- Email: [email protected]
An enzyme-linked immunosorbent assay (ELISA) for detection of Fcα receptors on isotype-specific T cells
Journal oflmmunological Methods, 85 (1985) 269 277 Elsevier
269
JIM 03728
An Enzyme-Linked Immunosorbent Assay (ELISA) for Detection of Fcc Receptors on Isotype-Specific T Cells 1 T o m o k o Kurita *, Hiroshi K i y o n o **, Suzanne M. M i c h a l e k * and Jerry R. M c G h e e ,.2 Departments of * Microbiology, ** Oral Biology and Preventive Dentistry, The Comprehensible Cancer Center, The Institute of Dental Research, The Universi O, of Alabama at Birmingham, University Station, Birmingham, A L 35294, U.S.A. (Received 16 August 1985, accepted 4 September 1985)
A sensitive and reproducible enzyme-linked immunosorbent assay (ELISA) has been developed using T cell hybridomas as coating antigen, for detection of Fc receptors for IgA (Fcc~R). T-T hybridomas were generated from fusions of F c a R + T cell clones from mouse Peyer's patches with the F c a R Rt.I T lymphoma cell line. The 2 T-T hybridomas (designated Th HA) used here express F e a R as determined by a rosette method and by ELISA. Th HA cells were cultured under conditions for maximum F c a R expression, were added to individual wells of 96-well EIA plates, and were fixed in situ with glutaraldehyde. Plates were incubated with purified mouse monoclonal IgA, IgM or IgGl and were developed with #-galactosidase-coupled goat IgG antibodies specific for mouse heavy chains. Using the ELISA, both Th HA cell lines were shown to express significant levels of FcaR, lower but detectable Fc~tR, and no discernible FcyiR. Interestingly, the rosette assay only allowed detection of receptors for lgA. When splenic lymphocytes were used, good Fc/zR and less Fcc~R expression occurred on these cells as determined by ELISA and rosetting; however, no Fcy~ R cells were detected by either method. Thus, the ELISA is sensitive and reproducible, and allows an objective measurement of FcR expressed on T cells. Key words: E L I S A -
Fc receptors- immunoc~ytoadherence a s s a y - T-Tl~vbridomas
Introduction T lymphocytes may express Fc receptors (FcR) for all 5 immunoglobulin classes, and T cells bearing FcyR, Fc~R or FcaR have been shown to regulate their
i Supported by U.S. Public Health Service Research Grants AI 18958, AI 19674, DE 04217, DE 02670 and New Investigator Research Award AI 21032 (to HK). 2 To whom all correspondence should be sent at the Department of Microbiology, The University of Alabama at Birmingham, University Station, Birmingham, AL 35294, U.S.A. Abbreviations: DNP, dinitrophenyl; ELISA, enzyme-linked immunosorbent assay; FcR, Fc receptors for immunoglobulins; FCS, fetal calf serum; IBF, lg-binding factor; Th HA, T helper cell hybridoma which regulates IgA responses; TNP, trinitrophenyl. 0022-1759/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
270
respective B cell isotype response. With respect to IgA, it has been shown in both mice and humans that T cell subpopulations which bind IgA via a specific Fc receptor (FcaR) occur (Strober et al., 1978; Lum et al., 1979). In some cases, F c a R + T cells exhibit a suppressor phenotype and inhibit the IgA response. For example, mice bearing IgA myeloma have increased numbers of F c a R + T cells of a suppressor phenotype (Hoover and Lynch, 1980), and these T cells inhibit IgA synthesis in vitro (Hoover and Lynch, 1983). Furthermore, an F c a R + T hybridoma cell line and concanavalin A-activated F c a R + T cells, upon culture with IgA, release IgA-binding factor(s) (IBFa) which suppress !gA synthesis in pokeweed mitogen (PWM)-stimulated cultures (Yodoi et al., 1983; Adachi et al., 1984). In other studies, F c a R + T cells from human peripheral blood mononuclear cell cultures selectively enhanced IgA responses when added to B cell cultures and stimulated with PWM (Endoh et al., 1981). Our group has characterized cloned T helper (Th) cells from mouse Peyer's patches, which mainly promote antigen-specific IgA responses (PP Th A cells) (Kiyono et al., 1982). These PP Th A cells express Fcc~R and collaborate with committed, surface IgA + B cells for IgA responses (Kiyono et al., 1984a). The F c a R is important in the IgA response, since preincubation of PP Th A cells with purified IgA, but not IgM or IgG, completely blocked this isotype-specific response (Kiyono et al., 1984b). We have generated 85 T-T hybridoma cell lines by fusion of PP Th A cell clones with the R1.1 lymphoma. Approximately 20% of these cell lines were Fct~R +, as determined by a rosette assay, and the cells produced IgA binding factors (IBFo0 which regulated IgA responses (Kiyono et al., 1985). Classically, FcR are detected by immunocytoadherence using immunoglobulincoated erythrocytes, or by modifications of this procedure. Although this rosetting method is sensitive and reproducible, it is subjective and cumbersome to perform when large numbers of samples require analysis. It would be of benefit to have alternate assays for detection of cells with FcR, which could be performed on a large number of samples with a minimum of experimental variation. In the present study, we report a sensitive and reproducible ELISA, which employs Fcc~R + T cell coated plates for detection of isotype-specific FcR expression.
Materials and Methods
T cell lines, cell fusions and growth conditions The RI.! T lymphoma cell line, which expresses Thy 1.2, TL and H-2 k, was maintained in RPMI 1640 (Grand Island Biological Co. (Gibco), Grand Island, NY) containing 2 mM L-glutamine, antibiotics (penicillin 100 U / m l , streptomycin 100 ~ g / m l and gentamicin 50 ~ g / m l ) and 10% fetal calf serum (FCS) (complete medium). The 2 T-T hybridomas used in this study were derived from a fusion of RI.1 T lymphoma with a cloned PP Th A cell (clone no. 1; Thy 1.2 +, Lyt-1 +, 2 , H-2 k) which supported IgM and mainly IgA responses (Kiyono et al., 1982, 1984a). Briefly, equal numbers of PP Th A no. 1 and RI.1 cells were pelleted together and fused by adding 1 ml of 50% polyethylene glycol in RPMI 1640. Following 1 min of
271
incubation (37°C), the cell suspension was slowly diluted, washed and resuspended in medium containing 0.01 M hypoxanthine, 4 × 10 -7 M aminopterin and 1.6 × 10 6 M thymidine. After 3-4 weeks of growth, the media was gradually replaced with complete medium. Two hybrids, designated Th HA t no. 9 and no. 10 (Thy 1.2 +, H-2k), which were shown previously to express F c a R (Kiyono et al., 1985), were used in the studies reported here.
