Measurement of IgE anti, radioailergosurbent test I. Technical
considerations
Gerald J. Gleich, M.D., Roch4xter, Ninn.
in the perfarrnance
and Richurd
T. Jones,
of the test
B.S.
Since its intro&&tin by Wide, Ben&h, rind Johansson in, 1967, the ra&allergoaorbent test (RAST) has been increasingly used for measurement of sped@ IgR antibodies. In this study we have varied the oonditioolzs &wing the perf&vuwx of the RAST and have investigated the effect of altering: (1) the speed of rotation of tubes, (a) the numbw of uwah.es in both steps of the test, (5) the length. of robatiwb of tubes, (4) the temperature, (5) the wnditiow during centrifwga&n, (6) the quantity of antibody to IgE in the second step of the best, (7) the aanap&tim& of buffers in the reactions, (8) the type of tube employed in the test, (9) the quant4ty of antigen employed &ring the cozcpling procedwe, and, #nally, (10) the ioWe isotope, 1251 or 1311, wed to mtlialabrl anti-IgE. Of these variables the most significant were the qwntity of anti-IgE used in the second step of the test and thp quantity of antigen wed in the coupling procedure. Other variabtes of impin-tawe were the choice of radkoiodine, whether the tubes were rotated or not, the type of tube used, and finally th.e diluent employed.
In 1967 Wide, Bennich, and Johansson’ introduced the radioallergosorbent test (RAST) as a method to measure IgE antibodies in serum. In principle the test was remarkably simple and was similar to the indirect Coombs test.%,8 Tn the first step of the R.AST, crude extracts of allergens covalently linked to solid-phase particles arc incubated with serum and washed. In the second step these particles are exposed to radiolabeled antibody to IgE, rewashed, and the radioactivity measured in a gamma scintillation counter. The number of counts present after the second wash is taken as a measure of the quantity of IgE antibodies that reacted with the solid-phase allergen in the first step of the reaetion. Subsequently, RAST has been used to measure IgE antibodies to a va,riety of antigens including animal danders, pollen antigens, and even food antigens.4-1”However, Sew investigations of the conditions important in the reaction have been r~ported.~~~lR In this study we have systematically analyzed many of the variables in the JUST procedure, and of these the quantity of antigen used in the coupling procedure and the quantity of antibody to IgE used were most important. From the Departments of Medicine, Microbiolo y, and Immunology, The Allergic Research Laboratory, Mayo Clinic and Xayo 4 mm&&ion, and the Mayo I@%& Supported by grants from the National Tn&tutes of Health, Division of Biologic NIH-DBS-72-2087. and tho National Institute of Allerm .__ and fnfwtious Diseases. Received for publioakon May 22, 1974. Reprint requests to : Dr. Gerald .J. Gleich, Mayo Clinic, Rochester, ZvIinn. 55901. Vol.
55, No.
5, pp. 394-345
Diseases S&c&. Standards, AT-flrlS3.
VOLUME NUMBER
55 5
Radioallergosorbent
test.
I
335
MATERIALS AND METHODS Extracts of short ragweed The preparation of the short ragweed extract has been briefly described previously.rs Short ragweed pollen was purchased from Greer and Greer Laboratories, Princeton, West Virginia, Lot No. G-91. Dry pollen, 148 Gm., was washed five times with 750-ml. portions of diethyl ether and dried under partial vacuum. The defatted pollen was suspended in 1 1. of distilled water and extracted for 5 hours at room temperature with magnetic stirring. The extract was filtered and the pollen cake was washed twice with 112-ml. portions of distilled water. The pH of reddish-brown filtrate was adjusted to 7 with 3 N NH,OH. The extract was concentrated to a volume of 62 ml. by ultrafiltration using a PM-10 Diaflo membrane (Amicon Corporation, Lexington, Massachusetts) and centrifuged at 7,000 relative centrifugal force (rcf) for 10 minutes at 4” C. The supernatant was applied to a 4 x 98 cm. column of Sephadex G-25 equilibrated with 0.05 M pH 7.9 Tris-HCl. Ten-milliliter fractions were collected and their absorbance at 280 nm. was measured. Two major peaks emerged from the column, one at the void volume and the second at approximately thrice the void volume. The bulk of the absorbance was contained in the first peak. The fractions in the first peak were pooled and concentrated to 60 ml., and the material was reapplied to the same G-25 column. On refiltration only a single peak eluted at the void volume of the column. Tubes containing significant absorbance at 280 nm. were pooled, concentrated, and dialyzed against 0.15 M NaCl. The final preparation (132 ml.) contained 24 mg. protein per milliliter when analyzed by the biuret method using human serum albumin as standard. The total amount of dissolved solid was 26 mg. per milliliter.19 Lyophilized crude short ragweed extract was obtained from Center Laboratories, Inc., Port Washington, New York, Lot No. 00802 FD.
Preparation
of solid-phase
ragweed
antigens
Ragweed extracts were covalently linked to microcrystalline cellulose to provide a solidphase ragweed antigen. One gram of microcrystalline cellulose (Avicel, Brinkman Instruments, Inc., Westbury, New Jersey, catalog No. 66-00-550-o) was transferred to a 50-ml. beaker, placed in an ice bath, and adjusted to pH 11 with 4 M NaOH. Four grams of cyanogen bromide (CNBr, Fisher Scientific, Chicago, Illinois) dissolved in 4 ml. of N, N-dimethylformamide was added and the pH maintained at 11 for a period of 45 minutes by addition of 4 M NaOH. During this time a total of 8.4 mEq. of NaOH was consumed. The activated cellulose was washed three times with 0.02 M pH 8 HsBOs-0.04 M NaOH-0.16 M NaCl (borate saline) and suspended in 50 ml. of borate saline in a 50-ml. Erlenmeyer flask. Following addition of 10 ml. (260 mg.) of partially purified ragweed extract, the flask was rotated overnight at 4’ C. At the completion of the reaction, the pH was 8. The solid-phase ragweed was repeatedly washed with borate saline until the absorbance at 280 nm. was less than 0.03, and it was suspended at a concentration of 1 mg. per milliliter in 0.1 M pH 7.4 KH,PO,-KzHPO, containing 0.2 per cent bovine serum albumin (BSA), 1 per cent Tween 20, and 0.1 per cent sodium azide. Hereafter, this buffer will be referred to as RAST diluent.
Analysis
of the solid-phase
ragweed
antigen
The ragweed extract and the solid-phase ragweed antigen were analyzed for their content of ragweed antigen E (AgE) by radioimmunoassay.*o The extract contained 2.95 mg. AgE per milliliter, and in the radioimmunoassay this AgE activity yielded an inhibition curve with a slope not different from that of purified AgE (IV C) of King, Norman, and Connell.zr The solid-phase ragweed antigen was analyzed by radioimmunoassay at 6 different concentrations and the particles were rotated during both steps of the radioimmunoassay. An average value of 32.6 ng. AgE was found per milligram of cellulose. Although the slopes of the resulting inhibition curves did differ statistically by the Scheffe test, they were remarkably similar, indicating that relatively little alteration had occurred to AgE antigenic determinants during the solid-phase coupling procedure.20
336
Gleich
TABLE
and
I. Effect
J. ALLERGY
Jones
of varying
the
speed
of rotation
Experiment 6.1
KS Normal Allergic, Allergic, Allergic,
0.46 0.59 21.4 11.8 3.9
undiluted l-10 l- 100
IMMUNOL. MAY 1975
of tubes* Per
Serum
CLIN.
cent
of counte
1 (rpm) 1
bound Expedment
12.6
0
0.54 0.62 20.3 12.3 4.0
0.23 0.27 16.6 4.4 0.9
2 (rpm) t
6.4
0.22 0.23 28.5 9.2 1.5
*The tubes not rotated were mixed thoroughly at the beginning of Step 1 and of Step 2. PCS refers to fetal calf serum (Grand Island Biological Co., Grand Island, New York). The normal serum was from a subject who was free of allergic symptoms and had a negative skin test to ragweed extract. Tt contained 85 ng. per milliliter of IgE as measured by radioimmunoassay.as The allergic serum was from a subject with ragweed hay fever who had positive tests to ragweed extract, and it contained 39,000 ng. per milliliter of IgE.
RAST procedure The RAST was performed essentially as described previous1y.r Briefly, 0.5 mg. of solidphase ragweed antigen in 0.5 ml. RAST diluent, 0.05 ml. of allergic serum, and 0.2 ml. XAST diluent were added to 10 x 44 glass tubes. After incubation the solid-phase antigen was washed with 1.0 ml. of 0.1 M pH 7.4 K,HPO,-KH,PO, containing 1 per cent Tween 20. In the second step of the RAST, 0.1 ml. of radiolabeled affinity chromatography purified sheep antibody to IgE containing approximately 1.2 ng. protein was added and the final volume adjusted to 0.75 ml. with RAST diluent. Unless otherwise noted, tubes were rotated for approximately 18 hours at room t.emperature. Dilutions of allergic serum were made with 0.1 M pH 7.4 KaHPO,-KH2P04, 0.1 per cent NaS, containing 1 per cent BSA (Sigma Chemical Company, St. Louis, Missouri). In ail experiments results are expressed as the number of counts bound to solid-phase antigen divided by the total counts added x 100. All tests were performed in duplicate and values in the tables are means. The actual number of counts per minute associated with anti-IgE in the different experiments varied from 50,900 to 150,000.
