- Email: [email protected]
Quantitative determination of rheumatoid factor by an enzyme-labeled immunoassay
Journal o f Immunological Methods, 32 (1980) 115--126 © Elsevier/North-Holland Biomedical Press
115
Q U A N T I T A T I V E D E T E R M I N A T I O N O F R H E U M A T O I D F A C T O R BY AN ENZYME-LABELED IMMUNOASSAY
J. KARSH, S.P. HALBERT, E. KLIMA and A.D. STEINBERG Clinical Immunology Section, National Institute o f Dental Research; The Arthritis Branch, National Institute o f Arthritis, Metabolism and Digestive Diseases, National Institutes o f Health, Bethesda, MD 20205; and the Cordis Laboratories, Miami, FL 33127, U.S.A.
(Received 2 July 1979, accepted 19 September 1979)
The presence and quantity of rheumatoid factor (RF) in human serum were determined by an indirect enzyme-linked immunoabsorbent assay (ELISA). A human rheumatoid factor control serum, standardized against the WHO reference rheumatoid arthritis serum preparation, was used to derive a standard curve in each assay. The results of unknowns were estimated from the standard curve and reported in International Units (IU) per milliliter (ml). The ELISA assay was compared with the bentonite flocculation test. The overall coefficient of correlation between the two assays was 87%; it was 93.3% for patients with rheumatoid arthritis but only 67.5% for patients with undiagnosed conditions. The error of the ELISA assay (coefficient of variation) was generally less than 10% at both high and low concentrations of rheumatoid factor. The quantitative reproducible nature of the assay allows the detection of small variations of rheumatoid factor level and could be useful in the serial evaluation of patients.
INTRODUCTION R h e u m a t o i d f a c t o r s are a n t i b o d i e s d i r e c t e d a g a i n s t t h e F c p o r t i o n o f IgG. T h e i r d e t e c t i o n in b i o l o g i c a l f l u i d s has b e c o m e a c o m m o n l a b o r a t o r y p r o c e dure; the information derived regarding both the presence and amount of r h e u m a t o i d f a c t o r is o f t e n u s e d t o s u p p o r t or r e f u t e clinical d i a g n o s i s a n d in the management of individual patients. T h e s o u r c e o f I g G has l i t t l e a p p a r e n t c l i n i c a l s p e c i f i c i t y a n d c o n v e n t i o n a l methods take advantage of the broad reactivity of circulating rheumatoid factors with i m m u n o g l o b u l i n s of m a n y species (Butler and Vaughan, 1964). T h e c o m m o n assays m e a s u r e o n l y IgM r h e u m a t o i d f a c t o r s a n d are b a s e d on the ability of these polyvalent immunoglobulins to agglutinate particles c o a t e d w i t h IgG, such as r e d b l o o d cells ( C a t h c a r t a n d O ' S u l l i v a n , 1 9 6 5 ) , l a t e x ( S i n g e r a n d P l o t z , 1 9 5 6 ) , o r b e n t o n i t e ( B o z i c e v i t c h e t al., 1 9 5 8 ) . T h e s e a s s a y s have c e r t a i n d i s a d v a n t a g e s in r o u t i n e c l i n i c a l use. T h e y are g e n e r a l l y n o t standardized and considerable d a y - t o - d a y variability can occur (Chandler a n d R o b i n s o n , 1 9 7 4 ) . S e m i - q u a n t i t a t i o n is a c h i e v e d b y serial d i l u t i o n o f
116 sera and selection of the last positive reaction. However, the process of serial dilution is time consuming, and the precision of end point titration is poor. In an effort to overcome these problems, several new second generation assays for the measurement of rheumatoid factors have been reported (Maiolini et al., 1978; Willems and De Kort, 1978). They offer the advantages of using standardized reagents and of being quick, reproducible and quantitative. We have developed an indirect enzyme-linked immunosorbent assay (ELISA) to detect and quantitate IgM rheumatoid factor in serum. The substrate is goat immunoglobulin G covalently bound to plastic discs. A human rheumatoid factor control serum pool was standardized against the WHO reference rheumatoid arthritis serum preparation, and was used to derive a standard curve in each assay. The results of single dilutions of the unknowns were estimated from the standard curves and reported in International Units/ml (IU/ml). The accuracy of the m e t h o d was confirmed by multiple replications of individual sera as well as serial dilution of several sera. METHODS AND MATERIALS
Preparation of goat immunoglobulin and linkage to immunosorbent Goat immunoglobulin G (Cohen Fraction II) was obtained from ICN Pharmaceuticals (Cleveland, OH). It was dissolved in 0.1 M sodium carbonate buffer, pH 9.6, and clarified by centrifugation at 9000 rev/min at 4°C. The goat IgG was covalently coated onto small plastic discs (13 mm diameter, 0.3 mm thick) via isothiocyanate groups introduced into their surface (Halbert and Anken, 1977). Exposure at the rate of 7 pg/disc at 2--8°C was permitted for 16--20 h with continuous gentle agitation. After washing the discs with buffered saline, they were rinsed in a bovine serum albumin solution (2 mg/ml in 0.01 sodium phosphate pH 8), drained and lyophilized. In this state, they are stable for at least 14 m o n t h s when stored at 2--8°C in the presence of a desiccant. Preliminary studies were carried out with discs coated with nonaggregated and with heat-aggregated goat immunoglobulin G (heated to 60°C for 10 min), but the results indicated that no significant difference in the ELISA read-out was evident when non-aggregated and aggregated goat IgG were compared.