Preparation of spleen cells Spleens were aseptically removed from C 3 H / H e N mice and single cell suspensions were prepared as previously described (Kiyono et al., 1985). Briefly, spleens were gently teased through sterile stainless steel screens into minimal essential medium (MEM; Gibco) containing sodium bicarbonate, nonessential amino acids, sodium pyruvate, L-glutamine and gentamicin. Cell clumps and tissue debris were removed by gravity sedimentation. Supernatant fluid containing single cells was removed, cells were washed in MEM and then were resuspended in appropriate media. To obtain the splenic lymphocyte population, spleen cells were further purified by separation on a Ficoll-Hypaque gradient.
Purified mouse immunoglobulins and anti-heavy chain antibodies IgA (MOPC 315), IgM (H5-3) and IgG1 (H35-3) anti-dinitrophenyl (DNP) monoclonal antibodies (the latter 2 antibodies were kindly provided by Drs. J.F. Kearney and J.O. Phillips, UAB) were purified by use of DNP-affinity columns (Kiyono et al., 1984b). MOPC 104E IgM (anti-dextran) was purified by affinity chromatography on a dextran B1355 Sepharose column (Kiyono et al., 1984b). Additional purified MOPC 104E IgM was kindly provided by Dr. V. Ghanta, UAB. Purified mouse IgG1 (MOPC 21) was purchased from Bionetics Laboratory Products (Kensington, MD). All purified monoclonal immunoglobulins were tested for purity by Ouchterlony, immunoelectrophoresis and radioimmunoassay prior to their use. Goat anti-mouse heavy chain reagents (anti-a, anti-/~ or anti-),) were obtained commercially (Bionetics), and tested with a battery of mouse myeloma immunoglobulins (Igs) and kappa and lambda chains for their specificity for heavy (H) chain determinants. These goat IgG anti-H chain antibodies were coupled with /3-galactosidase (O'Sullivan et al., 1979) for use in the ELISA (see below).
Immunocytoadherence assay An immunocytoadherence assay was performed in parallel with the ELISA for analysis of FcR expression on Th HA 1 no. 9 and no. 10 cells and on spleen cells. The RI.1 T lymphoma does not express FcR, and served as a negative control. Trinitrophenyl (TNP)-sensitized sheep erythrocytes (TNP-SRBC) were prepared as previously described (Rittenberg and Pratt, 1969). Splenic lymphocytes or T cells ( 5 ) < 10 6 cells/ml) were incubated with 100-200 ~ g / m l of monoclonal anti-DNP antibodies (MOPC 315 IgA, H5-3 IgM or H35-3 IgG1) at 4°C for 1 h. Cells were washed 3 times with RPMI 1640 containing 2% FCS and the pellet was gently resuspended with a 0.5% suspension of TNP-SRBC. Following incubation at 4°C for 30 rain, cells binding 5 or more red cells were counted as positive. Between 1000 and 1500 cells were scored in each experiment (representing 10-15 microscope fields).
272
Enzyme-linked immunosorbent assay for FeaR To obtain Th HA cells expressing maximum Fca receptors, cultures were washed (1200 r p m / 1 0 min), the cells were fractionated on Ficoll-Hypaque gradients and viable cells were recultured overnight in RPMI 1640 with 20% FCS. Th HA cells or spleen cells were extensively washed (4-6 times) with phosphate-buffered saline (PBS, pH 7.4), and appropriate cell numbers (in 100 /.tl) were added to Linbro flat-bottom EIA plates (Flow Laboratories, McLean, VA). Plates were centrifuged (1000 r p m / 5 rain), and 100 ~1 of 0.01% glutaraldehyde in PBS was added (final glutaraldehyde concentration was 0.005%) and were incubated for 30 min at 25°C. Wells were extensively washed with PBS (5-6 times) and then incubated with PBS containing 1% bovine serum albumin (BSA) for 90 min at 37°C. The contents of the wells were expelled and the wells were washed with PBS (2 times). To each well was added 100 /zl of either purified IgA, IgM or IgG1 (1-20 # g / m l ) and plates were incubated for 90 rain at 37°C. In 1 series of experiments, different amounts of IgA (i 20 /~g/ml) were added to each well as described above. Wells were then extensively washed (4-6 times) with PBS and 100 /.tl of appropriately diluted, /3-galactosidase-labeled anti-a, or anti-~ or anti-], were added to the wells. Anti-a, -~ and -], antibodies diluted to 1 : 250 gave maximum ELISA values, and these antisera dilutions were used in all experiments. Following overnight incubation at 4°C, the plates were washed 4 additional times with PBS. Substrate was prepared by dissolving o-nitrophenyl-fl-D-galactopyranoside in Alpase buffer (pH 7.5) (1 m g / m l ) and 100/~1 was added to each well. Plates were incubated at 37°C (in the dark) and color development was read at 414 nm using a Titertek Multiscan photometer (Flow) after 30, 60 and 120 min. Statistics Rosette formation is expressed as the range per triplicate determinations. The significance of difference between means of ELISA values was determined by the Student's t-test.
Results
Optimal conditions for FeaR detection by ELISA In our initial experiments, Th HA cells were fixed with various concentrations of glutaraldehyde (0.001-0.1%) prior to their use as coating antigen for plates. Generally, the fixed cells were poorly adherent, and more consistent results were obtained when living cells were used to coat the wells. Subsequently, Th HA cells were added to plates, spun at low speed and then fixed with 0.01% glutaraldehyde in PBS. Optimal ELISA values were obtained using 1 x l0 s Th HA cells/well (Fig. l); however, as few as 5 x 10 4 ceils also gave consistent and reproducible values. Generally, 5 x 104 to 5 x 10 s cells/well yielded ELISA values ranging from 0.1 to 0.15, when analyzed with lgA and developed with labelled anti-a (Fig. 1). In order to determine the optimal concentration of IgA required for maximum detection of F c a R by ELISA, microplates were coated with Th HA~ no. 9 or no. 10
273
~=
0.20 L Th HA, # 9
-
0.15
w
010 0.05 i
OI
05
i
i
I0
20
Th HA C E L L S / W E L L ( X l O
,J 0
5)
Fig. 1. Concentration of Th HA cells used as coating antigen for ELISA. EIA wells were coated with cells at the concentrations indicated and were fixed with 0.005% glutaraldehyde in situ. Wells were incubated with purified MOPC 315 IgA (0.5 fig/well) and color was developed after addition of labeled anti-a antibodies.
cells (1 x 105 cells/well) and different amounts of MOPC 315 IgA were added (0.1-2.0 ~g/well). A linear relationship between IgA added and increase in absorbance was observed until 0.5 fig/well of IgA was used (Fig. 2). Similar titrations were performed with IgM and IgG1 in wells coated with splenic lymphocytes (see below), and optimal values were also obtained with 0.5 ~g of Ig. In all subsequent experiments, the ELISA was performed with wells coated with 1 x 105 cells and with addition of 0.5/~g of the appropriate Ig.