RESULTS
The steps involved in the performance of RAST were systematically analyrRd. Speed
of rotation
of tubes
In two experiments we varied the speed of rotation of the tubes during both steps of the RAST, in one experiment by a factor of two and in a second exprriment by not rotating one set of tubes. The results are shown in Table 1. In the first experiment there was no difference in the result when speed of rotation was either 6 or 12 revolutions per minute (rpm), whereas in the second experiment there was a clear-cut drop in per cent of counts bound when the tubes were not rotated. RlPect of the number
of we&es
in the twu sfaps of -tke R&ST
To evaluate the importance of repetitive washing, we varied the number of washesin the two steps of the procedure between 2 and 4. The results indicated that two washes in both steps were adequate to reduce the oounts bound with
VOLUME NUMBER
55 5
Radioallergosorbent
TABLE II. Effect of variations
in the
incubation
periods Per
in the cent
of counts Step
Serum
FCS Normal Allergic, Allergic, Allergic, FCS Normal Allergic, Allergic, Allergic, FCS Normal Allergic, Allergic, Allergic, FCS Normal Allergic, Allergic, Allergic,
step
undiluted I-IO l-100 undiluted I-IO l-100 undiluted I-IO l-100 undiluted I-IO l-100
1
2 hr. 2 hr. 2 hr. 2 hr. 2 hr. 6 hr. 6 hr. 6 hr. 6 hr. 6 hr. I8 hr. I8 hr. I8 hr. I8 hr. I8 hr. 42 hr. 42 hr. 42 hr. 42 hr. 42 hr.
2 hr.
6 hr.
0.28 0.36 14.28 4.18 0.99
0.33 0.39 16.50 5.53 1.35
0.32 0.33 19.70 8.69 2.30
two
steps
test.
of the
I
337
RAST*
bound
2 16 hr.
42 hr.
0.37 0.66 18.86 8.27 2.57 0.54 0.42 19.99 10.02 3.11 0.45 0.49 19.85 I I .26 3.83 0.31 0.47 19.29 II.62 3.88
0.61 0.86 20.96 12.09 4.76 0.50 0.62 20.37 13.40 4.93
*The same reagents described in the footnote to Table I were employed.
fetal calf serum (FCS) or serum from a nonallergic subject to levels comparable to those found after 4 washesin both steps. Variation
in the length
of rotation
in the two steps of the RAST
In this experiment the effect of different time periods for rotation of tubes in the two steps of the RAST was investigated. The results are summarized in Table II and indicate that equilibrium is reached in the first step of the RAST by 2 to 6 hours when the second step is also incubated for 18 hours. Although there was a slight increase in binding of anti-IgE when the rotation period was prolonged to 42 hours in both steps, this was not marked and amounted to less than 1 per cent. On the other hand, it is evident that even a 6-hour incubation period in the second step of the RAST results iti a remarkable efficiency of binding, as indicated by final values only slightly less than those achieved by 1% and 48-hour incubation periods. It is evident that one could use quite brief incubation times in the RAST if desired. Variation
in temperature
In one experiment, RAST (both steps) was performed at 4, 23, and 37O C. and there was little difference between the binding at the various temperatures with the allergic serum. However, the per cent of counts bound to the negative controls, namely FCS and normal human serum, increased directly with temperature: 0.28 per cent at 4O C., 0.43 per cent at 23O C., and 0.66 per cent at 37O c.
338
Gleich
TABLE
and
III. Effect
J. ALLERGY
Jones
of variation
in the quantity
of anti-IgE Per
Quantity Serum
0.99
FCS
Normal Allergic, Allergic, Allergic, KS Normal Allergic, Allergic, Allergic,
Variation
0.283
undiluted l- 10 I-100
0.89 19.71 10.56 4.08
undiluted I-10 I-100
in centGfw@on
added cent
of anti-IgE I
in Step
of counts in Step
CLIN.
2 of the
tMMUNOL. MAY 1975
RAST
bound 2 of RAST.
ng. added
2.63
26.1
261
0.53 0.57 19.36 10.69 3.83
0.35 0.37 39.73 24.09 7.06
0.30 0.34 40.48 16.51 2.90
0.49 0.47 18.15 9.65 3.48
0.36 0.44 38.98 23.76 7.01
0.3 I 0.35 38.53
15.95 2.97
conditions
The results of variation in centrifugation conditions, including temperature, length of centrifugatiun, and speed, also were studied. In all of our prior experiments we had centrifuged tubes at 3,420 rcf for 15 minutes at ho C. In these experiments we systematically altered these variables, analyzing the results at A0 and 23O C., 3,420 ref and 1,470 ref, and 5 minutes and 15 minutes. The results indicated that there was little difference among tubes in the per cent of counts bound by the positive and negative controls. Centrifugation for 5 minutes at 23O C. and 1,470 rcf gave results not different from any of the other combinations. Variations
in the qw&iiy
of a&i-IgE
in the sacaIMt N
of ,IMT
In this experiment the quantity of anti-IgE utilized in the second step of the RAST was varied over a l,OOO-fold range in order to determine whether addition of increasing amounts of anti-&E might drive the reaction between the solid-phase allergen-IgE complex and anti-IgR to completion. The results are presented in Table III and show the data from two experiments conducted 5 days apart. In all of our prior experiments we had added from 1 to 2 ng. antiIgE per tube in the second step of the RAST, and with this quantity of anti-&E we usually found between 19 and 20 per cent binding with the particular allergic serum employed as a positive control in these experiments. As shown in Table III, reducing the quantity of anti-IgE by a factor of 10 to 0.26 ng. per tube did not reduce the per cent of counts bound. Increasing the quantity of anti-IgE to 26 ng. per tube resulted in an approximate doubling of the counts associated with solid-phase antigen, and this was found in both experiments with quite good agreement between them. Increasing the quantity of anti-&E by another factor of 10 to 261 ng. per tube did not increase the counts associated with the solid-phase antigen. Indeed, in the case of the 1:IO and l:lQo dilutions of allergic serum, there was lsss binding at 261 ng. anti-IgE than found with the 26 ng. anti-1gE. The nonspecific binding with the negative control serums was less with the larger amounts of anti-IgE. These results indicate that quantities of
VOLUME NUMBER
55 5
Radioallergosorbent
TABLE IV. Diluents
used
in Steps
1 and
I
339
2 of RAST* Per cent
of counts
I
bound
Diluent
Serum
RAST
FCS
undiluted l-100
0.76 0.52 17.5 3.5
0.33 0.38 19.0 4.0
None Normal Allergic, Allergic,
test.
Dextran
20
Dextran
1.0 0.43 18.7 4.2
80
Dextren
1.03 0.57 19.9 4.7
250
1.14 0.57 19.7 4.8
*All
diluents contained 1 per cent Tweet 20 and 0.1 per cent sodium azide dissolved in pH 7.4 0.1 M K,HPO,-KH,PO,. RAST diluent contained 0.2 per cent BSA. FCS contained 4 per cent FCS while the other buffers contained 0.2 per cent of the respective dextrans. These buffers were used as additives to bring the volume to 0.75 ml. in both steps of the RAST, and the solid-phase antigens were also suspended in the buffers. In all cases the solid-phase ragweed was washed with 0.1 m pR 7.4 K,HPO,-KHIPO, containing 1 per cent Tween 20.
TABLE V. Effect of varying
concentrations
of polyethylene Polyethylene
glycol
glycol
concentration
in the RAST* (%I
I Serum
I 1
0
Normal :
0.31 0.22
: 3
31.8 20.8 16.8
2
3
4
0.34 0.35
0.40 0.36
0.56 0.50
35.8 24.8 21.1
37.0 26.7 23.0
38.2 27.8 24.0
Allergic 33.6 22.8 18.1
*Polyethylene glycol was added only to buffers used in the second step of the RAST. In this experiment the second-step buffer was the same as RAST diluent except that the BSA was replaced by FCS, 4%. Also 26 ng. of anti-IgE rather than 1.2 ng. was employed in the second step of the RAST.
anti-IgE of the order of 26 ng. per tube produce optimal binding of anti-IgE at least when tested with the particular allergic serum utilized in these experiments. Composition
of buffers
used
in RAST
In all of the studies described above, solid-phase ragweed antigen was suspended in RAST diluent containing 0.2 per cent BSA and 1 per cent Tween 20. In the experiment shown in Table IV, the composition of the various buffers was varied to test the effects of different additives. The use of dextran seemedto increase binding of anti-IgE, and this was especially apparent with the negative controls. Addition of FCS reduced the per cent of counts associated with negative controls. In additional experiments we used either the FCS or RAST diluent in the two steps of the RAST and found that the effect of FCS in lowering nonspecific binding of counts was exerted in Step 2 of the reaction. Incorporation of potassium iodide (KI), up to 0.1 M, and normal sheep serum, either separately or together with KI, failed to affect nonspecific binding in the second step of the RAST. Finally, we tested the effect of addition of polyethylene glycol (20 M,
340
TABLE
Gleich
VI.
and
Jones
Use of different
tubes
and
caps
for
--I Fisher Sarum
FCS Normal Allergic, Allergic, Allergic,
0.55 0.46 16.96 9.03 3.40
undiluted I-10 l-100
*A
of counts
Experiment
2
0.54 0.54 16.62 9.66 3.43
counts Expartment
IMMUWL. MAY 1975
bound
Fisher SC-
CWbO
0.40 0.43 16.99 9.45 3.36
I Capfromtube
cent
1
Luckham Ltd.
t3CbtfItC
CLIN.
