- Email: [email protected]
An avidin-biotin ELISA assay for the measurement of de novo human IgE synthesis in culture supernatants
Journal of Immunological Methods, 87 (1986) 273-281 Elsevier
273
JIM 03830
An avidin-biotin ELISA assay for the measurement of de novo human IgE synthesis in culture supernatants Yael G. Alevy * and C a t h e r i n e M. Blynn Immunology Research Section, Health Care Division, Monsanto Company, Life Sciences Research Center, 700 ChesterfieM Village Pkwy., Chesterfield, MO 63198, U.S.A. (Received 5 August 1985, accepted 13 November 1985)
A very sensitive (100 pg/ml) solid-phase enzyme immunoassay (ELISA) for the determination of human IgE has been developed. This assay incorporates the avidin-biotin system to increase sensitivity and can detect as little as 100 p g / m l (10 pg/test) of human IgE. The assay is highly specific and allows quantitative determination of human IgE in supernatants of peripheral blood lymphocytes as well as in serum. The very high sensitivity of the assay was accomplished by optimizing concentrations of the following reagents: (1) affinity-purified rabbit anti-human IgE coating antibodies; (2) biotin-conjugated goat anti-human IgE; (3) avidin-horseradish peroxidase (HRP) conjugate. In summary, the assay described is rapid (6 h), reproducible, isotype specific, and has the sensitivity of radioimmunoassays usually employed for the quantification of IgE. This assay may be utilized in establishing concentrations of in vitro IgE levels synthesized by human peripheral blood lymphocyte (PBL). Key words: ELISA; IgE synthesis," Avidin-biotin
Introduction
IgE antibodies have been shown to mediate immediate type hypersensitivities (Ishizaka and Ishizaka, 1967). Serum levels of IgE in normal individuals vary over a wide range, with levels in most individuals between 50-300 n g / m l (Adkinson, 1976). The elevated IgE levels to a particular antigen are employed in the clinical diagnosis of allergic diseases. The measurement of total serum * Correspondence to be addressed to: Dr. Yael G. Alevy, Immunology Research Section, Monsanto Company AA4G, Health Care Division, Life Sciences Research Center, 700 Chesterfield Village Pkwy., Chesterfield, MO 63198, U.S.A. Abbreviations: ELISA, enzyme-linked immunosorbent assay; PRIST, paper radioimmunosorbent test; RAST, radioallergosorbent test; RIA, radioimmunoassay; PBL, peripheral blood lymphocytes; OPD, o-phenylenediamine; BSA, bovine serum albumin; HRD, horseradish peroxidase.
IgE levels is accomplished via the paper radioimmunosorbent test (PRIST) while the measurement of antigen-specific serum IgE is done via the radioallergosorbent test (RAST) (Wide et al., 1967; Ahlstedt et al., 1974; Pepys et al., 1975; Ortolani et al., 1981). The regulation of human IgE synthesis in vitro has recently been studied by a number of investigators. In asmuch as the amounts of IgE produced in vitro are in the pg range the radioimmunoassay (RIA) for total IgE in supernatants of PBL has been the assay of choice. However the hazards and cost associated with the use of radioactivity prompted us to set up a very sensitive avidin-biotin ELISA assay for the measurement of IgE in supernatants of human peripheral blood lymphocytes. Guesdon et al. (1979) as well as Kendall et al. (1983) have reported on an ELISA assay based on
0022-1759/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)
274
avidin-biotin interaction. Avidin, a glycoprotein from egg white binds to biotin via non-covalent but extremely strong interactions ( K D = 1 0 15 M 1) (Green, 1963). Based on this strong interaction, the avidin-biotin complex has been widely used in many fields of biology (Bayer and Wilcheck, 1978, 1980; Wilcheck and Bayer, 1984) and was first used in immunoassays by Guesdon et al. (1979). We report here on the establishment of a very sensitive ELISA assay for human IgE. Using commercially available reagents and by optimizing the concentrations of these reagents as well as incubation times, we were able to reach sensitivity levels in the 100 p g / m l range. This assay was applied to the measurement of human IgE synthesized in vitro by PBL.
Materials and methods Reagents The following reagents were used: BSA (bovine serum albumin) RIA grade (Sigma Chemical Company, St. Louis, MO); horseradish peroxidase avidin-D (HRP-AD, Vector Laboratories, Burlington, CA); o-Phenylenediamine (OPD, Zymed Laboratories, San Francisco, CA); and a substrate buffer for peroxidase (Zymed Laboratories, San Francisco, CA). A ntisera The following antibodies were used to coat plates: monoclonal mouse anti-human IgE clones 4.15 and 7.12 (gift from Dr. Andrew Saxon, Division of Clinical Immunology and Allergy, UCLA School of Medicine, Los Angeles, CA); affinitypurified goat anti-human IgE (TAGO, Burlingame, CA); and affinity-purified rabbit anti-human IgE (CalBiochem-Behring). Affinity-purified goat anti-human IgE-biotin conjugate was employed as the second antibody (TAGO, Burlingame, CA). Standards and controls The IgE preparation from WHO (National Biological Standards Board, London, England 75/502) was dissolved in 0.5 ml distilled water and used as the primary standard. The W H O
standard was used to calibrate a secondary standard which was obtained from Pharmacia Fine Chemicals, Piscataway, NJ. Controls for IgE levels (Phadexact serum) were obtained from Pharmacia Diagnostics. These controls are 3 human sera with different concentrations of IgE. The following controls were used: (1) low control, 24 + 1.5 kU IgE/1 (batch no. 5203); (2) medium control, 214 _+ 10 kU IgE/1 (batch no. 5204); and (3) high control, 362_+ 20 kU IgE/1 (batch no. 5205). Purified human IgA and IgG myeloma proteins were obtained from Tago, Inc., Burlingame, CA. All data is expressed as average of triplicate samples. IgE avidin-biotin E L I S A assay Immulon 2 plates (U bottom, Dynatech Laboratories, Alexandria, VA) were coated with varying concentrations of coating antibodies specific for human IgE. The plates were incubated overnight at 4°C and then washed 5 times with phosphate-buffered saline (PBS), pH 7.4 using the Titertek Microplate washer 120 (Flow Laboratories, McLean, VA). The unbound sites were then blocked with 1% BSA in PBS (250 /~l/well) and the plates were then incubated at room temperature for 1 h. Following that incubation the plates were washed 5 times with PBS and the standards or supernatants to be assayed were added to the plates and incubated for 2 h at 37°C. Following that the goat anti-human IgE-biotin conjugate was added and the plates were incubated for an additional 2 h at 37°C, washed 5 times with PBS and the avidin-D-HRP was then added at 100 ~l/well after which the plates were incubated for 30 min at 37°C. Plates were then washed with PBS and the OPD substrate was added to wells. De novo IgE synthesis by human P B L In vitro immunoglobulin synthesis studies were conducted using a modification of the method described by Waldmann et al. (1974), and Sampson and Buckley (1981). Unfractionated PBL (2 × 106) were placed in 24-well plates in 1 ml of RPMI 1640 medium containing N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid buffer (Hepes 2.5 mM), gentamycin (4 mg/100 ml), and L-glutamine (2 mM) and supplemented with 10%
275 fetal calf serum (FCS). Cultures were either stimulated or unstimulated with 5 ~tg/ml PWM (Gibco, G r a n d Island, NY) and then incubated for 10 days at 37°C in a humidified 5% CO2 atmosphere. Supernatants were collected from cultures after centrifugation at 900 × g for 10 min. A number of the cultures were frozen and thawed 5 times and then incubated for 10 days. These cultures were used to determine background release of IgE.