020 Th HA, # 9
c ,i k3
Z <:[ rt~ rr 0 o9 m
015
010
T
005
o,
o12
015
,io
20
o
tgA CONCENTRATION (~g/well)
Fig. 2. Determination of optimal IgA levels required for maximum detection of FcaR. Wells were coated with Th HA cells (1 x 105 cells/well), and were incubated with the indicated amount of purified MOPC 315 lgA. Color was developed after the addition of fl-galactosidase-labeled anti-a antibodies.
274 TABLE I DETECTION OF Fca RECEPTORS BY ELISA a T cell hybridomas were used as coating antigen and purified mouse IgA and fl-galactosidase-labeled anti-a were added for measurement of binding of IgA to Fca receptors. Experiment number
Optical density (414 nm) b (T cell line used) Th HA 1 no. 9
Th HA I no. 10
RI.I
1 2 3 4 5 6 7 8 9 10
0.170 0.127 0.166 0.143 0.152 0.197 0.184 0.150 0.171 0.149
0.100 0.121 0.158 0.150 0.124 0.107 0.143 0.155 0.143 0.129
0 0 0 0.046 0.083 0 0 0.078 0 0.046
Mean _+SEM
0.161 _+0.016 (P < 0.001)
0.133 _+0.006 (P < 0.001)
0.026 + 0.001
EIA plates were coated with Th HA cells (1 × l0 s cells/well), fixed with 0.005% glutaraldehyde and next incubated with purified MOPC 315 IgA (0.5 tLg/well) for 90 min at 37°C. /3-Galactosidase-labeled goat anti-mouse a chain was added to each well, the plates were incubated overnight at 4°C, and then were extensively washed. Color was developed with substrate and the OD determined at 414 nm. h Significance of values for Fca receptors on Th HA hybridomas when compared with the non-FcaR bearing RI.1 T cell line as determined by the Student's t test.
Comparison of ELISA and rosette assay for FcR expression T h e E L I S A was quite reproducible for detection of F c a R expression, and both Th HA] no. 9 a n d no. 10 consistently exhibited these receptors (Table I). In contrast, the R1.1 T l y m p h o m a does not exhibit F c R by E L I S A (Table I). In all experiments, the E L I S A values for F c a R expression on Th H A cells were significantly higher ( P < 0.001) when c o m p a r e d with RI.1 T cells. O u r previous studies ( K i y o n o et al., 1985) with Th H A cells consistently showed that 30-40% of these cells express F c a R as d e t e r m i n e d by rosetting. However, neither Fc/~R nor Fc'c]R could be detected by this m e t h o d , and these results were verified by i m m u n o c y t o a d h e r e n c e in the present study (Table II). As indicated above, the E L I S A clearly establishes the presence of F c a R on these Th H A cell lines. In addition, when wells coated with Th H A cells were i n c u b a t e d with I g M and d e v e l o p e d with fl-galactosidase-labeled anti-/~, low, but significant a b s o r b a n c e values were noted. In all e x p e r i m e n ts with I g M and anti-#, plates coated with RI.1 T l y m p h o m a cells failed to develop a color reaction. This study clearly implies that Th H A cells also express low levels of F c ~ R , and this finding has been verified by flow c y t o m e t r y using purified IgM and F I T C - l a b e l e d anti-~ ( m an u scr i p t in preparation). O n the other hand, Th H A cells did not exhibit F c y R as d e t e r m i n e d by E L I S A or rosetting using either purified IgG1 or serum IgG.
275 TABLE I1 COMPARISON OF IMMUNOCYTOADHERENCE AND ELISA TO DETECT Fc RECEPTOR EXPRESSION a T cell hybridomas and splenic lymphocytes were examined for FcR by incubation with IgA or IgM or IgG1 (anti-DNP) and rosette formation with TNP-SRBC, or by ELISA using either T cells or splenic lymphocytes as coating antigen. Cells used
lsotype-specificityof Fc receptors Rosette formation (%) ~
Optical density (414 nm) b IgM
IgG1
IgA
Th HA t no. 9 Th HA I no. 10
0.025_+0.003 0.024_+0.004
0 0
0.136_+0.008 0.119+_0.006
Splenic lymphocytes
0.125 _+0.009
0
0.056 -+0.001
R1.1
0
0
0
IgM 0 0 19-26 0
IgG1
IgA
0 0
31-34 32-39
0
1- 5
0
0
See Table I legend for details of the ELISA. For the immunocytoadherenceassay, Th HA cells or spleen cells were reacted with purified monoclonal lgM or IgG1 or IgA anti-DNP antibody for 60 min at 37°C. Rosettes were formed after addition of TNP-SRBC. Th HA or spleen cells binding 5 or more erythrocytes were counted as positive. b Mean values were obtained from quadruplicate determinations/experiment and 3 separate experiments. ~"The range of triplicate determinations of 3 separate experiments is presented.
I n order to confirm that E L I S A detects FcR, plates coated with splenic l y m p h o cytes were i n c u b a t e d with purified IgM, IgG1 or IgA a n d were developed. Using spleen cells, higher E L I S A values were seen for Fc/~R expression than for F c a R (Table II). As detected by i m m u n o c y t o a d h e r e n c e , splenic lymphocytes displayed Fc/~R and small percentages of F c a R (Table II), clearly c o n f i r m i n g the E L I S A results. Fc~,IR were not detectable o n splenic lymphocytes with either assay, suggesting the absence of cells which could b i n d IgG1. Nevertheless, splenic lymphocytes exhibited F c T R as determined by E L I S A (0.114_+0.002) when lymphocyte-coated wells were i n c u b a t e d with purified serum I g G a n d developed with fl-galactosidase-labeled anti-~'.
