DERIVATION OF A MONOCLONAL ANTI-MURINE IL-4 ANTIBODY SPECIFIC FOR AN EPITOPE EXPRESSED ON DENATURED MOLECULES Ulus Atasoy,
Cynthia
Watson,
William
E. Paul
Monoclonal antibodies to interleukin 4 (IL-4) were generated by immunization with recombinant IL-4 and screening by binding to IL-4 adsorbed to plastic surfaces. Three antibodies were obtained that scored well in this assay and one, 13E1, was studied in detail. It was very effective in detecting IL-4 by Western blotting whereas a neutralizing anti-IL-4 antibody, llB11, was 50-lOO-fold less sensitive as a detecting agent. Sequential immunoprecipitation and biosynthetic labelling studies indicated that the llBl1 and 13El epitopes are largely expressed on different forms of IL-4.13El was able to detect cytosolic IL-4 both by immunohistochemical and flow cytometric analysis of fixed cells. This was routinely successful in an insect cell line (Sf9) expressing large amounts of IL-4 as a result of infection with a recombinant ‘IL-4 baculovirus’. Although stimulated DlO cells could also be shown to express IL-4 in their cytosol, positive results were not obtained in all such studies and we have failed to detect IL-4 production by normal T cells using this method. This antibody may have substantial value in detecting IL-4 by Western blots and as a tool to analyze the biosynthesis of IL-4. With suitable improvement in sensitivity, it also may prove valuable in the detection of IL-4 in the cytosol of individual cells.
The analysis of the function of murine IL-4 has been strikingly clarified by the use of a monoclonal antibody (11Bll) capable of blocking its functions both in vitro and in vivo.l-4 Unfortunately, this antibody has been relatively insensitive as a reagent for immunoblotting and, in our hands, has been of limited use for cytosolic localization of IL-4 in fixed IL-4-producing cells. To this end, it would be very helpful to have additional monoclonal antibodies that were specific for IL-4 but more suitable for the recognition of epitopes found on surface adsorbed IL-4 (for immunoblotting) and in fixed IL-4-producing cells. It has been reported that screening hybridoma cell lines with antigens adsorbed onto plastic surfaces will sometimes result in the selection of monoclonal antibodies that recognize epitopes associated with ‘denatured’
proteins.s-* Such antibodies might be quite useful for the purposes described above. Here we report the derivation of a monoclonal antibody to mouse IL-4 that is extremely efficient in detecting IL-4 by immunoblotting and can bind to IL-4 in the cytosol of insect cells infected with an ‘IL-4-baculovirus’ (Ac.MNPV.IL-4). This antibody fails to neutralize IL-4 activity and recognizes an epitope found on denatured IL-4 but also identifies a unique biosynthetically labelled form of IL-4. It may be of substantial value both in detecting denatured IL-4 and in following the biosynthesis of this molecule.
RESULTS 13El Detects IL-4 coated on Plastic Wells
From the Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, USA. Correspondence to: William E. Paul. Received 13 April 1992; revised and accepted for publication 15 July 1992 @ 1992 Academic Press Limited 1043-4666/92/060537+08 $08.00/O KEY WORDS: monoclonal denatured epitope
CYTOKINE,
Vol.
4, No.
antibody/interleukin
6 (November),
1992:
4, lymphokine,
pp 537-544
Lewis rats were immunized with recombinant IL-4 derived from a yeast expression system. Hybridomas emerging from a fusion of spleen cells from these animals with SP2/0 were screened for their capacity to bind to Immulon 2 plates that had been coated with recombinant IL-4 (0.5 pg/ml). Binding was detected by subsequent addition of an alkaline phosphatase-labelled murine anti-rat Ig and development with 4-methylumbelliferylphosphate. Three 531
538
I Atasoy
CYTOKINE,
et al.
hybridomas (13E1, 7Fl and 6HlO) were selected for further study. After cloning, these were tested for specificity for IL-4 using Immunlon 2 plates that had either not been coated or had been coated with preparations of recombinant IL-4 derived from E. coli, yeast and baculovirus expression systems. Each of the antibodies was positive on all the IL-4 coated plates as detected by development with alkaline phosphatase conjugated goat anti-rat Ig (Fig. 1). In the experiment illustrated in Fig. 1, 13El was tested as a purified Ig while 7Fl and 6HlO were used as tissue culture supernatants. However, in other experiments in which each of the antibodies was tested as tissue culture supernatants, 13El was always superior to 7Fl and 6HlO in binding to IL-4-coated plates. Strikingly, llB11, a monoclonal antibody known to be capable of neutralizing and immunoprecipitating IL-4, failed to bind to IL-6coated plates; it gave no greater signal in the ELISA than either anti-CD8 or anti-TNP. The failure to detect llBl1 binding to IL-4-coated plates was not due to an inability of the alkaline phosphatase goat anti-rat Ig to detect llB11, since it bound well to Immunlon 2 plates coated with 11Bll directly (data not shown). These results indicate that 13El recognizes plate-bound IL-4 while llBl1 does not. In keeping with this, 13El gave a strong signal in detecting purified recombinant baculovirus-derived IL-4 by immunoblotting. It was 50-loo-fold more sensitive than 11Bll; 5 pg/ml of 13El detected 10 ng of IL-4 in a 1 h exposure whereas 11B 11, under the same conditions, was only positive with the 500 and 1000 ng samples of IL-4 (Fig. 2). With longer exposures, 13El could easily detect 1 ng of IL-4. To unequivocally establish that 13El recognized IL-4, recombinant baculovirus-derived IL4 was immunoprecipitated with 11Bll. The immunoprecipitate was boiled in 2.3% SDS under reducing conditions and the resultant solubilized sample was electrophoresed on a 12.5% SDSpolyacrylamide gel. The sample was transferred to
Detection Source
of IL4 by indirect
Recombinant
IL-4
(W3)
Antibody
J4.1
Figure
2.
13El
detects
IL-4
by immunoblotting.
Differing amounts of recombinant IL-4 were electrophoresed in a 12.5% polyacrylamide-SDS gel under reducing conditions, transferred to nitrocellulose and developed with 13E1, llBl1 and the control antibody J4.1 and detected with rabbit anti-rat IgG followed by W-Protein A, as described in Materials and Methods.
nitrocellulose and exposed to 13E1, llBl1 and a control antibody (54.1); 13El clearly detected the material that had been immunoprecipitated by llB11, whereas, under the conditions of this assay, neither llBl1 nor 54.1 gave a detectable signal (Fig. 3).
13El and llBl1 RecognizeEpitopes Not Found on the Same Molecules of IL-4 13El efficiently recognizes IL-4 bound to plastic surfaces or adsorbed onto nictrocellulose membranes whereas 11Bll either fails to bind to such material or does so very inefficiently. By contrast, 11Bll is a potent inhibitor of IL-4 function; 4 pg/ml completely inhibited the capacity of 1000 U/ml (c.O.5 @ml) of IL-4 to stimulate DNA synthesis by CT.4R cells. 13El
ELISA
of rlL4: PBS
E. coli
Yeast
Baculovirus
a;;ij;KK
Figure
o~g~o~g~o~g~o”o”o d
Vol. 4, No. 6 (November 1992: 537-544)
cj
0.
0.
0’
0.
0. 0.D
*denotes purified antIbodIes; rest are supermatants
0.
d
N 0.
m d
-’ ‘:0
1.
13El
detects
plate bound
IL-4
by ELBA.
A series of rat anti-IL-4 and control monoclonal antibodies were incubated on Immulon 2 plates that had been coated with recombinant IL-4 (0.5 pg/ml) derived from E. coli., yeast and baculovirus expression systems, as described in Materials and Methods. Binding was detected with an alkaline phosphatase conjugated goat anti-rat Ig and the fluorescent substrate 4methylumbelliferyl phosphate.
Monoclonal antibody to an epitope on denatured IL-4 I 539
prominently displayed on surface adsorbed IL-4, it is likely that it is a feature of some forms of denatured IL-4. The data in Fig. 5 do show that both 13El and 1lBll pre-precipitated IL-4 contain fewer cpm of 1251that can be precipitated with 11Bll and 13E1, respectively, than does 54.1 pre-precipitated 1251-IL-4.
lmmunopreclpitoting antibody 2 z =!?
lmmunoblotting antibody
: 3
400000 cells
t
Figure 3. Immunoblotting immunoprecipitated with
with 11Bll.