Statistical analysis of data The data was collected using an IBM-PC computer connected to an ELISA reader (MR600, Dynatech Laboratories, Alexandria, VA). The data was analyzed using the ' I m m u n o s o f t ' statistical software obtained from Dynatech Laboratories. The software has the ability to store raw data, average replicates and subtract a blank from each value. In addition, the software is capable of removing outliers and drawing standard curves based on regression analysis. The IgE levels were automatically calculated from a standard curve run on each plate. All samples were run in triplicates.
Results
Determination of optimal concentrations of coating antibody' In order to determine the appropriate coating antibody and the optimal concentration of this antibody, plates were incubated overnight with varying concentrations of either mouse anti-human IgE monoclonal antibodies (clones 4.15 and 7.12) or affinity-purified goat anti-human IgE antibodies. When the plates were coated with varying concentrations of mouse anti-human IgE monoclonal antibodies (Fig. 1), a combination of the 2 monoclonal antibodies provided the best linear standard curve (Fig. 1C). We next compared varying concentrations of a mixture of the monoclonal antibodies to the affinity-purified goat anti-human IgE antibodies. Fig. 2 demonstrates that the best linear curve drawn by the Immunosoft software was obtained when the mixture of 4.15 and 7.12 antibodies at a concentration of 0.7 /~g/ml was used to coat the plates. These assays were performed with a goat anti-human IgE-biotin con-
jugate concentration of 1/6000 and an avidin-DH R P concentration of 1/3000.
Determination of goat anti-human IgE-biotin conjugate and avidin-D-HRP optimal concentrations To establish the optimal concentration of biotin-conjugated antibodies and avidin-D-HRP, varying concentrations of avidin-D-HRP (1/500, 1/1000, 1/2000 and 1/3000) and biotin-conjugated antibodies (1/2000, 1/4000 and 1/6000) were used in combination. Fig. 3 demonstrates that when avidin-D-HRP is used at 1/1000 and biotin-conjugated antibodies at 1/4000 a good linear curve is obtained. In addition, when avidinD - H R P is used at 1/1000 and biotin-conjugated antibodies at 1/4000, background O D are very low (0.041-0.184). We then decided to compare this mixture of mouse anti-human monoclonal antibodies at an avidin-D-HRP concentration of 1/1000 and a biotin-conjugated antibodies concentration of 1/4000 to varying concentrations of rabbit anti-human IgE-coating antibodies in order to see which produced the best linear curve (as drawn by the Immunosoft software) and the best sensitivity. Our results indicate (Fig. 4) that the best linear curve with the lowest O D background is obtained when the affinity-purified rabbit anti-human IgE is used at 0.5/~g/ml. Fig. 5 demonstrates that when plates are coated with rabbit anti-human IgE at 0.5/~g/ml the standard curve is linear from 78 p g / m l to 5000 pg/ml.
Reproducibility of the assay The standard curves shown in Fig. 6 were all run on different plates on the same day. Although they are all linear at the range of 78 p g / m l to 2500 p g / m l the absolute O D readings vary from plate to plate. In addition the background O D values also vary from 0.108 to 0.184. It was thus concluded that a standard curve should be included on each plate in order to insure accuracy. In addition, outside wells on the plate cannot be used (data not shown).
Specificity of the assay To determine whether this assay is specific for IgE, varying concentrations of human IgG or human IgA were substituted for IgE. Our results
276 0,9'
/e 4.15 + 7.12 1 i-tg,ml . . . . . . 3 ~g,ml - - - - -- 5 ~.tg,ml
0.8"
0.7. 0.6-
,/~
4.15
~ ' .~,// ~,~/
7.12 1 ~tg/ml . . . . . . . 3 ~g/ml .... 5 ~g/ml
- 1 p.g/ml . . . . . . 3 ~tg/ml -- -- -- 5 ~g/ml
/,/ //
0.5"
/'
//z~/
0.4.
0.3"
1t
/ t
.
~"r I"I
A
~
..~ ~
~.I"
IgE concentration
//"
B
picograms/ml
Fig. 1. Determination of optimal concentrations of coating antibodies. Varying concentrations of mouse anti-human antibodies were used with 1 / 6 0 0 0 biotin and 1 / 3 0 0 0 avidin.
A
M o u s e anti H u m a n IgE 4.15 + 7.12 ...... 1.1
B
0.3 txg/ml
- - - -- 0.5 txg/ml .......... 0.7 ~tg/ml
------0.5 ixg/ml .......... 0.7 ~tg/ml
1.0
lgE monoclonal
G o a t anti H u m a n IgE .....
0.3 ~ g / m l
/~ ~1
/A /
I, ~='~"
~
/
/
.!