Discussion In this study, we have described the c o n d i t i o n s for use of an E L I S A to measure F c R expression o n T cell lines. The method is sensitive, since wells coated with 5 x 104-1 × 105 T h H A cells, when i n c u b a t e d with 0.5/~g of IgA, consistently gave E L I S A absorbances of 0.1 or greater, which indicate that F c a R can be reliably detected by this method. The E L I S A could also be used to measure F c a R present on other cells such as splenic lymphocytes. The assay could also easily be a d a p t e d to measure soluble F c a R , by p r e i n c u b a t i o n with purified IgA, prior to its a d d i t i o n to microplate wells. I n addition, the method will allow screening of m o n o c l o n a l
276 antibodies to FcaR, by addition of supernatants of putative clones to wells prior to development with IgA and labeled anti-a antibody. The Th HA cell lines used in this study had previously been shown to bear F c a R as determined by the rosette assay (Kiyono et al., 1985). By using immunocytoadherence, we failed to detect Fc/~R or FcTR on these Th HA cell lines. Somewhat surprisingly, however, the ELISA allowed detection of low levels of Fc#R on both T cell lines, a finding which was subsequently confirmed by FACS analysis with cells incubated with IgM and stained with FITC-anti-/,t reagent (manuscript in preparation). Although we have not directly compared the sensitivity of the ELISA with that of the immunocytoadherence assay, we feel that the ELISA is at least as sensitive as the rosette method. It was important to establish that the ELISA could detect FcR of all 3 major isotypes. To this end, splenic lymphocytes were used as coating antigen, and F c # R and to a lesser extent FcaR, were readily detectable both by ELISA and immunocytoadherence. Splenic lymphocytes did not bear FcT~R; however, purified serum I g G allowed detection of FcTR, suggesting that these receptors bind to other IgG subclasses. Collectively, our results with Th HA cells and splenic lymphocytes suggest that the ELISA can be reliably used for detection of FcR for all Ig classes. Generally, 2 types of methods have been used to analyze FcR on lymphocytes, i.e., methods which detect the receptor and those which both detect FcR and allow enrichment of these cells from total lymphoid cell populations. Antigen-antibody complexes readily bind FcR on both T and B cells (Stout and Herzenberg, 1975; Dickler, 1976) and aggregated immunoglobulin has also been widely used (Dickler, 1976). Both methods are especially useful for FcTR detection, since native IgG has a lower affinity for FcTR than multivalent forms of Fc present in immune complexes or in aggregates. This may explain why spleen cells bind purified serum IgG, which may contain multivalent molecules, versus a lack of binding of cells to purified monoclonal IgG1. It should be noted that use of FITC-conjugated Igs with these methods, provides a useful way to separate FcR + cells by use of flow cytometry (Stout and Herzenberg, 1975). Rosette formation allows detection of FcR for particular isotypes as well as enrichment of this cell type by separation on Ficoll density gradients. The rosette method has been effectively adapted for enrichment of F c ~ R (Moretta et al., 1976), FcTR (Fridman et al., 1981), Fc~R (Spiegelberg, 1981), and F c a R (Hoover et al., 1981). The ELISA, as reported here would have more limited application than the methods described above. However, the ELISA could be quantitative if the average numbers of Fcc~R could be determined on each cell. We are presently using radiolabeled IgA bound to Th HA cells to determine the number of molecules which bind to F c a R + Th HA cells. In subsequent experiments we will then use the Th HA cell preparations in the ELISA to estimate the total numbers of FcctR present/well. This should allow us to establish a quantitative method for measurement of FcR expression on T cell populations by the ELISA. Experiments of this nature are underway, and will be the subject of a future paper.
277
Acknowledgements W e t h a n k D r s . D a w n C o l w e l l , J o h n E l d r i d g e a n d K a z u o K o m i y a m a for a s s e s s m e n t o f this w o r k a n d m a n u s c r i p t , a n d Ms. B e t t y W e l l s f o r t y p i n g t h e p a p e r .
References Adachi, M., J. Yodoi. N. Noro, T. Masuda and H. Uchino, 1984, J. lmmunol. 133, 65. Dickler, H.B., 1976, Adv. lmmunol. 24, 167. Endoh. M., H. Sakai, Y. Nomoto, Y. Tomino and H. Kaneshige, 1981, J. Immunol. 127, 2612. Fridman, W.H., C. Rabourdin, C. Neauport-Sautes and R.H. Gisler, 1981, lmmunol. Rev. 56, 51. Hoover, R.G. and R.G. Lynch, 1980, J. Immunol. 125, 1280. Hoover, R.G. and R.G. Lynch, 1983, J. Immunol. 130, 521. Hoover, R.G., H.M. Gebel, B.K. Dieckgraefe, S. Hickman, N.F. Rebbe, N. Hirayama, Z. Ovary and R.G. Lynch, 1981, Immunol. Rev. 56, 115. Kiyono, H., J.R. McGhee, L.M. Mosteller, J.H. Eldridge, W.J. Koopman, J,F. Kearney and S.M. Michalek, 1982, J. Exp. Med. 156, 1115. Kiyono, H.. M.D. Cooper, J.F. Kearney, L.M. Mosteller, S.M. Michalek, W.J. Koopman and J.R. McGhee, 1984a, J. Exp. Med. 159, 798. Kiyono, H., J.O. Phillips, D.E. Colwell, S.M. Michalek, W.J. Koopman and J.R. McGhee, 1984b, J. lmmunol. 133, 1087. Kiyono, H., L.M. Mosteller-Barnum, A.M. Pitts, S.I. Williamson, S.M. Michalek and J.R. McGhee, 1985, J. Exp. Med. 161, 731. Lure, R.G., A.V. Muchmore, D. Keren, J. Decker, I. Koski, W. Strober and R.M. Blaese, 1979, J. lmmunol. 122, 65. Moretta, L.M., M. Ferrarini, M.C. Mingari, A. Moretta and S.R. Webb, 1976, J. Immunol. 117, 2171. O'Sullivan, M.J., E. Gnemmi, A.D. Simmonds, G. Chieregatti, E. Heyderman, J.W. Bridges and V. Marks, 1979, J. Immunol. Methods 31,247. Rittenberg, M.B. and K.L. Pratt, 1969, Proc. Soc. Exp. Biol. Med. 132, 575. Spiegelberg, H.L., 1981, Immunol. Rev. 56, 199. Stout, R.D. and L.A. Herzenberg, 1975, J. Exp. Med. 142, 1041. Strober, W.. N.E. Hague, L.G. Lum and P.A. Henkart, 1979, J. Immunol. 121, 2440. Yodoi, J., M. Adachi, K. Teshigawara, M. Miyama-lnaba, T. Masuda and W.H. Fridman, 1983, J. Immunol. 131, 303.