RAST* Per
Expariment
J. ALLERGY
Par minuta 1
Ludcham Lti.
0.53 0.69 20.64 10.92 4.13 assoclatad
0.70 0.88 22.06 11.96 4.19
CWbO
0.85 0.99 19.95 11.39 3.96
with caps Exp&Wtt2
I
1
176
254
i
378 348
:;
2,007
403
iz 64
4 2 ;I
246 189 279 2::
18 38 18 ::
i: 1,7z 4
4”: 930 495 599 734
1:: 104 121 77
8;! 5,748 92
9 10
153 190
39 224
ii
117 614
142 201
i: 160
Experiment
1 and
total of 64,000 cpm in Experiment 2.
wa.~ associated
with
19
the
anti-IgE
590
in
70 5,4;:
141,000
opm
Fisher Scientific, Chicago, Illinois) and confirmed the observation of Ceska and Vargal’ that this material enhanced binding of anti-TgE to the solid-phase allergen complex. The results of an experiment are shown in Table V. Additivn of polyethylene glycol increased both the specific binding with a&r&o serr;l’ms and the nonspecific binding with the normal serums. Use of diffwent
tubs
for perfmmnce
of RAST
In all of our prior experiments glass tubes obtained from Fisher Scientific, Chicago, Illinois, were employed. These tubes were 10 x 44 mm. in size and were stoppered with a plastic cap obtained from Niagara Plastics, Erie, Pennsylvania (blue plastic cap for 10 x 75 tube). We tested two other tubes, one obtained from Luckham Limited, Victoria Gardens, Burgess Hills, Sussex, England (LT/3 polystyrene test tubes and LTJ3S polythene press-on stoppers) and the other from Cerbo, Trollhattan, Sweden (Ellerman tube, No. 19180 with plug No 21040). This latter tube has been used by Wide, Bennieh, and J&arm&m1 and Berg, Bennich, and Johansson4 in their studies with RAST. The res&s comparing these various tubes are shown in Table VI. These experiments were performed on 2 days approximately a week apart, and there was little d&renee among the tubes in the binding of anti-IgE to solid-phase ragweed. To determine the amount of radioactivity retained by the caps, we counted the caps at the conclusion of the procedure and these results are also shown in Table VI. There was a tendency for the Luckham caps to retain less radioaotitity than the other caps. In a second experiment essentially the same results were vb&.&o& Cf r&e is the occasional Ellerman tube whose cap was associated with rather cotiderable
VOLUME NUMBER
55 5
Radioallergosorbent
TABLE VII. Comparison
of ‘“‘I
and
lz51 for
radioiodination
cent “‘1
Serum
Serum
I
of counts anti-IgE
bound (date)
G/21
5131
e/7
6114
0.53 0.69 20.64 10.92 4.13
0.69 0.64 19.51 11.00 3.96
0.82 0.73 18.78 10.98 4.31
0.99 0.66 18.79 10.03 3.90
98.3
96.5
95.6
95.5
I
341
of anti-lgE* Per
FCS Normal Allergic, undiluted Allergic, l-10 Allergic, 1- 100 % counts precipitated by 10% tungstic acid
test.
I
1*al enWoE
FCS 0.29 0.57 0.85 0.80 0.45 0.73 0.56 Normal 0.25 20.52 18.15 16.21 Allergic, undiluted 23.58 Allergic, l-10 12.93 11.81 10.43 9.50 3.78 Allergic, l-100 4.44 4.32 4.27 % counts precipitated 93.5 88.7 85.7 by 10% tungstic acid 98.2 *Both radioiodines were obtained from Cambridge Nuclear Corporation, Cambridge, Mass. After radioiodination both preparations were diluted to a concentration of 30 ng. per milliliter in 0.1 M pH 7.4 phosphate buffer containing 1 per cent BSA and frozen in aliquots at -20” C. Each assay was performed with a freshly thawed aliquot of each preparation, and 3 ng. of anti-IgE was added in the second step of the RAST. In the experiment on 5/81 approximately 140,000 epm were present with both radiolabeled preparations.
amounts of radioactivity of the order of 2,000 to 5,000 counts per minute (cpm). The most likely reason for this is the wedge shape of the Ellerman cap, which permits the trapping of small amounts of solid-phase antigen particles between the cap and the tube. We concluded from this experiment that the Luckham tubes were superior primarily because of a lesser tendency to trap radioactivity in the cap. Also the Ellerman tubes frequently split when the cap was vigorously inserted. With the Fisher tubes the blue caps obtained from Niagara Plastics sometimessplit, although this was not common. Finally, in the case of the Luckham tubes the cap was the most difficult to remove. Comparison
of radiolabeling
anti-IgE
with
“‘I and
lz51
The use of lz51 (half-life, 60 days) as a radiolabel for anti-IgE would offer considerable advantage over Is11 (half-life, 8 days). However, in earlier experiments we had noted a rather marked decrease in the binding of Y-labeled anti-IgE after storage at 4O C. for several weeks. This decrease was often pronounced and resulted in a drop from a peak of 20 per cent binding with the same positive control as employed in these experiments to approximately 11 per cent 6 weeks later. In this experiment we compared anti-IgE radiolabeled with Is11 and 1251,respectively; 17.3 pg affinity chromatography purified sheep anti-IgE was labeled with 2.2 me. lSII or with 2.0 me. lz51. In both instances the radiolabeling procedure was carried out by a modification of the technique of Greenwood, Hunter, and Glover2* as described previously.23 The radiolabeled purified antibody was separated from unreacted radioiodine by gel filtration on
342
Gleich
TABLE
VIII.
and
Jones
Effect
of
varying
the
quantity
of
antigen
in
J. ALLERGY
ctIt+.
IMMUNOL. MAY t975
reaction
with
activated
the
cellulose* Per Ragwead
FCS Normal Allergic, Allergic, Allergic, Allergic, *In
this anti-IgE
Serum
1 mg.
undiluted I - 10 l-100 l- 1000
1.54 0.91 21.25 15.21 5.89 3.71
cent
antigan
of counts used
bound
in coupling
reaetkn
10mg.
I
1oQ mg.
1.69 0.89 35.68 26.99 11.74 4.22
experiment 0.5 mg. of the designated solid-phase was used in the second step of the RAST.
0.40 0.68 39.63 28.56 9.15 1.91 cellulose
was
tested
and
26 ng.
Sephadex G-100 columns approximately 1.2 x 12 cm. in size. In both experiments the assumption was made that all the protein was recovered in the first peak eluting from the column, and calculations of radioactive iodine introduced into the protein were based on this assumption. According to the manufacturer’s is lz51, while only 30 estimate, 98 per cent of the material in lzaI preparation per cent of the lnlI preparation is 1311. Using these figures and knowing the activity of the preparations for iodination, we calculated the number of iodine atoms introduced into the anti-IgE. With 1311, the labeled protein had a specific activity of 75.4 PC per microgram and a total of 2.4 atoms of iodine were introduced per antibody molecule.* Anti-IgE labeled with 1251 had a specific activity of 66.5 pc per microgram, and a total of 5.0 atoms of iodine were introduced per antibody molecule. During the succeeding month we determined the reactivity in the RAST and the radioactivity precipitable by 10 per cent tungstic acid. The results are shown in Table VII. Initially the reactivity of the anti-TgE labeled with lz51 was greater and the blank values were also somewhat lower than with lSII anti-IgE. However, by the end of the first week the results comparing the two showed little difference; by the third week the peak amount of binding by the 1251 anti-IgE was somewhat lower than with 13*1 anti-IgE. By the fourth week there was a clear-cut difference between the two preparations, and the In11 anti-IgE showed greater binding. The per cent of counts in the two preparations precipitable by 10 per cent tungstic acid was equal at the beginning of the experiment, but during the 3-week period the lz31 anti-IgE dropped from an initial value of 98.2 per cent to 85.7 per cent, while with the ljlI anti-IgE the per cent of counts precipitated by 10 per cent tung&ic acid remained relatively stable. Effect
of
varying
the
quantity
of
ant@m
in the
rerran
with-
@batd
cdkhe
We also tested whether the binding of IgE antibodies is particularly affected by changes in that quantity of antigen used in the reaction between activated cellulose and ragweed. Lyophilieed crude short ragweed extract was dissolved *The 2.4 moles I per per mole was 13x1.
mole
protein
pertains
to total
I of which
only
30 per
cent
or 0.7 mole
VOLUME NUMBER
55 5
Radioallergosorbent
test.
