MolecularImmunolog,,Vol.

27. No. 8, pp. 745-750,

1990

0161-5890/90 $3.00 + 0.00 Pergamon Press plc

Printed in Great Britain.

MONOCLONAL ANTIBODIES AGAINST THE HUMAN INTERFERON-y RECEPTOR(S) ERIK DEPLA

Laboratorium

voor Biochemie,

and

MARC DE LEY*

KULeuven,

B-3000 Leuven,

Belgium

(First received 8 November 1989; accepted in revised form 23 January 1990) Abstract-Enriched human interferon-y (HuIFN-y) receptor preparations were obtained by affinity chromatography of non-ionic detergent solubilized COLO 205 cell membranes on immobilized recombinant HuIFN-y. The active fractions, identified by a competition ELISA, were used as the immunogen in a BALB/c mouse. Fusion of its splenocytes with myeloma cells yielded several hybrids secreting antibodies that inhibit the antiviral activity of HulFN-y; the two most active ones were selected for further characterization. This blocking activity was restricted to both the human species and the y type of IFN. Affinity purification of cell membrane extracts on the immobilized monoclonal antibodies resulted in the visualization of a major protein band with an M, of 90,000, which is in good agreement with the results obtained by other authors [Aguet M. and Merlin G. (1987) J. exp. Med. 165, 988-999; Novick D., Orchansky P., Revel M. and Rubinstein M. (1987) J. biol. Chem. 262. 8483-8487: Sheehan K. C. F.. Calderon J. and Schreiber R. D. (1988) J. Immun. 140, 423142371.

INTRODUCTION

tion by affinity chromatography on monoclonal antireceptor antibodies revealed a single receptor with an M, of 90,000, which was readily degraded by proteolytic enzymes to a species with an M, of 55,000 (Aguet and Merlin, 1987; Sheehan et al., 1988; Calderon et al., 1988; Rashidbaigi et al., 1988). The isolation of a cDNA for a HuIFN-y receptor present on B-lymphoblastoid Raji cells resulted in the characterization of a protein with an M, of 54,000, containing seven potential N-glycosylation sites and several threonine- and serine-rich regions that are indicative of O-glycosylation (Aguet et al., 1988). In the present study, we report the generation and characterization of monoclonal antibodies (mAbs) directed against an HuIFN-y receptor present on COLO 205 cells. Using these antibodies we also initiated the isolation of this receptor, resulting in an enriched preparation which had a prominent broad protein band with an M, of + 90,000 in SDS-PAGE. To our knowledge, this is the first report of the production of mAbs directed against a HuIFN-y receptor present on a cell line of non-hematopoietic origin. These mAbs, in combination with others, will be of interest in the discussion on the existence of different HuIFN-y receptors on different cell types.

Interferon-y (IFN-r) is endowed not only with antiviral activity (as are IFN-c( and -/J), but more importantly with a variety of immunoregulatory functions [for a review, see Pestka et al. (1987)]. A prerequisite in these multiple actions is its contact with a cellular receptor. Since Branca and Baglioni (1981) proved that the receptor for HuIFN-y is different from that for IFN-a or -p, the former has been extensively investigated both on various diploid cells and on cell lines, resulting in the elucidation of a complex IFN-y receptor system. Peripheral blood monocytes seem to have a receptor distinct from that observed on nonhematopoietic cells (Orchansky et al., 1986; Rubinstein et al’., 1987; Fischer et al., 1988). Differences were noticed even in the receptor population of resting compared with those of activated Tlymphocytes (Faltynek and Princler, 1986), and also between monocytes and macrophages (Yoshida et al., 1988). These results suggest the existence of a third type of IFN-;J receptor, but more evidence is needed to establish this hypothesis. Cross-linking of HuIFN-y to its receptor resulted in the appearance of complexes with an M, ranging from 70,000 to 160,000 as visualized by SDS-PAGE (Sarkar and Gupta, 1984; Ucer et al., 1986; Anderson et al., 1982; Celada et al., 1985). Recently, purifica-

MATERIALS AND METHODS Cells. COLO 205 cells were cultured, as monolayers, in RPM1 1640 supplemented with 0.25 g/l glucose, 2 mM glutamine, 50 mg/l gentamycin, 5% (v/v) newborn and 5% (v/v) fetal calf serum (media and supplements were purchased from Gibco, Paisley, Scotland). A549 and L929 cells were cultured as monolayers in Minimum Essential Medium (MEM) supplemented with 0.25 g/l glucose, 2 mM glutamine,

*Author to whom correspondence should be addressed at: Laboratorium voor Biochemie. Dekenstraat 6, B-3000 Leuven, Belgium. Abbreviations: HuIFN-7, human interferon-y; MuIFN-y, murine interferon-y; mAb, monoclonal antibody; MEM, minimum essential medium; PBS, Dulbecco’s phosphate-buffered saline; pfu, plaque-forming units. 74s

146

E.

