Journal of Immunological Methods, 132 (1990) 287-295 Elsevier
287
JIM 05693
A bioassay for the measurement of human interleukin-4 Jay P. Siegel and H o w a r d S. Mostowski Laboratory of Cellular Immunology, Division of Cytokine Biology, Centerfor Biologics Evaluation and Research, FDA, Bethesda, MD, U.S.A. (Received 2 February 1990, revised received 7 June 1990, accepted 8 June 1990)
We have developed a bioassay for human IL-4 based upon its ability to upregulate CD23 (low affinity IgE receptor) expression. Ramos, a B lymphocyte line derived from a Burkitt lymphoma, was repetitively subcloned yielding a clone, Ramos.G6.C10, which is several fold more sensitive to this effect of IL-4. In microtiter plates cells were cultured for 48 h in the presence of dilutions of recombinant human IL-4 or samples, and then stained with murine anti-human CD23 and goat anti-mouse IgG-FITC. IL-4 induced an eight-fold increase (60 channel shift) in fluorescence intensity as measured by flow cytometry. Significant effects were observed at an IL-4 concentration of 50-100 pg/ml and increased with concentrations up to 800 pg/ml. Inter- and intra-assay coefficients of variation were 10% and 11% respectively. The bioassay showed good specificity for IL-4; however, tumor necrosis factors a and r , at optimal concentrations, gave readings barely at the threshold of detection. Key words: Interleukin-4; Bioassay; B cell stimulatory factor; Flow cytometry
Introduction
Interleukin-4 (IL-4) is a lymphokine with profound effects on the growth and differentiation of immunologically competent cells. While best known for its various effects on B lymphocytes including induction of immunoglobulin class switching, IL-4 has also been described to have regulatory effects on T cell growth and differentiation, null cell activation, monocyte/macrophage differentiation, mast cell activation and hemato-
Correspondence to: J.P. Siegel,Laboratory of Cellular Immunology, Division of Cytokine Biology, Center for Biologics Evaluation and Research, FDA, Bethesda, MD, U.S.A. Abbreviations: IL, interleukin; IFN, interferon; TNF, tumor necrosis factor; GAM-FITC, goat anti-mouse immunogiobulin conjugated with fluorescein isothiocyanate; MIgG1, murine immunoglobulin G subclass 1; SD, standard deviation; CV, coefficient of variation.
poietic cell activation (reviewed by Ohara, 1989). Furthermore, data suggest that in vivo IL-4 production may be an important factor in allergic and parasitic diseases (Finkelman et al., 1988; Heinzel et al., 1989). For these reasons there has been considerable interest in the detection of IL-4 in in vitro and in vivo systems and in the administration of IL-4 to humans. Such studies have been hampered by the difficulties inherent in bioassays currently available for human IL-4. Although murine cell lines have been employed in a convenient and relatively specific murine IL-4 bioassay (Hu-Li et al., 1989), murine cells will not respond to human IL-4. The most commonly used bioassays for human IL-4 involve coproliferative effects on B cells stimulated with anti-/~ chain antibodies or on T cells activated with lectins (Yakota et al., 1986). The B cell bioassays suffer from the limited availability of human tonsiUar B cells, the technical difficul-
0022-1759/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
288 ties of obtaining purified B cells from circulating blood, and the inconsistencies resulting from the fact that freshly purified B cells may show considerable variation in their state of activation and responsiveness to IL-4. Bioassays based on the proliferative effects of IL-4 on lectin activated human T cells (Yakota et al., 1986) also suffer from lack of consistency due to donor variability and low signal to background ratios. Furthermore, assays of T or B cell proliferation generally will register other lymphokines in addition to IL-4. The Ramos cell line consists of Burkitt lymphoma-derived human B lymphocytes which respond to IL-4 with upregulation of expression of CD23, the low affinity receptor for IgE (Rousset et al., 1988). In this study, we generated successive subclones of the Ramos line with increased magr~i'tude of this response to IL-4. We describe herein a sensitive, accurate, and relatively specific bioassay for human IL-4 using a Ramos subclone with the convenience and reproducibility inherent in the use of a cell line.
Materials and methods
Development of an 1L-4 hyperresponsive clone Cells of the Ramos cell line were obtained from American Type Culture Collection. The Ramos cell line consists of human B lymphocytes, derived from a Burkitt lymphoma which respond to IL-4 with upregulation of expression of low affinity IgE receptors (CD23) (Rousset et al., 1988). Ramos cells were fluorescently labelled using anti-CD23 monoclonal antibody (The Binding Site, San Diego, CA), and fluorescein-labelled goat anti-mouse immunoglobulin (GAM-FITC, Cappel Laboratories). The dullest 1% of cells were then sorted as single cells into wells of a 96-well flat-bottomed plate using a Coulter Epics 752 flow cytometer with Auto-Clone (Coulter Electronics, Hialeah, FL) and grown into subclones. After clonal cell counts exceeded 104 cells per clone, they were split three ways and grown in the absence or presence of recombinant human IL-4 (Immunex, Seattle, WA) at 20 ng/ml. After 3 days, CD23 expression with and without IL-4 was measured fluorescently on approximately 30 clones and the subclone demonstrating the greatest IL-4 effect on a loga-
rithmic scale of fluorescence was selected and designated Ramos.G6. Ramos.G6 cells, not previously exposed to IL-4, were further expanded. After approximately 2 months of culturing, Ramos.G6 cells were stained for CD23 expression and the brightest 1% were subcloned (precisely one cell/well) and approximately 80 clones were tested for IL-4 responsiveness as above. The subclone demonstrating the greatest IL-4 effect was designated Ramos.G6.C10. Ramos.G6.C10 cells were expanded, frozen in multiple aliquots, and used in the bioassay described below. They have maintained their IL-4 responsiveness in tissue culture (grown without IL-4) for over 6 months. Saturation binding studies show Ramos, Ramos.G6, and Ramos.G6.C10 all to have approximately 1500 IL-4 binding sites per cell.
IL-4 bioassay Human recombinant IL-4 provided by Immunex was used as our IL-4 standard. IL-4 was diluted in RPMI 1640 medium supplemented with gentamicin, amphotericin B, and 15 % fetal calf serum (complete medium). Aliquots of 100 gl of diluted standard were placed into 96-well fiat-bottomed microtiter plates in triple replicates whose final concentrations (after addition of assay cells) ranged by two-fold dilutions from 1.6 or 3.2 ng/ml to 25 or 50 pg/ml. Samples were also added to the plate in various dilutions in 100 gl aliquots in triplicate. Negative control, medium-only samples were placed at the beginning, middle, and end of each plate to test for flow cytometer stability during the assay and other sources of variation dependent upon sample position. Edge wells were not used. Assay cells were harvested 2-3 days after splitting and suspended at 3-5 × 105 cells/ml in complete medium. 100 gl of cell suspension were added to each well containing dilutions of standard or samples. Plates were then placed in a humidified incubator containing room air supplemented with 7% CO 2. After 48 h, cells were transferred to round-bottomed plates, washed three times, serially labelled with anti-CD23 (3 gl/well) and GAM-FITC (4 gl/well), washed twice, and fixed in 1% paraformaldehyde. Fluorescence was
289
measured using a FACScan (Becton Dickinson, Mt. View, CA). A histogram of green fluorescence was measured for each well on a logarithmic scale in which channels 0 through 255 represent 4 logs base 10. While the position of these histograms was a function of IL-4 concentration, they remained unimodal at all concentrations. The fluorescent channel representing the 75 th percentile of cells was determined for each well and is referred to as the fluorescence channel for that sample. The 50 th percentile (median) fluorescence channel gave similar results but a slightly smaller difference in channels between cells treated with and without IL-4 than did use of the 75 th percentile. A standard curve of fluorescence channel as a function of IL-4 concentration was determined using the computer program, ALLFIT v2.7 (De Lean et al., 1988). IL-4 concentration in each sample was determined by comparison with the standard curve.
Reference preparations Reference samples contMning known concentrations of recombinant human IL-4 were prepared by dilution of rIL-4 with complete medium. Aliquots of these preparations were stored at - 70oC.
