 BRIC 127 LICR/LON R 10 CLB-ery-1 EL235 IgG2b CB6 kG3

treating elkctor % Rosette Daudi

u937 2*2 0 38 i_ 5 0 0 0 0 0 312 3i2 14Fh 75 + 3

cells on rosette formation forming cells B-U937”

7s + 3 0 85 i 2 0 0 3+3 13+5 35 +3 36 * 1 27ilI 77 * 7 ND 0

85 5 2 0 91+2 6lk3 0 0 22 & 2 35 2 2 31 * 1 22 * 2 73 i_ 3 x4 * 1 92 + 2

B Daudi” 86 f 0 83 f 0 551 lOi 85 i 92 5 87 * 91 t 92 * ND I+1

“B--U937 and B-Daudi = bromelin treated U937 and Daudi ‘RBC sensitized with 2O/(g anti-NIP IgG per 2 x IO’ RBC. ‘ND = not determined.

2 3

I

AC (rosette

formation)

= AC (AbR

interaction)

+ AC (non-specific cells.

DISCUSSION

The differences in affinity of human FcRI and FcRII for IgG necessitates that different approaches are adopted in the study of these two molecules. The high affinity of FcRI for monomeric IgG has allowed this interaction to be probed in the simplest manner, by studying the binding of monomer IgG to the

blood

receptor. In energy terms, the only consideration here would seem to be the binding affinity of the FcRI for the particular IgG molecule. With FcRII, however, since affinity for monomeric IgG is not detectable under normal experimental conditions, the IgG must be presented to the receptor in an aggregated form. In this study we have employed chimeric anti-NIP antibodies, aggregated by their binding to NIP coated onto the surface of RBC. Binding is now measured in terms of rosette formation. This is a complex phenomenon which, as evidenced in the data above, is dependent on a variety of factors. To rationalize the data, we suggest that rosette formation in energetic terms can be seen as the resultant of two opposing terms (for the sake of simplicity the red cell is considered to be precoated with antibody):

2 2 2 I I

was observed with IgG3 anti-Rhesus (D) CB6. BRIG 89 and 93 sensitized RBC mediated 5510% rosettes with bromelin treated Daudi cells, whereas values of 85592% were obtained using McAb BRIC 117, 119, 127, LICR/LON RIO and CLB-ery-1 sensitized RBC. In contrast, using control Daudi cells, significant rosettes (13-77%) were formed using McAb BRIC 177, 119, 127, LICR/LON RIO and CLB-ery-1 sensitized RBC, but not with McAbs BRIC 89 or 93 (Table 2).

Red

et al.

cellcell

interaction),

where AG(AbR interaction) is the free energy associated with the occupation of Fc receptor sites by antibody molecules in the rosette. The magnitude of the term will depend on: (1) the nature of the antibody; (2) the nature of the Fc receptor; and (3) the number of antibody-receptor interactions formed in the rosette. AG (non-specific cell
cell

Bromelin treated effector cell Effector

cell

Fig. 3. Schematic representation of host cell-target cell interaction. (m) IgG binding site on Fc receptors, (mm)mouse IgGl anti-glycophorin A antibody, BRIC 89, (0) human IgG3 anti-D antibody, CB6 or anti-NIP

chimeric

antibody.

lateraction of human Kg0 chimeric antibodies the factors contributing into play in determining considered in turn.