0.2. 0.1.
//
e
1 ~g/ml
. ..'" °
- -
1 i~g/ml
.....
3 ~g/ml
0.90.8-
°°°°°°°°°•oO°°°•°°•°°°°°°'°° 0.7"
E
•°°° °°.°°
" 0.604 O~
~
0.5-
~ I'
.,.0 •°°o°° f" °°o° ~. • °•°°• f ° °o°o° ° °°°°•
°o" °°°o°
O 0.4"
°o°°°°°°
f' ~'7
IJ
• f
°
•.•°°
•
0.30.2-
°°°°°'°
jf
0.1-
•. . " - ' 2 . f ' "
~ " ~ ' "
........
v
,.
!
!
IgE concentration picograms/ml Fig. 2. Determination of optimal concentrations of coating antibodies. Varying concentrations of mouse anti-human IgE monoclonal antibodies and goat anti-human antibodies were used with 1 / 6 0 0 0 biotin and 1 / 3 0 0 0 avidin.
277 1.4. 1.3-
°..°. °.,1~ °°o,°°°°°"
1.2-
.......y""
1.11.0-
.."
0.6-
°*°°
0.5-
:"
`. ./ t t
i
i
., .,.,
I
1/4000
`./
°..
`.
.°°
1/2ooo
0.20.1-
.,o°°'°'°'"
.it.,
Avidin 1/500 Biotin
~.~f/~
0.3-/..,p:,._ 1.31.2.
,, o°.,°o-°'°°°'" .,,,°.,°-"
t""
:, :°
0,6-
i "~
/i
"
0.7-
0.4-
.°o
°.°°°."
0,9-
O
B
A
°j,'°
/ ~ , s I¢"
D
Avidinl/2000 Biotin 1/2000 1/4000
1.1-
1,0.
`.i
1 / /'~
Avidin 1/1000 Biotin 1/2000 1/4000 1/6000
° ° • "o" s ~ "
-- -- -- 1/6000 C
i
Avidin 1/3000 Biotin 1/2000 1/4000 1/6000
/
0.90.80.70.60.50.40.30.20.1i
IgE concentration picograms/ml
Fig. 3. D e t e r m i n a t i o n of goat a n t i - h u m a n I g E - b i o t i n c o n j u g a t e and a v i d i n - D - H R P o p t i m a l concentrations.
indicate (data not shown) that the assay is highly specific for IgE as 100 ng/ml of IgG give an OD reading of 0.047 while 10 ng/ml of IgE give an OD reading greater than 2.0. In a similar manner 100 ng/ml of IgA gives a reading of but 0.04.
Detection of de novo human IgE synthesis in vitro
Human peripheral blood lymphocytes synthesize minute amounts (in the pg/ml range) of T A B L E II REPRODUCIBILITY A S S A Y F O R IgE
No mitogen
PWM 5/~g/ml
5 × freeze and t h a w
De novo IgE
1 2 3 4 5 6
4000 11 500 2500 100 1000 750
3750 2500 2750 1 600 2000 2500
2250 100 150 1 200 100 100
1750 11400 2350 0 900 650
AVIDIN-BIOTIN
IgE serum c o n c e n t r a t i o n a (ng/ml) Low
Medium
High
1 2 3 4 5
54.197 52.301 47.761 52.323 ND
530.5 502.16 514.23 516.14 573.26
ND 587.44 663.26 671.39 ND
OF MONONUCLEAR
Donor no.
THE
Plate No. TABLE I I g E L E V E L S IN S U P E R N A T A N T S CELLS (pg/ml)
OF
ELISA
a Three sera (low, m e d i u m and high, Phadexact, P h a r m a c i a ; see m e t h o d s section for c o n c e n t r a t i o n s ) were diluted and assayed for lgE. The samples were placed on 5 different plates, each plate had a s t a n d a r d curve. The plates were all assayed on the same day.
1.5 /
1.4
. . . . . 0.5 ~g/ml . . . . 0.7 ~g/ml ......... 1 ~g/ml 2 ~g/ml
1.3 1.2
E t-
. . . . . . . 0.5 0.g/ml ............ 1 ~xg/ml ..... 2 ~g/ml
A
///"
a
,// p /.°" ,/ :" / F
/ 1/.//~ ..:" / /
4 ~g/ml
1.1
/ :: ////
1.0
' / " ~ /.."
0.9-
,/
~P 0.8Q 0 0.7"
,,.:7
0.6-
/
0.5-
/,°
. -"*"
0.4-
f y..S22
0.3-
22" 7"
0.2-
~
o.ff"
I , ~'°°°°°° ~ "t o°.°,*° ,~ °°°°
0.14000
,/
1
20,000
20,000 '
40bo 36,000 IgE concentration picograms,,m!
36,000 '
Fig. 4. Determination of optimal concentrations of coating antibodies. Varying concentrations of mouse anti-human monoclonal antibodies (7.12 + 4.15 mixture, A) and rabbit anti-human IgE coating antibodies (B) were used with 1/4000 biotin and 1/1000 avidin.
B
A
1.0-
0.81.5"
°°o*
0.71.3-
..-'"'"'""
E c e~
1.1
a0
0.9
rE 0.6-
°°°°°° o°°°°°,°"
~" 0.5-
,z.,°°
0.4-
0.70.3-
0.5-
°°.,..°°'°~
0.2-
0.30.1-
0.1-
s6o
2s,boo
sobo
IgE concentration picograms/ml Fig. 5. Sensitivity of the assay. Two representative standard curves.
160
1500 560 1000 IgE concentration picograms/ml
279 2.0 1.9 =
1.8........
0
1.7-
1.6j¢
/
1.5-
1.4-
J
1.3-
/ j
1.2-
E "
04 O~ ,I¢
c~ 0
A/"
j 1.1-
~/
j
0.9
/ J ,,'2.."
II / ,/~7"I" i"..°°'°" 1/ ././~...
0.8 0.7 ¸
/
i¢1
°°
• o°OO°
0.6
/~ .." / tN" .."
A/ 0.5
, ~
..:
~
,
7,,"-
///" J j -I /
L/
~
/
/
,,,,¢,, = 7/"
/./...."