Cytokines Human recombinant IL-4 and IL-6 were provided by Immunex (Seattle, WA). Human recombinant IL-2 and M-CSF were provided by Cetus (Emeryville, CA). Human recombinant interferon (IFN)-3, and tumor necrosis factors (TNFs), TNF-
a and -fl, were provided by Genentech (So. San Francisco, CA). Human recombinant IFN-a was provided by Hoffman LaRoche (Nutley, N J). Human recombinant GM-CSF was provided by Schering-Plough (Kenilworth, N J). Human recombinant G-CSF was provided by Amgen (Thousand Oaks, CA). IL-lfl was provided by Otsuka (Rockville, MD).
Resd~
IL-4 responses of the cell lines used Parental Ramos cells and the subclones Ramos.G6 and Ramos.G6.C10 were cultured in parallel for 3 days in the absence or presence of IL-4 10 ng/ml. Flow cytometry was performed on unstained cells, and cells stained with either antiCD23 and GAM-FITC or a irrelevant control antibody and GAM-FITC. Fluorescence channels were determined as described above and are shown in Table I; data from the subclone used in our bioassay, Ramos.G6.C10 are also shown converted to a linear scale. Growth in IL-4 resulted in increase in log fluorescence channels of approximately ten channels for parental Ramos, 48 channels for Ramos.G6, and 60 channels for Ramos. G6.C10. Since data are generated on a scale in which 256 channels cover 4 logs base 10, these differences represents IL-4 induced increases in brightness of 43%, 462%, and 766%, respectively. Thus, as a result of the selection used, the clones became progressively more responsive tO IL-4.
TABLE I DEVELOPMENT OF AN IL-4 HYPERRESPONSIVE CLONE Fluorescence channel (log scale) Ramos
Unstained MIgG1 + GaM-FITC a aCD23 + GaM-FITC b
Relative fluorescence (linear scale) Ramos.G6
Ramos.G6.C10
Ramos.G6.C10
-I~4
+Ib4
-IL-4
+IL~
-I~4
+IL~
-
71 100 105
76 105 115
59 87 95
78 103 143
58 83 103
75 110 163
1.1 2.5 5.0
IL-4
+
IL-4
1.8 6.5 43.7
a 1st antibody: routine immunoglobulin G1 against irrelevant antigen. 2nd antibody: fluorescein-conjugated goat anti-murine immunoglobulin G. 2 1st antibody: murine immtmoglobulin G1 against human CD23. 2 nd antibody: fluorescein-conjugated goat-anti-murine immunoglobulin G.
290
l~
-~
3.2 ng/ml
=
/~
1.6 ng/ml
800 pg/ml ?.;~,.~
t'~x',L
¢
,e'"
400 pg/ml
j.,\
\%,.
~
~tI 200 pg/ml
0
".~
100
pg/ml
,.,,"~
0
•
~']
~ ..
.
50 pg/ml
25 pg/ml
~
M_"';~ .
..
0
\\.. Log green fluorescence Fig. 1. Flow cytometric profiles from the h u m a n IL-4 bioassay. The dashed and two dotted lines in each panel are profiles from cells cultured in three replicate wells supplemented with IL-4 at the concentrations indicated. The solid line is one of the profiles from cells cultured in the absence of IL-4. Each profile is on a logarithmic x axis on which full scale covers 4 logs base 10.
Two additional findings are of note in these data. The first subcloning (Ramos.G6) selected a subclone with less background autofluorescence A. IL-4ASSAY 1
(channel 71 vs. 59) allowing a more sensitive detection of CD23 dependent fluorescence. Also, while the linear scale data make it clear that the
B. IL-4 ASSAY 2 160
170
C. IL-4 ASSAY 3 160]
~ 160-
o
D
o
o
1501 a
140
1401
130
1301
~= 140 f-) 130
8
n
150 a
150
0
n
o
0 0
1201
;
120
~ 120
110
110 l
10c
u
I00 o 12.5
~i0t
50
200
800
IL-4 (pg/ml)
IOO t -'~, 012.5 3200
, 9~ 50 200 800 3200 IL-4 (pg/ml)
0 12.5
50 200 800 3200 IL-4 (pg/ml)
Fig. 2. Representative standard curves from the h u m a n IL-4 bioassay. Shown are data from testing the IL-4 standard in triplicate at serial two-fold dilutions from 12.5 to 3200 p g / m l in three separate assays. Fluorescence channels reflect CD23 expression on a logarithmic scale determined as described in the materials and methods SeCtion. Curves shown are hand-drawn; but calculations for samples are based on computer-fitted curves.
291 TABLE II DESCRIPTION OF STANDARD CURVES Assay
Maximal response a (Chan)
B a c k g r o u nbd (Chan)
MaxBkg. c (Chart)
Mean SD d (Chan)
Slope e (Chan/Dil)
EDs0 f (pg/ml)
1 2 3 4 5
161 156 159 155 158
104 103 99 103 101
57 53 60 52 57
1.9 1.8 1.9 2.1 1.7
17 12 14 14 14
274 191 173 136 143
a Cells cultured in IL-4, 3.2 ng/ml. b Cells cultured without IL-4. ¢ Maximal response minus background. d Mean standard deviation among three replicates at each dilution of the standard. e Change in fluorescence channel per two-fold dilution of IL-4 over range of 100-400 pg/ml. f [IL-4] leading to fluorescence channel half way between minimal and maximal response channels.
p r i n c i p a l IL-4 effect o n fluorescence i n this bioassay was u p r e g u l a t i o n of CD23, IL-4 also upregulated autofluorescence a n d nonspeciflc a n t i b o d y b i n d i n g o n these cells.
Dose-response of the IL-4 effect Fig. 1 shows flow cytometric profiles from the s t a n d a r d curve of a representative IL-4 assay. As c a n b e seen, the histograms are u n i m o d a l a n d cells cultured i n IL-4 c o n c e n t r a t i o n s of 50 p g / m l or greater showed a d o s e - d e p e n d e n t increase i n fluorescence. Fig. 2, panels A, B, a n d C, show three representative s t a n d a r d curves of fluorescence c h a n n e l vs. IL-4 c o n c e n t r a t i o n generated i n the m a n n e r described i n the materials a n d m e t h o d s section. As can b e seen, the lowest c o n c e n t r a t i o n of IL-4 i n d u c i n g significant fluorescence above b a c k g r o u n d is 50 or 100 p g / m l . T h e steepest part
of the response curves occurs at IL-4 c o n c e n t r a tions b e t w e e n 100 a n d 400 p g / m l . T h e IL-4 effect begins to reach a p l a t e a u at a c o n c e n t r a t i o n of 800 p g / m l . T a b l e II shows d a t a from five s t a n d a r d curves. T h e fluorescence difference b e t w e e n cells g r o w n with vs. w i t h o u t IL-4 (3200 p g / m l ) ranged from 52 to 60 channels, r e p r e s e n t i n g a 6 . 5 - 8 . 7 - f o l d increase i n fluorescence a t t r i b u t a b l e to IL-4. A s t a n d a r d d e v i a t i o n of a p p r o x i m a t e l y two c h a n n e l s was n o t e d a m o n g replicates dilutions of the standard at all IL-4 c o n c e n t r a t i o n s i n all assays representing replicate v a r i a t i o n i n fluorescence of app r o x i m a t e l y 7%. A t the steepest p o r t i o n of the s t a n d a r d curve each two-fold d i l u t i o n of the s t a n d a r d ( 4 0 0 - 2 0 0 a n d 2 0 0 - 1 0 0 p g / m l final conc e n t r a t i o n ) resulted i n a m e a n fluorescence difference of 14 channels. Thus, the difference in readings b e t w e e n two adjacent dilutions of the
TABLE III IL-4 READINGS OF REFERENCE PREPARATIONS Assay
300 pg/ml
no.
Mean + SD
CV
Mean + SD
200 pg/ml CV
Mean 4-SD
CV
Mean 4-SD
CV
1 2 3 4
3024-41 a 2974-23 286 + 18 ND
14% 8% 6%
1924- 3 1704- 0 ND 168 4-11
2% 0%
18% 22% 11%
ND 91 4-10 684- 9 ND
11% 13%
7%
165 ± 30 1894-41 140+15 ND
Mean
2954- 8 b
3%
1774-13
7%
165 4-25
15%
80 + 16
20%
150 pg/ml
Intra-assay mean + SD of three replicate wells within the assay. b Interassay mean + SD of the results of the assays shown. a
75 pg/rnl
292 standard in the steep portion of the curve (mean difference 14 channels) considerably exceeded the variation among replicates at a single dilution (standard deviation less than two channels). The concentration of IL-4 resulting in a fluorescence readout half way (in channels) between that seen with 0 and 3200 p g / m l (i.e., the EDs0 ) was typically approximately 200 p g / m l .