269
JIM 03728
An Enzyme-Linked Immunosorbent Assay (ELISA) for Detection of Fcc Receptors on Isotype-Specific T Cells 1 T o m o k o Kurita *, Hiroshi K i y o n o **, Suzanne M. M i c h a l e k * and Jerry R. M c G h e e ,.2 Departments of * Microbiology, ** Oral Biology and Preventive Dentistry, The Comprehensible Cancer Center, The Institute of Dental Research, The Universi O, of Alabama at Birmingham, University Station, Birmingham, A L 35294, U.S.A. (Received 16 August 1985, accepted 4 September 1985)
A sensitive and reproducible enzyme-linked immunosorbent assay (ELISA) has been developed using T cell hybridomas as coating antigen, for detection of Fc receptors for IgA (Fcc~R). T-T hybridomas were generated from fusions of F c a R + T cell clones from mouse Peyer's patches with the F c a R Rt.I T lymphoma cell line. The 2 T-T hybridomas (designated Th HA) used here express F e a R as determined by a rosette method and by ELISA. Th HA cells were cultured under conditions for maximum F c a R expression, were added to individual wells of 96-well EIA plates, and were fixed in situ with glutaraldehyde. Plates were incubated with purified mouse monoclonal IgA, IgM or IgGl and were developed with #-galactosidase-coupled goat IgG antibodies specific for mouse heavy chains. Using the ELISA, both Th HA cell lines were shown to express significant levels of FcaR, lower but detectable Fc~tR, and no discernible FcyiR. Interestingly, the rosette assay only allowed detection of receptors for lgA. When splenic lymphocytes were used, good Fc/zR and less Fcc~R expression occurred on these cells as determined by ELISA and rosetting; however, no Fcy~ R cells were detected by either method. Thus, the ELISA is sensitive and reproducible, and allows an objective measurement of FcR expressed on T cells. Key words: E L I S A -
Fc receptors- immunoc~ytoadherence a s s a y - T-Tl~vbridomas
Introduction T lymphocytes may express Fc receptors (FcR) for all 5 immunoglobulin classes, and T cells bearing FcyR, Fc~R or FcaR have been shown to regulate their
i Supported by U.S. Public Health Service Research Grants AI 18958, AI 19674, DE 04217, DE 02670 and New Investigator Research Award AI 21032 (to HK). 2 To whom all correspondence should be sent at the Department of Microbiology, The University of Alabama at Birmingham, University Station, Birmingham, AL 35294, U.S.A. Abbreviations: DNP, dinitrophenyl; ELISA, enzyme-linked immunosorbent assay; FcR, Fc receptors for immunoglobulins; FCS, fetal calf serum; IBF, lg-binding factor; Th HA, T helper cell hybridoma which regulates IgA responses; TNP, trinitrophenyl. 0022-1759/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
270
respective B cell isotype response. With respect to IgA, it has been shown in both mice and humans that T cell subpopulations which bind IgA via a specific Fc receptor (FcaR) occur (Strober et al., 1978; Lum et al., 1979). In some cases, F c a R + T cells exhibit a suppressor phenotype and inhibit the IgA response. For example, mice bearing IgA myeloma have increased numbers of F c a R + T cells of a suppressor phenotype (Hoover and Lynch, 1980), and these T cells inhibit IgA synthesis in vitro (Hoover and Lynch, 1983). Furthermore, an F c a R + T hybridoma cell line and concanavalin A-activated F c a R + T cells, upon culture with IgA, release IgA-binding factor(s) (IBFa) which suppress !gA synthesis in pokeweed mitogen (PWM)-stimulated cultures (Yodoi et al., 1983; Adachi et al., 1984). In other studies, F c a R + T cells from human peripheral blood mononuclear cell cultures selectively enhanced IgA responses when added to B cell cultures and stimulated with PWM (Endoh et al., 1981). Our group has characterized cloned T helper (Th) cells from mouse Peyer's patches, which mainly promote antigen-specific IgA responses (PP Th A cells) (Kiyono et al., 1982). These PP Th A cells express Fcc~R and collaborate with committed, surface IgA + B cells for IgA responses (Kiyono et al., 1984a). The F c a R is important in the IgA response, since preincubation of PP Th A cells with purified IgA, but not IgM or IgG, completely blocked this isotype-specific response (Kiyono et al., 1984b). We have generated 85 T-T hybridoma cell lines by fusion of PP Th A cell clones with the R1.1 lymphoma. Approximately 20% of these cell lines were Fct~R +, as determined by a rosette assay, and the cells produced IgA binding factors (IBFo0 which regulated IgA responses (Kiyono et al., 1985). Classically, FcR are detected by immunocytoadherence using immunoglobulincoated erythrocytes, or by modifications of this procedure. Although this rosetting method is sensitive and reproducible, it is subjective and cumbersome to perform when large numbers of samples require analysis. It would be of benefit to have alternate assays for detection of cells with FcR, which could be performed on a large number of samples with a minimum of experimental variation. In the present study, we report a sensitive and reproducible ELISA, which employs Fcc~R + T cell coated plates for detection of isotype-specific FcR expression.
Materials and Methods
T cell lines, cell fusions and growth conditions The RI.! T lymphoma cell line, which expresses Thy 1.2, TL and H-2 k, was maintained in RPMI 1640 (Grand Island Biological Co. (Gibco), Grand Island, NY) containing 2 mM L-glutamine, antibiotics (penicillin 100 U / m l , streptomycin 100 ~ g / m l and gentamicin 50 ~ g / m l ) and 10% fetal calf serum (FCS) (complete medium). The 2 T-T hybridomas used in this study were derived from a fusion of RI.1 T lymphoma with a cloned PP Th A cell (clone no. 1; Thy 1.2 +, Lyt-1 +, 2 , H-2 k) which supported IgM and mainly IgA responses (Kiyono et al., 1982, 1984a). Briefly, equal numbers of PP Th A no. 1 and RI.1 cells were pelleted together and fused by adding 1 ml of 50% polyethylene glycol in RPMI 1640. Following 1 min of
271
incubation (37°C), the cell suspension was slowly diluted, washed and resuspended in medium containing 0.01 M hypoxanthine, 4 × 10 -7 M aminopterin and 1.6 × 10 6 M thymidine. After 3-4 weeks of growth, the media was gradually replaced with complete medium. Two hybrids, designated Th HA t no. 9 and no. 10 (Thy 1.2 +, H-2k), which were shown previously to express F c a R (Kiyono et al., 1985), were used in the studies reported here.