0.4
~// /// /
/ I / .JI • ..."°°°
/.'**
/'°.°°° /" ..°° /" .°°"
,J/ //.." j I ." / /'..."
1.0-
/~. °°°
z.,...":/%" 0.3
• H? .." ." 9"5" /
0.2
0.1 u
i
,
,
!
i
IgE concentration picogramstml
Fig. 6. Reproducibility of the assay. Six different standard curves run at the s a m e time on 6 different plates.
IgE antibodies in vitro. One of the major obstacles in studying the regulation of human IgE synthesis in vitro has been the absence of a sensitive immunoassay for human IgE. Investigators currently use IRA assays for the measurement of human IgE (Sampson and Buckley, 1981; Zuraw et al., 1981; Hassner and Saxon, 1983). Using the ELISA assay described here we are able to detect de novo IgE synthesis by human peripheral blood lymphocyte in vitro (Table I). De novo IgE synthesis is determined by subtracting values obtained from cul-
tures frozen and thawed 5 times on day 0 and then cultured for 10 days (Sampson and Buckley, 1981) from the values obtained when PBL are cultured for 10 days with no mitogen. The addition of PWM to the cultures had no significant effect on the de novo synthesis. Although this assay was optimized for the~measurement of IgE in culture supernatants, I g E levels in serum can also be detected (data not shown). Because of the assay sensitivity, the serum has to be diluted. The degree of dilution depends upon the IgE concentration in the serum.
280
Accuracy of the assay To establish the reproducibility and accuracy of the assay, 3 sera of known IgE concentration (Pharmacia Phadexact sera) were assayed for the IgE content. The results shown in Table II indicate that when the sera are assayed on the same day on 5 different plates (with a standard curve on each plate) good reproducibility is observed. It was thus concluded that data can be compared from plate to plate provided that each plate has a standard curve. In addition our results indicate that when the same sample is placed at varying locations on the same plate, very good reproducibility is obtained (data not shown). It was therefore concluded that position on the plate is not a factor as long as outside wells are not utilized.
Discussion
The avidin-biotin ELISA assay described here for the measurement of human IgE can be used for the measurement of human de novo IgE synthesis in vitro by PBL. The assay is very sensiTABLE II1 W I T H I N ASSAY VARIATIONS ~l Plate no.
Location on plate
IgE serum concentration (ng/ml) Low
Medium
High
1 2 3
49.7 43.7
542.1 537.2 571.8
467.3 484.7 -
1 2 3 4
42.3 39.7 44.8
622.6 562.7 576.8
479.8 506.9 533.3 488.5
1 2 3 4
49.9 54.0 50.7 54.5
566.8 522.4 632.9
436.4 495.1
1 2 3
63.3 60.4
742.3 608.6 795.2
529.8 506.5 -
Three sera (low, medium and high, Phadexact, Pharmacia) were diluted and assayed for lgE. The samples were placed on 4 different plates. On each plate the same sample was assayed 2 - 4 times. Each plate had a standard curve.
tive (100 p g / m l - 1 0 pg/test) and it utilizes commercially available reagents. A sensitive and reliable assay for the measurement of human IgE in culture supernatants is very important because of the minute amounts of IgE synthesized by PBL and because of the variabilities observed between laboratories in the quantification of IgE in cell culture supernatants (Helm et al., 1985). The assay commonly used for the detection of IgE in culture supernatants is the RIA (Sampson and Buckley, 1981; Zuraw et al., 1981; Hassner and Saxon, 1983). The avidin-biotin ELISA assay described herein utilizes affinity-purified rabbit anti-human IgE antibodies to coat the plates. The optimal concentration for coating was determined to be 0.5 ~ g / m l . At that concentration one has to use antibodies coupled to biotin at 1/4000 and avidin-D-HRP at 1/1000. Because of the assay sensitivity it is important to include a standard curve on each 96-well plate. The standard curve is linear between 78 p g / m l to 2500 p g / m l . The assay is highly reproducible as demonstrated by the data obtained when samples of known concentrations are assayed using different plates with a standard curve on each plate (Table III). In summary, the assay described here is rapid (6 h), reproducible, isotype specific and has the sensitivity of radioimmunoassays usually employed for quantification of IgE. This assay can be utilized to establish concentrations of IgE synthesized by human PBL in vitro.
References Adkinson, N.F., 1976, Manual of Clinical Immunology, eds. N.R. Rose and H. Friedman (American Society for Microbiology, Washington, DC) p. 590. Ahlstedt, S., N. Eriksson, S. Lindgren and A. Roth, 1974, Clin. Allergy 4, 131. Bayer, E.A. and M. Wilcheck, 1978, Trends Biochem. Sci. 3, N257. Bayer, E.A. and M. Wilcheck, 1980, Methods Biochem. Anal. 26, 1. Green, N.M., 1963, Biochem. J. 89, 585. Guesdon, J., T. Ternyck and S. Avrameas, 1979, J. Histochem. Cytochem. 27, 1131. Hassner, A. and A. Saxon, 1983, J. Immunol. 130, 1567. Helm, R.M., R.H. Buckley, N.F. Adkinson and D.L. Squillance, 1985, J. Allergy Clin. Immunol. 75, 138.
281 Ishizaka, K. and T. Ishizaka, 1967, J. Immunol. 99, 1187. Kendall, C., I. Ionescu-Matiu and G.R. Dressman, 1983, J. Immunol. Methods 56, 329. Ortolani, C., A. Miadonna, R. Adomi, M. Restuccia and C. Zanussi, 1981, Clin. Allergy 11,249. Pepys, J., A. Roth and K.B. Carroll, 1975, Clin. Allergy 5, 431. Sampson, H.A. and R.H. Buckley, 1981, J. Immunol. 127, 3.
Waldmann, T.A., M. Duam, S. Broder, M. Blackman, R.M. Blease and W. Strober, 1974, Lancet ii, 609. Wide, L., H. Bennich and S.G.O. Johansson, 1967, Lancet ii, 1105. Wilcheck, M. and E.A. Bayer, 1984, Immunol. Today 5, 39. Zuraw, B.L., M. Nonaka, C. O'Hair and D.H. Katz, 1981, J. Immunot. 127, 1169.