Precision and reproducibility of the IL-4 bioassay In order to determine the accuracy and reproducibility of the bioassay on samples at various concentrations, reference samples of IL-4 were prepared as described above and tested in triplicate in single concentrations over a 2 month period. Reference samples were tested at final concentrations of 300 p g / m l , 200 p g / m l , 150 p g / m l , and 75 p g / m l . The bioassay results were then used to determine IL-4 content for each concentration tested and the results are presented in Table III. The means of the values determined for reference samples tested at 300, 200, 150, and 75 p g / m l were 295, 177, 165, and 80 p g / m l respectively. The 11 determinations on the four samples had a mean interassay coefficient of variation of 10%. The mean intra-assay coefficient of variation for the four samples was 11%. The difference between measured and actual value for the 11 determinations averaged 11%, exceeded 20% in only two cases, and never exceeded 33%.
Cytokine specificity In order to test the specificity of the response measured for IL-4, several other recombinant human cytokines were tested in the bioassay. A cytokine which has no direct effect in the bioassay might, in the presence of IL-4, have synergistic or antagonistic effects which interfere with the determination of IL-4 levels. Therefore, cytokines were tested not only alone, but also in the presence of 200 p g / m l IL-4, a concentration giving approximately half-maximal activity. As can be seen in Table IV, IL-lfl, IL-2, IL-6, IL-7, G-CSF, M-CSF, and G M - C S F all had no effect at a concentration of 100 n g / m l alone and they did not significantly alter the effect of IL-4 at 200 p g / m l . IFN-t~ and IFN-T had no effect by themselves but antagonized the effects of IL-4. T N F - a and, to a lesser extent, TNF-fl registered weakly
TABLE IV EFFECTS OF OTHER CYTOKINES ON THE IL-4 BIOASSAY Cytokine
Conc. (ng/ml)
G-CSF M-CSF GM-CSF IL-I• IL-2 IL-6
100 100 100 100 100 100 100 50 100 100
IFN-a IFN-T TNF-a TNF-fl
Result (IL-4 pg/ml) a No IL-4 With IL-4 200 pg/ml < 25 237 < 25 187 < 25 195 < 25 187 < 25 196 < 25 217 < 25 122 < 25 145 46 > 800 36 320
a Calculated apparent IL-4 concentrations for samples containing other cytokines in the concentrations indicated together with IL-4 at 0 or 200 pg/ml.
in the bioassay when tested alone but had more significant augmenting effects when used in combination with IL-4. In order to further assess the effects of IFN-a, IFN-T, T N F - a , and TNF-fl in this bioassay, these cytokines were tested at multiple concentrations ranging from 100 p g / m l to 100 n g / m l in the presence of IL-4 at 200 p g / m l . Results are shown in Fig. 3. IFN-a partially inhibited CD23 induction at all concentrations tested. However, even at the highest concentration of IFN-a (500-fold excess over IL-4), the apparent IL-4 level was still approximately 50% of the actual level. IFN-T had a lesser inhibitory effect and no effect of IFN-T was observed at concentration of 1 n g / m l or less. T N F - a and TNF-fl showed a synergy, with IL-4 with effects first observed at concentrations of 100 p g / m l of T N F - a or 1 n g / m l of TNF-fl a n d increasing through the highest concentration tested (100 n g / m l ) . TNF-ct was also tested at the above concentrations in the absence of IL-4 (data not shown). At 100 p g / m l , no effect of T N F - a was observed while at 1, 10, and 100 n g / m l , gradually increasing effects were observed which were significantly different from background but remained less than that seen with 50 p g / m l IL-4. Thus, TNF-c~ and TNF-fl strongly synergize with IL-4 in this bioassay but by themselves have mea-
293 800
700. b~ c~
600.
4--4 A--A
TNFa TNF~
•--$
IFN 7
0--0
IFNa
•
4oo. oo.
i
200t t
0
0
0.1
1
10
100
0ytokine Concentration, ng/rnl Fig. 3. Dose-respo~e relationship of the interaction of IFNs and TNFs with IL-4. IL-4 was tested in the bioassay at a concentration of 200 pg/ml in the presence or absence of varying concentration of TNF-a or -fl or IFN-a or -V in triplicate wells. By comparison of CD23-dependent fluorescence channel with a standard curve generated using serial dilution of IL-4 alone as describedin the materials and methods section, the apparent level of IL-4 in each sample was determined and is shown above.
surable effects below the stated sensitivity threshold of the bioassay (50-100 pg/ml).
Discussion We have described herein a sensitive, reproducible, relatively simple and moderately specific bioassay for human IL-4. The sensitivity of 50-100 p g / m l (3-6 pM) is excellent in light of the K d of 50 pM for the binding of IL-4 to its receptor on human cells. The inter- and intra-assay coefficients of variation of this bioassay are both near 10% and compare quite favorably with those of other bioassays in general and bioassays for cytokines in particular. Of a broad range of cytokines tested, only the T N F s had a direct effect on the bioassay and this effect was minimal. Compared with the commonly used bioassays for human IL-4 involving costimulation of tonsillar or purified blood B cells or peripheral blood T cells, our bioassay, being based on a cell line, offers obvious advantages in the convenience of obtaining cells and the reproducibility of cell behavior. Additionally, the coefficients of variation of bioassays on
freshly isolated cells are much higher and the specificity is usually poor. The ability of some IFNs and TNFs to alter the bioassay results when IL-4 is present might be problematic in some settings. The use of specific neutralizing antibodies to these other cytokines may be of value in minimizing this effect. In comparison with other bioassays commonly in use for human IL-4, the specificity of our assay is excellent. Under some circumstances, proliferation of lectin~stimulated T cells may be supported by IL-2, IL-6 (Garman et al., 1987), or IL-7 (Morissey et al., 1989) in addition to IL-4. Similarly, proliferation of activated B cells may be supported by IL-2 (Zubler et al., 1984), and interferon-a or -7 (Romagnani et al., 1986; Morikawa et al., 1987). Indeed, the two cytokines other than IL-4 found to register weakly in our assay, T N F - a and -fl, have more potent effects on B cell proliferation (Kehrl et al., 1987a,b). And while IFN-7 partially inhibited IL-4 induced CD-23 upregulation on our assay clone, it completely blocked the IL-4 effect on other B cell lines (Rousset et al., 1988). The interactions of IFNs and T N F s with IL-4 may have interesting implications regarding the roles of these cytokines, in conjunction with IL-4, in the regulation of the differentiation and the activation of B cells. In this regard, the ability of IFN-7 to antagonize CD23 induction by IL-4 has already been studied (Defrance et al., 1987; Rousset et al., 1988). T N F - a has been shown to be a growth factor for some B cell malignancies and IFN-a opposed this effect (Cordingley et al., 1988). TNF-fl has been reported not to upregulate CD23 on resting B cells either alone or with anti-IgM or IL-4 but it did increase the induction on B cells by IgM and IL-4 of the activation marker recognized by Leu 21 (Zola and Nikoloutsopolous, 1989). In light of the .potent synergism of T N F s with IL-4 on Ramos.G6.C10 cells, we are currently studying their combined effects on the expression of CD23 and other differentiation markers on resting B cells. While the fluorimetric readout of this bioassay may makes inaccessible to some laboratories, our use of an automated 96-well plate sampler on the flow cytometer has markedly diminished our operator time in performing the assay. Murine cell lines which exhibit a proliferative response to