DEPLA and

50 mg:‘l gentamycinc, 0.1% (wiv) NaHCO,-HCI (pH 7.0) and 10% (v/v) fetal calf serum. virus. Encephalomyocarditis virus for use in the antiviral assays was propagated on L929 cells and stored at -80 C. In terfivvtts nnd mA h directed ugainst HuIFN -y. HuIFN-a (kindly provided by Dr H. Claeys of the Blood Transfusion Center, Leuven) was prepared on “buffy coat” leukocytes with Sendai virus and purified as previously described (Cantell and Hirvonen, 1978). HuIFN-fi (kindly provided by Dr J. Van Dammc, Rega Institute for Medical Research. Leuven) was prepared according to the superinduction scheme and purified on controlled pore glass (Van Damme and Billiau, 1981) followed by zinc chclate chromatography (Heine and Billiau, 1981). Recombinant glycosylated HuIFN-7 (kindly provided by Professor W. Fiers, State University of Gent) from CHO cells (Scahill ct rd., 1983) was purified on silicic acid, ConA-Sepharose 48 and monoS in an FPLC system. as previously described (De Ley, 1986). Esc~herirhiu &i-derived HulFN-7 for immobilization on CNBr-activated Sepharose 4B and for USCin the inhibition ELISA was kindly donated by Dr G. R. Adolf (Ernst Boehringer lnstitut fiir Arzneimittclforschung, Wien. Austria) and had sp. act. +2 x 10’ IUimg protein. Recombinant MulFN:’ (kindly provided by Dr R. Dijkmans, Rega Institute for Medical Research. Leuven) was purified on silicic acid (Dijkmans ct cd., 1985). The mAbs D9DlO. DIG2 and Dl3C8, directed against HulFN-;, have been described previously (Sandvig et cd.. 1986). The mAbs 3~3 and M7G4, directed against haemocyanin. were used as negative controls. SoIuhili~ution of’ wll mrmhranes. COLO 205 ceils were collected with a rubber policeman. washed twice with Dulbccco’s phosphate-buffered saline (PBS) and resuspended in extraction buffer [lo mM Tris-HCI. pH 7.5. 150 mM NaCI, 1% (v/v) Triton X-100] to 30-50 x IO” cells:ml (Novick et d.. 1987). They were kept at 4 C for l-2 hr. with regular shaking. The extracts were centrifuged for 15 min at 3000~. followed by 1.5 hr at 150.000~, both at 4 C. The supcrnatants were pooled and stored at -80 C. A,fittit?. r,hrot?lcttoRrupl~,. Solubilizcd proteins were subjected to affinity chromatography on immobilized HuIFN-;, as well as on immobilized mAb. Therefore 6 mg of E. coli derived HuIFN-7 or 10 mg of mAb were coupled to 2 g of CNBr-activated Sepharose 4B according to the manut:acturer’s protocol (Pharmacia, Uppsala. Sweden). For affinity purification, pooled cell extracts were passed over the matrix (I .8 x 3.6 cm) at a flow rate of 10 mlihr. The column was washed with Tris-HCI [IO mM, pH 7.5, 150 mM NaCI, 0.1% (v/v) Triton X-100] and subsequently eluted with 4.5 M MgC& The MgCl,-eluted fractions were concentrated by dialysis against PEG 20,000 and finally dialysed against PBS.

M.