273
JIM 03830
An avidin-biotin ELISA assay for the measurement of de novo human IgE synthesis in culture supernatants Yael G. Alevy * and C a t h e r i n e M. Blynn Immunology Research Section, Health Care Division, Monsanto Company, Life Sciences Research Center, 700 ChesterfieM Village Pkwy., Chesterfield, MO 63198, U.S.A. (Received 5 August 1985, accepted 13 November 1985)
A very sensitive (100 pg/ml) solid-phase enzyme immunoassay (ELISA) for the determination of human IgE has been developed. This assay incorporates the avidin-biotin system to increase sensitivity and can detect as little as 100 p g / m l (10 pg/test) of human IgE. The assay is highly specific and allows quantitative determination of human IgE in supernatants of peripheral blood lymphocytes as well as in serum. The very high sensitivity of the assay was accomplished by optimizing concentrations of the following reagents: (1) affinity-purified rabbit anti-human IgE coating antibodies; (2) biotin-conjugated goat anti-human IgE; (3) avidin-horseradish peroxidase (HRP) conjugate. In summary, the assay described is rapid (6 h), reproducible, isotype specific, and has the sensitivity of radioimmunoassays usually employed for the quantification of IgE. This assay may be utilized in establishing concentrations of in vitro IgE levels synthesized by human peripheral blood lymphocyte (PBL). Key words: ELISA; IgE synthesis," Avidin-biotin
Introduction
IgE antibodies have been shown to mediate immediate type hypersensitivities (Ishizaka and Ishizaka, 1967). Serum levels of IgE in normal individuals vary over a wide range, with levels in most individuals between 50-300 n g / m l (Adkinson, 1976). The elevated IgE levels to a particular antigen are employed in the clinical diagnosis of allergic diseases. The measurement of total serum * Correspondence to be addressed to: Dr. Yael G. Alevy, Immunology Research Section, Monsanto Company AA4G, Health Care Division, Life Sciences Research Center, 700 Chesterfield Village Pkwy., Chesterfield, MO 63198, U.S.A. Abbreviations: ELISA, enzyme-linked immunosorbent assay; PRIST, paper radioimmunosorbent test; RAST, radioallergosorbent test; RIA, radioimmunoassay; PBL, peripheral blood lymphocytes; OPD, o-phenylenediamine; BSA, bovine serum albumin; HRD, horseradish peroxidase.
IgE levels is accomplished via the paper radioimmunosorbent test (PRIST) while the measurement of antigen-specific serum IgE is done via the radioallergosorbent test (RAST) (Wide et al., 1967; Ahlstedt et al., 1974; Pepys et al., 1975; Ortolani et al., 1981). The regulation of human IgE synthesis in vitro has recently been studied by a number of investigators. In asmuch as the amounts of IgE produced in vitro are in the pg range the radioimmunoassay (RIA) for total IgE in supernatants of PBL has been the assay of choice. However the hazards and cost associated with the use of radioactivity prompted us to set up a very sensitive avidin-biotin ELISA assay for the measurement of IgE in supernatants of human peripheral blood lymphocytes. Guesdon et al. (1979) as well as Kendall et al. (1983) have reported on an ELISA assay based on
0022-1759/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)
274
avidin-biotin interaction. Avidin, a glycoprotein from egg white binds to biotin via non-covalent but extremely strong interactions ( K D = 1 0 15 M 1) (Green, 1963). Based on this strong interaction, the avidin-biotin complex has been widely used in many fields of biology (Bayer and Wilcheck, 1978, 1980; Wilcheck and Bayer, 1984) and was first used in immunoassays by Guesdon et al. (1979). We report here on the establishment of a very sensitive ELISA assay for human IgE. Using commercially available reagents and by optimizing the concentrations of these reagents as well as incubation times, we were able to reach sensitivity levels in the 100 p g / m l range. This assay was applied to the measurement of human IgE synthesized in vitro by PBL.
Materials and methods Reagents The following reagents were used: BSA (bovine serum albumin) RIA grade (Sigma Chemical Company, St. Louis, MO); horseradish peroxidase avidin-D (HRP-AD, Vector Laboratories, Burlington, CA); o-Phenylenediamine (OPD, Zymed Laboratories, San Francisco, CA); and a substrate buffer for peroxidase (Zymed Laboratories, San Francisco, CA). A ntisera The following antibodies were used to coat plates: monoclonal mouse anti-human IgE clones 4.15 and 7.12 (gift from Dr. Andrew Saxon, Division of Clinical Immunology and Allergy, UCLA School of Medicine, Los Angeles, CA); affinitypurified goat anti-human IgE (TAGO, Burlingame, CA); and affinity-purified rabbit anti-human IgE (CalBiochem-Behring). Affinity-purified goat anti-human IgE-biotin conjugate was employed as the second antibody (TAGO, Burlingame, CA). Standards and controls The IgE preparation from WHO (National Biological Standards Board, London, England 75/502) was dissolved in 0.5 ml distilled water and used as the primary standard. The W H O
standard was used to calibrate a secondary standard which was obtained from Pharmacia Fine Chemicals, Piscataway, NJ. Controls for IgE levels (Phadexact serum) were obtained from Pharmacia Diagnostics. These controls are 3 human sera with different concentrations of IgE. The following controls were used: (1) low control, 24 + 1.5 kU IgE/1 (batch no. 5203); (2) medium control, 214 _+ 10 kU IgE/1 (batch no. 5204); and (3) high control, 362_+ 20 kU IgE/1 (batch no. 5205). Purified human IgA and IgG myeloma proteins were obtained from Tago, Inc., Burlingame, CA. All data is expressed as average of triplicate samples. IgE avidin-biotin E L I S A assay Immulon 2 plates (U bottom, Dynatech Laboratories, Alexandria, VA) were coated with varying concentrations of coating antibodies specific for human IgE. The plates were incubated overnight at 4°C and then washed 5 times with phosphate-buffered saline (PBS), pH 7.4 using the Titertek Microplate washer 120 (Flow Laboratories, McLean, VA). The unbound sites were then blocked with 1% BSA in PBS (250 /~l/well) and the plates were then incubated at room temperature for 1 h. Following that incubation the plates were washed 5 times with PBS and the standards or supernatants to be assayed were added to the plates and incubated for 2 h at 37°C. Following that the goat anti-human IgE-biotin conjugate was added and the plates were incubated for an additional 2 h at 37°C, washed 5 times with PBS and the avidin-D-HRP was then added at 100 ~l/well after which the plates were incubated for 30 min at 37°C. Plates were then washed with PBS and the OPD substrate was added to wells. De novo IgE synthesis by human P B L In vitro immunoglobulin synthesis studies were conducted using a modification of the method described by Waldmann et al. (1974), and Sampson and Buckley (1981). Unfractionated PBL (2 × 106) were placed in 24-well plates in 1 ml of RPMI 1640 medium containing N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid buffer (Hepes 2.5 mM), gentamycin (4 mg/100 ml), and L-glutamine (2 mM) and supplemented with 10%
275 fetal calf serum (FCS). Cultures were either stimulated or unstimulated with 5 ~tg/ml PWM (Gibco, G r a n d Island, NY) and then incubated for 10 days at 37°C in a humidified 5% CO2 atmosphere. Supernatants were collected from cultures after centrifugation at 900 × g for 10 min. A number of the cultures were frozen and thawed 5 times and then incubated for 10 days. These cultures were used to determine background release of IgE.