294 m u r i n e b u t n o t h u m a n I L - 4 have b e e n in use in m u r i n e I L - 4 bioassays. H u - L i et al. (1989) h a v e described such a cell line with the a d v a n t a g e of being relatively insensitive to IL-2. By contrast, cell lines are n o t in w i d e s p r e a d use for b i o a s s a y for h u m a n IL-4. M a e d a et al. (1988) r e p o r t e d the d e v e l o p m e n t of a b i o a s s a y for h u m a n I L - 4 emp l o y i n g a n IL-2 d e p e n d e n t T cell line, Sez 627, which proliferates in r e s p o n s e to h u m a n IL-4. W h i l e their s t u d y p r e s e n t s t o o little d a t a to det e r m i n e the sensitivity, r e p r o d u c i b i l i t y , or accur a c y of t h a t bioassay, it is of n o t e that the rep o r t e d m a x i m a l p r o l i f e r a t i v e responses of 6589 + 411 a n d 9409 + 1101 differed f r o m b a c k g r o u n d b y 13 a n d 7 s t a n d a r d deviations c o m p a r e d with an average of 30 s t a n d a r d d e v i a t i o n s b e t w e e n m a x i m a l r e s p o n s e a n d b a c k g r o u n d in o u r b i o a s s a y ( a p p r o x i m a t e l y 60 channels, two c h a n n e l s p e r SD). Thus, that b i o a s s a y is likely to b e c o n s i d e r a b l y less r e p r o d u c i b l e o r a c c u r a t e t h a n the one rep o r t e d herein. F u r t h e r m o r e , the Sez 627 cells res p o n d e d to IL-2 at 0.01 U / m l to an extent which exceeds their o p t i m a l r e s p o n s e to IL-4. Thus, very small a m o u n t s of IL-2 w o u l d fully m a s k the I L - 4 response. T h e a b i l i t y o f cytokines to i n t e r a c t w i t h I L - 4 in their b i o a s s a y was n o t studied. A n i m m u n o a s s a y has b e e n r e p o r t e d for h u m a n I L - 4 (Chretien et al., 1989) a n d such assays are c o m m e r c i a l l y a v a i l a b l e ( G e n z y m e , 1989). T h e s e i m m u n o a s s a y s have r e p o r t e d sensitivities ( b o t h slightly less t h a n 100 p g / m l ) c o m p a r a b l e to t h a t o b s e r v e d in o u r bioassay. I m m u n o a s s a y s for cytokines m a y be generally a d v a n t a g e o u s over b i o a s says in terms of speed, convenience, a n d specificity with respect to u n r e l a t e d cytokines. W h i l e such assays can b e o f c o n s i d e r a b l e value, i m m u n o a s says in general c a n n o t s u p p l a n t b i o a s s a y s d u e to their i n a b i l i t y to distinguish b i o a c t i v e molecules f r o m m a n y m e t a b o l i t e s , d e g r a d a t i o n p r o d u c t s , den a t u r e d o r otherwise i n a c t i v a t e d molecules, precursors, aggregates, a n d i n c o r r e c t l y synthesized molecules. I n s u m m a r y , we have d e v e l o p e d a cell line for use in a b i o a s s a y for h u m a n IL-4. T h e biological a n d statistical characteristics o f this b i o a s s a y indicate that it will b e highly useful in a v a r i e t y of settings.
References Chretien, I., Van Kimmenade, A., Pearce, M.K., Banchereau, J. and Abrams, J.S. (1989) Development of polyclonal and monoclonal antibodies for immunoassay and neutralization of human interleukin-4. J. Immunol. Methods 117, 67-81. Cordingley, F.T., Bianchi, A., Hoffbrand, A.V., Reittie, J.E., Heslop, H.E., Vyakarnam, A., Turner, M., Meager, A. and Brenner, M.K. (1988) Tumour necrosis factor as an autocfine tumour growth factor for chronic B-cell malignancies. Lancet i, 969-971. Defrance, T., Aubry, J.P., Rousset, F., Vanbervliet, B., Bonnefoy, J.Y., Arai, N., Takebe, Y., Yokota, T., Lee, F., Arai, K., De Vries, J.E. and Bancherean, J. (1987) Human recombinant interleukin 4 induces Fc epsilon receptors (CD23) on normal human B lymphocytes. J. Exp. Med. 165, 14591467. De Lean, A., Mnnson, P.J. and Rodbard, D. (1978) Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay, and physiological dose-response curves. Am. J. Physiol. 235, E97-E102. Finkelman, F.D., Katona, I.M., Urban, Jr., J.F., Holmes, J., Ohara, J., Tung, A.S., Sample, J.vG. and Paul, W.E. (1988) IL-4 is required to generate and sustain in vivo IgE responses. J. Immunol. 141, 2335-2341. Garman, R.D., Jacobs, K.A., Clark, S.C. and Raulet, D.H. (1987) B-cell stimulatory factor 2 (Beta 2 interferon) functions as a second signal for IL-2 production by mature routine T cells. Proc. Natl. Acad. Sci. U.S.A. 84, 7629. Genzyme Corporation (1989) Human Interleukin-4 ELISA: Intertest-4. Product Instruction Booklet. Boston, MA. Heinzel, F.P., Sadick, M.D., Holaday, B.J., Coffman, R.L. and Locksley, R.M. (1989) Reciprocal expression of interferon-T or interleukin 4 during the resolution or progression of murine leishmaniasis. J. Exp. Med. 169, 59-72. Hu-Li, J., Ohara, J., Watson, C., Tsang, W. and Paul, W.E. (1989) Derivation of a T cell line that is highly responsive to IL-4 and IL-2 (CT.4R) and of an IL-2 hyporesponsive mutant of that line (CT.4S). J. Immunol. 142, 800-807. Kehrl, J.H., Miller, A. and Fauci, A.S. (1987a) Efffect of tumor necrosis factor a on mitogen-activated human B cells. J. Exp. Med. 166, 786-791. Kehrl J.H., Alvarez-Mon, M., Delsing, G.A. and Fauci A.S. (1987b) Lymphotoxin is an important T cell-derived growth factor for human B cells. Science 238, 1144-1146. Maeda, M., Noma, T., Hama, K. and Honjo, T. (1988) Application of a human T cell line derived from a Sezary syndrome patient for human interleukin 4 assay. Immunol. Lett. 18, 247-254. Morikawa, K., Kubagawa, H., Suzuki, T. and Cooper, M.D. (1987) Recombinant interferon-a, -fl, and -'t enhance the proliferative response of human B cells. J. Immunol. 139, 761. Morrissey, P.J., Goodwin, R.G., Nordan, R.P., Anderson, D., Grabstein, K.H., Cosman, D., Sims, J., Lupton, S., Acres, B., Reed, S.G., Mochizuki, D., Eisenman, J., Conlon, P.J.
295 and Namen, A.E. (1989) Recombinant interleukin 7, pre-B cell growth factor, has costimulatory activity on purified mature T cells. J. Exp. Med. 169, 707-716 Ohara, J. (1989) Interleukin-4: molecular structure and biochemical characteristics, biological function and receptor expression. Year Immunol. 5, 126-159 Romagnani, S., Giudizi, M.G., Bagotti, R., Almerigogna, F., Mingara, C., Maggi, E., Liang, C. and Moretta, L. (1986) B cell growth factor activity of interferon-gamma: Recombinant human interferon-gamma promotes proliferation of anti-~activated human B lymphocytes. J. Immunol. 136, 3513. Rousset, F., De Waal Malefijt, R., Slierendregt, B., Aubry, J.-P., Bormefoy, J.-Y., Defrance, T., Banchereau, J. and De Vries, J.E. (1988) Regulation of Fc receptor for IgE (CD23) and class II MHC antigen expression on Burkitt's
lymphoma cell lines by human IL-4 and IFN-gamma. J. Immunol. 140, 2625-2632 Yakota, T. Otsuka, T., Mosmann, T., Banchereau, J., Defrance, T., Blanchard, D., De Vries, J.E., Lee, F. and Arai, K. (1986) Isolation and characterization of a human cDNA clone homologous to mouse B-cell stimulatory factor 1, that expresses B-cell- and T-cell-stimulating activities. Proc. Natl. Acad. Sci. U.S.A. 83, 5894-5898. Zola, H. and Nikoloutsopoulos, A. (1989) Effect of recombinant human turnout necrosis factor fl (TNF-fl) on activation, proliferation and differentiation of human B lymphocytes. Immunology 67, 231-236. Zubler, R.H., Lowenthul, J.W., Erard, F., Hashimoto, N., Devos, R. and Macdonald, H.R. (1984) Activated B cells express receptors for and proliferate in response to pure interleukin-2. J. Exp. Med. 160, 1170.