DE LEY

AntiGul assuys. Hybridoma supernatants or immunoglobulins isolated from ascitcs fluid were sequentially diluted in a 96-well plate with MEM containing 2% (v/v) newborn calf serum. A549 cells were added (50,00O/well). followed by HuIFN-7 (0.42 IU/well). After a 16-hr incubation at 37 C the medium was replaced and cnccphalomyocarditis virus added (2.4 x IO4 pfu:‘well). The cell viability was assessed microscopically and when the cytopathogenie effect had exceeded 95%. the medium was removed by inverted shaking, and the cells were fixed with formaldehyde and colored with Naphthol Blue Black. After repeated washing with tap water. the dye was extracted from the cells using 50 mM NaOH and the plates were read with a multichannel spectrophotometer at 620 nm. The observed IFN-7 activities were corrected relative to an international standard (Gg-23-901-530-y). Except for the applied virus concentration (1 .I x 10J pfu/well) the assay used for L929 cells was performed as described above. Using COLO 205 cells as a monolayer WC had to replace the Naphthol Blue Black staining by an enzymatic assay based on the endogenous enzyme hexosaminidase (EC 3.2. I .30) (Landegrcn. 1984). Therefore we added, after removal of the medium by inverted shaking, 20 /tl of 7.5 mM 4-nitrophenyl-2acetamido-2-deoxy$-D-glucopyranoside dissolved in 50 mM citrate buffer (pH 5.0) containing 0.25% (v/v) Triton X-100. After a 15-min incubation at 37 C the reaction was stopped and visualized by adding I50 /cl of 50 mM NaOH, and the plate was read at 405 nm. Inhibition ELISA. Microtiterplates (96-well) were coated overnight with 1% (w/v) ovalbumin in phosphate buffer (10 mM, pH 7.2) at 4 C. After washing [three times with 10 mM phosphate buffer. pH 7.2, 0.05% (v/v) Triton X-100] HulFN-;I was added (125 IU/well in 100 /tl phosphate buffer) and incubated for 2 hr. Plates were washed again and sequentially diluted samples of suspected receptor preparations were added. incubated for 2 hr and washed repeatedly. followed by the addition of antiHuIFN-y mAb in a 1’5000 dilution (starting from ascites fluid). The detection was carried out with a rabbit-anti-mouse IgG serum conjugated to horseradish peroxidase (EC I. I 1. I .7) (Dako, Denmark) using orthophenylenediaminc as a substrate. The plates were read spectrophotometrically at 492 nm (MacDonald cut ul.. 1979). Immunixnion ctnd ,fu.sion. A 6-week-old female BALBlc mouse was immunized with a receptor preparation originating from a total of 3 x IO’ COLO 205 cells by four consecutive i.p. injections. The first was given with Freund’s complete adjuvant on day 39 before fusion. and the next 2 weeks later with incomplete Freund’s adjuvant. Finally. the animal was boosted on days 4 and 1 before fusion, the injection on day -4 being iv. as well as i.p. The animal was killed by cervical dislocation and the spleen removed aseptically. The fusion was carried out by the method of Kiihler and Milstein (1975). Brietly. suspended

mAbs against HuIFN-7 receptor splenocytes and myelomas (P3X63-Ag-8.563) were mixed and fused by means of PEG 4000. The fusion products were distributed over 15 96-well plates in RPM1 1640 supplemented with 0.25 g/l glucose, 2 mM glutamine. 50 mg/l gentamycin and 10% (v/v) fetal calf serum. After HAT selection, hybrids of interest were cloned by limiting dilution and the immunoglobulins were produced as ascites fluid in BALB/c mice pretreated with pristane. Pur$cation of‘ mAb. Monoclonal antibodies produced as ascites fluid were purified by (NH,)$O, precipitation. Ascites was mixed with an equal volume of saturated (NH,)$O, (pH 7.0) and magnetically stirred for 15 min. After centrifugation (10.000 8, 15 min, 25 C) the precipitate was dissolved in borate-buffered saline (150 mM NaCl, 170 mM HIBO,, 25 mM NaOH) and half a volume of saturated (NH&SO, was added and stirred again for 15 min. The precipitate was collected by centrifugation (lO.OOO~, 15 min. 25 C), dissolved and dialysed against PBS. The presence and the degree of purity of the immunoglobulins were controlled by cellulose acetate electrophoresis. SDS-PAGE. SDS-PAGE was performed according to the method of Laemmli (1970). using a 10% (w/v) running and a 4% (w/v) stacking polyacrylamide gel. Bands were visualized by Coomassie Brilliant Blue or silver staining. Induction und detection of HLA class II antigens. COLO 205 cells were seeded over 96-well plates (15.000 cells/well) and incubated at 37 C in the presence of 1 IU HuIFN-:,‘ml and of mAb at different concentrations. After 48 hr. cells were harvested using EDTA [0.03% (w/v) EDTA, 0.1% (w/v) BSA, 0.01% NaN, (w,‘v) in PBS. 15 min. 37 C]. The ceils were transferred to V-shaped 96-well plates, washed [0.1% (w/v) BSA. 0.01% (w/v) NaN, in PBS) and incubated (: hr. 4 C) with the mAb 7.5.10.1 directed against human HLA class II antigens (Koning et al.. 1984). After repeated washing a goat-anti-mouse antiserum conjug.ated to FITC was added (i hr, 4 C). Finally. the cells were observed under a fluorescence microscope. Depending on the number and intensity of fluorescing cells, the induction was ranked on scale from 0 to 3. The nonsense mAb 3~3. directed against He1i.u potnuticr hemocyanin. was used as a blank.