Statistical analysis of data The data was collected using an IBM-PC computer connected to an ELISA reader (MR600, Dynatech Laboratories, Alexandria, VA). The data was analyzed using the ' I m m u n o s o f t ' statistical software obtained from Dynatech Laboratories. The software has the ability to store raw data, average replicates and subtract a blank from each value. In addition, the software is capable of removing outliers and drawing standard curves based on regression analysis. The IgE levels were automatically calculated from a standard curve run on each plate. All samples were run in triplicates.
Results
Determination of optimal concentrations of coating antibody' In order to determine the appropriate coating antibody and the optimal concentration of this antibody, plates were incubated overnight with varying concentrations of either mouse anti-human IgE monoclonal antibodies (clones 4.15 and 7.12) or affinity-purified goat anti-human IgE antibodies. When the plates were coated with varying concentrations of mouse anti-human IgE monoclonal antibodies (Fig. 1), a combination of the 2 monoclonal antibodies provided the best linear standard curve (Fig. 1C). We next compared varying concentrations of a mixture of the monoclonal antibodies to the affinity-purified goat anti-human IgE antibodies. Fig. 2 demonstrates that the best linear curve drawn by the Immunosoft software was obtained when the mixture of 4.15 and 7.12 antibodies at a concentration of 0.7 /~g/ml was used to coat the plates. These assays were performed with a goat anti-human IgE-biotin con-
jugate concentration of 1/6000 and an avidin-DH R P concentration of 1/3000.
Determination of goat anti-human IgE-biotin conjugate and avidin-D-HRP optimal concentrations To establish the optimal concentration of biotin-conjugated antibodies and avidin-D-HRP, varying concentrations of avidin-D-HRP (1/500, 1/1000, 1/2000 and 1/3000) and biotin-conjugated antibodies (1/2000, 1/4000 and 1/6000) were used in combination. Fig. 3 demonstrates that when avidin-D-HRP is used at 1/1000 and biotin-conjugated antibodies at 1/4000 a good linear curve is obtained. In addition, when avidinD - H R P is used at 1/1000 and biotin-conjugated antibodies at 1/4000, background O D are very low (0.041-0.184). We then decided to compare this mixture of mouse anti-human monoclonal antibodies at an avidin-D-HRP concentration of 1/1000 and a biotin-conjugated antibodies concentration of 1/4000 to varying concentrations of rabbit anti-human IgE-coating antibodies in order to see which produced the best linear curve (as drawn by the Immunosoft software) and the best sensitivity. Our results indicate (Fig. 4) that the best linear curve with the lowest O D background is obtained when the affinity-purified rabbit anti-human IgE is used at 0.5/~g/ml. Fig. 5 demonstrates that when plates are coated with rabbit anti-human IgE at 0.5/~g/ml the standard curve is linear from 78 p g / m l to 5000 pg/ml.
Reproducibility of the assay The standard curves shown in Fig. 6 were all run on different plates on the same day. Although they are all linear at the range of 78 p g / m l to 2500 p g / m l the absolute O D readings vary from plate to plate. In addition the background O D values also vary from 0.108 to 0.184. It was thus concluded that a standard curve should be included on each plate in order to insure accuracy. In addition, outside wells on the plate cannot be used (data not shown).
Specificity of the assay To determine whether this assay is specific for IgE, varying concentrations of human IgG or human IgA were substituted for IgE. Our results
276 0,9'
/e 4.15 + 7.12 1 i-tg,ml . . . . . . 3 ~g,ml - - - - -- 5 ~.tg,ml
0.8"
0.7. 0.6-
,/~
4.15
~ ' .~,// ~,~/
7.12 1 ~tg/ml . . . . . . . 3 ~g/ml .... 5 ~g/ml
- 1 p.g/ml . . . . . . 3 ~tg/ml -- -- -- 5 ~g/ml
/,/ //
0.5"
/'
//z~/
0.4.
0.3"
1t
/ t
.
~"r I"I
A
~
..~ ~
~.I"
IgE concentration
//"
B
picograms/ml
Fig. 1. Determination of optimal concentrations of coating antibodies. Varying concentrations of mouse anti-human antibodies were used with 1 / 6 0 0 0 biotin and 1 / 3 0 0 0 avidin.
A
M o u s e anti H u m a n IgE 4.15 + 7.12 ...... 1.1
B
0.3 txg/ml
- - - -- 0.5 txg/ml .......... 0.7 ~tg/ml
------0.5 ixg/ml .......... 0.7 ~tg/ml
1.0
lgE monoclonal
G o a t anti H u m a n IgE .....
0.3 ~ g / m l
/~ ~1
/A /
I, ~='~"
~
/
/
.!
0.2. 0.1.
//
e
1 ~g/ml
. ..'" °
- -
1 i~g/ml
.....