Preparation and labeling o f goat anti-human IgM Human immunoglobulin M (IgM) was isolated from a patient with WaldenstrSm's macroglobulinemia by precipitation with a m m o n i u m sulfate (1.37 M), followed by Sepharose 6B gel filtration. The purified IgM was shown to be free of detectable IgG by gel diffusion methods, and was used to immunize goats following emulsification with complete Freund's adjuvant. The resulting p o t e n t antiserum was absorbed with a solid-phase im-
117 munoadsorbent prepared from specially selected human cord sera found to contain very low levels of IgM. The cord serum was insolubilized to a gel by glutaraldehyde cross-linking using the m e t h o d of Avrameas et al. (1969). After the antiserum was made monospecific for IgM, it was passed through a column of solid-phase purified IgM, and the antibodies were eluted with glycine buffer at pH 2.5. After neutralization, dialysis and centrifugation, the immunospecifically purified antibodies were lyophilized for storage. The purified anti-IgM was coupled to purified calf intestinal alkaline phosphatase (PL Biochemicals, Milwaukee, WI) with glutaraldehyde by the m e t h o d of Engvall and Perlmann (1971). The stock conjugate was diluted to the appropriate concentration in a buffer at pH 8.0 containing 0.15 M NaC1, 0.0025 M Tris, 0.2% bovine serum albumin, 0.001 M magnesium chloride, 0.03 M sodium phosphate, 10 /~g/ml of purified human IgG, and 0.02% sodium azide. The sterile reagent is stable for at least 13 months at 2--8 ° C. The enzyme-labeled anti-IgM was tested for specificity by exposing it to discs coated with purified human IgG, IgM or normal goat serum, then assaying the discs for enzyme activity. Based on the absorbance of the final substrate solution with the IgM discs (33.5), the IgG discs (0.15), and the goat serum discs (0.08), it was estimated that the enzyme-labeled a n t i b o d y was 99.8% specific for IgM. The rheumatoid factor positive control represented a pool from 12 sera containing elevated levels of rheumatoid factor. This was diluted appropriately in normal sera negative for rheumatoid factor. Its potency in IU/ml was established by assaying it by the ELISA technic described below, while using the World Health Organization (WHO) 1st reference preparation (1970) of human rheumatoid arthritis serum as the primary standard (WHO Biological Substances, International Standards, 1977). This has been arbitrarily assigned a p o t e n c y of 100 IU/ampoule by the WHO. The lyophilized powder of 1 ampoule was dissolved in 1 ml of water, and serial 2-fold dilutions of this were made in the diluent for specimens, to represent 100, 50, 25 and 12.5 IU/ml. Using the logarithms of the observed absorbance readings and the corresponding IU/ml values for the WHO standard, a regression line was established in the m e m o r y of a hand-held calculator (Texas Instruments TIo55). From the straight line obtained, the IU/ml p o t e n c y of the ELISA positive control was established based on its absorbance value. It was found to have 96 IU/ml and was positive in the bentonite floculation test at a titer of 1/256.
Buffers and solutions Diluent for specimens. The test sample is diluted in 6% bovine serum albumin in phosphate-buffered saline (pH = 7.2, 0.15 M ) c o n t a i n i n g 0.05% Tween-20 and 0.1% sodium azide. Washing buffer. During the assay, the plastic discs were washed with 0.15 M sodium chloride buffered to pH 7.2--8.0 with 0.002 M Tris. When the absorbance at 405 nm was greater than the accurate readability of the
118 p h o t o m e t e r (> 1.5), dilutions of the substrate reaction solution were made in the washing buffer. Carbonate buffer. The a m o u n t of alkaline phosphatase activity per disc were d e t e r m i n e d using P- ni t r ophe nyl phosphate as substrate at a dilution of 1 mg/ml in a buf f er containing 0.016 M sodium bicarbonate, 0.012 M sodium carbonate, and 0.001 M magnesium chloride, pH 9.8 -+ 0.1.