147

ovalbumin. IFN bound in this way to ovalbumincoated microtiterplates could be easily detected down to 10 W/well (contained in 100 ~1) using mAbs directed against HuIFN-y. This detection was inhibited by preincubation with various receptor preparations in a dose-dependent way between 0.2 and 25 pg/ml (Fig. 1). Taking into account the volume of the samples, a total IFN-binding capacity of about 5000 IU was calculated, corresponding to I O-20% of what could be predicted theoretically as originating from 3 x 10’ COLO 205 cells [20,000 receptors/cell (Ucer et al.. 1986)]. Based on an M, of 90,000 for the HuIFN-y receptor (Aguet and Merlin, 1987; Sheehan et al., 1988), on the protein content as determined by the Bradford method (1976), and on its IFN-; binding capacity, 0.7% of the protein material was estimated to represent active receptor. This heterogeneity in protein composition was confirmed by SDS-PAGE and could be partially explained by aspecific adsorption, partially by potential proteolytic degradation. Detection qf’hybrids secreting anti-receptor

antibodies

The serum of the BALB/c mouse immunized with the aforementioned preparation blocked the antiviral activity of HuIFN-y (2 IU/ml on A549 cells) by 50% at a l/100 dilution. The fusion products, seeded over microtiterplates, showed more than 90% cellular outgrowth in HAT medium. Hybridoma supernatants of 900 wells were screened for their ability to block the antiviral activity of HuIFN-;l (0.4 IUiwell contained in 200 ~1). About 50 such hybrids were initially withheld; the two most inhibiting ones, R1 GlO and R9D3, were selected for further investigation.

d4921------

RESULTS

Prepration

und characterixztion

of the imnunogen

Extracts of 3 )( 10’ COLO 205 cells were pooled and purified on immobilized recombinant nonglycosylated HulFN-g. These crude receptor preparations were characterized with respect to their HuIFN;’ binding capacity and their protein composition as revealed by SDS--PAGE. To obtain an idea of the IFN-binding capacity, an inhibition ELISA was carried out. Advantage was taken of the property of human non-glycosylated IFN-7 to adsorb aspecifically to certain proteins. like

I-

m -

/cy2 dilution

Fig. I. Inhibition of the immunochemical detection [with three different anti-HuIFN-y mAbs: D9DlO (a), D13CX (m) and DlG2 (+)I of HuIFN-y (125 IU/well) by an enriched receptor preparation (starting protein concentration was 50pg/ml).

E. DEPLA and M. DE LEI

748

Table I. Comparison of the antiwral activity blocking rates of anti-HuIFN-~-receptor (RIG10 and R9D3) and anti-HuIFN-;I mAbs (D9Dl0, D13C8 and DlG2) on two human cell lines peg IgG/ml inhibiting 50% of the antiviral activity of 10 IU/ml HuIFN-y on: mAb

A549 cells*

COLO 205 cells’

RlGlO R9D3 D9DlO DlG2 Dl3C8

2.6 5 1.O 8.2 i 2.9 0.02 f 0.01 0.29 + 0.05 0.12&0.02

I8 +4 40 * 8 0.13 i_ 0.1 1.0+04 0.X k 0.6

n =4. *No preincubation.