I
343
in borate saline, and 1 mg., 10 mg., or 100 mg. of the crude material was reacted with 0.5 Gm. of activated cellulose. These solid-phase antigens were repeatedly washed and then tested for their reactivity in the RAST. The results are shown in Table VIII and indicate that under these conditions the solid-phase antigen prepared with 10 mg. of antigen had essentially the same reactivity as that prepared with 100 mg. of antigen. In contrast, there was a clear-cut diminution of counts bound with the solid-phase antigen prepared with 1 mg. antigen. Finally, the counts bound by the negative controls were less when more antigen was used in the coupling reaction. DISCUSSION
In this study we have analyzed certain variables in the performance of the RAST procedure. Changes in many of the conditions appeared to make little difference in the final result. Specifically, we noted little advantage in increasing the number of washes above two in the two steps of the RAST, nor was there any effect noted by a twofold variation in the speed of rotation of the tubes. The maximum binding is reached rather quickly in both steps of the RAST, and one could perform the test with a 6-hour incubation period in the first step and an B-hour incubation period in the second and achieve essentially the same binding as with longer periods of incubation in both steps of the procedure. Also there was little variation with changes in temperature between 4O and 37O C. The centrifugation conditions were not critical, and one could centrifuge the solidphase ragweed antigens for 5 minutes at 1,500 g at room temperature and achieve the same degree of binding as achieved under a variety of other conditions with longer periods of centrifugation, higher speeds,or lower temperatures. Finally, the type of tube used was not a critical variable, although there were differences in the number of counts trapped with the caps. Variables of importance in the RAST were the quantity of anti-IgE used in Step 2 of the RAST, the diluent employed, the isotope used to label the anti-IgE, and the quantity of antigen coupled to cellulose. We noted a clear-cut rise in the per cent of counts bound to the solid-phase ragweed IgE complex when the quantity of anti-IgE in the second step of the RAST was increased by tenfold from 2.6 to 26 ng. per tube. We attribute this increase to a mass action effect in which more of the available sites on the IgE antibodies react with anti-IgE. The effect of altering the diluent was less pronounced, but the use of 4 per cent FCS in the second step of the RAST seemedto decrease the per cent of counts bound to the ragweed solid-phase antigens by the negative controls. Addition of polyethylene glycol enhanced binding of anti-IgE to the IgE allergen complex.17 Although it would seem that *25I has theoretical advantages for the labeling of anti-IgE because of its longer half-life, in actual practice we found that binding of 125I anti-IgE dropped significantly over a 3-week period. The cause of the deterioration of binding by anti-IgE labeled with lz51 and the reduction of per cent of counts precipitable by 10 per cent tungstic acid is not clear but presumably results from radiation damage induced by the 1251.24 We utilized approximately 2 me. to label the anti-IgE, and it is possible that lowering the quantity of 125Irelative to that of anti-IgE might result in a longer useful life-
344
Gleich
J. ALLERGY
and Jones
CLIN.
IMMUNOL. MAY 155’5
span for the anti-IgE. Also, the finding that quantities of anti-IgE of the order of 20 to 30 ng. per tube are optimal further encourages belief that ‘9 anti-W of considerably lower specific activity could be used. Wc also found that the RAST tubes must be rotated during the test, although we found no difference in results between 6 rpm and 12 rpm. Finally, we noted a definite differcncc in the per cent of counts bound as a result of altering the quantity of antigen coupled to cellulose. When only 1 mg. of crude extract was employed, there was a clearcut diminution of binding. As a result of these studies, we have modified our procedure for performance of the RAST. First we now use the Luckham tubes and suspend our solid-phase antigen at a. concentration of 1 mg. per milliliter in RAST diluent. Following addition of test serums and solid-phase antigen, RAST diluent is added to bring the total volume up to 0.75 ml., although one may use 1.0 ml. with essentially no difference in results. Tubes are rotated overnight at room temperature at, 6 rpm, centrifuged at 1,470 rcf in an International Model K centrifuge at room temperature for 5 minutes, and washed twice with 1 ml. of 0.1 M pH 7.4 KH,l%,K,HPO, containing 1 per cent Tween 20. In the second step of the R,AST, about 20 ng. 13111 anti-TgE is added and the volume is again adjusted to 0.75 ml. with RAST diluent substituting 4 per cent FCS for the bovine serum albumin. WC then rotate the tubes overnight at room temprraturc, repeat the washing proccdurc, and finally measure t,hc radioactivity in a gamma spectrometer. We have not routinely added polyethylene glycol to our di1uent.s as WCflntl that it often increases binding associated with negative cwntrols. This rffcct was slight in the case of ragweed antigens but was marked with solid-phase gra,ss antipenzz I’inally, our results pertain to thr use of allergens coupled to microcr~stallinr ccllulose, and we do not know how the critical variables influence the rttsults when allergens are coupled to paper discs.‘” We thank
Dr. John
Yunginger
for
his critical
review
of this
manuscript.
REFERENCES 1 Wide, L., Bonnich, II., and Johansson, 5. G. 0.: Diagnosis of allergy by an in-vitro test for allergen antibodies, Lancet 2: 1105, 1967. 2 Johansson, S. G. O., Bennich, H. H., and Berg, T.: The clinical significance of IgE, in Schwartz, R. S., editor: Progress in Clinical lmmunology, New York, 1972, Grune & Stratton, Tne., vol. 1, p. 157. 3 Gleich, G. J., and Yunginger, J. W.: Jmmunoglobulin F: and atopic diseases, South Dakota J. Med. 26: 39, 1973. 4 Berg, T., Bennich, H., and Johansson, S. G. 0.: In-vitro diagnosis of atopic allergy. 1. A ncmparisonbetiveenprovocationtests and the radiollergosorbent test, Int. Arch. Allergy Appl. Immunol. 40: 770, 1971. 5 Berg, T., Bennich, H., and Johansson, 8. G. 0.: In-vitro diagnosis of atopic allergy. IV. Seasonal variations of IgE antibodies in children allergic to pollens, lnt. Arch. Allergy Appl. Immunol. 41: 452, 1971. 6 Lichtenstein, L. I&, Ishizaka, K., Norman, P. S., Robotka, A. K., and Hill, R. M.: IgE antibody measurements in ragweed hay fever. Relationship to clinical severity and the results of immunotherapy, J. Clin. Invest. 52: 478, 1973. 7 Yunginger, J. W., and Gleich, G. J.: Seasonal changes in IgE antibodies and their relationship to IgG antibodies during immunotherapy for ragweed hay fever, J. Clin. Invest. 62: 1268, 1973.
VOLUME NUMBER
55 5
Radioallergosorbent
test.
I
345
8 Fagerberg, E., and Wide, L.: Diagnosis of hypersensitivity to dog epithelium in patients with asthma bronchiale, Int. Arch. Allergy Appl. Immunol. 39: 301, 1970. 9 Evans, R., Reisman, R. E., Wypych, J. I., and Arbesman, C. E.: An immunologic evaluation of ragweed-sensitive patients by newer techniques, J. ALLERGY CLIN. IMMUNOL. 49: 285, 1972. 10 Wide, L., and Juhlin, L.: Detection of penicillin allergy of the immediate type by radioimmunoassay of reagins (IgE) to penicilloyl conjugates, Clin. Allergy 1: 171, 1971. 11 Aas, K., and Johansson, S. G. 0.: The radioallergosorbent test in the in-vitro diagnosis of multiple reaginic allergy, J. ALLERGY CLIN. IYMUNOL. 48: 134, 1971. 12 Stenius, B., and Wide, L.: Reaginic antibody (IgE), skin and provocation tests to Dermatophagoides mlinae and house dust in respiratory allergy, Lancet 2: 455, 1969. 13 Holford-Strevens, V., Wide, L., Milne, J. F., and Pepys, J.: Allergens and antigens of .Dermatophagoides farinae, Clin. Exp. Immunol. 6: 49, 1970. 14 Foucard, T., Aas, K., and Johansson, S. G. 0.: Concentration of IgE antibodies, PK titers, and chopped lung titers in sera from children with hypersensitivity to cod, J. ALLERGY CLIN. hfMUNOL. 51: 39, 1973. 15 Reisman, R. E., and Arbesman, C. E.: Clinical and immunological studies following immunotherapy with aqueous and alum precipitated ragweed fraction A, Int. Arch. Allergy Appl. Immunol. 44: 161, 1973. 16 Huldt, G., Johansson, S. G. Cl., and Lantto, S.: Echinococeosis in Northern Scandinavia, Arch. Environ. Health 26: 36, 1973. 17 Ceska, M., and Varga, J. M.: Effect of non-ionic polymers on radioallergoimmunosorbent reactions, Eur. J. Immunol. 2: 58, 1972. 18 Sarsfield, J. K., and Gowland, G.: A modified radioallergosorbent test for the in-vitro detection of allergen antibodies, Clin. Exp. Immunol. 13: 619, 1973. 19 Gleich, G. J., Larson, J. B., Jones, R. T., and Baer, H.: Measurement of the potency of allergy extracts by their inhibitory capacities in the radioallergosorbent test, J. ALLERGY CLIN. IMMUNOL. 53: 158, 1974. 20 Yunginger, J. W., and Gleich, G. J.: Measurement of ragweed antigen-E by double antibody radioimmunoassay,J. ALLERGY CLIN. IMMUNOL. 50: 326,1972. 21 King, T. P., Norman, P. S., and Connell, J. T.: Isolation and characterization of allergens from ragweed pollen. II, Biochemistry 3: 458, 1964. of IslI-labelled 22 Greenwood, F. C., Hunter, W. M., and Glover, J. S.: The preparation human growth hormone of high specific radioactivity, Biochem. J. 89: 114, 1963. 23 Gleich, G. J., Averbeck, A. K., and Swedlund, H. A.: Measurement of 1gE in normal and allergic serum by radioimmunoassay, J. Lab. Clin. Med. 77: 690, 1971. 24 Yalow, R. S.: Preparation and stability of iodine labelled compounds in protein and polypeptide hormones, Excerpta Medica Foundation, Int. Congress Series No. 161, 1969, p. 605. 25 Gleich, G. J. and Jones, R. T.: Unpublished observations. 26 Ceska, M., Eriksson, R., and Varga, J. M.: Radioimmunosorbent assay of allergens, J. ALLERGY CLIN. IMMUNOL. 49: 1,1972.
considerations
Gerald J. Gleich, M.D., Roch4xter, Ninn.
in the perfarrnance
and Richurd
T. Jones,
of the test
B.S.