13El
detects
IL-4
that
had
stimulated
with
IL-4
t
7FI
been
IL-4 was immunoprecipitated with llB11, 13El or 54.1. The dissolved immunoprecipitate was electrophoresed on an SDSpolyacrylamide gel, transferred to nitrocellulose and detected by blotting with llB11, 13El or 54.1, followed by rabbit anti-rat IgG and 12s1-Protein A. The arrow indicates the band with the size expected for IL-4. The other bands represent Ig H and L chains of the antibodies used for immunoprecipitation.
at the same concentration failed to inhibit stimulation by as little as 10 U/ml of IL-4; similarly 7Fl and 6HlO also failed to inhibit (Fig. 4). Taken together, these results indicate that 13El reacts with a conformer of IL-4 that is not recognized by llBl1 and, since 13El fails to inhibit IL-4, 13El may not recognize epitopes on active IL-4. To examine in more detail the possibility that the two monoclonal antibodies recognize epitopes on distinct forms of IL-4, IL-4 was labelled with 1251and immunoprecipitated with either 13E1, llBl1 or 54.1. The residual material was divided into three portions, each of which was reprecipitated with one of the monoclonal antibodies. As shown in Fig. 5, llBl1 and 13El precipitated approximately equal amounts of IL-4 that had been precleared with 54.1. By contrast, 11Bll was superior to 13El in precipitating 1*51-IL-4 that had been pre-precipitated with 13El and 13El was superior to 11Bll in precipitation of material that had been pre-precipitated with 11Bll. This result indicates that molecules of IL-4 exist that express the 13El but not the 11Bll epitope and that others express the 11Bll but not the 13El epitopes. In view of the capacity of 11Bll adsorbed to Affigel to remove all biologically active IL-4, molecules that express the 13El but not the 11Bll epitope must be biologically inactive. Since the 13El epitope is
IL-4
Figure IL-4.
4.
13El
to inhibit
fails
(U/ml
1
stimulation
of CT.4R
cells
by
CT.4R cells (5000iwell) were cultured with varying amounts of IL-4 in the presence of 4 pg/ml of various monoclonal anti-IL-4 antibodies or an anti-class II MHC antibody. Cells were pulsed with 3H-thymidine 48 h later and uptake of radioactivity measured after an additional 6 h.
Precipitating Prrnary
ontlbody Secondary I 34.1
13EI
13EI IlBll
J4.1 13EI
11811
IlBll
J4.1 J4.1
l3EI IlBll I 0
I
I IO
Percent
Figure 5. molecules.
13El
and
llBl1
identify
I 20
I
lz51-IL-4
IL-4
preclpltoted
epitopes
on
distinct
W-IL-4 was immunoprecipitated with protein A-Sepharose beads coated with 13E1, 1lBll or 54.1. The residual material was divided into three portions and immunoprecipitated with beads coated with 54.1, 13El or llBl1 and the amounts of IL-4 bound determined by measuring the percent of the 12sI-IL-4 remaining after the first precipitation that was precipitated by the secondary antibody.
540 I Atasoy
CYTOKINE,
et al.
This may reflect the presence of some molecules that express both epitopes or may be due to non-specific losses in the course of immunoprecipitation. To further examine the forms of IL-4 that were recognized by 11Bll and 13E1, we infected Sf9 cells with Ac.MNPV.IL-4 for 48 h, washed the cells, placed them in methionine-free medium for 1 h and then pulsed them with 35S-methionine for 30 mins. An excess of cold methionine was then added. Samples were taken at the end of the pulse and after 15 or 60 mins of chase and cellular lysates were immunoprecipitated with 13E1, llBl1 or 54.1 (Fig. 6). 54.1 failed to precipitate any detectable material. 13El brought down three bands (18 000 Da, 14 500 Da and 13 900 Da) of which the 14 500 Da band was clearly dominant. Each of these bands was prominently seen at the end of the pulse and had diminished considerably in intensity by 60 mins of chase. By contrast llBl1 immunoprecipitated mainly the 18 000 Da band, the intensity of which did not appreciably vary in the course of the chase. In addition, 11Bll also brought down the 13 900 Da band and small amounts of a band of c.16 000, detectable only after 1 h of chase. These results indicate that 13El principally recognizes epitopes on biosynthetically labelled IL-4 with molecular weight of 14 500 while the 11Bll epitope is principally found on a 18 000 Da form of IL-4. The analysis does not allow us to determine whether the form of IL-4 displaying the 13El epitope is processed into a form expressing the 11Bll epitope. The diminished signal seen with 13El after 60 min of pulse could represent further processing of this form of IL-4 or could be due to degradation of this species.
Vol. 4, No. 6 (November 1992: 537-544)
and fixed in suspension with buffered formaldehyde (9.25%)/acetone (45%). They were exposed to 13El or J4.1 followed by biotinylated rabbit anti-rat Ig and FITC-streptavidin. As shown in Fig. 8, the cells infected with IL-4 baculovirus had greater fluorescence when stained with 13El than with 54.1 whereas the cells infected with the IL-5 baculovirus showed little or no enhanced fluorescence with 13El. Similar experiments were done with DlO cells that had been stimulated with PMA and A23187 for 12 h or had been unstimulated. In this case, PE-streptavidin was used in an effort to enhance the fluorescence. Stimulated DlO cells incubated with 13El gave somewhat more fluorescence than cells incubated with 54.1 whereas no difference was observed in the unstimulated cells. However, the staining of fixed D 10 cells with 13El was only found in a proportion of the experiments (2/4) that were carried out. The variables of stimulation or staining procedures that may explain these differences have not been determined.
DISCUSSION We report here the derivation of a monoclonal anti-murine IL-4 antibody that may be of considerable value in the analysis of IL-4 both for its detection by immunoblotting, for its potential to analyze expression of IL-4 immunohistochemically and for its possible use in the analysis of IL-4 biosynthesis. This reagent
IlBll
Immunohistochemical
Detection
of IL-4 with 13El
The capacity of 13El to recognize certain biosynthetically labelled forms of IL-4 as well as non-active IL-4 bound to plastic surfaces suggested that it might detect IL-4 within fixed cells. Accordingly, we infected Sf9 cells with Ac.MNPV.IL-4 or with Ac.MNPV.IL-5. After 48 h, the cells were washed and deposited on slides by cytocentrifugation. They were fixed in 95% ethanol/5% acetic acid on ice and then stained with 13El. To detect the bound 13E1, the cells were exposed to biotinylated rabbit anti-rat Ig, washed and incubated with alkaline phosphatase labelled avidin. After development, a striking deposition of substrate was seen in the Sf9 cells infected with the ‘IL-4 baculovirus’ but not with the ‘IL-5 baculovirus’ (Fig. 7). In these experiments, 54.1 failed to give a detectable signal (data not shown). The immunohistochemical detection of cytosolic IL-4 in Sf9 cells infected with the ‘IL-4 baculovirus’ could also be detected by flow cytometry using 13El. In this case, the infected cells were washed
0’
Figure 6. of IL-4.
13El
15’
identifies
13EI 60’
a distinct
0’
15’
J41 60’
biosynthetically
0’
15’
labelled
60’
form
Sf9 cells were infected with Ac.MNPV.IL-4. 48 h later, the cells were washed, placed in methionine free medium for 1 h, and then pulsed with 3sS-methionine for 30 mins after which an excess of cold methionine was added. Samples were taken at the end of the pulse and 15 and 60 mins later. Cellular lysates were immunoprecipitated with llB11, 13El or 54.1 and analyzed by SDS-PAGE. The lane between the llBl1 and 13El groups represents molecular weight standards.
Monoclonal
Figure
7.
13El detects
IL-4
in fixed
Sf9 cells infected
with
antibody
to an epitope
on denatured
IL-4
i 541
Ac.MNPV.IL-4.
Sf9 cells were infected with Ac.MNPV.IL-4 or with Ac.MNPV.IL-5. 48 h later, the cells were deposited on slides by cytocentrifugation, fixed with 95% ethanoli5% acetic acid and stained with 13E1 as described in Materials and Methods. Panels A and B are low and high power magnifications of cells infected with Ac.MNPV.IL-5. Panels C and D are comparable magnifications of cells infected with Ac.MNPV.IL-4.
A
Sf9 - AcMNPV.
i
IL-4
Sf9
AcMNPV.
IL-5
No 10 Ab 5;: ;' YantiLTNP ;: .j
;::
.:~3E1
c
..
:.
Figure 8. Flow cytometric cells stained with 13El.
DlO Stimulated with PMA + A23187 No 1”Ab
m
100
101 RELATIVE
102
103
FLUORESCENCE
100
10' RELATIVE
102
103
FLUORESCENCE
104
analysis
of IL-4-producing
Sf9 cells infected with Ac.MNPV.IL-4 (Panel A) or with Ac.MNPV.IL-5 (Panel B) and DlO cells stimulated with PMA and A23187 (Panel C) or unstimulated (Panel D) were fixed in fixation buffer (see Materials and Methods) and incubated with 5 vg/ml of 13El or a monoclonal anti-TNP antibody (54.1) or nothing. After 30 mins, the cells were washed, incubated with biotinylated rabbit anti-rat IgG, washed again and stained with FITC-streptavidin (for the Sf9 cells) or phycoerythrinstreptavidin (for the D10 cells). The cells were then analyzed on a Facscan.