287
JIM 05693
A bioassay for the measurement of human interleukin-4 Jay P. Siegel and H o w a r d S. Mostowski Laboratory of Cellular Immunology, Division of Cytokine Biology, Centerfor Biologics Evaluation and Research, FDA, Bethesda, MD, U.S.A. (Received 2 February 1990, revised received 7 June 1990, accepted 8 June 1990)
We have developed a bioassay for human IL-4 based upon its ability to upregulate CD23 (low affinity IgE receptor) expression. Ramos, a B lymphocyte line derived from a Burkitt lymphoma, was repetitively subcloned yielding a clone, Ramos.G6.C10, which is several fold more sensitive to this effect of IL-4. In microtiter plates cells were cultured for 48 h in the presence of dilutions of recombinant human IL-4 or samples, and then stained with murine anti-human CD23 and goat anti-mouse IgG-FITC. IL-4 induced an eight-fold increase (60 channel shift) in fluorescence intensity as measured by flow cytometry. Significant effects were observed at an IL-4 concentration of 50-100 pg/ml and increased with concentrations up to 800 pg/ml. Inter- and intra-assay coefficients of variation were 10% and 11% respectively. The bioassay showed good specificity for IL-4; however, tumor necrosis factors a and r , at optimal concentrations, gave readings barely at the threshold of detection. Key words: Interleukin-4; Bioassay; B cell stimulatory factor; Flow cytometry
Introduction
Interleukin-4 (IL-4) is a lymphokine with profound effects on the growth and differentiation of immunologically competent cells. While best known for its various effects on B lymphocytes including induction of immunoglobulin class switching, IL-4 has also been described to have regulatory effects on T cell growth and differentiation, null cell activation, monocyte/macrophage differentiation, mast cell activation and hemato-
Correspondence to: J.P. Siegel,Laboratory of Cellular Immunology, Division of Cytokine Biology, Center for Biologics Evaluation and Research, FDA, Bethesda, MD, U.S.A. Abbreviations: IL, interleukin; IFN, interferon; TNF, tumor necrosis factor; GAM-FITC, goat anti-mouse immunogiobulin conjugated with fluorescein isothiocyanate; MIgG1, murine immunoglobulin G subclass 1; SD, standard deviation; CV, coefficient of variation.
poietic cell activation (reviewed by Ohara, 1989). Furthermore, data suggest that in vivo IL-4 production may be an important factor in allergic and parasitic diseases (Finkelman et al., 1988; Heinzel et al., 1989). For these reasons there has been considerable interest in the detection of IL-4 in in vitro and in vivo systems and in the administration of IL-4 to humans. Such studies have been hampered by the difficulties inherent in bioassays currently available for human IL-4. Although murine cell lines have been employed in a convenient and relatively specific murine IL-4 bioassay (Hu-Li et al., 1989), murine cells will not respond to human IL-4. The most commonly used bioassays for human IL-4 involve coproliferative effects on B cells stimulated with anti-/~ chain antibodies or on T cells activated with lectins (Yakota et al., 1986). The B cell bioassays suffer from the limited availability of human tonsiUar B cells, the technical difficul-
0022-1759/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
288 ties of obtaining purified B cells from circulating blood, and the inconsistencies resulting from the fact that freshly purified B cells may show considerable variation in their state of activation and responsiveness to IL-4. Bioassays based on the proliferative effects of IL-4 on lectin activated human T cells (Yakota et al., 1986) also suffer from lack of consistency due to donor variability and low signal to background ratios. Furthermore, assays of T or B cell proliferation generally will register other lymphokines in addition to IL-4. The Ramos cell line consists of Burkitt lymphoma-derived human B lymphocytes which respond to IL-4 with upregulation of expression of CD23, the low affinity receptor for IgE (Rousset et al., 1988). In this study, we generated successive subclones of the Ramos line with increased magr~i'tude of this response to IL-4. We describe herein a sensitive, accurate, and relatively specific bioassay for human IL-4 using a Ramos subclone with the convenience and reproducibility inherent in the use of a cell line.
Materials and methods
Development of an 1L-4 hyperresponsive clone Cells of the Ramos cell line were obtained from American Type Culture Collection. The Ramos cell line consists of human B lymphocytes, derived from a Burkitt lymphoma which respond to IL-4 with upregulation of expression of low affinity IgE receptors (CD23) (Rousset et al., 1988). Ramos cells were fluorescently labelled using anti-CD23 monoclonal antibody (The Binding Site, San Diego, CA), and fluorescein-labelled goat anti-mouse immunoglobulin (GAM-FITC, Cappel Laboratories). The dullest 1% of cells were then sorted as single cells into wells of a 96-well flat-bottomed plate using a Coulter Epics 752 flow cytometer with Auto-Clone (Coulter Electronics, Hialeah, FL) and grown into subclones. After clonal cell counts exceeded 104 cells per clone, they were split three ways and grown in the absence or presence of recombinant human IL-4 (Immunex, Seattle, WA) at 20 ng/ml. After 3 days, CD23 expression with and without IL-4 was measured fluorescently on approximately 30 clones and the subclone demonstrating the greatest IL-4 effect on a loga-
rithmic scale of fluorescence was selected and designated Ramos.G6. Ramos.G6 cells, not previously exposed to IL-4, were further expanded. After approximately 2 months of culturing, Ramos.G6 cells were stained for CD23 expression and the brightest 1% were subcloned (precisely one cell/well) and approximately 80 clones were tested for IL-4 responsiveness as above. The subclone demonstrating the greatest IL-4 effect was designated Ramos.G6.C10. Ramos.G6.C10 cells were expanded, frozen in multiple aliquots, and used in the bioassay described below. They have maintained their IL-4 responsiveness in tissue culture (grown without IL-4) for over 6 months. Saturation binding studies show Ramos, Ramos.G6, and Ramos.G6.C10 all to have approximately 1500 IL-4 binding sites per cell.
IL-4 bioassay Human recombinant IL-4 provided by Immunex was used as our IL-4 standard. IL-4 was diluted in RPMI 1640 medium supplemented with gentamicin, amphotericin B, and 15 % fetal calf serum (complete medium). Aliquots of 100 gl of diluted standard were placed into 96-well fiat-bottomed microtiter plates in triple replicates whose final concentrations (after addition of assay cells) ranged by two-fold dilutions from 1.6 or 3.2 ng/ml to 25 or 50 pg/ml. Samples were also added to the plate in various dilutions in 100 gl aliquots in triplicate. Negative control, medium-only samples were placed at the beginning, middle, and end of each plate to test for flow cytometer stability during the assay and other sources of variation dependent upon sample position. Edge wells were not used. Assay cells were harvested 2-3 days after splitting and suspended at 3-5 × 105 cells/ml in complete medium. 100 gl of cell suspension were added to each well containing dilutions of standard or samples. Plates were then placed in a humidified incubator containing room air supplemented with 7% CO 2. After 48 h, cells were transferred to round-bottomed plates, washed three times, serially labelled with anti-CD23 (3 gl/well) and GAM-FITC (4 gl/well), washed twice, and fixed in 1% paraformaldehyde. Fluorescence was
289
measured using a FACScan (Becton Dickinson, Mt. View, CA). A histogram of green fluorescence was measured for each well on a logarithmic scale in which channels 0 through 255 represent 4 logs base 10. While the position of these histograms was a function of IL-4 concentration, they remained unimodal at all concentrations. The fluorescent channel representing the 75 th percentile of cells was determined for each well and is referred to as the fluorescence channel for that sample. The 50 th percentile (median) fluorescence channel gave similar results but a slightly smaller difference in channels between cells treated with and without IL-4 than did use of the 75 th percentile. A standard curve of fluorescence channel as a function of IL-4 concentration was determined using the computer program, ALLFIT v2.7 (De Lean et al., 1988). IL-4 concentration in each sample was determined by comparison with the standard curve.
Reference preparations Reference samples contMning known concentrations of recombinant human IL-4 were prepared by dilution of rIL-4 with complete medium. Aliquots of these preparations were stored at - 70oC.