Preparation of spleen cells Spleens were aseptically removed from C 3 H / H e N mice and single cell suspensions were prepared as previously described (Kiyono et al., 1985). Briefly, spleens were gently teased through sterile stainless steel screens into minimal essential medium (MEM; Gibco) containing sodium bicarbonate, nonessential amino acids, sodium pyruvate, L-glutamine and gentamicin. Cell clumps and tissue debris were removed by gravity sedimentation. Supernatant fluid containing single cells was removed, cells were washed in MEM and then were resuspended in appropriate media. To obtain the splenic lymphocyte population, spleen cells were further purified by separation on a Ficoll-Hypaque gradient.
Purified mouse immunoglobulins and anti-heavy chain antibodies IgA (MOPC 315), IgM (H5-3) and IgG1 (H35-3) anti-dinitrophenyl (DNP) monoclonal antibodies (the latter 2 antibodies were kindly provided by Drs. J.F. Kearney and J.O. Phillips, UAB) were purified by use of DNP-affinity columns (Kiyono et al., 1984b). MOPC 104E IgM (anti-dextran) was purified by affinity chromatography on a dextran B1355 Sepharose column (Kiyono et al., 1984b). Additional purified MOPC 104E IgM was kindly provided by Dr. V. Ghanta, UAB. Purified mouse IgG1 (MOPC 21) was purchased from Bionetics Laboratory Products (Kensington, MD). All purified monoclonal immunoglobulins were tested for purity by Ouchterlony, immunoelectrophoresis and radioimmunoassay prior to their use. Goat anti-mouse heavy chain reagents (anti-a, anti-/~ or anti-),) were obtained commercially (Bionetics), and tested with a battery of mouse myeloma immunoglobulins (Igs) and kappa and lambda chains for their specificity for heavy (H) chain determinants. These goat IgG anti-H chain antibodies were coupled with /3-galactosidase (O'Sullivan et al., 1979) for use in the ELISA (see below).
Immunocytoadherence assay An immunocytoadherence assay was performed in parallel with the ELISA for analysis of FcR expression on Th HA 1 no. 9 and no. 10 cells and on spleen cells. The RI.1 T lymphoma does not express FcR, and served as a negative control. Trinitrophenyl (TNP)-sensitized sheep erythrocytes (TNP-SRBC) were prepared as previously described (Rittenberg and Pratt, 1969). Splenic lymphocytes or T cells ( 5 ) < 10 6 cells/ml) were incubated with 100-200 ~ g / m l of monoclonal anti-DNP antibodies (MOPC 315 IgA, H5-3 IgM or H35-3 IgG1) at 4°C for 1 h. Cells were washed 3 times with RPMI 1640 containing 2% FCS and the pellet was gently resuspended with a 0.5% suspension of TNP-SRBC. Following incubation at 4°C for 30 rain, cells binding 5 or more red cells were counted as positive. Between 1000 and 1500 cells were scored in each experiment (representing 10-15 microscope fields).
272
Enzyme-linked immunosorbent assay for FeaR To obtain Th HA cells expressing maximum Fca receptors, cultures were washed (1200 r p m / 1 0 min), the cells were fractionated on Ficoll-Hypaque gradients and viable cells were recultured overnight in RPMI 1640 with 20% FCS. Th HA cells or spleen cells were extensively washed (4-6 times) with phosphate-buffered saline (PBS, pH 7.4), and appropriate cell numbers (in 100 /.tl) were added to Linbro flat-bottom EIA plates (Flow Laboratories, McLean, VA). Plates were centrifuged (1000 r p m / 5 rain), and 100 ~1 of 0.01% glutaraldehyde in PBS was added (final glutaraldehyde concentration was 0.005%) and were incubated for 30 min at 25°C. Wells were extensively washed with PBS (5-6 times) and then incubated with PBS containing 1% bovine serum albumin (BSA) for 90 min at 37°C. The contents of the wells were expelled and the wells were washed with PBS (2 times). To each well was added 100 /zl of either purified IgA, IgM or IgG1 (1-20 # g / m l ) and plates were incubated for 90 rain at 37°C. In 1 series of experiments, different amounts of IgA (i 20 /~g/ml) were added to each well as described above. Wells were then extensively washed (4-6 times) with PBS and 100 /.tl of appropriately diluted, /3-galactosidase-labeled anti-a, or anti-~ or anti-], were added to the wells. Anti-a, -~ and -], antibodies diluted to 1 : 250 gave maximum ELISA values, and these antisera dilutions were used in all experiments. Following overnight incubation at 4°C, the plates were washed 4 additional times with PBS. Substrate was prepared by dissolving o-nitrophenyl-fl-D-galactopyranoside in Alpase buffer (pH 7.5) (1 m g / m l ) and 100/~1 was added to each well. Plates were incubated at 37°C (in the dark) and color development was read at 414 nm using a Titertek Multiscan photometer (Flow) after 30, 60 and 120 min. Statistics Rosette formation is expressed as the range per triplicate determinations. The significance of difference between means of ELISA values was determined by the Student's t-test.
Results
Optimal conditions for FeaR detection by ELISA In our initial experiments, Th HA cells were fixed with various concentrations of glutaraldehyde (0.001-0.1%) prior to their use as coating antigen for plates. Generally, the fixed cells were poorly adherent, and more consistent results were obtained when living cells were used to coat the wells. Subsequently, Th HA cells were added to plates, spun at low speed and then fixed with 0.01% glutaraldehyde in PBS. Optimal ELISA values were obtained using 1 x l0 s Th HA cells/well (Fig. l); however, as few as 5 x 10 4 ceils also gave consistent and reproducible values. Generally, 5 x 104 to 5 x 10 s cells/well yielded ELISA values ranging from 0.1 to 0.15, when analyzed with lgA and developed with labelled anti-a (Fig. 1). In order to determine the optimal concentration of IgA required for maximum detection of F c a R by ELISA, microplates were coated with Th HA~ no. 9 or no. 10
273
~=
0.20 L Th HA, # 9
-
0.15
w
010 0.05 i
OI
05
i
i
I0
20
Th HA C E L L S / W E L L ( X l O
,J 0
5)
Fig. 1. Concentration of Th HA cells used as coating antigen for ELISA. EIA wells were coated with cells at the concentrations indicated and were fixed with 0.005% glutaraldehyde in situ. Wells were incubated with purified MOPC 315 IgA (0.5 fig/well) and color was developed after addition of labeled anti-a antibodies.
cells (1 x 105 cells/well) and different amounts of MOPC 315 IgA were added (0.1-2.0 ~g/well). A linear relationship between IgA added and increase in absorbance was observed until 0.5 fig/well of IgA was used (Fig. 2). Similar titrations were performed with IgM and IgG1 in wells coated with splenic lymphocytes (see below), and optimal values were also obtained with 0.5 ~g of Ig. In all subsequent experiments, the ELISA was performed with wells coated with 1 x 105 cells and with addition of 0.5/~g of the appropriate Ig.