542 I Atasoy et al.
(13El), and two other less well characterized monoclonal antibodies (7Fl and 6HlO), emerged from the immunization of rats with recombinant IL-4 derived from a yeast expression system and subsequent screening using IL-4 adsorbed to plastic plates. Interestingly, 11B 11, a potent neutralizing anti-IL-4 antibody, failed to detect plastic adsorbed IL-4 suggesting that the llBl1 epitope had been destroyed by surface adsorption. Whether this is due to general unfolding of the IL-4 molecule when it adsorbs to surfaces is unknown but others have observed that screening for monoclonal antibodies to other antigens with plastic adsorbed proteins may lead to the detection of reagents that recognize epitopes not observed on the undenatured protein.5 We have concluded that 13El recognizes an epitope expressed on IL-4 molecules that lack biological activity and that fail to express the 11Bll epitope. Our reason for this conclusion is that pre-precipitation of t251-labelled IL-4 with llBl1 depletes molecules that can be subsequently precipitated with 11Bll and removes all biologically active IL-4. However, some molecules can still be precipitated with 13El. Similarly, pre-precipitation with 13El removed molecules that could be precipitated with 13El but some were retained that could be precipitated with 11Bll. Two points need to be made here. Firstly, pre-precipitation with either 13El or 11Bll diminished the fraction of r*sI-IL-4 that could be precipitated with the alternative monoclonal, compared to the precipitation observed in preparations pre-precipitated with the control antibody 54.1. This suggests that some IL-4 molecules may bear both epitopes. Secondly, it should be pointed out that the IL-4 that was used for labelling with 1251 in these experiments had been purified on an 11Bll affinity column. Thus, the molecules of r*sI-IL-4 that were 13El+, llBllwere derived from molecules that had been llBll+. The concept that the 11Bll and 13El epitopes are largely expressed on distinct molecules is strongly supported by studies of biosynthesis of IL-4 in Ac.MNPV.IL-6infected Sf9 cells. In these cells, 13El mainly precipitates an IL-4 species with a molecular size of c.14 500 Da; only a small amount of the 18 000 Da species is immunoprecipitated. By contrast, llBl1 principally brings down a form of IL-4 of 18 000 with virtually no 14 500 material. It should be pointed out that 13El will strongly detect the 18 000 Da form on a Western blot indicating that it does recognize an epitope that can be generated in the 18 000 Da form. Although these results suggest that different conformers of IL-4 express distinct IL-4 epitopes, we do not know what structural changes are important for the expression of the 13El epitope and the loss of the 11Bll epitope. It is unlikely that differences
CYTOKINE,
Vol.
4, No.
6 (November 1992:537-544)
in expression of carbohydrates are critical as 13El and 11Bll can both react with E. c&-derived IL-4; the former in immunoblots and with plate-adsorbed IL-4; the latter by neutralization of IL-4 biological activity. It is striking that differences in expression in different biosynthetically labelled form of IL-4 exist. It is possible that 13El recognizes an incompletely folded version of IL-4 and that it precipitates molecules with such a conformation that exist in the cell; alternatively, it is possible that the 13El epitope is not present in biosynthetically labelled material in situ but that the epitope is generated by denaturation of the molecule upon extraction and that the 14 500 Da form is more susceptible to such denaturation than is the 18 000 Da ‘native’ molecule. We have had some success in using 13El to detect cytosolic IL-4 in both immunofluorescence and alkaline phosphatase assays of fixed cells. In our experience, llBl1 has not given positive results by either method. However, it must be pointed out that others have reported success in the identification of cells with cytoplasmic IL-4 using llB11.9~10 The differences between our studies and theirs might be due to precise conditions of fixation. In our hands, 13El is quite effective in detecting IL-4 expressed by the insect cell line Sf9 infected with Ac.MNPV.IL-4. We have had success both in staining of cytocentrifuge preparations and in flow cytometric analysis. However, analysis of T cells of the DlO line, which produce relatively large amounts of IL-4 in response to stimulation of phorbol esters and calcium ionophore, have yielded variable results. Furthermore, we have thus far not succeeded in detecting IL-4 in normal cell populations enriched for IL-6producing capacity and stimulated with anti-CD3 plus IL-2, which causes secretion of substantial amounts of IL-4. Whether this simply reflects a limitation in the sensitivity of this method or has other explanations has not yet been determined.
MATERIALS AMI METHODS IL-4 Preparations Recombinant murine IL-4 prepared in an E. coli expression system was kindly provided by Dr Alan Levine, Monsanto Central Research, St. Louis, MO. Yeast-derived recombinant IL-4 was provided by Dr Stephen Gillis of Immunex Corp., Seattle, WA. Baculovirus derived-IL-4 was prepared by MS Cynthia Watson, Laboratory of Immunology, NIAID, using the recombinant IL-4 baculovirus Ac.MNPV.IL-4. Amounts of IL-4 are reported in terms of U/ml. Ten U/ml was initially defined as the concentration of IL-4 required for a half-maximal response in an anti-IgM B cell co-stimulation assayiiJ2 and is equivalent to c.O.5 pgiml.
Monoclonal antibody to an epitope on denatured IL-4 I 543
Production of Monoclonal Antibodies Two male Lewis rats were immunized i.p. with 20 pg of purified recombinant yeast-derived murine IL-4 in 0.5 ml of complete Freund’s adjuvant. Two weeks later, they were boosted i.p. with the same amount of IL-4 in incomplete Freund’s adjuvant. One week later, the animals were boosted with 20 pg of IL-4 in PBS. Three days later, the spleens were removed and cell suspensions prepared. Approximately 3.6 x 10srat splenocytes were fused with an equal number of cells of the murine myeloma cell line SP2/0 using 50% polyethylene glycol m.w. 1300 to 1600 (catalog No. P7777, Sigma). Costar 96-well, flat bottom plates were seeded with a cell density of 3 x 105 cells per well in 100 ~1 of RPM1 1640 supplemented with 20% FCS, 2 mM glutamine, 50 PM 2-mercaptoethanol, 100 U/ml penicillin and 100 pgiml streptomycin. The next day, 100 ~1of hypoxanthine-aminopterin-thymidine (Sigma) was added to each well. On day 3 post-fusion 100 ~1 of medium was removed from each well and replaced with an equal amount of fresh 1 X HAT medium. This was repeated on alternating days until day 11 when 100 ~1 of supernatant was removed from each well to assay for the presence of anti-IL-4 antibody. Cells from those wells that were found to be positive by ELISA were cloned by limiting dilution and rescreened for anti-IL-4 activity. Clones from these wells that were still found to be positive were expanded in 24-well plates and then frozen.
Indirect ELBA Immulon-2 microtiter plates (Dynatech, Chantilly, VA) were coated with 50 ~1 of yeast-derived recombinant IL-4 at a concentration of 0.5 pgiml at 37°C in PBS.13 These plates were then washed with 0.15 M NaCl, 0.05% Tween-20 (saline-Tween wash buffer) and then blocked with 100 ~1 per well of a solution consisting of PBS with the following additives: 1% FCS, 10 mM HEPES and 0.025% sodium azide (1% CP), for 60 mins at room temperature. After washing the plates with the saline-Tween wash buffer, 50 ~1 of either hybridoma supernatant, diluted anti-sera or purified antibodies were added for 1 h at room temperature. The plates were then washed and 75 ~1 of a l/2000 dilution of goat-anti-rat IgG (heavy and light chain) alkaline phosphatase conjugate (catalog No. 112-055-003, Jackson Immunoresearch, Avondale, PA) in 1% CP were added to the wells for 1 h at room temperature. After another washing step, 100 1.11of substrate solution containing 4methylumbelliferyl phosphate (5 mg of per 100 ml in the following buffer: 3.71 g Na,CO,, 0.142 g MgC1,.6H,O, and 3.5 ml 10% Na azide in 700 ml dH,O was added and the OD was measured at 450 nm .
Iodination of IL-4 One millicurie of 1251 was added to the following mixture: 20 pg murine rIL-4 (in 50 pl), 50 1.11 Enzymobeads (BioRad), 50 yl 0.104 M PO, buffer pH 7.2, 20 l.112.5% p-D-g!ucose.14 The reaction was carried out at 25°C for 15 mins, after which it was quenched with 25 ~1of 10% sodium azide. The reaction mixture was then applied to a G-25
Sephadex column (prepared beforehand by washing with PBS and blocking non-specific binding with 0.1% BSA) and eluted with PBS. Fractions 4, 5 and 6 had the bulk of radioactivity and were pooled to obtain the 1251-IL-4. BSA (200 pgiml) was added for stabilization and the material was stored at -20°C.
Affinity Column Purification of 13El Rabbit anti-rat Ig (H & L) (Cappel-Organon-Teknika, West Chester, PA) was covalently linked to SepharoseCyanogen Bromide coupled beads. 13El supernatant was then passed through this column, washed and eluted with a Tris-HCl pH 4.0 buffer.