Cytokines Human recombinant IL-4 and IL-6 were provided by Immunex (Seattle, WA). Human recombinant IL-2 and M-CSF were provided by Cetus (Emeryville, CA). Human recombinant interferon (IFN)-3, and tumor necrosis factors (TNFs), TNF-
a and -fl, were provided by Genentech (So. San Francisco, CA). Human recombinant IFN-a was provided by Hoffman LaRoche (Nutley, N J). Human recombinant GM-CSF was provided by Schering-Plough (Kenilworth, N J). Human recombinant G-CSF was provided by Amgen (Thousand Oaks, CA). IL-lfl was provided by Otsuka (Rockville, MD).
Resd~
IL-4 responses of the cell lines used Parental Ramos cells and the subclones Ramos.G6 and Ramos.G6.C10 were cultured in parallel for 3 days in the absence or presence of IL-4 10 ng/ml. Flow cytometry was performed on unstained cells, and cells stained with either antiCD23 and GAM-FITC or a irrelevant control antibody and GAM-FITC. Fluorescence channels were determined as described above and are shown in Table I; data from the subclone used in our bioassay, Ramos.G6.C10 are also shown converted to a linear scale. Growth in IL-4 resulted in increase in log fluorescence channels of approximately ten channels for parental Ramos, 48 channels for Ramos.G6, and 60 channels for Ramos. G6.C10. Since data are generated on a scale in which 256 channels cover 4 logs base 10, these differences represents IL-4 induced increases in brightness of 43%, 462%, and 766%, respectively. Thus, as a result of the selection used, the clones became progressively more responsive tO IL-4.
TABLE I DEVELOPMENT OF AN IL-4 HYPERRESPONSIVE CLONE Fluorescence channel (log scale) Ramos
Unstained MIgG1 + GaM-FITC a aCD23 + GaM-FITC b
Relative fluorescence (linear scale) Ramos.G6
Ramos.G6.C10
Ramos.G6.C10
-I~4
+Ib4
-IL-4
+IL~
-I~4
+IL~
-
71 100 105
76 105 115
59 87 95
78 103 143
58 83 103
75 110 163
1.1 2.5 5.0
IL-4
+
IL-4
1.8 6.5 43.7
a 1st antibody: routine immunoglobulin G1 against irrelevant antigen. 2nd antibody: fluorescein-conjugated goat anti-murine immunoglobulin G. 2 1st antibody: murine immtmoglobulin G1 against human CD23. 2 nd antibody: fluorescein-conjugated goat-anti-murine immunoglobulin G.
290
l~
-~
3.2 ng/ml
=
/~
1.6 ng/ml
800 pg/ml ?.;~,.~
t'~x',L
¢
,e'"
400 pg/ml
j.,\
\%,.
~
~tI 200 pg/ml
0
".~
100
pg/ml
,.,,"~
0
•
~']
~ ..
.
50 pg/ml
25 pg/ml
~
M_"';~ .
..
0
\\.. Log green fluorescence Fig. 1. Flow cytometric profiles from the h u m a n IL-4 bioassay. The dashed and two dotted lines in each panel are profiles from cells cultured in three replicate wells supplemented with IL-4 at the concentrations indicated. The solid line is one of the profiles from cells cultured in the absence of IL-4. Each profile is on a logarithmic x axis on which full scale covers 4 logs base 10.
Two additional findings are of note in these data. The first subcloning (Ramos.G6) selected a subclone with less background autofluorescence A. IL-4ASSAY 1
(channel 71 vs. 59) allowing a more sensitive detection of CD23 dependent fluorescence. Also, while the linear scale data make it clear that the
B. IL-4 ASSAY 2 160
170
C. IL-4 ASSAY 3 160]
~ 160-
o
D
o
o
1501 a
140
1401
130
1301
~= 140 f-) 130
8
n
150 a
150
0
n
o
0 0
1201
;
120
~ 120
110
110 l
10c
u
I00 o 12.5
~i0t
50
200
800
IL-4 (pg/ml)
IOO t -'~, 012.5 3200
, 9~ 50 200 800 3200 IL-4 (pg/ml)
0 12.5
50 200 800 3200 IL-4 (pg/ml)
Fig. 2. Representative standard curves from the h u m a n IL-4 bioassay. Shown are data from testing the IL-4 standard in triplicate at serial two-fold dilutions from 12.5 to 3200 p g / m l in three separate assays. Fluorescence channels reflect CD23 expression on a logarithmic scale determined as described in the materials and methods SeCtion. Curves shown are hand-drawn; but calculations for samples are based on computer-fitted curves.
291 TABLE II DESCRIPTION OF STANDARD CURVES Assay
Maximal response a (Chan)
B a c k g r o u nbd (Chan)
MaxBkg. c (Chart)
Mean SD d (Chan)
Slope e (Chan/Dil)
EDs0 f (pg/ml)
1 2 3 4 5
161 156 159 155 158
104 103 99 103 101
57 53 60 52 57
1.9 1.8 1.9 2.1 1.7
17 12 14 14 14
274 191 173 136 143
a Cells cultured in IL-4, 3.2 ng/ml. b Cells cultured without IL-4. ¢ Maximal response minus background. d Mean standard deviation among three replicates at each dilution of the standard. e Change in fluorescence channel per two-fold dilution of IL-4 over range of 100-400 pg/ml. f [IL-4] leading to fluorescence channel half way between minimal and maximal response channels.
p r i n c i p a l IL-4 effect o n fluorescence i n this bioassay was u p r e g u l a t i o n of CD23, IL-4 also upregulated autofluorescence a n d nonspeciflc a n t i b o d y b i n d i n g o n these cells.
Dose-response of the IL-4 effect Fig. 1 shows flow cytometric profiles from the s t a n d a r d curve of a representative IL-4 assay. As c a n b e seen, the histograms are u n i m o d a l a n d cells cultured i n IL-4 c o n c e n t r a t i o n s of 50 p g / m l or greater showed a d o s e - d e p e n d e n t increase i n fluorescence. Fig. 2, panels A, B, a n d C, show three representative s t a n d a r d curves of fluorescence c h a n n e l vs. IL-4 c o n c e n t r a t i o n generated i n the m a n n e r described i n the materials a n d m e t h o d s section. As can b e seen, the lowest c o n c e n t r a t i o n of IL-4 i n d u c i n g significant fluorescence above b a c k g r o u n d is 50 or 100 p g / m l . T h e steepest part
of the response curves occurs at IL-4 c o n c e n t r a tions b e t w e e n 100 a n d 400 p g / m l . T h e IL-4 effect begins to reach a p l a t e a u at a c o n c e n t r a t i o n of 800 p g / m l . T a b l e II shows d a t a from five s t a n d a r d curves. T h e fluorescence difference b e t w e e n cells g r o w n with vs. w i t h o u t IL-4 (3200 p g / m l ) ranged from 52 to 60 channels, r e p r e s e n t i n g a 6 . 5 - 8 . 7 - f o l d increase i n fluorescence a t t r i b u t a b l e to IL-4. A s t a n d a r d d e v i a t i o n of a p p r o x i m a t e l y two c h a n n e l s was n o t e d a m o n g replicates dilutions of the standard at all IL-4 c o n c e n t r a t i o n s i n all assays representing replicate v a r i a t i o n i n fluorescence of app r o x i m a t e l y 7%. A t the steepest p o r t i o n of the s t a n d a r d curve each two-fold d i l u t i o n of the s t a n d a r d ( 4 0 0 - 2 0 0 a n d 2 0 0 - 1 0 0 p g / m l final conc e n t r a t i o n ) resulted i n a m e a n fluorescence difference of 14 channels. Thus, the difference in readings b e t w e e n two adjacent dilutions of the
TABLE III IL-4 READINGS OF REFERENCE PREPARATIONS Assay
300 pg/ml
no.
Mean + SD
CV
Mean + SD
200 pg/ml CV
Mean 4-SD
CV
Mean 4-SD
CV
1 2 3 4
3024-41 a 2974-23 286 + 18 ND
14% 8% 6%
1924- 3 1704- 0 ND 168 4-11
2% 0%
18% 22% 11%
ND 91 4-10 684- 9 ND
11% 13%
7%
165 ± 30 1894-41 140+15 ND
Mean
2954- 8 b
3%
1774-13
7%
165 4-25
15%
80 + 16
20%
150 pg/ml
Intra-assay mean + SD of three replicate wells within the assay. b Interassay mean + SD of the results of the assays shown. a
75 pg/rnl
292 standard in the steep portion of the curve (mean difference 14 channels) considerably exceeded the variation among replicates at a single dilution (standard deviation less than two channels). The concentration of IL-4 resulting in a fluorescence readout half way (in channels) between that seen with 0 and 3200 p g / m l (i.e., the EDs0 ) was typically approximately 200 p g / m l .