Characterization

cf the mAbs R lGl0

and R9D3

The effect of the mAbs RIG10 (IgM) and R9D3 (IgG2b/IgM) on the establishment of the antiviral state by HuIFN-y in A549 as well as in COLO 205 cells was investigated and their neutralizing activity was compared with those of anti-HuIFN-y mAbs. The amount of RlGlO necessary to block the activity of 10 III/ml HuIFN-y by 50% was about a quarter of that of R9D3, both being significantly higher than those for the anti-HuIFN-y mAbs (Table 1). The amounts needed on COLO 205 cells were on average six times higher than those for A549 cells, which is in agreement with the observation that COLO 205 monolayers are five times more sensitive to the antiviral action of HuIFN-y than are A549 cells. When murine L929 cells were used in combination with MuIFN-y no effect at all was observed, neither with the anti-receptor nor with anti-HuIFN-y mAbs (data not shown). The effect of RIG10 and R9D3 as well as that of a nonsense mAb (3~3) on the induction of an antiviral state by HuIFN-x, -1 and -y in A549 cells is presented in Fig. 2. The anti-receptor antibodies had no effect at all on the activity of HuIFN-LX and a small, but reproducible. blocking effect towards HuIFN-P. A 5-hr preincubation with the mAbs before adding the HuIFNs resulted in somewhat higher RlGlO

25

50 25

12 6

15

5025

12 6 3 15 ug W/ml

Fig. 2. Effect (mean of two experiments) of the anti-receptor mAbs RIG10 (0) and R9D3 (0) on the establishment of the antiviral state by HuIFN-u (bottom), -/J’ (middle), and -7 (top) in A549 cells, without (left figures) and after 5 hr preincubation (right figures) with the mAbs. The nonsense mAb 3~3 (0, anti-hemocyanin) was used as a control.

blocking activity for the anti-receptor mAbs (about two-fold for RlGlO; Fig. 2). A second well-known biological effect of IFN-y is its ability to induce HLA class II antigens. Using the anti-receptor mAbs RlGlO and R9D3 we were able to block this effect as well; again, in a quantitative way. As for the blocking of the antiviral activity, the

t@&% RSD3

12.5

3

3c3

6.2

3.1

1.6

microgram

lglml

Fig. 3. Influence of the anti-receptor mAbs RIG10 and R9D3 on the HuIFN-1; induced expression of HLA class 11 antigens by COLO 305 cells. The nonsense mAb 3~3 was used as a negative control.

mAbs

12

against

3

Fig. 4. SDS-PAGE (stained by Coomassie Brilliant Blue) of HulFN-)? receptor preparations isolated by affinity chromatography on immobilized mAb RlGlO (lane 1) and R9D3 (lane 2) as compared with the total cell extract before chromatography (lane 3). Well-defined hemocyanin fragments (M, 155,000, 110,000 and 90,000), bovine serum albumin (M, 68,000). ovalbl:min (M, 43,000), and myoglobin (M, 17,000) were used as molecular weight markers on a 10% (w/v) polyacrylamide gel under reducing conditions.

effect of R9D3 was smaller than that observed RIGIO, but was significantly higher than that tained using a nonsense mAb (Fig. 3). Afin@

for ob-

pur~jication

To build affinity columns, purified RIGIO, R9D3 or M7G4 (as a negative control) IgG was coupled to CNBr-activated Sepharose 4B. After passing the pooled extracts of 3 x IO’ COLO 205 cells over this matrix a highly enriched receptor preparation could be eluted, which again inhibited the immunochemical detection of HulFN-7 adsorbed to ovalbumin (results not shown). Analysis of the peak fractions, eluted from both anti-receptor columns by SDS-PAGE under reducing conditions, revealed a major broad protein band with an M, of k90,OOO and several minor bands (Fig. 4).

DISCUSSION

Triton X-100 membrane extracts from COLO 205 cells, enriched on immobilized recombinant HuIFN7, were used as an immunogen to produce mAbs directed against the HuIFN-y receptor. The aspecific adsorption of the IFN-1; to proteins such as ovalbumin was used to develop an assay for the detection of its receptor. Adsorbed to microtiterplates coated with