Immunoprecipitation of 12T-IL-4 Protein A-Sepharose (PAS) beads were coated with rabbit anti-rat IgG (H & L) at a final concentration of 2 pg/ml of the polyclonal rabbit antibody for 60 min at room temperature and then washed. The beads were then placed in a 15 ml volume of purified antibodies (13El; 1lBll (anti-IL-4), or J4.1 (anti-trinitrophenyl [TNP]) at 5 pg/ml in 1% CP. After overnight incubation, the beads were washed extensively with PBSiTween. Conjugated beads were then incubated overnight with a lz51-IL-4. After the samples were spun down, the supernatants were removed and stored at -20°C. These supernatants were then divided into three portions and mixed with each of the antibody-bead complexes.
SDS-PAGE Electrophoresisand Immunoblottkg The reactivity of the different antibodies was determined by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and immunoblotting. Differing amounts of recombinant-baculovirus derived IL-4 was treated under reducing conditions in sample buffer containing: 0.062 M Tris, 2.3% SDS, 5% 2-mercaptoethanol, 10% glycerol and electrophoresed on a continuous 12.5% polyacrylamide gel according to the method of Laemmli.15 The proteins were transferred electrophoretically to nitrocellulose (Schleicher and Schuell, Keene, NH) at 4°C in the following buffer: 3.03 g Tris- HCl, 14.4 g glycine, 200 ml methanol and 800 ml dH,O. The nitrocellulose was then incubated for l-2 h in the blocking buffer: 5% non-fat dry milk, 0.01% antifoam A (Sigma) and 0.1% azide. Individual strips of nitrocellulose were then incubated for 1 h at room temperature in llB11, 13El or J4.1 antibody solutions at 5 pg/ml. After extensive washing in PBS/Tween, all the strips were incubated in 5 pg/ml rabbit anti-rat IgG (H & L), (Cappel-Organon Teknika, West Chester, PA, Catalog No. 02134082). Each of the antibody solutions were prepared in the blocking buffer. The protein complexes we& detected by using 1251-Protein A (Amersham, Arlington Heights, IL) at a 1:lOOOdilution in PBS-Tween 20-BSA for 1 h at room temperature. They were then washed with blocking buffer and exposed to X-ray film (Eastman Kodak, Rochester, NY) for 1 h.
IL-4 Proliferation Studies To test the ability of antibodies to block IL-4 stimulation
544 I Atasoy
CYTOKINE,
et al.
Vol. 4, No. 6 (November 1992: 537-544)
of DNA synthesis by the IL-4 responsive T cell line CT.4R, 50 ~1of antibody was added to an equal amount of varying concentrations of recombinant murine IL-4 and incubated in Costar 96-well plates (Cambridge, MA) at 37°C for 1 h. Fifty ~1 of this mixture was then added to Costar plate to which 5 x 103 CT.4R cells per well were added. After 48 h, the plates were then pulsed with3H-thymidine and the cells were harvested 6 h later.
on ice. After washing x2, streptavidin-phycoerythrin or streptavidin-fluorescein (Tago Laboratories, Burlingame, CA) was added. The cells were then washed and analyzed on the FACSCAN (Becton-Dickinson).
Infection of Sfs Cells with Ac.MNPV.IL-4
1. Ohara .I, Paul WE (1985) Production of a monoclonal antibody to and molecular characterization of B-cell stimulatory factor-l. Nature 315:333. 2. Ohara J, Coligan J, Zoon K, Maloy W, Paul W (1987) High-efficiency purification and chemical characterization of B cell stimulatory factor-l/Interleukin 4. J Immunol139:1127-1134. 3. Howard M, Farrar J, Hilfiker B, Johnson K, Takatsu T, Hamaoka T, Paul W (1982) Identification of a T cell-derived B cell growth factor distinct from interleukin-2. J Exp Med 155:914. 4. Ohara J, Lahet S, Inman J, Paul W (1985) Partial purification of murine B-cell stimulatory factor (BSF-1) by HPLC. J Immunol135:2518. 5. Prud’homme JF, Jolivet A, Pichon MR, Savouret JF, Milgrom E (1990) Monoclonal antibodies against native and denatured forms of estrogen-induced breast cancer protein (BCEIlpS2) obtained by expression in Escherichia coli. CancerRes50:2390. 6. Lane RD, Mellgren R, Hegazy M, Gonzalez S, Nepomuceno V, Reimann E, Schlender K (1989) The grid-blot: a procedure for screening large numbers of monoclonal antibodies for specificity to native and denatured proteins. Hybridoma 8:661. 7. Smith J, Olson J, Klapper D (1988) Monoclonal antibodies to denatured ragweed pollen allergan Amb a I: characterization, specificity for the denatured allergen, and utilization for the isolation of immunogenic peptides of Amb a I. Molecular Immunology
Four X 106 Sf9 cells (at 106 cells/ml) were plated in 60 mm Petri dishes in TNM-FH medium containing 10% fetal bovine serum. Cells were allowed to attach for 1 h at 27°C. After 1 h, the medium was removed and 4 x 107 plaque forming units of AcMNPV.IL-4 or AcMNPV.IL-5 was added to each dish. The plates were gently rocked to distribute the viral inoculum and incubated at 27°C for 1 h. The viral inoculum was gently removed after 1 h and 4 ml of TNM-FH medium with 10% fetal bovine serum was added. The cells were incubated at 27°C for 24 or 48 h and then harvested by washing twice in the TNM-FH medium and gently spinning down at 1000 RPM.
Immunocytoplasmic Staining of Sj9 Cells AcMNPV.IL-4 or AcMNPV.IL-5-infected cells were washed X2 in a medium consisting of RPM1 1640 with 2 mM glutamine, 50 FM 2-mercaptoethanol, 100 U/ml penicillin and 100 pgiml streptomycin + 10% FCS (CRPMI). The cells were resuspended in the same medium and cytospins were prepared (c. 50 000 - 2.0 x 105 cells per slide), air dried and fixed with cold 95% ethanol/5% acetic acid for 60 s. The cells were then stained by the Vector ABC Method (Burlingame, CA) using 13El (5 pgiml) for 3c60 mins followed by a biotinylated rabbit anti-rat IgG (H & L). The slides were developed with the Vector Brown-black substrate kit specific for avidin-alkaline phosphatase.
Flow Cytometric Detection of IL-4 Producing Cells One x 107 DlO.G4.1 cells were not stimulated or were stimulated for 12 h at 37°C with phorbol myristate acetate (PMA) (10 r&ml) and the calcium ionophore A23187 (100 rig/ml). Infected Sf9 cells were prepared as outlined above. After being washed with PBS x2, both cell types were centrifuged and resuspended in 250 1.11of a pH 6.8 fixation buffer (BFA) composed of the following: 0.2 mg Na,HP0,/2 H,O ml, 1 mg KH,POdml, 45% (v/v) acetone, 9.25% (v/v) formaldehyde and 45.75% distilled water. After 2 s in BFA at room temperature, the cells were washed twice with chilled (4°C) PBYBSA (0.2% bovine serum albumin in PBS). Primary antibody at a concentration of 5 pg/ml was added and the cells were incubated on ice for 30 min. After washing ~2, a biotinylated secondary antibody, rabbit anti-rat IgG (H & L), 5 @ml (Vector Laboratories, Burlingame, CA), was added for 30 mins
REFERENCES
25:355. 8. Overall ML, Marzuki S, Hertzog PJ (1989) Comparison of different ELISAs for the detection of monoclonal antibodies to human interferon-a. J Immunol Methods 119:27. 9. Kupfer A, Mossmann TR (1991) Polarized expression of cytokinesin cell conjugatesof helper T cells and splenic B cells. Proc Nat1 Acad Sci USA 88:775. 10. SanderB, Cardell S, Heremans H, Andersson U, Moller G (1989) Detection of individual IL-4 and gamma interferon-
producing murine spleencellsafter activation with T-cell mitogens. Scan J Immun 30:315.
11. Howard M, Farrar J, Hilfiker M, Johnson B, Takatsu K, Hamaoka T, Paul, WE (1982) Identification of a T cell derived B cell growth factor distinctfrom interleukin 2. J Exp Med 155:914. 12. Grabstein K, Eiseman J, Mochizuki D, Shanebeck K, Conlon P, Hopp T, March C, Gillis S (1986) Purification to homogeneity of B cell stimulatory factor: a molecule that stimulates proliferation of multiple lymphokine-dependent cell lines. J Exp Med 163:1405. 13. Abrams J, Pearce M (1988) Development of rat anti-mouse interleukin 3 monoclonal antibodies which neutralize bioactivity in vitro. J 1mmuno1140:131. 14. Park L, Friend D, Gillis S, Urdal D (1986) Characterization of the cell surface receptor for granulocyte-macrophage colonystimulating factor. J Biol Chem 261:4177. 15. Laemmli U (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680.