020 Th HA, # 9
c ,i k3
Z <:[ rt~ rr 0 o9 m
015
010
T
005
o,
o12
015
,io
20
o
tgA CONCENTRATION (~g/well)
Fig. 2. Determination of optimal IgA levels required for maximum detection of FcaR. Wells were coated with Th HA cells (1 x 105 cells/well), and were incubated with the indicated amount of purified MOPC 315 lgA. Color was developed after the addition of fl-galactosidase-labeled anti-a antibodies.
274 TABLE I DETECTION OF Fca RECEPTORS BY ELISA a T cell hybridomas were used as coating antigen and purified mouse IgA and fl-galactosidase-labeled anti-a were added for measurement of binding of IgA to Fca receptors. Experiment number
Optical density (414 nm) b (T cell line used) Th HA 1 no. 9
Th HA I no. 10
RI.I
1 2 3 4 5 6 7 8 9 10
0.170 0.127 0.166 0.143 0.152 0.197 0.184 0.150 0.171 0.149
0.100 0.121 0.158 0.150 0.124 0.107 0.143 0.155 0.143 0.129
0 0 0 0.046 0.083 0 0 0.078 0 0.046
Mean _+SEM
0.161 _+0.016 (P < 0.001)
0.133 _+0.006 (P < 0.001)
0.026 + 0.001
EIA plates were coated with Th HA cells (1 × l0 s cells/well), fixed with 0.005% glutaraldehyde and next incubated with purified MOPC 315 IgA (0.5 tLg/well) for 90 min at 37°C. /3-Galactosidase-labeled goat anti-mouse a chain was added to each well, the plates were incubated overnight at 4°C, and then were extensively washed. Color was developed with substrate and the OD determined at 414 nm. h Significance of values for Fca receptors on Th HA hybridomas when compared with the non-FcaR bearing RI.1 T cell line as determined by the Student's t test.
Comparison of ELISA and rosette assay for FcR expression T h e E L I S A was quite reproducible for detection of F c a R expression, and both Th HA] no. 9 a n d no. 10 consistently exhibited these receptors (Table I). In contrast, the R1.1 T l y m p h o m a does not exhibit F c R by E L I S A (Table I). In all experiments, the E L I S A values for F c a R expression on Th H A cells were significantly higher ( P < 0.001) when c o m p a r e d with RI.1 T cells. O u r previous studies ( K i y o n o et al., 1985) with Th H A cells consistently showed that 30-40% of these cells express F c a R as d e t e r m i n e d by rosetting. However, neither Fc/~R nor Fc'c]R could be detected by this m e t h o d , and these results were verified by i m m u n o c y t o a d h e r e n c e in the present study (Table II). As indicated above, the E L I S A clearly establishes the presence of F c a R on these Th H A cell lines. In addition, when wells coated with Th H A cells were i n c u b a t e d with I g M and d e v e l o p e d with fl-galactosidase-labeled anti-/~, low, but significant a b s o r b a n c e values were noted. In all e x p e r i m e n ts with I g M and anti-#, plates coated with RI.1 T l y m p h o m a cells failed to develop a color reaction. This study clearly implies that Th H A cells also express low levels of F c ~ R , and this finding has been verified by flow c y t o m e t r y using purified IgM and F I T C - l a b e l e d anti-~ ( m an u scr i p t in preparation). O n the other hand, Th H A cells did not exhibit F c y R as d e t e r m i n e d by E L I S A or rosetting using either purified IgG1 or serum IgG.
275 TABLE I1 COMPARISON OF IMMUNOCYTOADHERENCE AND ELISA TO DETECT Fc RECEPTOR EXPRESSION a T cell hybridomas and splenic lymphocytes were examined for FcR by incubation with IgA or IgM or IgG1 (anti-DNP) and rosette formation with TNP-SRBC, or by ELISA using either T cells or splenic lymphocytes as coating antigen. Cells used
lsotype-specificityof Fc receptors Rosette formation (%) ~
Optical density (414 nm) b IgM
IgG1
IgA
Th HA t no. 9 Th HA I no. 10
0.025_+0.003 0.024_+0.004
0 0
0.136_+0.008 0.119+_0.006
Splenic lymphocytes
0.125 _+0.009
0
0.056 -+0.001
R1.1
0
0
0
IgM 0 0 19-26 0
IgG1
IgA
0 0
31-34 32-39
0
1- 5
0
0
See Table I legend for details of the ELISA. For the immunocytoadherenceassay, Th HA cells or spleen cells were reacted with purified monoclonal lgM or IgG1 or IgA anti-DNP antibody for 60 min at 37°C. Rosettes were formed after addition of TNP-SRBC. Th HA or spleen cells binding 5 or more erythrocytes were counted as positive. b Mean values were obtained from quadruplicate determinations/experiment and 3 separate experiments. ~"The range of triplicate determinations of 3 separate experiments is presented.
I n order to confirm that E L I S A detects FcR, plates coated with splenic l y m p h o cytes were i n c u b a t e d with purified IgM, IgG1 or IgA a n d were developed. Using spleen cells, higher E L I S A values were seen for Fc/~R expression than for F c a R (Table II). As detected by i m m u n o c y t o a d h e r e n c e , splenic lymphocytes displayed Fc/~R and small percentages of F c a R (Table II), clearly c o n f i r m i n g the E L I S A results. Fc~,IR were not detectable o n splenic lymphocytes with either assay, suggesting the absence of cells which could b i n d IgG1. Nevertheless, splenic lymphocytes exhibited F c T R as determined by E L I S A (0.114_+0.002) when lymphocyte-coated wells were i n c u b a t e d with purified serum I g G a n d developed with fl-galactosidase-labeled anti-~'.