Technique of ELISA assay To establish the standard curve in each assay, the r h e u m a t o i d factor positive control p o ol was diluted 1 : 2, 1 : 4, and 1 : 8 with the diluent for specimens. Ten pl of the undiluted c ontrol pool and each dilution were added to 500 pl of the diluent for specimens in f l a t - b o t t o m e d vials. Similarly, 10 pl o f each u n k n o w n test specimen was added to 500 pl of the diluent (1 : 51 final dilution for each sample). Using the CORDIA ® Disc Transfer Device (Cordis, Miami, FL), one plastic disc coated with goat IgG was added to each vial, and the vials were incubated for 45 min at 37°C in a shaking water bath. The vials were then transferred to an ice-water bath and each disc was washed 5 times with 2.5 ml of wash solution using the CORDIA ® Disc Washing Device and the CORDIA ® 250 Wash Dispenser. Each wash cycle was 1.3 sec. After washing the discs, 500 pl of the alkaline phosphate-labeled anti-IgM were added to each vial, and t h e y were again incubated for 45 min at 37°C in a shaking water bath. The vials were then immersed in an ice-water bath, and each disc was washed 5 times. The washed discs were then transferred to clean unused vials with the CORDIA ® Disc Transfer Device, and 1000 pl of cold freshly prepared p - n i t r o p h e n y l p h o s p h ate substrate solution were added to each vial. After 30 min incubation at 37°C in a shaking water bath, the chromogenic reaction was terminated by the addition of 2 drops of 3 M sodium h y d r o x i d e . The absorbance o f each sample at 405 n m was read against a distilled water blank. When the absorbance of a specimen was greater than 1.5, appropriate dilutions were made and the true absorbance determined.
In terpretation of results The straight line standard curve was obtained by plotting the logarithms o f the absorbance values o f each dilution of the r h e u m a t o i d factor positive control against the logarithms of its p o t e n c y in IU/ml. Although this could be accomplished by the use of log-log graph paper, it is much more conveniently and precisely done by establishing the regression line. The coefficient o f correlation of the points in the standard curve was usually 0.995 or greater. The r h e u m a t o i d factor level in IU/ml of each u n k n o w n was established f r o m the standard curve.
Inhibition experiments Hu man IgG was dissolved in 0.2 M NaC1 (0.1 M Tris, pH 8.3) and heat aggregated at 63°C for 10 min. The supernatant was ultracentrifuged and
119 soluble ag~egates separated on a Sepharose 6B column. Goat IgG was aggregated by heating at 63°C for 20 min. Aggregated and non-aggregated human IgG and goat IgG were diluted in normal human serum devoid of rheumatoid factor activity. Ten pl of each aggregate at various concentrations and 10 pl of serum containing rheumatoid factor were each added to 500 gl of the incubation solution and allowed to react at 37°C for 30 min. The ELISA assay was then performed as described. Bentonite flocculation test The bentonite flocculation test was performed according to the m e t h o d of Bozicevitch et ah (1958). The substrate for the test is human IgG and serum specimens giving agglutination titers/> 1/32 are considered positive. Samples tested A total of 360 serum samples from separate persons were tested by both the bentonite flocculation test and the ELISA assay. All sera tested had been stored frozen at --20°C for less than 2 years. A standard curve of the reference RF positive control was established for each assay, which always included a negative control. The samples consisted of 162 normal controls, 60 from patients with rheumatoid arthritis, 66 from patients with systemic lupus erythematosus, 41 from patients with other diseases (23 Sj5gren's syndrome, 8 scleroderma, 5 mixed connective tissue disease, 2 osteoarthritis and 1 each from polymyositis, psoriatic arthritis, and thalassemia arthritis), as well as 24 from patients with undiagnosed conditions. RESULTS Standard curve Two-fold dilutions of the reference standard were made in the diluent for specimens. Fig. 1 shows a typical standard curve on log-log paper. By plotting the absorbance against concentration, a straight line is obtained between 12 and 96 IU/ml. Specificity o f the assay Three lines of evidence support the specificity of the assay for measuring rheumatoid factor. (1) Fig. I shows that the standard curves obtained either with dilutions of the positive control rheumatoid factor serum pool or with dilutions of a representative rheumatoid factor serum sample are parallel. This has been repeated with 5 other samples with identical results (data not shown). The straight line relationship holds over an extended range of dilutions. (2) With two exceptions, all sera from patients negative by the BFT were negative by the ELISA assay. These samples were from two patients with systemic lupus erythematosus and may represent examples of non-agglutinating rheumatoid factors.