Cynthia
Watson,
William
E. Paul
Monoclonal antibodies to interleukin 4 (IL-4) were generated by immunization with recombinant IL-4 and screening by binding to IL-4 adsorbed to plastic surfaces. Three antibodies were obtained that scored well in this assay and one, 13E1, was studied in detail. It was very effective in detecting IL-4 by Western blotting whereas a neutralizing anti-IL-4 antibody, llB11, was 50-lOO-fold less sensitive as a detecting agent. Sequential immunoprecipitation and biosynthetic labelling studies indicated that the llBl1 and 13El epitopes are largely expressed on different forms of IL-4.13El was able to detect cytosolic IL-4 both by immunohistochemical and flow cytometric analysis of fixed cells. This was routinely successful in an insect cell line (Sf9) expressing large amounts of IL-4 as a result of infection with a recombinant ‘IL-4 baculovirus’. Although stimulated DlO cells could also be shown to express IL-4 in their cytosol, positive results were not obtained in all such studies and we have failed to detect IL-4 production by normal T cells using this method. This antibody may have substantial value in detecting IL-4 by Western blots and as a tool to analyze the biosynthesis of IL-4. With suitable improvement in sensitivity, it also may prove valuable in the detection of IL-4 in the cytosol of individual cells.
The analysis of the function of murine IL-4 has been strikingly clarified by the use of a monoclonal antibody (11Bll) capable of blocking its functions both in vitro and in vivo.l-4 Unfortunately, this antibody has been relatively insensitive as a reagent for immunoblotting and, in our hands, has been of limited use for cytosolic localization of IL-4 in fixed IL-4-producing cells. To this end, it would be very helpful to have additional monoclonal antibodies that were specific for IL-4 but more suitable for the recognition of epitopes found on surface adsorbed IL-4 (for immunoblotting) and in fixed IL-4-producing cells. It has been reported that screening hybridoma cell lines with antigens adsorbed onto plastic surfaces will sometimes result in the selection of monoclonal antibodies that recognize epitopes associated with ‘denatured’
proteins.s-* Such antibodies might be quite useful for the purposes described above. Here we report the derivation of a monoclonal antibody to mouse IL-4 that is extremely efficient in detecting IL-4 by immunoblotting and can bind to IL-4 in the cytosol of insect cells infected with an ‘IL-4-baculovirus’ (Ac.MNPV.IL-4). This antibody fails to neutralize IL-4 activity and recognizes an epitope found on denatured IL-4 but also identifies a unique biosynthetically labelled form of IL-4. It may be of substantial value both in detecting denatured IL-4 and in following the biosynthesis of this molecule.
RESULTS 13El Detects IL-4 coated on Plastic Wells
From the Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, USA. Correspondence to: William E. Paul. Received 13 April 1992; revised and accepted for publication 15 July 1992 @ 1992 Academic Press Limited 1043-4666/92/060537+08 $08.00/O KEY WORDS: monoclonal denatured epitope
CYTOKINE,
Vol.
4, No.
antibody/interleukin
6 (November),
1992:
4, lymphokine,
pp 537-544
Lewis rats were immunized with recombinant IL-4 derived from a yeast expression system. Hybridomas emerging from a fusion of spleen cells from these animals with SP2/0 were screened for their capacity to bind to Immulon 2 plates that had been coated with recombinant IL-4 (0.5 pg/ml). Binding was detected by subsequent addition of an alkaline phosphatase-labelled murine anti-rat Ig and development with 4-methylumbelliferylphosphate. Three 531
538
I Atasoy
CYTOKINE,
et al.
hybridomas (13E1, 7Fl and 6HlO) were selected for further study. After cloning, these were tested for specificity for IL-4 using Immunlon 2 plates that had either not been coated or had been coated with preparations of recombinant IL-4 derived from E. coli, yeast and baculovirus expression systems. Each of the antibodies was positive on all the IL-4 coated plates as detected by development with alkaline phosphatase conjugated goat anti-rat Ig (Fig. 1). In the experiment illustrated in Fig. 1, 13El was tested as a purified Ig while 7Fl and 6HlO were used as tissue culture supernatants. However, in other experiments in which each of the antibodies was tested as tissue culture supernatants, 13El was always superior to 7Fl and 6HlO in binding to IL-4-coated plates. Strikingly, llB11, a monoclonal antibody known to be capable of neutralizing and immunoprecipitating IL-4, failed to bind to IL-6coated plates; it gave no greater signal in the ELISA than either anti-CD8 or anti-TNP. The failure to detect llBl1 binding to IL-4-coated plates was not due to an inability of the alkaline phosphatase goat anti-rat Ig to detect llB11, since it bound well to Immunlon 2 plates coated with 11Bll directly (data not shown). These results indicate that 13El recognizes plate-bound IL-4 while llBl1 does not. In keeping with this, 13El gave a strong signal in detecting purified recombinant baculovirus-derived IL-4 by immunoblotting. It was 50-loo-fold more sensitive than 11Bll; 5 pg/ml of 13El detected 10 ng of IL-4 in a 1 h exposure whereas 11B 11, under the same conditions, was only positive with the 500 and 1000 ng samples of IL-4 (Fig. 2). With longer exposures, 13El could easily detect 1 ng of IL-4. To unequivocally establish that 13El recognized IL-4, recombinant baculovirus-derived IL4 was immunoprecipitated with 11Bll. The immunoprecipitate was boiled in 2.3% SDS under reducing conditions and the resultant solubilized sample was electrophoresed on a 12.5% SDSpolyacrylamide gel. The sample was transferred to
Detection Source
of IL4 by indirect
Recombinant
IL-4
(W3)
Antibody
J4.1
Figure
2.
13El
detects
IL-4
by immunoblotting.
Differing amounts of recombinant IL-4 were electrophoresed in a 12.5% polyacrylamide-SDS gel under reducing conditions, transferred to nitrocellulose and developed with 13E1, llBl1 and the control antibody J4.1 and detected with rabbit anti-rat IgG followed by W-Protein A, as described in Materials and Methods.
nitrocellulose and exposed to 13E1, llBl1 and a control antibody (54.1); 13El clearly detected the material that had been immunoprecipitated by llB11, whereas, under the conditions of this assay, neither llBl1 nor 54.1 gave a detectable signal (Fig. 3).
13El and llBl1 RecognizeEpitopes Not Found on the Same Molecules of IL-4 13El efficiently recognizes IL-4 bound to plastic surfaces or adsorbed onto nictrocellulose membranes whereas 11Bll either fails to bind to such material or does so very inefficiently. By contrast, 11Bll is a potent inhibitor of IL-4 function; 4 pg/ml completely inhibited the capacity of 1000 U/ml (c.O.5 @ml) of IL-4 to stimulate DNA synthesis by CT.4R cells. 13El
ELISA
of rlL4: PBS
E. coli
Yeast
Baculovirus
a;;ij;KK
Figure
o~g~o~g~o~g~o”o”o d
Vol. 4, No. 6 (November 1992: 537-544)
cj
0.
0.
0’
0.
0. 0.D
*denotes purified antIbodIes; rest are supermatants
0.
d
N 0.
m d
-’ ‘:0
1.
13El
detects
plate bound
IL-4
by ELBA.
A series of rat anti-IL-4 and control monoclonal antibodies were incubated on Immulon 2 plates that had been coated with recombinant IL-4 (0.5 pg/ml) derived from E. coli., yeast and baculovirus expression systems, as described in Materials and Methods. Binding was detected with an alkaline phosphatase conjugated goat anti-rat Ig and the fluorescent substrate 4methylumbelliferyl phosphate.
Monoclonal antibody to an epitope on denatured IL-4 I 539
prominently displayed on surface adsorbed IL-4, it is likely that it is a feature of some forms of denatured IL-4. The data in Fig. 5 do show that both 13El and 1lBll pre-precipitated IL-4 contain fewer cpm of 1251that can be precipitated with 11Bll and 13E1, respectively, than does 54.1 pre-precipitated 1251-IL-4.
lmmunopreclpitoting antibody 2 z =!?
lmmunoblotting antibody
: 3
400000 cells
t
Figure 3. Immunoblotting immunoprecipitated with
with 11Bll.