to the two terms can come rosette formation. They are

(1) Nature of the antibody. Rosettes via FcRII were only formed with human IgGl and IgG3 of the human subclasses. The weak ability of IgG4 paraprotein preparations to inhibit such rosettes can be attributed to trace contamination of IgGl and/or IgG3. Thus, in our view, IgG4 has no significant affinity for FcRII. Previous studies of FcRII interaction with human IgG subclasses, using chemically or heat-aggregated human IgG paraproteins, have suggested that the specificity of FcRII for human IgG encompasses all the four s&&asses (Pfuefler and Luscher, 1972; Henson and Spiegelberg, 1973; Martin et d., 1978), whilst inhibition experiments suggested that FcRII bound IgG with affinities ranked IgGl = IgG3 >> IgGZ = IgG4 (Karas et al., 1982). However, these studies are also subject to the question of paraprotein purity. The difference in specificity for IgG4 between FcRl and FcRII shown here suggests that the binding site on human IgG of FcRI and FcRII, and perhaps the architecture of the binding site of these two receptors, is different. The binding specificities of the mouse subclasses IgG2b and IgGl have also been investigated. Murine IgG2b has been generally accepted to bind to FcRII, but not to FcRI (Anderson and Looney, 1986). However, using murine IgG2b anti-NIP sensitized RBC, no rosette formation was observed with U937, Daudi or KS62 cells (Table 3), although significant rosetting was observed with the positive control of meus peritoneal macrophages. Using this system, therefore, we found no evidence of mouse IgG2b interacting with FcRII. Exchanging Glu for Leu at position 235 (Eu index), within the area suggested by Woof et al. (1986) to be the interaction site for FcRI, resulted in resetting of RBC sensitized with this protein and U937 cells, in agreement with the result of direct binding studies using this mutant mouse IgG2b (Duncan et af., 1988). The interaction ofmurine IgGl with human FcRII has also been examined using McAb recognizing defined epitopes on the glycophorin A molecule expressed by human O+ RBC (Anstee and Edwards, 1982; Dr D. Anstee, personal communication). Sig nificant rosetting was observed with RBC sensitized with several of these McAb and Daudi and K562 cells (Table 2), confirming the previously described interaction of murine IgGl with human FcRIf (Jones et al., 1985; Looney et cl., 1986). However, none ofthese proteins mediated rasette formation with the FcRII’ U937 or HL60 cells. This can be seen to be largely independent of the density of FcRII on these cell types, since stimulation by Bt,cAMP, which up-regulates F&II whilst down-regulating FcRI, produced only a

409

small increase in the rosetting of McAb CLB-cry-1 sensitized RBC (from 5 to 15%). It has previously been demonstrated that there are ~lymo~hisms in FcRII receptors, with some individuals expressing FcRII receptors capable of binding murine IgGl (termed responders) and others with FcRII receptors incapable of binding murine IgGl (non-responders) (Tax et al., 1984; Anderson et ai., 1987). Since both responders and non-responders bind human IgG, it is &ear that human IgG and mu&e IgGl recognize FcRfI differently. (2) Nature of the Fc receptor. It is clear from many studies that the affinity of FcRX for appropriate monomer IgG is much higher than that of FcRII (Anderson and Looney, 1986). (3) Numfier of antibody-& receptor i~te~o~t~o~~ The effect of this parameter on rosette formation is illustrated for the human IgG3-FcRI interaction in Fig. 1 c, e and f. Up-regulation of FcRI by interferony increased rosette formation considerably, whereas down-regulation by CAMP had the opposite effect.

(1) Natu+e of two cell sq%ces. Bromelin treatment of cells results in the cleavage of surface glycoproteins, thereby reducing the net surface charge on the cells (Miale, 1977). The effect of this treatment on U937 was to produce a dramatic increase in rosette formation via FcRI with IgGI and IgG3 sensitized red cells. Furthermore, rosettes, not formed with untreated U937 cells and IgG4 sensitized red cells, are formed with bromelin treated U937 cells. This pattern correlates with the IO-fold lower F&I affinity of IgG4 than IgGl or IgG3. (2) Geomerry of‘ ant&e=. An example of the importance of this factor is provided by compa~son of FcRII-mediated rosetting of anti-D and anti-NIP antibodies. Thus, although rosetting could be achieved in this paper with anti-NIP antibodies, it was not shown with anti-D antibodies, even if the effector (Daudi) cell was treated with bromelin. NIP coating wifl yield antigen at varying distances from the red cell membrane, whereas the D antigen is attached to the cytoskeleton and relatively close to the membrane. This situation is represented schematically in Fig. 3. A second example is provided by the FcRII rosetting ability of the panel of mouse IgGl monoclonal anti-glycophorin A antibodies. Rosetting is found not to correlate with the level of sensitization of red cells achieved, since McAbs BRIC 89 and 93 recognize the most sites per red cell but mediate the least number of rosettes. Rather the extent of rosette formation correlates with the proximity of the epitope recognized to the red cell membrane. Thus (compare Table 2), BRIG 89 and 93 have been mapped to recognize a site on glycophorin A close to the transmembrane domain (residues 62-72), LICR/LON RIO and BRIC 117 recognize an area further from the membrane surface, encompassing residues 26-39, whilst BRIC 119 and 127 bind to a similar but distinct epitope (Anstee and