Discussion In this study, we have described the c o n d i t i o n s for use of an E L I S A to measure F c R expression o n T cell lines. The method is sensitive, since wells coated with 5 x 104-1 × 105 T h H A cells, when i n c u b a t e d with 0.5/~g of IgA, consistently gave E L I S A absorbances of 0.1 or greater, which indicate that F c a R can be reliably detected by this method. The E L I S A could also be used to measure F c a R present on other cells such as splenic lymphocytes. The assay could also easily be a d a p t e d to measure soluble F c a R , by p r e i n c u b a t i o n with purified IgA, prior to its a d d i t i o n to microplate wells. I n addition, the method will allow screening of m o n o c l o n a l
276 antibodies to FcaR, by addition of supernatants of putative clones to wells prior to development with IgA and labeled anti-a antibody. The Th HA cell lines used in this study had previously been shown to bear F c a R as determined by the rosette assay (Kiyono et al., 1985). By using immunocytoadherence, we failed to detect Fc/~R or FcTR on these Th HA cell lines. Somewhat surprisingly, however, the ELISA allowed detection of low levels of Fc#R on both T cell lines, a finding which was subsequently confirmed by FACS analysis with cells incubated with IgM and stained with FITC-anti-/,t reagent (manuscript in preparation). Although we have not directly compared the sensitivity of the ELISA with that of the immunocytoadherence assay, we feel that the ELISA is at least as sensitive as the rosette method. It was important to establish that the ELISA could detect FcR of all 3 major isotypes. To this end, splenic lymphocytes were used as coating antigen, and F c # R and to a lesser extent FcaR, were readily detectable both by ELISA and immunocytoadherence. Splenic lymphocytes did not bear FcT~R; however, purified serum I g G allowed detection of FcTR, suggesting that these receptors bind to other IgG subclasses. Collectively, our results with Th HA cells and splenic lymphocytes suggest that the ELISA can be reliably used for detection of FcR for all Ig classes. Generally, 2 types of methods have been used to analyze FcR on lymphocytes, i.e., methods which detect the receptor and those which both detect FcR and allow enrichment of these cells from total lymphoid cell populations. Antigen-antibody complexes readily bind FcR on both T and B cells (Stout and Herzenberg, 1975; Dickler, 1976) and aggregated immunoglobulin has also been widely used (Dickler, 1976). Both methods are especially useful for FcTR detection, since native IgG has a lower affinity for FcTR than multivalent forms of Fc present in immune complexes or in aggregates. This may explain why spleen cells bind purified serum IgG, which may contain multivalent molecules, versus a lack of binding of cells to purified monoclonal IgG1. It should be noted that use of FITC-conjugated Igs with these methods, provides a useful way to separate FcR + cells by use of flow cytometry (Stout and Herzenberg, 1975). Rosette formation allows detection of FcR for particular isotypes as well as enrichment of this cell type by separation on Ficoll density gradients. The rosette method has been effectively adapted for enrichment of F c ~ R (Moretta et al., 1976), FcTR (Fridman et al., 1981), Fc~R (Spiegelberg, 1981), and F c a R (Hoover et al., 1981). The ELISA, as reported here would have more limited application than the methods described above. However, the ELISA could be quantitative if the average numbers of Fcc~R could be determined on each cell. We are presently using radiolabeled IgA bound to Th HA cells to determine the number of molecules which bind to F c a R + Th HA cells. In subsequent experiments we will then use the Th HA cell preparations in the ELISA to estimate the total numbers of FcctR present/well. This should allow us to establish a quantitative method for measurement of FcR expression on T cell populations by the ELISA. Experiments of this nature are underway, and will be the subject of a future paper.
277
Acknowledgements W e t h a n k D r s . D a w n C o l w e l l , J o h n E l d r i d g e a n d K a z u o K o m i y a m a for a s s e s s m e n t o f this w o r k a n d m a n u s c r i p t , a n d Ms. B e t t y W e l l s f o r t y p i n g t h e p a p e r .
References Adachi, M., J. Yodoi. N. Noro, T. Masuda and H. Uchino, 1984, J. lmmunol. 133, 65. Dickler, H.B., 1976, Adv. lmmunol. 24, 167. Endoh. M., H. Sakai, Y. Nomoto, Y. Tomino and H. Kaneshige, 1981, J. Immunol. 127, 2612. Fridman, W.H., C. Rabourdin, C. Neauport-Sautes and R.H. Gisler, 1981, lmmunol. Rev. 56, 51. Hoover, R.G. and R.G. Lynch, 1980, J. Immunol. 125, 1280. Hoover, R.G. and R.G. Lynch, 1983, J. Immunol. 130, 521. Hoover, R.G., H.M. Gebel, B.K. Dieckgraefe, S. Hickman, N.F. Rebbe, N. Hirayama, Z. Ovary and R.G. Lynch, 1981, Immunol. Rev. 56, 115. Kiyono, H., J.R. McGhee, L.M. Mosteller, J.H. Eldridge, W.J. Koopman, J,F. Kearney and S.M. Michalek, 1982, J. Exp. Med. 156, 1115. Kiyono, H.. M.D. Cooper, J.F. Kearney, L.M. Mosteller, S.M. Michalek, W.J. Koopman and J.R. McGhee, 1984a, J. Exp. Med. 159, 798. Kiyono, H., J.O. Phillips, D.E. Colwell, S.M. Michalek, W.J. Koopman and J.R. McGhee, 1984b, J. lmmunol. 133, 1087. Kiyono, H., L.M. Mosteller-Barnum, A.M. Pitts, S.I. Williamson, S.M. Michalek and J.R. McGhee, 1985, J. Exp. Med. 161, 731. Lure, R.G., A.V. Muchmore, D. Keren, J. Decker, I. Koski, W. Strober and R.M. Blaese, 1979, J. lmmunol. 122, 65. Moretta, L.M., M. Ferrarini, M.C. Mingari, A. Moretta and S.R. Webb, 1976, J. Immunol. 117, 2171. O'Sullivan, M.J., E. Gnemmi, A.D. Simmonds, G. Chieregatti, E. Heyderman, J.W. Bridges and V. Marks, 1979, J. Immunol. Methods 31,247. Rittenberg, M.B. and K.L. Pratt, 1969, Proc. Soc. Exp. Biol. Med. 132, 575. Spiegelberg, H.L., 1981, Immunol. Rev. 56, 199. Stout, R.D. and L.A. Herzenberg, 1975, J. Exp. Med. 142, 1041. Strober, W.. N.E. Hague, L.G. Lum and P.A. Henkart, 1979, J. Immunol. 121, 2440. Yodoi, J., M. Adachi, K. Teshigawara, M. Miyama-lnaba, T. Masuda and W.H. Fridman, 1983, J. Immunol. 131, 303.