HuIFN-y

receptor

749

ovalbumin, HuIFN-y could be quantitated by antiHuIFN-y mAbs, the reaction being inhibited by the presence of receptor preparations. Quantitative evaluation of the IFN-binding capacity of these isolates allowed us to calculate a recovery of IO-20%, a figure similar to that observed by other authors. Splenocytes of a seropositive BALB/c mouse, immunized with this enriched preparation, 0.7% (w/w) of the protein present being active HuIFN-y receptors, were used for fusion with P3X63-Ag8.653 myeloma cells. An initial screening of hybridoma supernatants, based on the inhibition of the antiviral effect of HuIFN-g on AS49 cells, yielded about 50 positive cultures, two of which (RIG10 and R9D3) were selected, on the basis of their high neutralizing activity, for investigation in more detail. RlGlO shows the highest blocking activity (IO IU/ml of HuIFN-y inhibited by 50% at 2.6 icg IgG/ml), whereas R9D3 exhibited a somewhat lower blocking activity (8.2 /lg IgG/ml for the same ir hibition). No effect at all was observed against IFN-x IS an agonist and only a small, but reproducible, inhibitory effect towards IFN-B. This last observation is consistent to that of Anderson et a/. (1982). who found some competition between ‘251-labelled HuIFN-7 and HuIFN-B for binding to GM258 fibroblasts. The antiviral activity of HuIFN-y on COLO 205 cells could also be inhibited but at higher mAb concentrations; this is in agreement with the observed higher sensitivity of this cell line to the action of HuIFN-y. The establishment of the antiviral state by MuIFN-;I on a murine cell line (L929) could not be inhibited, proving that both mAbs are specific for the HuIFN-?; receptor. Extra evidence, proving that both mAbs are directed against a HulFN-y receptor, was offered using a different approach. Triton X- 100 extracts of COLO 205 cells were adsorbed on affinity columns based on the mAbs RIG10 or R9D3. SDS-PAGE of the eluted fractions revealed the presence of several protein bands, among which is a major broad band with an M, of 90,000; this figure is consistent with the observations of other authors (Aguet and Merlin. 1987; Sheehan et al.. 1988; Calderon et al., 1988). As these mAbs are directed against a HuIFN-y receptor present on a non-hematopoietic cell line, they are interesting tools to study the differences of IFN-y receptor populations on different cell types. Preliminary results. obtained by ELISA on fixed cells, revealed that the mAb R 1G IO reacts specifically with a receptor present on non-hematopoietic cells and on peripheral blood monocytes, but not on other hematopoietic cells (manuscript in preparation). We herein describe the production of two antiHuIFN-1; receptor mAbs which can be used for affinity purification of this receptor, its degradation products and/or subunits. Further purification to homogeneity, allowing its sequence determination and, ultimately, comparison with the amino acid sequence predicted from a cDNA clone. is under way.

750

E. DEPLA and M. DE LEY

A,knon,lrd~em~Mrs--We are grateful to the Instituut tot Aanmoediging van het Wetenschappelijk Onderzoek in Nijverheid en Landbouw for a research fellowship (E. Depia) and to Dr G. Adolf for providing us with recombinant human interferon-p.

REFERENCES Aguet M., Dembic Z. and Merlin G. (1988) Molecular cloning and expression of the human interferon-gamma receptor. Ceil 55, 273-280. Aguet M. and Merlin G. (1987) ~urifieation of human gamma interferon receptors by sequential affinity chromatography on immobilized monoclonal antireceptor antibodies and human gamma interferon. J. e.yp. Med. 165, 988-999. Anderson P.. Yip Y. K. and Vilcek J. (1982) Specific binding of ““l-human interferon-gamma to high affinity receptors on human fibroblasts. J. hioi. Chem. 257, 11,301~11,304. Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. .4nu/~,l. B&hem. 72, 248 -254. Branca A. A. and Baglioni C. (1981) Evidence that types I and II interferons have different receptors. Nature, Lo&. 294, 768%770. Calderon J.. Sheehan K. C. F.. Chance C., Thomas M. L. and Schreiher R. D. (1988) Purification and characterization of the human interferon-gamma receptor from placenta. P,oi,. ,~rrl,i. Acnd. Sci. L:.S.A. 85, 48374841. Cantell K. and Hirvonen S. (1978) Large-scale production of human leukocyte interferon containing lO(8) units per ml. J. ,qan. i’irol. 39, 541-543. Celada A., Allen R.. Esparza I., Gray P. W. and Schreiber R. D. (198%~eIn~)nst~dtion and partial characterization of the interferon-gamma receptor on human mononuclear phagocytes. J. c/in. freest. 76, 2196.-2205. De Ley M. (1986) OK-432-Stimulated production of human interferon-gamma, In In~munopl~ctmmrr~u/o~ical Aspecr.s of 01(-432 it! HU)IINXV(Edited by Hoshino T.), pp. 6048. Excerpla Medica. Tokyo. Dijkmans R.. Volckaert G.. Van Damme J.. De Ley M.. Billiau A. and De Somer P. (1985) Molecular cloning of murine interferon-gamma cDNA and its expression m hsterologous mammalian cells. J. IFN Rrs. 5, 51 I 510. Faltynek c‘. R. and Princlcr G. L. (1986) Modulation of interferon-a and interferon-gamma receptor expression during T-lyll~phocytc activation and proliferation. J. IFN Ru. 6, 639 65.1, Fischer D. G., Novick D., Orchansky P. and Rubinstein M. (1988) Two molecular forms of the human interferongamma receptor. J. hiol. C/WV. 263, 2632-2637. Heine K. J. W. and Billiau A. (1981) Purification of human fibroblast interferon by adsorption to controlled-pore glass and zinc-&elate chromatography. Mtatlr. .&:yrn. 78, 44x 456. Kbhlcr G. and Milstein C, (1975) Continuous cultures of