Precision and reproducibility of the IL-4 bioassay In order to determine the accuracy and reproducibility of the bioassay on samples at various concentrations, reference samples of IL-4 were prepared as described above and tested in triplicate in single concentrations over a 2 month period. Reference samples were tested at final concentrations of 300 p g / m l , 200 p g / m l , 150 p g / m l , and 75 p g / m l . The bioassay results were then used to determine IL-4 content for each concentration tested and the results are presented in Table III. The means of the values determined for reference samples tested at 300, 200, 150, and 75 p g / m l were 295, 177, 165, and 80 p g / m l respectively. The 11 determinations on the four samples had a mean interassay coefficient of variation of 10%. The mean intra-assay coefficient of variation for the four samples was 11%. The difference between measured and actual value for the 11 determinations averaged 11%, exceeded 20% in only two cases, and never exceeded 33%.
Cytokine specificity In order to test the specificity of the response measured for IL-4, several other recombinant human cytokines were tested in the bioassay. A cytokine which has no direct effect in the bioassay might, in the presence of IL-4, have synergistic or antagonistic effects which interfere with the determination of IL-4 levels. Therefore, cytokines were tested not only alone, but also in the presence of 200 p g / m l IL-4, a concentration giving approximately half-maximal activity. As can be seen in Table IV, IL-lfl, IL-2, IL-6, IL-7, G-CSF, M-CSF, and G M - C S F all had no effect at a concentration of 100 n g / m l alone and they did not significantly alter the effect of IL-4 at 200 p g / m l . IFN-t~ and IFN-T had no effect by themselves but antagonized the effects of IL-4. T N F - a and, to a lesser extent, TNF-fl registered weakly
TABLE IV EFFECTS OF OTHER CYTOKINES ON THE IL-4 BIOASSAY Cytokine
Conc. (ng/ml)
G-CSF M-CSF GM-CSF IL-I• IL-2 IL-6
100 100 100 100 100 100 100 50 100 100
IFN-a IFN-T TNF-a TNF-fl
Result (IL-4 pg/ml) a No IL-4 With IL-4 200 pg/ml < 25 237 < 25 187 < 25 195 < 25 187 < 25 196 < 25 217 < 25 122 < 25 145 46 > 800 36 320
a Calculated apparent IL-4 concentrations for samples containing other cytokines in the concentrations indicated together with IL-4 at 0 or 200 pg/ml.
in the bioassay when tested alone but had more significant augmenting effects when used in combination with IL-4. In order to further assess the effects of IFN-a, IFN-T, T N F - a , and TNF-fl in this bioassay, these cytokines were tested at multiple concentrations ranging from 100 p g / m l to 100 n g / m l in the presence of IL-4 at 200 p g / m l . Results are shown in Fig. 3. IFN-a partially inhibited CD23 induction at all concentrations tested. However, even at the highest concentration of IFN-a (500-fold excess over IL-4), the apparent IL-4 level was still approximately 50% of the actual level. IFN-T had a lesser inhibitory effect and no effect of IFN-T was observed at concentration of 1 n g / m l or less. T N F - a and TNF-fl showed a synergy, with IL-4 with effects first observed at concentrations of 100 p g / m l of T N F - a or 1 n g / m l of TNF-fl a n d increasing through the highest concentration tested (100 n g / m l ) . TNF-ct was also tested at the above concentrations in the absence of IL-4 (data not shown). At 100 p g / m l , no effect of T N F - a was observed while at 1, 10, and 100 n g / m l , gradually increasing effects were observed which were significantly different from background but remained less than that seen with 50 p g / m l IL-4. Thus, TNF-c~ and TNF-fl strongly synergize with IL-4 in this bioassay but by themselves have mea-
293 800
700. b~ c~
600.
4--4 A--A
TNFa TNF~
•--$
IFN 7
0--0
IFNa
•
4oo. oo.
i
200t t
0
0
0.1
1
10
100
0ytokine Concentration, ng/rnl Fig. 3. Dose-respo~e relationship of the interaction of IFNs and TNFs with IL-4. IL-4 was tested in the bioassay at a concentration of 200 pg/ml in the presence or absence of varying concentration of TNF-a or -fl or IFN-a or -V in triplicate wells. By comparison of CD23-dependent fluorescence channel with a standard curve generated using serial dilution of IL-4 alone as describedin the materials and methods section, the apparent level of IL-4 in each sample was determined and is shown above.
surable effects below the stated sensitivity threshold of the bioassay (50-100 pg/ml).
Discussion We have described herein a sensitive, reproducible, relatively simple and moderately specific bioassay for human IL-4. The sensitivity of 50-100 p g / m l (3-6 pM) is excellent in light of the K d of 50 pM for the binding of IL-4 to its receptor on human cells. The inter- and intra-assay coefficients of variation of this bioassay are both near 10% and compare quite favorably with those of other bioassays in general and bioassays for cytokines in particular. Of a broad range of cytokines tested, only the T N F s had a direct effect on the bioassay and this effect was minimal. Compared with the commonly used bioassays for human IL-4 involving costimulation of tonsillar or purified blood B cells or peripheral blood T cells, our bioassay, being based on a cell line, offers obvious advantages in the convenience of obtaining cells and the reproducibility of cell behavior. Additionally, the coefficients of variation of bioassays on
freshly isolated cells are much higher and the specificity is usually poor. The ability of some IFNs and TNFs to alter the bioassay results when IL-4 is present might be problematic in some settings. The use of specific neutralizing antibodies to these other cytokines may be of value in minimizing this effect. In comparison with other bioassays commonly in use for human IL-4, the specificity of our assay is excellent. Under some circumstances, proliferation of lectin~stimulated T cells may be supported by IL-2, IL-6 (Garman et al., 1987), or IL-7 (Morissey et al., 1989) in addition to IL-4. Similarly, proliferation of activated B cells may be supported by IL-2 (Zubler et al., 1984), and interferon-a or -7 (Romagnani et al., 1986; Morikawa et al., 1987). Indeed, the two cytokines other than IL-4 found to register weakly in our assay, T N F - a and -fl, have more potent effects on B cell proliferation (Kehrl et al., 1987a,b). And while IFN-7 partially inhibited IL-4 induced CD-23 upregulation on our assay clone, it completely blocked the IL-4 effect on other B cell lines (Rousset et al., 1988). The interactions of IFNs and T N F s with IL-4 may have interesting implications regarding the roles of these cytokines, in conjunction with IL-4, in the regulation of the differentiation and the activation of B cells. In this regard, the ability of IFN-7 to antagonize CD23 induction by IL-4 has already been studied (Defrance et al., 1987; Rousset et al., 1988). T N F - a has been shown to be a growth factor for some B cell malignancies and IFN-a opposed this effect (Cordingley et al., 1988). TNF-fl has been reported not to upregulate CD23 on resting B cells either alone or with anti-IgM or IL-4 but it did increase the induction on B cells by IgM and IL-4 of the activation marker recognized by Leu 21 (Zola and Nikoloutsopolous, 1989). In light of the .potent synergism of T N F s with IL-4 on Ramos.G6.C10 cells, we are currently studying their combined effects on the expression of CD23 and other differentiation markers on resting B cells. While the fluorimetric readout of this bioassay may makes inaccessible to some laboratories, our use of an automated 96-well plate sampler on the flow cytometer has markedly diminished our operator time in performing the assay. Murine cell lines which exhibit a proliferative response to