M. R. WALKER ei al

410

Edwards, 1982; Dr D. Anstee, personal communication). The epitope recognized by McAb CLB-ery-1 has not been determined (to our knowledge), but is most Iikely, according to our data, to be farthest from the membrane surface. (3) Geometry of the Fc receptor on the leucotyte cell surface. The effect of this factor is very difficult to separate from that due to the nature, i.e. binding affinity, of the Fc receptor described above. Thus, the greater propensity of FcRI than FcRIf to form rosettes, in comparable circumstances (Walker et af., 1988; this paper), could be due to both affinity and geometry differences. In Figure 3 we have speculatively represented FcRI as extending further out of the membrane than FcRII since FcRI is a 3- and FcRII a 2-domain structure (D. Simmons, private communization). (4) Geometry C$ the bridging antibody. The importance of this factor is seen in the greater tendency to form rosettes with IgG3 than IgGl, in comparable circumstances (Walker et al., 1988; this paper; Merry et nl., 1989; Dougherty et al., 1987; Wiener e6 uf., 1987). This presumably reflects the shorter hinge of IgGl compared to IgG3 (Gregory et al., 1987), necessitating closer cell-cell approach for antibody bridging. In conclusion, this study illustrates the variety of factors which can influence host cell-target cell interaction mediated by antibody. However, interaction is only the first stage in the process of target cell destruction and there is evidence that factors which favour interaction may, in some cases, be less advantageous for destruction (Bruggemann et al., 1987). We are currently extending our studies to a systematic investigation of the factors important in antibody-mediated target cell destruction. .4cknowlednements-The authors would like to thank Dr D. Anstee and Dr P. Edwards for the kind gift of murine IgGl anti-glycophorin A McAbs, Dr D. McDonald for the human red blood cells and assistance with the bromelin treatment of cells, Dr B. Kumpel for the anti-D (Rh) McAb CB6 and Peter Richardson and Julie Campbell for helping to establish the rosetting assays. This work was supported by a grant from the Wellcome Trust. D. R. Burton is a Jenner Fellow of the Lister Institute of Preventive Medicine. M. Briiggemann is a recipient of a Special Fellowship from the Leukaemia Society of America.

REFERENCES Anderson C. L., Guyre P. M., Whitin J. C.. Ryan D. H., Looney R. J. and Fanger M. W. (1986) Monoclonal antibodies to Fc receptors for IgG on human mononuclear phagocytes. Antibody characterisation and induction of superoxide production in a monocyte cell line. f. hial Chem. 261, 12,856.12,864. Anderson C. L. and Loonev R. J. (1986). Human leukocyte IgG Fc receptors. Itnmu~. T0da.i 7, 2&G-266. Anderson C. L.. Rvan D. II.. Loonev R. J. and Learv P. C. (1987) Structural polymorphism of the human monocyte 40 kilodalton Fc receptor for IgG. J. Immun. 138, 2254 2256. Anstee D. J. and Edwards P. A. W. (1982) Monoclonal