fused cells secreting antibody of predefined specificity. Nature, Land. 256, 495499. Koning F., Schreuder J., Giphart M. and Bruning H. (1984) A mouse monoclonal antibody detecting a DR-related MT-2 like specificity. Humus Immun. 9, 221. Laemmli U. K. (1970) Charge of structural proteins during the assembly of the head of the bacteriophage T4. Nnture, Lund. 227, 680--685. Landegren U. (1984) Measurement of cell numbers by means of the endogenous enzyme hexosaminidase. Applications to detection of lymphokines and cell surface antigens. J. fnmiun. Mettl.-67~ 379-388. MacDonaid D. J.. Belfild A.. Steele C. J.. Mack D. S, and Shah M. M. (lY79) The q~dntitation of pregnancy specitic /I,-glycoprotein by enzyme linked Immunoassay. Cl& C/Jim. Acts 94, 4149. Novick D., Orchansky P., Revel M. and Rubinstein M. (1987) The human interferon-gamma receptor. J. hinl. C/?em. 262, 8483-8487. Orchansky P., Rubinstein D. and Fischer D. G. (1986) The interferon-mamma receptor in human monocytes is different from the one in nonhematopoietic cells. J. Irnnzun. 136, 169-173. Pestka S., Langer J. A., Zoon K. C. and Samuel C. E. (1987) Interferons and their actions. A,ln. Rw. Biochcm 56, 727-777. Rashidbaigi A.. Stefanos S.. Jung V. and Langer J. A. (1988) Immune interferon receptor: chemical and enzymatic sensitivity. J. fFN Res. 8, 641-654. Rubinstein M., Novick D. and Fischer D. G. (1987) The human interferon-gamma receptor system. Imnrun. Rw. 97, 29-50. Sandvig S., Laskay T., Andersson J.. De Ley M. and Andersson U. (1986) Gamma-interferon is produced by CD3 + and CD3 - lymphocytes. In2nmn. Rev. 97, 5 l-65. Sarkar F. H. and Gupta S. L. (1984) Receptors for human gamma interferon: binding and crosslinking of “‘I-labeled recombinant human gamma interferon to receptors on WISH cells. Proc. nutn. Acud. S(,i. L’.S.A. 81, 5160.-5164. Scahill S. J., Devos R., Van der Heyden J. and Fiers W. (1983) Expression and characterization of the nroduct of a human immune interferon cDNA gene in Chinese hamster ovary cells. Proc. nrrfn. Acad Sci. C’.S..A. 80, 4654 4658. Sheehan K. C. F., Calderon J. and Schreiber R. D. (1988) Generation and characterization of monocional antibodies specific for the human interferon-gamma receptor. J. In~nw~. 140, 423 I- 4237. Ucer U., Bartsch H.. Scheurich P.. Berkovic D.. Ertel C. and Pfizenmaier K. (1986) Quantitation and characterization of ~drnrna-interferon receptors on human tumor cells. Cffnwr fzYS. 46, 5339-5343. Van Damme J. and Billiau A. ( I981) Large scale production of human fibroblast interferon. Mrth. Etz:jw. 78, 101-I 19. Yoshida R., Murray H. W. and Nathan C. F. (1988) Agonist and antagonist effects of interferon Y and /j on activation of human macrophages. Two classes of interfcron-ganirna receptors and blockade of the high-a~nity sites by interferon x and I(. .I. f.yp. M&. 167, 1171 -I 18s.