13El
detects
IL-4
that
had
stimulated
with
IL-4
t
7FI
been
IL-4 was immunoprecipitated with llB11, 13El or 54.1. The dissolved immunoprecipitate was electrophoresed on an SDSpolyacrylamide gel, transferred to nitrocellulose and detected by blotting with llB11, 13El or 54.1, followed by rabbit anti-rat IgG and 12s1-Protein A. The arrow indicates the band with the size expected for IL-4. The other bands represent Ig H and L chains of the antibodies used for immunoprecipitation.
at the same concentration failed to inhibit stimulation by as little as 10 U/ml of IL-4; similarly 7Fl and 6HlO also failed to inhibit (Fig. 4). Taken together, these results indicate that 13El reacts with a conformer of IL-4 that is not recognized by llBl1 and, since 13El fails to inhibit IL-4, 13El may not recognize epitopes on active IL-4. To examine in more detail the possibility that the two monoclonal antibodies recognize epitopes on distinct forms of IL-4, IL-4 was labelled with 1251and immunoprecipitated with either 13E1, llBl1 or 54.1. The residual material was divided into three portions, each of which was reprecipitated with one of the monoclonal antibodies. As shown in Fig. 5, llBl1 and 13El precipitated approximately equal amounts of IL-4 that had been precleared with 54.1. By contrast, 11Bll was superior to 13El in precipitating 1*51-IL-4 that had been pre-precipitated with 13El and 13El was superior to 11Bll in precipitation of material that had been pre-precipitated with 11Bll. This result indicates that molecules of IL-4 exist that express the 13El but not the 11Bll epitope and that others express the 11Bll but not the 13El epitopes. In view of the capacity of 11Bll adsorbed to Affigel to remove all biologically active IL-4, molecules that express the 13El but not the 11Bll epitope must be biologically inactive. Since the 13El epitope is
IL-4
Figure IL-4.
4.
13El
to inhibit
fails
(U/ml
1
stimulation
of CT.4R
cells
by
CT.4R cells (5000iwell) were cultured with varying amounts of IL-4 in the presence of 4 pg/ml of various monoclonal anti-IL-4 antibodies or an anti-class II MHC antibody. Cells were pulsed with 3H-thymidine 48 h later and uptake of radioactivity measured after an additional 6 h.
Precipitating Prrnary
ontlbody Secondary I 34.1
13EI
13EI IlBll
J4.1 13EI
11811
IlBll
J4.1 J4.1
l3EI IlBll I 0
I
I IO
Percent
Figure 5. molecules.
13El
and
llBl1
identify
I 20
I
lz51-IL-4
IL-4
preclpltoted
epitopes
on
distinct
W-IL-4 was immunoprecipitated with protein A-Sepharose beads coated with 13E1, 1lBll or 54.1. The residual material was divided into three portions and immunoprecipitated with beads coated with 54.1, 13El or llBl1 and the amounts of IL-4 bound determined by measuring the percent of the 12sI-IL-4 remaining after the first precipitation that was precipitated by the secondary antibody.
540 I Atasoy
CYTOKINE,
et al.
This may reflect the presence of some molecules that express both epitopes or may be due to non-specific losses in the course of immunoprecipitation. To further examine the forms of IL-4 that were recognized by 11Bll and 13E1, we infected Sf9 cells with Ac.MNPV.IL-4 for 48 h, washed the cells, placed them in methionine-free medium for 1 h and then pulsed them with 35S-methionine for 30 mins. An excess of cold methionine was then added. Samples were taken at the end of the pulse and after 15 or 60 mins of chase and cellular lysates were immunoprecipitated with 13E1, llBl1 or 54.1 (Fig. 6). 54.1 failed to precipitate any detectable material. 13El brought down three bands (18 000 Da, 14 500 Da and 13 900 Da) of which the 14 500 Da band was clearly dominant. Each of these bands was prominently seen at the end of the pulse and had diminished considerably in intensity by 60 mins of chase. By contrast llBl1 immunoprecipitated mainly the 18 000 Da band, the intensity of which did not appreciably vary in the course of the chase. In addition, 11Bll also brought down the 13 900 Da band and small amounts of a band of c.16 000, detectable only after 1 h of chase. These results indicate that 13El principally recognizes epitopes on biosynthetically labelled IL-4 with molecular weight of 14 500 while the 11Bll epitope is principally found on a 18 000 Da form of IL-4. The analysis does not allow us to determine whether the form of IL-4 displaying the 13El epitope is processed into a form expressing the 11Bll epitope. The diminished signal seen with 13El after 60 min of pulse could represent further processing of this form of IL-4 or could be due to degradation of this species.
Vol. 4, No. 6 (November 1992: 537-544)
and fixed in suspension with buffered formaldehyde (9.25%)/acetone (45%). They were exposed to 13El or J4.1 followed by biotinylated rabbit anti-rat Ig and FITC-streptavidin. As shown in Fig. 8, the cells infected with IL-4 baculovirus had greater fluorescence when stained with 13El than with 54.1 whereas the cells infected with the IL-5 baculovirus showed little or no enhanced fluorescence with 13El. Similar experiments were done with DlO cells that had been stimulated with PMA and A23187 for 12 h or had been unstimulated. In this case, PE-streptavidin was used in an effort to enhance the fluorescence. Stimulated DlO cells incubated with 13El gave somewhat more fluorescence than cells incubated with 54.1 whereas no difference was observed in the unstimulated cells. However, the staining of fixed D 10 cells with 13El was only found in a proportion of the experiments (2/4) that were carried out. The variables of stimulation or staining procedures that may explain these differences have not been determined.
DISCUSSION We report here the derivation of a monoclonal anti-murine IL-4 antibody that may be of considerable value in the analysis of IL-4 both for its detection by immunoblotting, for its potential to analyze expression of IL-4 immunohistochemically and for its possible use in the analysis of IL-4 biosynthesis. This reagent
IlBll
Immunohistochemical
Detection
of IL-4 with 13El
The capacity of 13El to recognize certain biosynthetically labelled forms of IL-4 as well as non-active IL-4 bound to plastic surfaces suggested that it might detect IL-4 within fixed cells. Accordingly, we infected Sf9 cells with Ac.MNPV.IL-4 or with Ac.MNPV.IL-5. After 48 h, the cells were washed and deposited on slides by cytocentrifugation. They were fixed in 95% ethanol/5% acetic acid on ice and then stained with 13El. To detect the bound 13E1, the cells were exposed to biotinylated rabbit anti-rat Ig, washed and incubated with alkaline phosphatase labelled avidin. After development, a striking deposition of substrate was seen in the Sf9 cells infected with the ‘IL-4 baculovirus’ but not with the ‘IL-5 baculovirus’ (Fig. 7). In these experiments, 54.1 failed to give a detectable signal (data not shown). The immunohistochemical detection of cytosolic IL-4 in Sf9 cells infected with the ‘IL-4 baculovirus’ could also be detected by flow cytometry using 13El. In this case, the infected cells were washed
0’
Figure 6. of IL-4.
13El
15’
identifies
13EI 60’
a distinct
0’
15’
J41 60’
biosynthetically
0’
15’
labelled
60’
form
Sf9 cells were infected with Ac.MNPV.IL-4. 48 h later, the cells were washed, placed in methionine free medium for 1 h, and then pulsed with 3sS-methionine for 30 mins after which an excess of cold methionine was added. Samples were taken at the end of the pulse and 15 and 60 mins later. Cellular lysates were immunoprecipitated with llB11, 13El or 54.1 and analyzed by SDS-PAGE. The lane between the llBl1 and 13El groups represents molecular weight standards.
Monoclonal
Figure
7.
13El detects
IL-4
in fixed
Sf9 cells infected
with
antibody
to an epitope
on denatured
IL-4
i 541
Ac.MNPV.IL-4.
Sf9 cells were infected with Ac.MNPV.IL-4 or with Ac.MNPV.IL-5. 48 h later, the cells were deposited on slides by cytocentrifugation, fixed with 95% ethanoli5% acetic acid and stained with 13E1 as described in Materials and Methods. Panels A and B are low and high power magnifications of cells infected with Ac.MNPV.IL-5. Panels C and D are comparable magnifications of cells infected with Ac.MNPV.IL-4.
A
Sf9 - AcMNPV.
i
IL-4
Sf9
AcMNPV.
IL-5
No 10 Ab 5;: ;' YantiLTNP ;: .j
;::
.:~3E1
c
..
:.
Figure 8. Flow cytometric cells stained with 13El.
DlO Stimulated with PMA + A23187 No 1”Ab
m
100
101 RELATIVE
102
103
FLUORESCENCE
100
10' RELATIVE
102
103
FLUORESCENCE
104
analysis
of IL-4-producing
Sf9 cells infected with Ac.MNPV.IL-4 (Panel A) or with Ac.MNPV.IL-5 (Panel B) and DlO cells stimulated with PMA and A23187 (Panel C) or unstimulated (Panel D) were fixed in fixation buffer (see Materials and Methods) and incubated with 5 vg/ml of 13El or a monoclonal anti-TNP antibody (54.1) or nothing. After 30 mins, the cells were washed, incubated with biotinylated rabbit anti-rat IgG, washed again and stained with FITC-streptavidin (for the Sf9 cells) or phycoerythrinstreptavidin (for the D10 cells). The cells were then analyzed on a Facscan.