antibodies to human erythrocytes. Eur. J. Immun. 12,228232. Briiggemann M., Williams G. T., Bindon C. I., Clark M. R., Walker M. R., Jefferis R., Waldmann H. and Neuberger M. S. (1987) Comparison of the effector functions of human immunogiobulins using a matched set of chimeric antibodies. J. rxp. Med. 166, 1351~-1361. Burton D. R. (1985) Immunoglobulin G: functional sites. M&c. Immun. 22, 161-206. Dougherty G. J., Selvendran Y., Murdoch S., Palmer D. G. and Hogg N. (I 987) The human mononuclear phagocyte high affinity Fc receptor FcRI defined by a monoclonal antibody, 10.1. Eur. J. rrnrnfi~. 17, 1453-1459. Duncan A. R., Woof J. M., Partridge L. J., Burton D. R. and Winter G. (1988) Localization of the binding site for the human high-affinity Fc receptor on IgG. Numrr 332, 563-564. Frank M. M., Hamburger M. I., Lawley T. J., Kimberly R. P. and Plotz P. H. (1979) Defective reticuloendotheliai system Fe-receptor function in systemic lupus erythematosus. New Ennl. J. Med. 300, 518-523. Fries L. F., MuGins W. W., Cho K. R., Plotz P. H. and Frank M. M. (1984) Monocyte receptors for the Fc portion of IgG are increased in systemic lupus erythematosus. J. Immun. 132, 695--700. Graziano R. F. and Fanger M. W. (1987) Fey RI and Fq RI1 on monocytes and granulocytes are cytotoxic trigger molecuies for tumor cells. f. ~mrnu~. 139, 35363541. Gregory L., Davis K. G., Sheth B., Boyd J., Jefferis R., Nave C. and Burton D. R. (1987) The solution conformations of the subclasses of human IgG deduced from sedimentation and small angle X-ray scattering studies. Moler Immun. 24, 821-829. Henson P. M. and Spiegelberg W. L. (‘1973) Release of serotonin from human platelets induced by aggregated immunogiobulins of different classes and subclasses. J. clin. inoesr. 52, 1282-1288. Hunter W. M. and Greenwood F. C. (1962) Preparation of iodine-131-labelled human growth hormone of high specific activity. Narure 194, 495496. Jones D. H., Looney R. J. and Anderson C. L. (1985) Two distinct classes of IgG Fc receptors on a human monocyte line (U937) defined by differences in binding of murine IgG subciasses at low ionic strength. J. Immun. 135,334~.3353. Karas S. P., Rosse W. F. and Kurlander R. J. (1982) Characterisation of the IgG-Fc receptor on human platelets. BIood 60, 1277-1282. Katayama S. (1981) Increased Fc receptor activity in monocytes from patients with rheumatoid arthritis. J. Immun. 127, 643-647. Lobatto S., Breedveld F. C., Camps J. A. J., Pauwels E. K. J., Westedt M.-L., Daha M. R. and Van Es L. A. (I 987) Mononuclear phagocyte system Fc-receptor function in patients with sero-positive rheumatoid arthritis. C/in. exp. Immun. 67, 461466. Looney R. J.. Abraham G, N. and Anderson C. L. (19X6) Human monocytes and U937 cells bear two distinct Fc receptors for IgG. .I. Immutt. 136, 1641- 1647. Lowe J. A., Bird P., Hardie D., Jefferis R. and Ling N. R. (1982) Monoclonal antibodies (McAbs) to determinants on human gamma chains: properties of antibodies showing subclass restriction or subclass specificity. Zmmunolog~ 47, 329-335. Martin S. E., Breckenridge R. T., Rosenfeid S. I. and Leddy J. P. (1978) Responses of human platelets to immuzlologic stimuli: independent roles for complement and IgG in zymosan activation. J. fmmun. 120, 9- 14. Merry A. H., Brojer E., Zupanska B., Hadley A. G., Kumpel B. M. and Hughes-Jones N. C. (1989) The ability of monoclonal anti-D antibodies to promote the binding of red cells to lymphocytes, granulocytes and monocytes. VOX Suna. (In press.)