3 ~g/ml
0.90.8-
°°°°°°°°°•oO°°°•°°•°°°°°°'°° 0.7"
E
•°°° °°.°°
" 0.604 O~
~
0.5-
~ I'
.,.0 •°°o°° f" °°o° ~. • °•°°• f ° °o°o° ° °°°°•
°o" °°°o°
O 0.4"
°o°°°°°°
f' ~'7
IJ
• f
°
•.•°°
•
0.30.2-
°°°°°'°
jf
0.1-
•. . " - ' 2 . f ' "
~ " ~ ' "
........
v
,.
!
!
IgE concentration picograms/ml Fig. 2. Determination of optimal concentrations of coating antibodies. Varying concentrations of mouse anti-human IgE monoclonal antibodies and goat anti-human antibodies were used with 1 / 6 0 0 0 biotin and 1 / 3 0 0 0 avidin.
277 1.4. 1.3-
°..°. °.,1~ °°o,°°°°°"
1.2-
.......y""
1.11.0-
.."
0.6-
°*°°
0.5-
:"
`. ./ t t
i
i
., .,.,
I
1/4000
`./
°..
`.
.°°
1/2ooo
0.20.1-
.,o°°'°'°'"
.it.,
Avidin 1/500 Biotin
~.~f/~
0.3-/..,p:,._ 1.31.2.
,, o°.,°o-°'°°°'" .,,,°.,°-"
t""
:, :°
0,6-
i "~
/i
"
0.7-
0.4-
.°o
°.°°°."
0,9-
O
B
A
°j,'°
/ ~ , s I¢"
D
Avidinl/2000 Biotin 1/2000 1/4000
1.1-
1,0.
`.i
1 / /'~
Avidin 1/1000 Biotin 1/2000 1/4000 1/6000
° ° • "o" s ~ "
-- -- -- 1/6000 C
i
Avidin 1/3000 Biotin 1/2000 1/4000 1/6000
/
0.90.80.70.60.50.40.30.20.1i
IgE concentration picograms/ml
Fig. 3. D e t e r m i n a t i o n of goat a n t i - h u m a n I g E - b i o t i n c o n j u g a t e and a v i d i n - D - H R P o p t i m a l concentrations.
indicate (data not shown) that the assay is highly specific for IgE as 100 ng/ml of IgG give an OD reading of 0.047 while 10 ng/ml of IgE give an OD reading greater than 2.0. In a similar manner 100 ng/ml of IgA gives a reading of but 0.04.
Detection of de novo human IgE synthesis in vitro
Human peripheral blood lymphocytes synthesize minute amounts (in the pg/ml range) of T A B L E II REPRODUCIBILITY A S S A Y F O R IgE
No mitogen
PWM 5/~g/ml
5 × freeze and t h a w
De novo IgE
1 2 3 4 5 6
4000 11 500 2500 100 1000 750
3750 2500 2750 1 600 2000 2500
2250 100 150 1 200 100 100
1750 11400 2350 0 900 650
AVIDIN-BIOTIN
IgE serum c o n c e n t r a t i o n a (ng/ml) Low
Medium
High
1 2 3 4 5
54.197 52.301 47.761 52.323 ND
530.5 502.16 514.23 516.14 573.26
ND 587.44 663.26 671.39 ND
OF MONONUCLEAR
Donor no.
THE
Plate No. TABLE I I g E L E V E L S IN S U P E R N A T A N T S CELLS (pg/ml)
OF
ELISA
a Three sera (low, m e d i u m and high, Phadexact, P h a r m a c i a ; see m e t h o d s section for c o n c e n t r a t i o n s ) were diluted and assayed for lgE. The samples were placed on 5 different plates, each plate had a s t a n d a r d curve. The plates were all assayed on the same day.
1.5 /
1.4
. . . . . 0.5 ~g/ml . . . . 0.7 ~g/ml ......... 1 ~g/ml 2 ~g/ml
1.3 1.2
E t-
. . . . . . . 0.5 0.g/ml ............ 1 ~xg/ml ..... 2 ~g/ml
A
///"
a
,// p /.°" ,/ :" / F
/ 1/.//~ ..:" / /
4 ~g/ml
1.1
/ :: ////
1.0
' / " ~ /.."
0.9-
,/
~P 0.8Q 0 0.7"
,,.:7
0.6-
/
0.5-
/,°
. -"*"
0.4-
f y..S22
0.3-
22" 7"
0.2-
~
o.ff"
I , ~'°°°°°° ~ "t o°.°,*° ,~ °°°°
0.14000
,/
1
20,000
20,000 '
40bo 36,000 IgE concentration picograms,,m!
36,000 '
Fig. 4. Determination of optimal concentrations of coating antibodies. Varying concentrations of mouse anti-human monoclonal antibodies (7.12 + 4.15 mixture, A) and rabbit anti-human IgE coating antibodies (B) were used with 1/4000 biotin and 1/1000 avidin.
B
A
1.0-
0.81.5"
°°o*
0.71.3-
..-'"'"'""
E c e~
1.1
a0
0.9
rE 0.6-
°°°°°° o°°°°°,°"
~" 0.5-
,z.,°°
0.4-
0.70.3-
0.5-
°°.,..°°'°~
0.2-
0.30.1-
0.1-
s6o
2s,boo
sobo
IgE concentration picograms/ml Fig. 5. Sensitivity of the assay. Two representative standard curves.
160
1500 560 1000 IgE concentration picograms/ml
279 2.0 1.9 =
1.8........
0
1.7-
1.6j¢
/
1.5-
1.4-
J
1.3-
/ j
1.2-
E "
04 O~ ,I¢
c~ 0
A/"
j 1.1-
~/
j
0.9
/ J ,,'2.."
II / ,/~7"I" i"..°°'°" 1/ ././~...
0.8 0.7 ¸
/
i¢1
°°
• o°OO°
0.6
/~ .." / tN" .."
A/ 0.5
, ~
..:
~
,
7,,"-
///" J j -I /
L/
~
/
/
,,,,¢,, = 7/"
/./...."