542 I Atasoy et al.
(13El), and two other less well characterized monoclonal antibodies (7Fl and 6HlO), emerged from the immunization of rats with recombinant IL-4 derived from a yeast expression system and subsequent screening using IL-4 adsorbed to plastic plates. Interestingly, 11B 11, a potent neutralizing anti-IL-4 antibody, failed to detect plastic adsorbed IL-4 suggesting that the llBl1 epitope had been destroyed by surface adsorption. Whether this is due to general unfolding of the IL-4 molecule when it adsorbs to surfaces is unknown but others have observed that screening for monoclonal antibodies to other antigens with plastic adsorbed proteins may lead to the detection of reagents that recognize epitopes not observed on the undenatured protein.5 We have concluded that 13El recognizes an epitope expressed on IL-4 molecules that lack biological activity and that fail to express the 11Bll epitope. Our reason for this conclusion is that pre-precipitation of t251-labelled IL-4 with llBl1 depletes molecules that can be subsequently precipitated with 11Bll and removes all biologically active IL-4. However, some molecules can still be precipitated with 13El. Similarly, pre-precipitation with 13El removed molecules that could be precipitated with 13El but some were retained that could be precipitated with 11Bll. Two points need to be made here. Firstly, pre-precipitation with either 13El or 11Bll diminished the fraction of r*sI-IL-4 that could be precipitated with the alternative monoclonal, compared to the precipitation observed in preparations pre-precipitated with the control antibody 54.1. This suggests that some IL-4 molecules may bear both epitopes. Secondly, it should be pointed out that the IL-4 that was used for labelling with 1251 in these experiments had been purified on an 11Bll affinity column. Thus, the molecules of r*sI-IL-4 that were 13El+, llBllwere derived from molecules that had been llBll+. The concept that the 11Bll and 13El epitopes are largely expressed on distinct molecules is strongly supported by studies of biosynthesis of IL-4 in Ac.MNPV.IL-6infected Sf9 cells. In these cells, 13El mainly precipitates an IL-4 species with a molecular size of c.14 500 Da; only a small amount of the 18 000 Da species is immunoprecipitated. By contrast, llBl1 principally brings down a form of IL-4 of 18 000 with virtually no 14 500 material. It should be pointed out that 13El will strongly detect the 18 000 Da form on a Western blot indicating that it does recognize an epitope that can be generated in the 18 000 Da form. Although these results suggest that different conformers of IL-4 express distinct IL-4 epitopes, we do not know what structural changes are important for the expression of the 13El epitope and the loss of the 11Bll epitope. It is unlikely that differences
CYTOKINE,
Vol.
4, No.
6 (November 1992:537-544)
in expression of carbohydrates are critical as 13El and 11Bll can both react with E. c&-derived IL-4; the former in immunoblots and with plate-adsorbed IL-4; the latter by neutralization of IL-4 biological activity. It is striking that differences in expression in different biosynthetically labelled form of IL-4 exist. It is possible that 13El recognizes an incompletely folded version of IL-4 and that it precipitates molecules with such a conformation that exist in the cell; alternatively, it is possible that the 13El epitope is not present in biosynthetically labelled material in situ but that the epitope is generated by denaturation of the molecule upon extraction and that the 14 500 Da form is more susceptible to such denaturation than is the 18 000 Da ‘native’ molecule. We have had some success in using 13El to detect cytosolic IL-4 in both immunofluorescence and alkaline phosphatase assays of fixed cells. In our experience, llBl1 has not given positive results by either method. However, it must be pointed out that others have reported success in the identification of cells with cytoplasmic IL-4 using llB11.9~10 The differences between our studies and theirs might be due to precise conditions of fixation. In our hands, 13El is quite effective in detecting IL-4 expressed by the insect cell line Sf9 infected with Ac.MNPV.IL-4. We have had success both in staining of cytocentrifuge preparations and in flow cytometric analysis. However, analysis of T cells of the DlO line, which produce relatively large amounts of IL-4 in response to stimulation of phorbol esters and calcium ionophore, have yielded variable results. Furthermore, we have thus far not succeeded in detecting IL-4 in normal cell populations enriched for IL-6producing capacity and stimulated with anti-CD3 plus IL-2, which causes secretion of substantial amounts of IL-4. Whether this simply reflects a limitation in the sensitivity of this method or has other explanations has not yet been determined.
MATERIALS AMI METHODS IL-4 Preparations Recombinant murine IL-4 prepared in an E. coli expression system was kindly provided by Dr Alan Levine, Monsanto Central Research, St. Louis, MO. Yeast-derived recombinant IL-4 was provided by Dr Stephen Gillis of Immunex Corp., Seattle, WA. Baculovirus derived-IL-4 was prepared by MS Cynthia Watson, Laboratory of Immunology, NIAID, using the recombinant IL-4 baculovirus Ac.MNPV.IL-4. Amounts of IL-4 are reported in terms of U/ml. Ten U/ml was initially defined as the concentration of IL-4 required for a half-maximal response in an anti-IgM B cell co-stimulation assayiiJ2 and is equivalent to c.O.5 pgiml.
Monoclonal antibody to an epitope on denatured IL-4 I 543
Production of Monoclonal Antibodies Two male Lewis rats were immunized i.p. with 20 pg of purified recombinant yeast-derived murine IL-4 in 0.5 ml of complete Freund’s adjuvant. Two weeks later, they were boosted i.p. with the same amount of IL-4 in incomplete Freund’s adjuvant. One week later, the animals were boosted with 20 pg of IL-4 in PBS. Three days later, the spleens were removed and cell suspensions prepared. Approximately 3.6 x 10srat splenocytes were fused with an equal number of cells of the murine myeloma cell line SP2/0 using 50% polyethylene glycol m.w. 1300 to 1600 (catalog No. P7777, Sigma). Costar 96-well, flat bottom plates were seeded with a cell density of 3 x 105 cells per well in 100 ~1 of RPM1 1640 supplemented with 20% FCS, 2 mM glutamine, 50 PM 2-mercaptoethanol, 100 U/ml penicillin and 100 pgiml streptomycin. The next day, 100 ~1of hypoxanthine-aminopterin-thymidine (Sigma) was added to each well. On day 3 post-fusion 100 ~1 of medium was removed from each well and replaced with an equal amount of fresh 1 X HAT medium. This was repeated on alternating days until day 11 when 100 ~1 of supernatant was removed from each well to assay for the presence of anti-IL-4 antibody. Cells from those wells that were found to be positive by ELISA were cloned by limiting dilution and rescreened for anti-IL-4 activity. Clones from these wells that were still found to be positive were expanded in 24-well plates and then frozen.
Indirect ELBA Immulon-2 microtiter plates (Dynatech, Chantilly, VA) were coated with 50 ~1 of yeast-derived recombinant IL-4 at a concentration of 0.5 pgiml at 37°C in PBS.13 These plates were then washed with 0.15 M NaCl, 0.05% Tween-20 (saline-Tween wash buffer) and then blocked with 100 ~1 per well of a solution consisting of PBS with the following additives: 1% FCS, 10 mM HEPES and 0.025% sodium azide (1% CP), for 60 mins at room temperature. After washing the plates with the saline-Tween wash buffer, 50 ~1 of either hybridoma supernatant, diluted anti-sera or purified antibodies were added for 1 h at room temperature. The plates were then washed and 75 ~1 of a l/2000 dilution of goat-anti-rat IgG (heavy and light chain) alkaline phosphatase conjugate (catalog No. 112-055-003, Jackson Immunoresearch, Avondale, PA) in 1% CP were added to the wells for 1 h at room temperature. After another washing step, 100 1.11of substrate solution containing 4methylumbelliferyl phosphate (5 mg of per 100 ml in the following buffer: 3.71 g Na,CO,, 0.142 g MgC1,.6H,O, and 3.5 ml 10% Na azide in 700 ml dH,O was added and the OD was measured at 450 nm .
Iodination of IL-4 One millicurie of 1251 was added to the following mixture: 20 pg murine rIL-4 (in 50 pl), 50 1.11 Enzymobeads (BioRad), 50 yl 0.104 M PO, buffer pH 7.2, 20 l.112.5% p-D-g!ucose.14 The reaction was carried out at 25°C for 15 mins, after which it was quenched with 25 ~1of 10% sodium azide. The reaction mixture was then applied to a G-25
Sephadex column (prepared beforehand by washing with PBS and blocking non-specific binding with 0.1% BSA) and eluted with PBS. Fractions 4, 5 and 6 had the bulk of radioactivity and were pooled to obtain the 1251-IL-4. BSA (200 pgiml) was added for stabilization and the material was stored at -20°C.
Affinity Column Purification of 13El Rabbit anti-rat Ig (H & L) (Cappel-Organon-Teknika, West Chester, PA) was covalently linked to SepharoseCyanogen Bromide coupled beads. 13El supernatant was then passed through this column, washed and eluted with a Tris-HCl pH 4.0 buffer.