294 m u r i n e b u t n o t h u m a n I L - 4 have b e e n in use in m u r i n e I L - 4 bioassays. H u - L i et al. (1989) h a v e described such a cell line with the a d v a n t a g e of being relatively insensitive to IL-2. By contrast, cell lines are n o t in w i d e s p r e a d use for b i o a s s a y for h u m a n IL-4. M a e d a et al. (1988) r e p o r t e d the d e v e l o p m e n t of a b i o a s s a y for h u m a n I L - 4 emp l o y i n g a n IL-2 d e p e n d e n t T cell line, Sez 627, which proliferates in r e s p o n s e to h u m a n IL-4. W h i l e their s t u d y p r e s e n t s t o o little d a t a to det e r m i n e the sensitivity, r e p r o d u c i b i l i t y , or accur a c y of t h a t bioassay, it is of n o t e that the rep o r t e d m a x i m a l p r o l i f e r a t i v e responses of 6589 + 411 a n d 9409 + 1101 differed f r o m b a c k g r o u n d b y 13 a n d 7 s t a n d a r d deviations c o m p a r e d with an average of 30 s t a n d a r d d e v i a t i o n s b e t w e e n m a x i m a l r e s p o n s e a n d b a c k g r o u n d in o u r b i o a s s a y ( a p p r o x i m a t e l y 60 channels, two c h a n n e l s p e r SD). Thus, that b i o a s s a y is likely to b e c o n s i d e r a b l y less r e p r o d u c i b l e o r a c c u r a t e t h a n the one rep o r t e d herein. F u r t h e r m o r e , the Sez 627 cells res p o n d e d to IL-2 at 0.01 U / m l to an extent which exceeds their o p t i m a l r e s p o n s e to IL-4. Thus, very small a m o u n t s of IL-2 w o u l d fully m a s k the I L - 4 response. T h e a b i l i t y o f cytokines to i n t e r a c t w i t h I L - 4 in their b i o a s s a y was n o t studied. A n i m m u n o a s s a y has b e e n r e p o r t e d for h u m a n I L - 4 (Chretien et al., 1989) a n d such assays are c o m m e r c i a l l y a v a i l a b l e ( G e n z y m e , 1989). T h e s e i m m u n o a s s a y s have r e p o r t e d sensitivities ( b o t h slightly less t h a n 100 p g / m l ) c o m p a r a b l e to t h a t o b s e r v e d in o u r bioassay. I m m u n o a s s a y s for cytokines m a y be generally a d v a n t a g e o u s over b i o a s says in terms of speed, convenience, a n d specificity with respect to u n r e l a t e d cytokines. W h i l e such assays can b e o f c o n s i d e r a b l e value, i m m u n o a s says in general c a n n o t s u p p l a n t b i o a s s a y s d u e to their i n a b i l i t y to distinguish b i o a c t i v e molecules f r o m m a n y m e t a b o l i t e s , d e g r a d a t i o n p r o d u c t s , den a t u r e d o r otherwise i n a c t i v a t e d molecules, precursors, aggregates, a n d i n c o r r e c t l y synthesized molecules. I n s u m m a r y , we have d e v e l o p e d a cell line for use in a b i o a s s a y for h u m a n IL-4. T h e biological a n d statistical characteristics o f this b i o a s s a y indicate that it will b e highly useful in a v a r i e t y of settings.
References Chretien, I., Van Kimmenade, A., Pearce, M.K., Banchereau, J. and Abrams, J.S. (1989) Development of polyclonal and monoclonal antibodies for immunoassay and neutralization of human interleukin-4. J. Immunol. Methods 117, 67-81. Cordingley, F.T., Bianchi, A., Hoffbrand, A.V., Reittie, J.E., Heslop, H.E., Vyakarnam, A., Turner, M., Meager, A. and Brenner, M.K. (1988) Tumour necrosis factor as an autocfine tumour growth factor for chronic B-cell malignancies. Lancet i, 969-971. Defrance, T., Aubry, J.P., Rousset, F., Vanbervliet, B., Bonnefoy, J.Y., Arai, N., Takebe, Y., Yokota, T., Lee, F., Arai, K., De Vries, J.E. and Bancherean, J. (1987) Human recombinant interleukin 4 induces Fc epsilon receptors (CD23) on normal human B lymphocytes. J. Exp. Med. 165, 14591467. De Lean, A., Mnnson, P.J. and Rodbard, D. (1978) Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay, and physiological dose-response curves. Am. J. Physiol. 235, E97-E102. Finkelman, F.D., Katona, I.M., Urban, Jr., J.F., Holmes, J., Ohara, J., Tung, A.S., Sample, J.vG. and Paul, W.E. (1988) IL-4 is required to generate and sustain in vivo IgE responses. J. Immunol. 141, 2335-2341. Garman, R.D., Jacobs, K.A., Clark, S.C. and Raulet, D.H. (1987) B-cell stimulatory factor 2 (Beta 2 interferon) functions as a second signal for IL-2 production by mature routine T cells. Proc. Natl. Acad. Sci. U.S.A. 84, 7629. Genzyme Corporation (1989) Human Interleukin-4 ELISA: Intertest-4. Product Instruction Booklet. Boston, MA. Heinzel, F.P., Sadick, M.D., Holaday, B.J., Coffman, R.L. and Locksley, R.M. (1989) Reciprocal expression of interferon-T or interleukin 4 during the resolution or progression of murine leishmaniasis. J. Exp. Med. 169, 59-72. Hu-Li, J., Ohara, J., Watson, C., Tsang, W. and Paul, W.E. (1989) Derivation of a T cell line that is highly responsive to IL-4 and IL-2 (CT.4R) and of an IL-2 hyporesponsive mutant of that line (CT.4S). J. Immunol. 142, 800-807. Kehrl, J.H., Miller, A. and Fauci, A.S. (1987a) Efffect of tumor necrosis factor a on mitogen-activated human B cells. J. Exp. Med. 166, 786-791. Kehrl J.H., Alvarez-Mon, M., Delsing, G.A. and Fauci A.S. (1987b) Lymphotoxin is an important T cell-derived growth factor for human B cells. Science 238, 1144-1146. Maeda, M., Noma, T., Hama, K. and Honjo, T. (1988) Application of a human T cell line derived from a Sezary syndrome patient for human interleukin 4 assay. Immunol. Lett. 18, 247-254. Morikawa, K., Kubagawa, H., Suzuki, T. and Cooper, M.D. (1987) Recombinant interferon-a, -fl, and -'t enhance the proliferative response of human B cells. J. Immunol. 139, 761. Morrissey, P.J., Goodwin, R.G., Nordan, R.P., Anderson, D., Grabstein, K.H., Cosman, D., Sims, J., Lupton, S., Acres, B., Reed, S.G., Mochizuki, D., Eisenman, J., Conlon, P.J.
295 and Namen, A.E. (1989) Recombinant interleukin 7, pre-B cell growth factor, has costimulatory activity on purified mature T cells. J. Exp. Med. 169, 707-716 Ohara, J. (1989) Interleukin-4: molecular structure and biochemical characteristics, biological function and receptor expression. Year Immunol. 5, 126-159 Romagnani, S., Giudizi, M.G., Bagotti, R., Almerigogna, F., Mingara, C., Maggi, E., Liang, C. and Moretta, L. (1986) B cell growth factor activity of interferon-gamma: Recombinant human interferon-gamma promotes proliferation of anti-~activated human B lymphocytes. J. Immunol. 136, 3513. Rousset, F., De Waal Malefijt, R., Slierendregt, B., Aubry, J.-P., Bormefoy, J.-Y., Defrance, T., Banchereau, J. and De Vries, J.E. (1988) Regulation of Fc receptor for IgE (CD23) and class II MHC antigen expression on Burkitt's
lymphoma cell lines by human IL-4 and IFN-gamma. J. Immunol. 140, 2625-2632 Yakota, T. Otsuka, T., Mosmann, T., Banchereau, J., Defrance, T., Blanchard, D., De Vries, J.E., Lee, F. and Arai, K. (1986) Isolation and characterization of a human cDNA clone homologous to mouse B-cell stimulatory factor 1, that expresses B-cell- and T-cell-stimulating activities. Proc. Natl. Acad. Sci. U.S.A. 83, 5894-5898. Zola, H. and Nikoloutsopoulos, A. (1989) Effect of recombinant human turnout necrosis factor fl (TNF-fl) on activation, proliferation and differentiation of human B lymphocytes. Immunology 67, 231-236. Zubler, R.H., Lowenthul, J.W., Erard, F., Hashimoto, N., Devos, R. and Macdonald, H.R. (1984) Activated B cells express receptors for and proliferate in response to pure interleukin-2. J. Exp. Med. 160, 1170.