Since its intro&&tin by Wide, Ben&h, rind Johansson in, 1967, the ra&allergoaorbent test (RAST) has been increasingly used for measurement of sped@ IgR antibodies. In this study we have varied the oonditioolzs &wing the perf&vuwx of the RAST and have investigated the effect of altering: (1) the speed of rotation of tubes, (a) the numbw of uwah.es in both steps of the test, (5) the length. of robatiwb of tubes, (4) the temperature, (5) the wnditiow during centrifwga&n, (6) the quantity of antibody to IgE in the second step of the best, (7) the aanap&tim& of buffers in the reactions, (8) the type of tube employed in the test, (9) the quant4ty of antigen employed &ring the cozcpling procedwe, and, #nally, (10) the ioWe isotope, 1251 or 1311, wed to mtlialabrl anti-IgE. Of these variables the most significant were the qwntity of anti-IgE used in the second step of the test and thp quantity of antigen wed in the coupling procedure. Other variabtes of impin-tawe were the choice of radkoiodine, whether the tubes were rotated or not, the type of tube used, and finally th.e diluent employed.
In 1967 Wide, Bennich, and Johansson’ introduced the radioallergosorbent test (RAST) as a method to measure IgE antibodies in serum. In principle the test was remarkably simple and was similar to the indirect Coombs test.%,8 Tn the first step of the R.AST, crude extracts of allergens covalently linked to solid-phase particles arc incubated with serum and washed. In the second step these particles are exposed to radiolabeled antibody to IgE, rewashed, and the radioactivity measured in a gamma scintillation counter. The number of counts present after the second wash is taken as a measure of the quantity of IgE antibodies that reacted with the solid-phase allergen in the first step of the reaetion. Subsequently, RAST has been used to measure IgE antibodies to a va,riety of antigens including animal danders, pollen antigens, and even food antigens.4-1”However, Sew investigations of the conditions important in the reaction have been r~ported.~~~lR In this study we have systematically analyzed many of the variables in the JUST procedure, and of these the quantity of antigen used in the coupling procedure and the quantity of antibody to IgE used were most important. From the Departments of Medicine, Microbiolo y, and Immunology, The Allergic Research Laboratory, Mayo Clinic and Xayo 4 mm&&ion, and the Mayo I@%& Supported by grants from the National Tn&tutes of Health, Division of Biologic NIH-DBS-72-2087. and tho National Institute of Allerm .__ and fnfwtious Diseases. Received for publioakon May 22, 1974. Reprint requests to : Dr. Gerald .J. Gleich, Mayo Clinic, Rochester, ZvIinn. 55901. Vol.
55, No.
5, pp. 394-345
Diseases S&c&. Standards, AT-flrlS3.
VOLUME NUMBER
55 5
Radioallergosorbent
test.
I
335
MATERIALS AND METHODS Extracts of short ragweed The preparation of the short ragweed extract has been briefly described previously.rs Short ragweed pollen was purchased from Greer and Greer Laboratories, Princeton, West Virginia, Lot No. G-91. Dry pollen, 148 Gm., was washed five times with 750-ml. portions of diethyl ether and dried under partial vacuum. The defatted pollen was suspended in 1 1. of distilled water and extracted for 5 hours at room temperature with magnetic stirring. The extract was filtered and the pollen cake was washed twice with 112-ml. portions of distilled water. The pH of reddish-brown filtrate was adjusted to 7 with 3 N NH,OH. The extract was concentrated to a volume of 62 ml. by ultrafiltration using a PM-10 Diaflo membrane (Amicon Corporation, Lexington, Massachusetts) and centrifuged at 7,000 relative centrifugal force (rcf) for 10 minutes at 4” C. The supernatant was applied to a 4 x 98 cm. column of Sephadex G-25 equilibrated with 0.05 M pH 7.9 Tris-HCl. Ten-milliliter fractions were collected and their absorbance at 280 nm. was measured. Two major peaks emerged from the column, one at the void volume and the second at approximately thrice the void volume. The bulk of the absorbance was contained in the first peak. The fractions in the first peak were pooled and concentrated to 60 ml., and the material was reapplied to the same G-25 column. On refiltration only a single peak eluted at the void volume of the column. Tubes containing significant absorbance at 280 nm. were pooled, concentrated, and dialyzed against 0.15 M NaCl. The final preparation (132 ml.) contained 24 mg. protein per milliliter when analyzed by the biuret method using human serum albumin as standard. The total amount of dissolved solid was 26 mg. per milliliter.19 Lyophilized crude short ragweed extract was obtained from Center Laboratories, Inc., Port Washington, New York, Lot No. 00802 FD.
Preparation
of solid-phase
ragweed
antigens
Ragweed extracts were covalently linked to microcrystalline cellulose to provide a solidphase ragweed antigen. One gram of microcrystalline cellulose (Avicel, Brinkman Instruments, Inc., Westbury, New Jersey, catalog No. 66-00-550-o) was transferred to a 50-ml. beaker, placed in an ice bath, and adjusted to pH 11 with 4 M NaOH. Four grams of cyanogen bromide (CNBr, Fisher Scientific, Chicago, Illinois) dissolved in 4 ml. of N, N-dimethylformamide was added and the pH maintained at 11 for a period of 45 minutes by addition of 4 M NaOH. During this time a total of 8.4 mEq. of NaOH was consumed. The activated cellulose was washed three times with 0.02 M pH 8 HsBOs-0.04 M NaOH-0.16 M NaCl (borate saline) and suspended in 50 ml. of borate saline in a 50-ml. Erlenmeyer flask. Following addition of 10 ml. (260 mg.) of partially purified ragweed extract, the flask was rotated overnight at 4’ C. At the completion of the reaction, the pH was 8. The solid-phase ragweed was repeatedly washed with borate saline until the absorbance at 280 nm. was less than 0.03, and it was suspended at a concentration of 1 mg. per milliliter in 0.1 M pH 7.4 KH,PO,-KzHPO, containing 0.2 per cent bovine serum albumin (BSA), 1 per cent Tween 20, and 0.1 per cent sodium azide. Hereafter, this buffer will be referred to as RAST diluent.
Analysis
of the solid-phase
ragweed
antigen
The ragweed extract and the solid-phase ragweed antigen were analyzed for their content of ragweed antigen E (AgE) by radioimmunoassay.*o The extract contained 2.95 mg. AgE per milliliter, and in the radioimmunoassay this AgE activity yielded an inhibition curve with a slope not different from that of purified AgE (IV C) of King, Norman, and Connell.zr The solid-phase ragweed antigen was analyzed by radioimmunoassay at 6 different concentrations and the particles were rotated during both steps of the radioimmunoassay. An average value of 32.6 ng. AgE was found per milligram of cellulose. Although the slopes of the resulting inhibition curves did differ statistically by the Scheffe test, they were remarkably similar, indicating that relatively little alteration had occurred to AgE antigenic determinants during the solid-phase coupling procedure.20
336
Gleich
TABLE
and
I. Effect
J. ALLERGY
Jones
of varying
the
speed
of rotation
Experiment 6.1
KS Normal Allergic, Allergic, Allergic,
0.46 0.59 21.4 11.8 3.9
undiluted l-10 l- 100
IMMUNOL. MAY 1975
of tubes* Per
Serum
CLIN.
cent
of counte
1 (rpm) 1
bound Expedment
12.6
0
0.54 0.62 20.3 12.3 4.0
0.23 0.27 16.6 4.4 0.9
2 (rpm) t
6.4
0.22 0.23 28.5 9.2 1.5
*The tubes not rotated were mixed thoroughly at the beginning of Step 1 and of Step 2. PCS refers to fetal calf serum (Grand Island Biological Co., Grand Island, New York). The normal serum was from a subject who was free of allergic symptoms and had a negative skin test to ragweed extract. Tt contained 85 ng. per milliliter of IgE as measured by radioimmunoassay.as The allergic serum was from a subject with ragweed hay fever who had positive tests to ragweed extract, and it contained 39,000 ng. per milliliter of IgE.
RAST procedure The RAST was performed essentially as described previous1y.r Briefly, 0.5 mg. of solidphase ragweed antigen in 0.5 ml. RAST diluent, 0.05 ml. of allergic serum, and 0.2 ml. XAST diluent were added to 10 x 44 glass tubes. After incubation the solid-phase antigen was washed with 1.0 ml. of 0.1 M pH 7.4 K,HPO,-KH,PO, containing 1 per cent Tween 20. In the second step of the RAST, 0.1 ml. of radiolabeled affinity chromatography purified sheep antibody to IgE containing approximately 1.2 ng. protein was added and the final volume adjusted to 0.75 ml. with RAST diluent. Unless otherwise noted, tubes were rotated for approximately 18 hours at room t.emperature. Dilutions of allergic serum were made with 0.1 M pH 7.4 KaHPO,-KH2P04, 0.1 per cent NaS, containing 1 per cent BSA (Sigma Chemical Company, St. Louis, Missouri). In ail experiments results are expressed as the number of counts bound to solid-phase antigen divided by the total counts added x 100. All tests were performed in duplicate and values in the tables are means. The actual number of counts per minute associated with anti-IgE in the different experiments varied from 50,900 to 150,000.