0.4
~// /// /
/ I / .JI • ..."°°°
/.'**
/'°.°°° /" ..°° /" .°°"
,J/ //.." j I ." / /'..."
1.0-
/~. °°°
z.,...":/%" 0.3
• H? .." ." 9"5" /
0.2
0.1 u
i
,
,
!
i
IgE concentration picogramstml
Fig. 6. Reproducibility of the assay. Six different standard curves run at the s a m e time on 6 different plates.
IgE antibodies in vitro. One of the major obstacles in studying the regulation of human IgE synthesis in vitro has been the absence of a sensitive immunoassay for human IgE. Investigators currently use IRA assays for the measurement of human IgE (Sampson and Buckley, 1981; Zuraw et al., 1981; Hassner and Saxon, 1983). Using the ELISA assay described here we are able to detect de novo IgE synthesis by human peripheral blood lymphocyte in vitro (Table I). De novo IgE synthesis is determined by subtracting values obtained from cul-
tures frozen and thawed 5 times on day 0 and then cultured for 10 days (Sampson and Buckley, 1981) from the values obtained when PBL are cultured for 10 days with no mitogen. The addition of PWM to the cultures had no significant effect on the de novo synthesis. Although this assay was optimized for the~measurement of IgE in culture supernatants, I g E levels in serum can also be detected (data not shown). Because of the assay sensitivity, the serum has to be diluted. The degree of dilution depends upon the IgE concentration in the serum.
280
Accuracy of the assay To establish the reproducibility and accuracy of the assay, 3 sera of known IgE concentration (Pharmacia Phadexact sera) were assayed for the IgE content. The results shown in Table II indicate that when the sera are assayed on the same day on 5 different plates (with a standard curve on each plate) good reproducibility is observed. It was thus concluded that data can be compared from plate to plate provided that each plate has a standard curve. In addition our results indicate that when the same sample is placed at varying locations on the same plate, very good reproducibility is obtained (data not shown). It was therefore concluded that position on the plate is not a factor as long as outside wells are not utilized.
Discussion
The avidin-biotin ELISA assay described here for the measurement of human IgE can be used for the measurement of human de novo IgE synthesis in vitro by PBL. The assay is very sensiTABLE II1 W I T H I N ASSAY VARIATIONS ~l Plate no.
Location on plate
IgE serum concentration (ng/ml) Low
Medium
High
1 2 3
49.7 43.7
542.1 537.2 571.8
467.3 484.7 -
1 2 3 4
42.3 39.7 44.8
622.6 562.7 576.8
479.8 506.9 533.3 488.5
1 2 3 4
49.9 54.0 50.7 54.5
566.8 522.4 632.9
436.4 495.1
1 2 3
63.3 60.4
742.3 608.6 795.2
529.8 506.5 -
Three sera (low, medium and high, Phadexact, Pharmacia) were diluted and assayed for lgE. The samples were placed on 4 different plates. On each plate the same sample was assayed 2 - 4 times. Each plate had a standard curve.
tive (100 p g / m l - 1 0 pg/test) and it utilizes commercially available reagents. A sensitive and reliable assay for the measurement of human IgE in culture supernatants is very important because of the minute amounts of IgE synthesized by PBL and because of the variabilities observed between laboratories in the quantification of IgE in cell culture supernatants (Helm et al., 1985). The assay commonly used for the detection of IgE in culture supernatants is the RIA (Sampson and Buckley, 1981; Zuraw et al., 1981; Hassner and Saxon, 1983). The avidin-biotin ELISA assay described herein utilizes affinity-purified rabbit anti-human IgE antibodies to coat the plates. The optimal concentration for coating was determined to be 0.5 ~ g / m l . At that concentration one has to use antibodies coupled to biotin at 1/4000 and avidin-D-HRP at 1/1000. Because of the assay sensitivity it is important to include a standard curve on each 96-well plate. The standard curve is linear between 78 p g / m l to 2500 p g / m l . The assay is highly reproducible as demonstrated by the data obtained when samples of known concentrations are assayed using different plates with a standard curve on each plate (Table III). In summary, the assay described here is rapid (6 h), reproducible, isotype specific and has the sensitivity of radioimmunoassays usually employed for quantification of IgE. This assay can be utilized to establish concentrations of IgE synthesized by human PBL in vitro.
References Adkinson, N.F., 1976, Manual of Clinical Immunology, eds. N.R. Rose and H. Friedman (American Society for Microbiology, Washington, DC) p. 590. Ahlstedt, S., N. Eriksson, S. Lindgren and A. Roth, 1974, Clin. Allergy 4, 131. Bayer, E.A. and M. Wilcheck, 1978, Trends Biochem. Sci. 3, N257. Bayer, E.A. and M. Wilcheck, 1980, Methods Biochem. Anal. 26, 1. Green, N.M., 1963, Biochem. J. 89, 585. Guesdon, J., T. Ternyck and S. Avrameas, 1979, J. Histochem. Cytochem. 27, 1131. Hassner, A. and A. Saxon, 1983, J. Immunol. 130, 1567. Helm, R.M., R.H. Buckley, N.F. Adkinson and D.L. Squillance, 1985, J. Allergy Clin. Immunol. 75, 138.
281 Ishizaka, K. and T. Ishizaka, 1967, J. Immunol. 99, 1187. Kendall, C., I. Ionescu-Matiu and G.R. Dressman, 1983, J. Immunol. Methods 56, 329. Ortolani, C., A. Miadonna, R. Adomi, M. Restuccia and C. Zanussi, 1981, Clin. Allergy 11,249. Pepys, J., A. Roth and K.B. Carroll, 1975, Clin. Allergy 5, 431. Sampson, H.A. and R.H. Buckley, 1981, J. Immunol. 127, 3.
Waldmann, T.A., M. Duam, S. Broder, M. Blackman, R.M. Blease and W. Strober, 1974, Lancet ii, 609. Wide, L., H. Bennich and S.G.O. Johansson, 1967, Lancet ii, 1105. Wilcheck, M. and E.A. Bayer, 1984, Immunol. Today 5, 39. Zuraw, B.L., M. Nonaka, C. O'Hair and D.H. Katz, 1981, J. Immunot. 127, 1169.