120 • RHEUMATOID FACTOR POSITIVECONTROL96 TU/ML 0 PATIENTRNEUMATI~DFACTOR(BFT 1 32,000 )
10.C
td < CO -CO <
PATIENT DILUTIONS UND I U / M L , P O S I T I V E CONTROL 9 6
112 48
118 1116 I~32
1':4 24
1164-
12
Fig. 1. Straight-line relationship b e t w e e n absorbance values for r h e u m a t o i d f a c t o r E L I S A assay and: the I U / m l values o f the positive control standard curve; serial dilutions of a serum f r o m a patient w i t h a high titer o f r h e u m a t o i d factor. Bars represent standard errors o f the m e a n for each point.
Two samples from patients with WaldenstrSm's macroglobulinema were tested. The IgM paraprotein in each case was devoid of RF activity by the BTF and, despite serum IgM levels of 10 and 50 mg/ml, were also negative by the ELISA method. Elevated levels of IgM thus did not produce false positive results by non-specific adherence to the discs. (3) Fig. 2 illustrates the ability of aggregated human and goat immunoglobulins to inhibit the binding of RF to the discs. Unaggregated human IgG showed no inhibition while unaggregated goat immunoglobulin provided a 20% decrease in absorbance at a final concentration in the assayed sample of HUMAN IgG • GOAT IgG
8L
D
,< ~3 d3 < < u) uJ
1
46 AGGREGATED
,~.2 IgG
3&.,~
4.8
CONC. ( p G / M L )
Fig. 2. I n h i b i t i o n o f b i n d i n g o f r h e u m a t o i d f a c t o r t o the discs by pre-incubation o f one test sample w i t h aggregated h u m a n or goat i m m u n o g l o b u l i n G.
121 154 pg/ml. R h e u m a t o i d factors f r om 6 patients with r h e u m a t o i d arthritis were subjected to inhibition experiments with aggregated human IgG. At a final c o n c e n t r a t i o n of 19.2 pg/ml absorbance decreased a mean of 46% (range 27--63%) while at a final c o n c e n t r a t i o n of 9.6 pg/ml absorbance decreased a mean of 23% (1--41%). At lower concentrations little or no inhibition occurred. The degree of inhibition was i n d e p e n d e n t of the initial c o n c e n t r a t i o n of r heum a t oi d f act or (absorbance) and implied an equilibrium between the immunoglobulin b o u n d to the discs and the aggregates in solution.
Reproducibility The intra-assay precision was estimated by establising the coefficient of variation for quadruplicates of samples at several dilutions. The results are summarized in Table 1. The mean coefficient of variation was 8% and was generally equal at b o t h higher and lower concentrations. After logarithmic transformation, the very high coefficients of correlation in the straight-line relationship between absorbance and IU/ml of these regression lines is evident (0.999, 0.999, 0.998). Comparison between the B F T and ELISA assay Normals. One hundr e d and sixty-two sera from normal controls were studied. The mean ELISA value (IU) was 13.0-+ 11.4 (S.D.). The ELISA value o f ~>36 IU (2 S.D. above the mean) was t herefore taken to represent the presence o f r h e u m a t o i d factor in the serum. Of the 162 normal sera, 4 (2.5%) exhibited positive B F T titers of 1/32, 1/64, 1/64, and 1/512. Three of the 4 B F T positive sera were ELISA negative, while the serum with a BFT titer o f 1/32 was ELISA positive at 42 IU. A total of 7 sera from the normal controls were ELISA positive (4.3%) at values ranging from 38 to 84 IU. Patients. A total of 191 sera from individual patients were tested for RF. The overall coefficient of correlation between the two assays was 87.0% (Table 2). For patients with r h e u m a t o i d arthritis it was 93.3%, for patients with systemic lupus e r yt hem a t os us it was 86.4%, for patients with ot her diseases it was 90.0%, and for patients with undiagnosed conditions it was only 67.5%. The relationships between B FT titer and RF concent rat i on determined by the ELISA m e t h o d for the above groups of patients are shown in Figs. 3, 4, 5 and 6. Several points stand out: (1) B F T titers of 1/32 are almost always negative by ELISA (8/9), while titers greater than 1/ 2048 are always positive. Titers between 1/64 and 1 / 1 0 2 4 are most oft en positive (77%; 49/ 64) with the majority of discordant results occurring in patients with systemic lupus e r y t h e m a t o s u s and u n k n o w n conditions. (2) For any specific BFT titer there are considerable differences in the level of RF as established by the ELISA assay. (3) For patients with r h e u m a t o i d arthritis and ot her diseases, increasing B F T titer is generally associated with higher levels in the ELISA assay. In contrast, patients with systemic lupus e r y t h e m a t o s u s
: 2 : 4 : 8 : 16 : 32 : 64 : 128
.