RESULTS
The steps involved in the performance of RAST were systematically analyrRd. Speed
of rotation
of tubes
In two experiments we varied the speed of rotation of the tubes during both steps of the RAST, in one experiment by a factor of two and in a second exprriment by not rotating one set of tubes. The results are shown in Table 1. In the first experiment there was no difference in the result when speed of rotation was either 6 or 12 revolutions per minute (rpm), whereas in the second experiment there was a clear-cut drop in per cent of counts bound when the tubes were not rotated. RlPect of the number
of we&es
in the twu sfaps of -tke R&ST
To evaluate the importance of repetitive washing, we varied the number of washesin the two steps of the procedure between 2 and 4. The results indicated that two washes in both steps were adequate to reduce the oounts bound with
VOLUME NUMBER
55 5
Radioallergosorbent
TABLE II. Effect of variations
in the
incubation
periods Per
in the cent
of counts Step
Serum
FCS Normal Allergic, Allergic, Allergic, FCS Normal Allergic, Allergic, Allergic, FCS Normal Allergic, Allergic, Allergic, FCS Normal Allergic, Allergic, Allergic,
step
undiluted I-IO l-100 undiluted I-IO l-100 undiluted I-IO l-100 undiluted I-IO l-100
1
2 hr. 2 hr. 2 hr. 2 hr. 2 hr. 6 hr. 6 hr. 6 hr. 6 hr. 6 hr. I8 hr. I8 hr. I8 hr. I8 hr. I8 hr. 42 hr. 42 hr. 42 hr. 42 hr. 42 hr.
2 hr.
6 hr.
0.28 0.36 14.28 4.18 0.99
0.33 0.39 16.50 5.53 1.35
0.32 0.33 19.70 8.69 2.30
two
steps
test.
of the
I
337
RAST*
bound
2 16 hr.
42 hr.
0.37 0.66 18.86 8.27 2.57 0.54 0.42 19.99 10.02 3.11 0.45 0.49 19.85 I I .26 3.83 0.31 0.47 19.29 II.62 3.88
0.61 0.86 20.96 12.09 4.76 0.50 0.62 20.37 13.40 4.93
*The same reagents described in the footnote to Table I were employed.
fetal calf serum (FCS) or serum from a nonallergic subject to levels comparable to those found after 4 washesin both steps. Variation
in the length
of rotation
in the two steps of the RAST
In this experiment the effect of different time periods for rotation of tubes in the two steps of the RAST was investigated. The results are summarized in Table II and indicate that equilibrium is reached in the first step of the RAST by 2 to 6 hours when the second step is also incubated for 18 hours. Although there was a slight increase in binding of anti-IgE when the rotation period was prolonged to 42 hours in both steps, this was not marked and amounted to less than 1 per cent. On the other hand, it is evident that even a 6-hour incubation period in the second step of the RAST results iti a remarkable efficiency of binding, as indicated by final values only slightly less than those achieved by 1% and 48-hour incubation periods. It is evident that one could use quite brief incubation times in the RAST if desired. Variation
in temperature
In one experiment, RAST (both steps) was performed at 4, 23, and 37O C. and there was little difference between the binding at the various temperatures with the allergic serum. However, the per cent of counts bound to the negative controls, namely FCS and normal human serum, increased directly with temperature: 0.28 per cent at 4O C., 0.43 per cent at 23O C., and 0.66 per cent at 37O c.
338
Gleich
TABLE
and
III. Effect
J. ALLERGY
Jones
of variation
in the quantity
of anti-IgE Per
Quantity Serum
0.99
FCS
Normal Allergic, Allergic, Allergic, KS Normal Allergic, Allergic, Allergic,
Variation
0.283
undiluted l- 10 I-100
0.89 19.71 10.56 4.08
undiluted I-10 I-100
in centGfw@on
added cent
of anti-IgE I
in Step
of counts in Step
CLIN.
2 of the
tMMUNOL. MAY 1975
RAST
bound 2 of RAST.
ng. added
2.63
26.1
261
0.53 0.57 19.36 10.69 3.83
0.35 0.37 39.73 24.09 7.06
0.30 0.34 40.48 16.51 2.90
0.49 0.47 18.15 9.65 3.48
0.36 0.44 38.98 23.76 7.01
0.3 I 0.35 38.53
15.95 2.97
conditions
The results of variation in centrifugation conditions, including temperature, length of centrifugatiun, and speed, also were studied. In all of our prior experiments we had centrifuged tubes at 3,420 rcf for 15 minutes at ho C. In these experiments we systematically altered these variables, analyzing the results at A0 and 23O C., 3,420 ref and 1,470 ref, and 5 minutes and 15 minutes. The results indicated that there was little difference among tubes in the per cent of counts bound by the positive and negative controls. Centrifugation for 5 minutes at 23O C. and 1,470 rcf gave results not different from any of the other combinations. Variations
in the qw&iiy
of a&i-IgE
in the sacaIMt N
of ,IMT
In this experiment the quantity of anti-IgE utilized in the second step of the RAST was varied over a l,OOO-fold range in order to determine whether addition of increasing amounts of anti-&E might drive the reaction between the solid-phase allergen-IgE complex and anti-IgR to completion. The results are presented in Table III and show the data from two experiments conducted 5 days apart. In all of our prior experiments we had added from 1 to 2 ng. antiIgE per tube in the second step of the RAST, and with this quantity of anti-&E we usually found between 19 and 20 per cent binding with the particular allergic serum employed as a positive control in these experiments. As shown in Table III, reducing the quantity of anti-IgE by a factor of 10 to 0.26 ng. per tube did not reduce the per cent of counts bound. Increasing the quantity of anti-IgE to 26 ng. per tube resulted in an approximate doubling of the counts associated with solid-phase antigen, and this was found in both experiments with quite good agreement between them. Increasing the quantity of anti-&E by another factor of 10 to 261 ng. per tube did not increase the counts associated with the solid-phase antigen. Indeed, in the case of the 1:IO and l:lQo dilutions of allergic serum, there was lsss binding at 261 ng. anti-IgE than found with the 26 ng. anti-1gE. The nonspecific binding with the negative control serums was less with the larger amounts of anti-IgE. These results indicate that quantities of
VOLUME NUMBER
55 5
Radioallergosorbent
TABLE IV. Diluents
used
in Steps
1 and
I
339
2 of RAST* Per cent
of counts
I
bound
Diluent
Serum
RAST
FCS
undiluted l-100
0.76 0.52 17.5 3.5
0.33 0.38 19.0 4.0
None Normal Allergic, Allergic,
test.
Dextran
20
Dextran
1.0 0.43 18.7 4.2
80
Dextren
1.03 0.57 19.9 4.7
250
1.14 0.57 19.7 4.8
*All
diluents contained 1 per cent Tweet 20 and 0.1 per cent sodium azide dissolved in pH 7.4 0.1 M K,HPO,-KH,PO,. RAST diluent contained 0.2 per cent BSA. FCS contained 4 per cent FCS while the other buffers contained 0.2 per cent of the respective dextrans. These buffers were used as additives to bring the volume to 0.75 ml. in both steps of the RAST, and the solid-phase antigens were also suspended in the buffers. In all cases the solid-phase ragweed was washed with 0.1 m pR 7.4 K,HPO,-KHIPO, containing 1 per cent Tween 20.
TABLE V. Effect of varying
concentrations
of polyethylene Polyethylene
glycol
glycol
concentration
in the RAST* (%I
I Serum
I 1
0
Normal :
0.31 0.22
: 3
31.8 20.8 16.8
2
3
4
0.34 0.35
0.40 0.36
0.56 0.50
35.8 24.8 21.1
37.0 26.7 23.0
38.2 27.8 24.0
Allergic 33.6 22.8 18.1
*Polyethylene glycol was added only to buffers used in the second step of the RAST. In this experiment the second-step buffer was the same as RAST diluent except that the BSA was replaced by FCS, 4%. Also 26 ng. of anti-IgE rather than 1.2 ng. was employed in the second step of the RAST.
anti-IgE of the order of 26 ng. per tube produce optimal binding of anti-IgE at least when tested with the particular allergic serum utilized in these experiments. Composition
of buffers
used
in RAST
In all of the studies described above, solid-phase ragweed antigen was suspended in RAST diluent containing 0.2 per cent BSA and 1 per cent Tween 20. In the experiment shown in Table IV, the composition of the various buffers was varied to test the effects of different additives. The use of dextran seemedto increase binding of anti-IgE, and this was especially apparent with the negative controls. Addition of FCS reduced the per cent of counts associated with negative controls. In additional experiments we used either the FCS or RAST diluent in the two steps of the RAST and found that the effect of FCS in lowering nonspecific binding of counts was exerted in Step 2 of the reaction. Incorporation of potassium iodide (KI), up to 0.1 M, and normal sheep serum, either separately or together with KI, failed to affect nonspecific binding in the second step of the RAST. Finally, we tested the effect of addition of polyethylene glycol (20 M,
340
TABLE
Gleich
VI.
and
Jones
Use of different
tubes
and
caps
for
--I Fisher Sarum
FCS Normal Allergic, Allergic, Allergic,
0.55 0.46 16.96 9.03 3.40
undiluted I-10 l-100
*A
of counts
Experiment
2
0.54 0.54 16.62 9.66 3.43
counts Expartment
IMMUWL. MAY 1975
bound
Fisher SC-
CWbO
0.40 0.43 16.99 9.45 3.36
I Capfromtube
cent
1
Luckham Ltd.
t3CbtfItC
CLIN.