Immunoprecipitation of 12T-IL-4 Protein A-Sepharose (PAS) beads were coated with rabbit anti-rat IgG (H & L) at a final concentration of 2 pg/ml of the polyclonal rabbit antibody for 60 min at room temperature and then washed. The beads were then placed in a 15 ml volume of purified antibodies (13El; 1lBll (anti-IL-4), or J4.1 (anti-trinitrophenyl [TNP]) at 5 pg/ml in 1% CP. After overnight incubation, the beads were washed extensively with PBSiTween. Conjugated beads were then incubated overnight with a lz51-IL-4. After the samples were spun down, the supernatants were removed and stored at -20°C. These supernatants were then divided into three portions and mixed with each of the antibody-bead complexes.
SDS-PAGE Electrophoresisand Immunoblottkg The reactivity of the different antibodies was determined by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and immunoblotting. Differing amounts of recombinant-baculovirus derived IL-4 was treated under reducing conditions in sample buffer containing: 0.062 M Tris, 2.3% SDS, 5% 2-mercaptoethanol, 10% glycerol and electrophoresed on a continuous 12.5% polyacrylamide gel according to the method of Laemmli.15 The proteins were transferred electrophoretically to nitrocellulose (Schleicher and Schuell, Keene, NH) at 4°C in the following buffer: 3.03 g Tris- HCl, 14.4 g glycine, 200 ml methanol and 800 ml dH,O. The nitrocellulose was then incubated for l-2 h in the blocking buffer: 5% non-fat dry milk, 0.01% antifoam A (Sigma) and 0.1% azide. Individual strips of nitrocellulose were then incubated for 1 h at room temperature in llB11, 13El or J4.1 antibody solutions at 5 pg/ml. After extensive washing in PBS/Tween, all the strips were incubated in 5 pg/ml rabbit anti-rat IgG (H & L), (Cappel-Organon Teknika, West Chester, PA, Catalog No. 02134082). Each of the antibody solutions were prepared in the blocking buffer. The protein complexes we& detected by using 1251-Protein A (Amersham, Arlington Heights, IL) at a 1:lOOOdilution in PBS-Tween 20-BSA for 1 h at room temperature. They were then washed with blocking buffer and exposed to X-ray film (Eastman Kodak, Rochester, NY) for 1 h.
IL-4 Proliferation Studies To test the ability of antibodies to block IL-4 stimulation
544 I Atasoy
CYTOKINE,
et al.
Vol. 4, No. 6 (November 1992: 537-544)
of DNA synthesis by the IL-4 responsive T cell line CT.4R, 50 ~1of antibody was added to an equal amount of varying concentrations of recombinant murine IL-4 and incubated in Costar 96-well plates (Cambridge, MA) at 37°C for 1 h. Fifty ~1 of this mixture was then added to Costar plate to which 5 x 103 CT.4R cells per well were added. After 48 h, the plates were then pulsed with3H-thymidine and the cells were harvested 6 h later.
on ice. After washing x2, streptavidin-phycoerythrin or streptavidin-fluorescein (Tago Laboratories, Burlingame, CA) was added. The cells were then washed and analyzed on the FACSCAN (Becton-Dickinson).
Infection of Sfs Cells with Ac.MNPV.IL-4
1. Ohara .I, Paul WE (1985) Production of a monoclonal antibody to and molecular characterization of B-cell stimulatory factor-l. Nature 315:333. 2. Ohara J, Coligan J, Zoon K, Maloy W, Paul W (1987) High-efficiency purification and chemical characterization of B cell stimulatory factor-l/Interleukin 4. J Immunol139:1127-1134. 3. Howard M, Farrar J, Hilfiker B, Johnson K, Takatsu T, Hamaoka T, Paul W (1982) Identification of a T cell-derived B cell growth factor distinct from interleukin-2. J Exp Med 155:914. 4. Ohara J, Lahet S, Inman J, Paul W (1985) Partial purification of murine B-cell stimulatory factor (BSF-1) by HPLC. J Immunol135:2518. 5. Prud’homme JF, Jolivet A, Pichon MR, Savouret JF, Milgrom E (1990) Monoclonal antibodies against native and denatured forms of estrogen-induced breast cancer protein (BCEIlpS2) obtained by expression in Escherichia coli. CancerRes50:2390. 6. Lane RD, Mellgren R, Hegazy M, Gonzalez S, Nepomuceno V, Reimann E, Schlender K (1989) The grid-blot: a procedure for screening large numbers of monoclonal antibodies for specificity to native and denatured proteins. Hybridoma 8:661. 7. Smith J, Olson J, Klapper D (1988) Monoclonal antibodies to denatured ragweed pollen allergan Amb a I: characterization, specificity for the denatured allergen, and utilization for the isolation of immunogenic peptides of Amb a I. Molecular Immunology
Four X 106 Sf9 cells (at 106 cells/ml) were plated in 60 mm Petri dishes in TNM-FH medium containing 10% fetal bovine serum. Cells were allowed to attach for 1 h at 27°C. After 1 h, the medium was removed and 4 x 107 plaque forming units of AcMNPV.IL-4 or AcMNPV.IL-5 was added to each dish. The plates were gently rocked to distribute the viral inoculum and incubated at 27°C for 1 h. The viral inoculum was gently removed after 1 h and 4 ml of TNM-FH medium with 10% fetal bovine serum was added. The cells were incubated at 27°C for 24 or 48 h and then harvested by washing twice in the TNM-FH medium and gently spinning down at 1000 RPM.
Immunocytoplasmic Staining of Sj9 Cells AcMNPV.IL-4 or AcMNPV.IL-5-infected cells were washed X2 in a medium consisting of RPM1 1640 with 2 mM glutamine, 50 FM 2-mercaptoethanol, 100 U/ml penicillin and 100 pgiml streptomycin + 10% FCS (CRPMI). The cells were resuspended in the same medium and cytospins were prepared (c. 50 000 - 2.0 x 105 cells per slide), air dried and fixed with cold 95% ethanol/5% acetic acid for 60 s. The cells were then stained by the Vector ABC Method (Burlingame, CA) using 13El (5 pgiml) for 3c60 mins followed by a biotinylated rabbit anti-rat IgG (H & L). The slides were developed with the Vector Brown-black substrate kit specific for avidin-alkaline phosphatase.
Flow Cytometric Detection of IL-4 Producing Cells One x 107 DlO.G4.1 cells were not stimulated or were stimulated for 12 h at 37°C with phorbol myristate acetate (PMA) (10 r&ml) and the calcium ionophore A23187 (100 rig/ml). Infected Sf9 cells were prepared as outlined above. After being washed with PBS x2, both cell types were centrifuged and resuspended in 250 1.11of a pH 6.8 fixation buffer (BFA) composed of the following: 0.2 mg Na,HP0,/2 H,O ml, 1 mg KH,POdml, 45% (v/v) acetone, 9.25% (v/v) formaldehyde and 45.75% distilled water. After 2 s in BFA at room temperature, the cells were washed twice with chilled (4°C) PBYBSA (0.2% bovine serum albumin in PBS). Primary antibody at a concentration of 5 pg/ml was added and the cells were incubated on ice for 30 min. After washing ~2, a biotinylated secondary antibody, rabbit anti-rat IgG (H & L), 5 @ml (Vector Laboratories, Burlingame, CA), was added for 30 mins
REFERENCES
25:355. 8. Overall ML, Marzuki S, Hertzog PJ (1989) Comparison of different ELISAs for the detection of monoclonal antibodies to human interferon-a. J Immunol Methods 119:27. 9. Kupfer A, Mossmann TR (1991) Polarized expression of cytokinesin cell conjugatesof helper T cells and splenic B cells. Proc Nat1 Acad Sci USA 88:775. 10. SanderB, Cardell S, Heremans H, Andersson U, Moller G (1989) Detection of individual IL-4 and gamma interferon-
producing murine spleencellsafter activation with T-cell mitogens. Scan J Immun 30:315.
11. Howard M, Farrar J, Hilfiker M, Johnson B, Takatsu K, Hamaoka T, Paul, WE (1982) Identification of a T cell derived B cell growth factor distinctfrom interleukin 2. J Exp Med 155:914. 12. Grabstein K, Eiseman J, Mochizuki D, Shanebeck K, Conlon P, Hopp T, March C, Gillis S (1986) Purification to homogeneity of B cell stimulatory factor: a molecule that stimulates proliferation of multiple lymphokine-dependent cell lines. J Exp Med 163:1405. 13. Abrams J, Pearce M (1988) Development of rat anti-mouse interleukin 3 monoclonal antibodies which neutralize bioactivity in vitro. J 1mmuno1140:131. 14. Park L, Friend D, Gillis S, Urdal D (1986) Characterization of the cell surface receptor for granulocyte-macrophage colonystimulating factor. J Biol Chem 261:4177. 15. Laemmli U (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680.