Interaction

of human

Miale J. B. (1977) Laboratory Medicine Hematology, 5th edition. C. V. Mosby Company, Saint Louis. Nik Jaafar M. I., Lowe J. A., Ling N. R. and Jefferis R. (1983) Immunogenic and antigenic epitopes on immunoglobulins-V. Reactivity of a panel of monoclonal antibodies with sub-fragments of human Fey and abnormal paraproteins having deletions. Mofec. Immun. 20, 679686. Ortaldo J. R., Woodhouse C., Morgan A. C., Herberman R. B., Cheresh D. A. and Reisfeld R. (1987) Analysis of effector cells in human antib~y-d~~ndent cellular cytotoxicitv with murine monoclonal antibodies. J. Immun. 138, 3&X&3572. Pfueller S. L. and Luscher E. F. (1972) The effects of \ aggregated immunoglobulins on human blood platelets in relation to their complement-fixing abilities. I. Studies of immunoglobulins of different types. J. Immun. 109, 517.-525. Rosenfeld S. I., Ryan D. H., Looney R. J., Anderson C. L., Abraham G. N. and Leddy J. P. (1987) Human Fc receptors: stabie inter-donor variation in quantitative expression on platelets correlates with functional responses. .I. Immun. 138, 2869-2873. Rouzer C. A., Scott W. A., Kempe J. and Cohn Z. A. (1980) Prostaglandin synthesis by macrophages requires a specific receptor-ligand interaction. Proc. natn. Acad. Sri. U.S.A. 17,4279-4282. Salmon J. E., Kimberly R. P., Gibofsky A. and Fotino M. (1984) Defective mononuclear phagocyte function in systemic lupus erythematosus: dissociation of Fc receptor-ligand binding and internalization. J. Immun. 133, 2525-253 1. Salmon J. E., Kimberly R. P., Gibofsky A. and Fotino M. (1986) Altered phagocytosis by monocytes from HLADR2 and DR3-positive healthy adults is Fc receptor specific. .I. Immun. 136, 362553630. Shen L., Guyre P. M. and Fanger M. W. (1987) Poiymorphonuclear leukocyte function triggered through the high afhnity Fc receptor for monomeric IgG. J. fmmun. 139,534538. I

IgG chimeric

antibodies

411

Sheth B., Dransfield I., Partridge L. K., Barker M. D. and Burton D. R. (1988) Dibutyryl cyclic AMP stimulation of a monocyte-like cell line, U937: a model for monocyte chemotaxis and Fc receptor-related functions. Immunology 63,483-490. Silverstein S. C., Steinman R. M. and Cohen 2. A. (1977) Endocytosis. A. Rev. B&hem. 46, 669-722. Tax W. J. M., Hermes F. F. M.. Willems R. W., Cape1 P. J. A. and Koene R. A. P. (1984) Fc receptors for mouse IgGl on human monocytes: polymorphism and role in antibody-induced T cell proliferation. J. Immun. 133, 1185-i 189. Walker M. R., Kumpel B. M., Thompson K., Woof J. M., Burton D. R. and Jefferis R. (1988) Immunogenic and antigenic epitopes of immunoglobulins. Binding of human monoclonal anti-D antibodies to FcRI on the monocytelike U937 cell line. VOX Sang. 55, 222-228. Walker M. R., Lee J. and Jefferis R. (1987) Immunogeniety and antigenicity of immunoglobulins: detection of human immunoglobulin light-chain carbohydrate, using concanavalin A. Riochim. biophys. Acta 915, 314-320. Walker M.R., Solomon A., Ling N. R., Brown B., Lowe J. A., Hardie D. and Jefferis R. (19861 Immunorrenic and antigenic epitopes of immunog~obu~s. XVIII. Subpopulations of human lambda chains defined with a panel of monoclonal antibodies. immunology 59, 467471. Wiener E., Atwal A., Thompson K. M., Meiamed M. D., Gorick B. and Hughes-Jones N. C. (1987) Differences between the activities of human monoclonal IgGl and IgG3 subclasses of anti-D (Rh) antibody in their ability to mediate red cell binding to macrophages. Immunology 62, 401404. Woof J. M., Nik Jaafer M. I., Jefferis R. and Burton D. R. (1984) Localisation of the monocyte-binding domain(s) on human immunoglobulin G. Molec Immun. 21, 523-527. Woof J. M., Partridge L. J., Jefferies R. and Burton D. R. (1986) Localisation of the monocyte-binding region on human immunoglobulin G. M&c. Immun. 23, 319-330. Yamamoto K. and Johnson R. B. (1984) Dissociation of phagocytosis from stimulation of the oxidative metabolic burst in macrophages. J. exp. Med. 159,405-416.