RAST* Per
Expariment
J. ALLERGY
Par minuta 1
Ludcham Lti.
0.53 0.69 20.64 10.92 4.13 assoclatad
0.70 0.88 22.06 11.96 4.19
CWbO
0.85 0.99 19.95 11.39 3.96
with caps Exp&Wtt2
I
1
176
254
i
378 348
:;
2,007
403
iz 64
4 2 ;I
246 189 279 2::
18 38 18 ::
i: 1,7z 4
4”: 930 495 599 734
1:: 104 121 77
8;! 5,748 92
9 10
153 190
39 224
ii
117 614
142 201
i: 160
Experiment
1 and
total of 64,000 cpm in Experiment 2.
wa.~ associated
with
19
the
anti-IgE
590
in
70 5,4;:
141,000
opm
Fisher Scientific, Chicago, Illinois) and confirmed the observation of Ceska and Vargal’ that this material enhanced binding of anti-TgE to the solid-phase allergen complex. The results of an experiment are shown in Table V. Additivn of polyethylene glycol increased both the specific binding with a&r&o serr;l’ms and the nonspecific binding with the normal serums. Use of diffwent
tubs
for perfmmnce
of RAST
In all of our prior experiments glass tubes obtained from Fisher Scientific, Chicago, Illinois, were employed. These tubes were 10 x 44 mm. in size and were stoppered with a plastic cap obtained from Niagara Plastics, Erie, Pennsylvania (blue plastic cap for 10 x 75 tube). We tested two other tubes, one obtained from Luckham Limited, Victoria Gardens, Burgess Hills, Sussex, England (LT/3 polystyrene test tubes and LTJ3S polythene press-on stoppers) and the other from Cerbo, Trollhattan, Sweden (Ellerman tube, No. 19180 with plug No 21040). This latter tube has been used by Wide, Bennieh, and J&arm&m1 and Berg, Bennich, and Johansson4 in their studies with RAST. The res&s comparing these various tubes are shown in Table VI. These experiments were performed on 2 days approximately a week apart, and there was little d&renee among the tubes in the binding of anti-IgE to solid-phase ragweed. To determine the amount of radioactivity retained by the caps, we counted the caps at the conclusion of the procedure and these results are also shown in Table VI. There was a tendency for the Luckham caps to retain less radioaotitity than the other caps. In a second experiment essentially the same results were vb&.&o& Cf r&e is the occasional Ellerman tube whose cap was associated with rather cotiderable
VOLUME NUMBER
55 5
Radioallergosorbent
TABLE VII. Comparison
of ‘“‘I
and
lz51 for
radioiodination
cent “‘1
Serum
Serum
I
of counts anti-IgE
bound (date)
G/21
5131
e/7
6114
0.53 0.69 20.64 10.92 4.13
0.69 0.64 19.51 11.00 3.96
0.82 0.73 18.78 10.98 4.31
0.99 0.66 18.79 10.03 3.90
98.3
96.5
95.6
95.5
I
341
of anti-lgE* Per
FCS Normal Allergic, undiluted Allergic, l-10 Allergic, 1- 100 % counts precipitated by 10% tungstic acid
test.
I
1*al enWoE
FCS 0.29 0.57 0.85 0.80 0.45 0.73 0.56 Normal 0.25 20.52 18.15 16.21 Allergic, undiluted 23.58 Allergic, l-10 12.93 11.81 10.43 9.50 3.78 Allergic, l-100 4.44 4.32 4.27 % counts precipitated 93.5 88.7 85.7 by 10% tungstic acid 98.2 *Both radioiodines were obtained from Cambridge Nuclear Corporation, Cambridge, Mass. After radioiodination both preparations were diluted to a concentration of 30 ng. per milliliter in 0.1 M pH 7.4 phosphate buffer containing 1 per cent BSA and frozen in aliquots at -20” C. Each assay was performed with a freshly thawed aliquot of each preparation, and 3 ng. of anti-IgE was added in the second step of the RAST. In the experiment on 5/81 approximately 140,000 epm were present with both radiolabeled preparations.
amounts of radioactivity of the order of 2,000 to 5,000 counts per minute (cpm). The most likely reason for this is the wedge shape of the Ellerman cap, which permits the trapping of small amounts of solid-phase antigen particles between the cap and the tube. We concluded from this experiment that the Luckham tubes were superior primarily because of a lesser tendency to trap radioactivity in the cap. Also the Ellerman tubes frequently split when the cap was vigorously inserted. With the Fisher tubes the blue caps obtained from Niagara Plastics sometimessplit, although this was not common. Finally, in the case of the Luckham tubes the cap was the most difficult to remove. Comparison
of radiolabeling
anti-IgE
with
“‘I and
lz51
The use of lz51 (half-life, 60 days) as a radiolabel for anti-IgE would offer considerable advantage over Is11 (half-life, 8 days). However, in earlier experiments we had noted a rather marked decrease in the binding of Y-labeled anti-IgE after storage at 4O C. for several weeks. This decrease was often pronounced and resulted in a drop from a peak of 20 per cent binding with the same positive control as employed in these experiments to approximately 11 per cent 6 weeks later. In this experiment we compared anti-IgE radiolabeled with Is11 and 1251,respectively; 17.3 pg affinity chromatography purified sheep anti-IgE was labeled with 2.2 me. lSII or with 2.0 me. lz51. In both instances the radiolabeling procedure was carried out by a modification of the technique of Greenwood, Hunter, and Glover2* as described previously.23 The radiolabeled purified antibody was separated from unreacted radioiodine by gel filtration on
342
Gleich
TABLE
VIII.
and
Jones
Effect
of
varying
the
quantity
of
antigen
in
J. ALLERGY
ctIt+.
IMMUNOL. MAY t975
reaction
with
activated
the
cellulose* Per Ragwead
FCS Normal Allergic, Allergic, Allergic, Allergic, *In
this anti-IgE
Serum
1 mg.
undiluted I - 10 l-100 l- 1000
1.54 0.91 21.25 15.21 5.89 3.71
cent
antigan
of counts used
bound
in coupling
reaetkn
10mg.
I
1oQ mg.
1.69 0.89 35.68 26.99 11.74 4.22
experiment 0.5 mg. of the designated solid-phase was used in the second step of the RAST.
0.40 0.68 39.63 28.56 9.15 1.91 cellulose
was
tested
and
26 ng.
Sephadex G-100 columns approximately 1.2 x 12 cm. in size. In both experiments the assumption was made that all the protein was recovered in the first peak eluting from the column, and calculations of radioactive iodine introduced into the protein were based on this assumption. According to the manufacturer’s is lz51, while only 30 estimate, 98 per cent of the material in lzaI preparation per cent of the lnlI preparation is 1311. Using these figures and knowing the activity of the preparations for iodination, we calculated the number of iodine atoms introduced into the anti-IgE. With 1311, the labeled protein had a specific activity of 75.4 PC per microgram and a total of 2.4 atoms of iodine were introduced per antibody molecule.* Anti-IgE labeled with 1251 had a specific activity of 66.5 pc per microgram, and a total of 5.0 atoms of iodine were introduced per antibody molecule. During the succeeding month we determined the reactivity in the RAST and the radioactivity precipitable by 10 per cent tungstic acid. The results are shown in Table VII. Initially the reactivity of the anti-TgE labeled with lz51 was greater and the blank values were also somewhat lower than with lSII anti-IgE. However, by the end of the first week the results comparing the two showed little difference; by the third week the peak amount of binding by the 1251 anti-IgE was somewhat lower than with 13*1 anti-IgE. By the fourth week there was a clear-cut difference between the two preparations, and the In11 anti-IgE showed greater binding. The per cent of counts in the two preparations precipitable by 10 per cent tungstic acid was equal at the beginning of the experiment, but during the 3-week period the lz31 anti-IgE dropped from an initial value of 98.2 per cent to 85.7 per cent, while with the ljlI anti-IgE the per cent of counts precipitated by 10 per cent tung&ic acid remained relatively stable. Effect
of
varying
the
quantity
of
ant@m
in the
rerran
with-
@batd
cdkhe
We also tested whether the binding of IgE antibodies is particularly affected by changes in that quantity of antigen used in the reaction between activated cellulose and ragweed. Lyophilieed crude short ragweed extract was dissolved *The 2.4 moles I per per mole was 13x1.
mole
protein
pertains
to total
I of which
only
30 per
cent
or 0.7 mole
VOLUME NUMBER
55 5
Radioallergosorbent
test.
I
343
in borate saline, and 1 mg., 10 mg., or 100 mg. of the crude material was reacted with 0.5 Gm. of activated cellulose. These solid-phase antigens were repeatedly washed and then tested for their reactivity in the RAST. The results are shown in Table VIII and indicate that under these conditions the solid-phase antigen prepared with 10 mg. of antigen had essentially the same reactivity as that prepared with 100 mg. of antigen. In contrast, there was a clear-cut diminution of counts bound with the solid-phase antigen prepared with 1 mg. antigen. Finally, the counts bound by the negative controls were less when more antigen was used in the coupling reaction. DISCUSSION
In this study we have analyzed certain variables in the performance of the RAST procedure. Changes in many of the conditions appeared to make little difference in the final result. Specifically, we noted little advantage in increasing the number of washes above two in the two steps of the RAST, nor was there any effect noted by a twofold variation in the speed of rotation of the tubes. The maximum binding is reached rather quickly in both steps of the RAST, and one could perform the test with a 6-hour incubation period in the first step and an B-hour incubation period in the second and achieve essentially the same binding as with longer periods of incubation in both steps of the procedure. Also there was little variation with changes in temperature between 4O and 37O C. The centrifugation conditions were not critical, and one could centrifuge the solidphase ragweed antigens for 5 minutes at 1,500 g at room temperature and achieve the same degree of binding as achieved under a variety of other conditions with longer periods of centrifugation, higher speeds,or lower temperatures. Finally, the type of tube used was not a critical variable, although there were differences in the number of counts trapped with the caps. Variables of importance in the RAST were the quantity of anti-IgE used in Step 2 of the RAST, the diluent employed, the isotope used to label the anti-IgE, and the quantity of antigen coupled to cellulose. We noted a clear-cut rise in the per cent of counts bound to the solid-phase ragweed IgE complex when the quantity of anti-IgE in the second step of the RAST was increased by tenfold from 2.6 to 26 ng. per tube. We attribute this increase to a mass action effect in which more of the available sites on the IgE antibodies react with anti-IgE. The effect of altering the diluent was less pronounced, but the use of 4 per cent FCS in the second step of the RAST seemedto decrease the per cent of counts bound to the ragweed solid-phase antigens by the negative controls. Addition of polyethylene glycol enhanced binding of anti-IgE to the IgE allergen complex.17 Although it would seem that *25I has theoretical advantages for the labeling of anti-IgE because of its longer half-life, in actual practice we found that binding of 125I anti-IgE dropped significantly over a 3-week period. The cause of the deterioration of binding by anti-IgE labeled with lz51 and the reduction of per cent of counts precipitable by 10 per cent tungstic acid is not clear but presumably results from radiation damage induced by the 1251.24 We utilized approximately 2 me. to label the anti-IgE, and it is possible that lowering the quantity of 125Irelative to that of anti-IgE might result in a longer useful life-
344
Gleich
J. ALLERGY
and Jones
CLIN.
IMMUNOL. MAY 155’5
span for the anti-IgE. Also, the finding that quantities of anti-IgE of the order of 20 to 30 ng. per tube are optimal further encourages belief that ‘9 anti-W of considerably lower specific activity could be used. Wc also found that the RAST tubes must be rotated during the test, although we found no difference in results between 6 rpm and 12 rpm. Finally, we noted a definite differcncc in the per cent of counts bound as a result of altering the quantity of antigen coupled to cellulose. When only 1 mg. of crude extract was employed, there was a clearcut diminution of binding. As a result of these studies, we have modified our procedure for performance of the RAST. First we now use the Luckham tubes and suspend our solid-phase antigen at a. concentration of 1 mg. per milliliter in RAST diluent. Following addition of test serums and solid-phase antigen, RAST diluent is added to bring the total volume up to 0.75 ml., although one may use 1.0 ml. with essentially no difference in results. Tubes are rotated overnight at room temperature at, 6 rpm, centrifuged at 1,470 rcf in an International Model K centrifuge at room temperature for 5 minutes, and washed twice with 1 ml. of 0.1 M pH 7.4 KH,l%,K,HPO, containing 1 per cent Tween 20. In the second step of the R,AST, about 20 ng. 13111 anti-TgE is added and the volume is again adjusted to 0.75 ml. with RAST diluent substituting 4 per cent FCS for the bovine serum albumin. WC then rotate the tubes overnight at room temprraturc, repeat the washing proccdurc, and finally measure t,hc radioactivity in a gamma spectrometer. We have not routinely added polyethylene glycol to our di1uent.s as WCflntl that it often increases binding associated with negative cwntrols. This rffcct was slight in the case of ragweed antigens but was marked with solid-phase gra,ss antipenzz I’inally, our results pertain to thr use of allergens coupled to microcr~stallinr ccllulose, and we do not know how the critical variables influence the rttsults when allergens are coupled to paper discs.‘” We thank
Dr. John
Yunginger
for
his critical
review
of this
manuscript.
REFERENCES 1 Wide, L., Bonnich, II., and Johansson, 5. G. 0.: Diagnosis of allergy by an in-vitro test for allergen antibodies, Lancet 2: 1105, 1967. 2 Johansson, S. G. O., Bennich, H. H., and Berg, T.: The clinical significance of IgE, in Schwartz, R. S., editor: Progress in Clinical lmmunology, New York, 1972, Grune & Stratton, Tne., vol. 1, p. 157. 3 Gleich, G. J., and Yunginger, J. W.: Jmmunoglobulin F: and atopic diseases, South Dakota J. Med. 26: 39, 1973. 4 Berg, T., Bennich, H., and Johansson, S. G. 0.: In-vitro diagnosis of atopic allergy. 1. A ncmparisonbetiveenprovocationtests and the radiollergosorbent test, Int. Arch. Allergy Appl. Immunol. 40: 770, 1971. 5 Berg, T., Bennich, H., and Johansson, 8. G. 0.: In-vitro diagnosis of atopic allergy. IV. Seasonal variations of IgE antibodies in children allergic to pollens, lnt. Arch. Allergy Appl. Immunol. 41: 452, 1971. 6 Lichtenstein, L. I&, Ishizaka, K., Norman, P. S., Robotka, A. K., and Hill, R. M.: IgE antibody measurements in ragweed hay fever. Relationship to clinical severity and the results of immunotherapy, J. Clin. Invest. 52: 478, 1973. 7 Yunginger, J. W., and Gleich, G. J.: Seasonal changes in IgE antibodies and their relationship to IgG antibodies during immunotherapy for ragweed hay fever, J. Clin. Invest. 62: 1268, 1973.
VOLUME NUMBER
55 5
Radioallergosorbent
test.
I
345
8 Fagerberg, E., and Wide, L.: Diagnosis of hypersensitivity to dog epithelium in patients with asthma bronchiale, Int. Arch. Allergy Appl. Immunol. 39: 301, 1970. 9 Evans, R., Reisman, R. E., Wypych, J. I., and Arbesman, C. E.: An immunologic evaluation of ragweed-sensitive patients by newer techniques, J. ALLERGY CLIN. IMMUNOL. 49: 285, 1972. 10 Wide, L., and Juhlin, L.: Detection of penicillin allergy of the immediate type by radioimmunoassay of reagins (IgE) to penicilloyl conjugates, Clin. Allergy 1: 171, 1971. 11 Aas, K., and Johansson, S. G. 0.: The radioallergosorbent test in the in-vitro diagnosis of multiple reaginic allergy, J. ALLERGY CLIN. IYMUNOL. 48: 134, 1971. 12 Stenius, B., and Wide, L.: Reaginic antibody (IgE), skin and provocation tests to Dermatophagoides mlinae and house dust in respiratory allergy, Lancet 2: 455, 1969. 13 Holford-Strevens, V., Wide, L., Milne, J. F., and Pepys, J.: Allergens and antigens of .Dermatophagoides farinae, Clin. Exp. Immunol. 6: 49, 1970. 14 Foucard, T., Aas, K., and Johansson, S. G. 0.: Concentration of IgE antibodies, PK titers, and chopped lung titers in sera from children with hypersensitivity to cod, J. ALLERGY CLIN. hfMUNOL. 51: 39, 1973. 15 Reisman, R. E., and Arbesman, C. E.: Clinical and immunological studies following immunotherapy with aqueous and alum precipitated ragweed fraction A, Int. Arch. Allergy Appl. Immunol. 44: 161, 1973. 16 Huldt, G., Johansson, S. G. Cl., and Lantto, S.: Echinococeosis in Northern Scandinavia, Arch. Environ. Health 26: 36, 1973. 17 Ceska, M., and Varga, J. M.: Effect of non-ionic polymers on radioallergoimmunosorbent reactions, Eur. J. Immunol. 2: 58, 1972. 18 Sarsfield, J. K., and Gowland, G.: A modified radioallergosorbent test for the in-vitro detection of allergen antibodies, Clin. Exp. Immunol. 13: 619, 1973. 19 Gleich, G. J., Larson, J. B., Jones, R. T., and Baer, H.: Measurement of the potency of allergy extracts by their inhibitory capacities in the radioallergosorbent test, J. ALLERGY CLIN. IMMUNOL. 53: 158, 1974. 20 Yunginger, J. W., and Gleich, G. J.: Measurement of ragweed antigen-E by double antibody radioimmunoassay,J. ALLERGY CLIN. IMMUNOL. 50: 326,1972. 21 King, T. P., Norman, P. S., and Connell, J. T.: Isolation and characterization of allergens from ragweed pollen. II, Biochemistry 3: 458, 1964. of IslI-labelled 22 Greenwood, F. C., Hunter, W. M., and Glover, J. S.: The preparation human growth hormone of high specific radioactivity, Biochem. J. 89: 114, 1963. 23 Gleich, G. J., Averbeck, A. K., and Swedlund, H. A.: Measurement of 1gE in normal and allergic serum by radioimmunoassay, J. Lab. Clin. Med. 77: 690, 1971. 24 Yalow, R. S.: Preparation and stability of iodine labelled compounds in protein and polypeptide hormones, Excerpta Medica Foundation, Int. Congress Series No. 161, 1969, p. 605. 25 Gleich, G. J. and Jones, R. T.: Unpublished observations. 26 Ceska, M., Eriksson, R., and Varga, J. M.: Radioimmunosorbent assay of allergens, J. ALLERGY CLIN. IMMUNOL. 49: 1,1972.