.
. .
.
.
96 48 24 12
. .
0.999
.
.
. .
71 62 21 11
8 12 7 6
0.999
526 253 132 66 33 16 8.2 4.1
5000 3400 2025 1325 830 490 330 225
Absorb. x 1000 (mean) 240 85 25 25 25 22 8.7 6.3
S.E.M.
9.6 5.0 2.4 3.8 6.0 8.9
Coeff. var. (%)
0.998
467 234 117 58 29 15 7 3.6
IU/ml a
4300 2700 2100 1200 770 470 310 210
Absorb. x 1000 (mean)
Serum patient 2
340 410 85 41 14 17 4.1 2.5
S.E.M.
16 30 8 7 4 7 3 2
Coeff. vat. (%)
a The I U / m l value for these specimens in this e x p e r i m e n t was d e t e r m i n e d by averaging the values f o u n d for each dilution t h r o u g h 1 : 32.
. .
.
.
1700 1000 590 380
Coeff. var. (%)
IU/ml a
S.E.M.
IU/ml a
Absorb. × 1000 (mean)
Serum p a t i e n t 1
ELISA R F positive c o n t r o l
Coefficient o f correlation (r)
0 1 1 1 1 1 1 1
Serum dilution
In addition, the coefficient o f correlation b e t w e e n the I U / m l value and the E L I S A a b s o r b a n c e for each s p e c i m e n was d e t e r m i n e d after logarithmic t r a n s f o r m a t i o n .
R E P R O D U C I B I L I T Y OF E L I S A F O R R H E U M A T O I D F A C T O R , E X P R E S S E D AS S T A N D A R D E R R O R OF THE MEAN (S.E.M.) AND C O E F F I C I E N T OF V A R I A T I O N (%) OF EACH A S S A Y P E R F O R M E D IN Q U A D R U P L I C A T E
TABLE 1
123 TABLE 2 C O M P A R I S O N OF R E S U L T S F O R THE D E T E C T I O N OF R H E U M A T O I D F A C T O R U S I N G THE B E N T O N I T E F L O C C U L A T I O N T E S T A N D E L I S A A total o f 1 6 6 / 1 9 1 sera (87%) gave c o m p a r a b l e results (positive or negative by b o t h assays). RA, r h e u m a t o i d arthritis; SLE, s y s t e m i c lupus e r y t h e m a t o s u s . B e n t o n i t e f l o c c u l a t i o n test
Total
4-
SLE
Other
Unknown
RA
53 16 S u b t o t a l 96
22
5
0 2 Subtotal
4. 7 S u b t o t a l 23
4
RA
Elisa + Elisa - -
Total
8
SLE
Other
Total
0
0
15
11
2
3 41 S u b t o t a l 70
119
Unknown
72
98 93 191
RHEUMATOID ARTHRITIS
91210
800 °
3 E
70C
i-t v
60C
o
5oc
D~ Ld T n-
40C
°•
•
°
°
°
~ 3oo w
200
• •
° °
0
•
100
• °°~ NEG 32
S
•
°
64 128 2 5 6 512 1024 2048 4 0 9 6 e19216000 32000 BENTONtTE FLOCCULATION T I T E R
Fig. 3. R e l a t i o n s h i p b e t w e e n E L I S A values and B F T in 60 p a t i e n t s with r h e u m a t o i d arthritis.
124 (900)
E SYSTEMIC LUPUS ERYTHEb4ATOSUS 300
2OC
|
~oc
NEG
32
64
128
256
512 1024 2048
32000
BENTONITE FLOCCULATION TITER
Fig. 4. Relationship between ELISA values and BFT in 66 patients with systemic lupus erythematosus.
a n d u n d i a g n o s e d c o n d i t i o n s s h o w n o a p p a r e n t relationship. T h e o n l y p a t i e n t s w i t h a negative B F T a n d a positive E L I S A were 2 p a t i e n t s w i t h systemic lupus erythematosus.
600
OTHER
CONNECTIVE
TISSUE DISEASES
50C H
n--
400
2 h
c~
30C
<
IE D
t.d "1-
200
< .-I Ld
100
NEG
32
~4
t
"
128
256 512 1024 204B 4 0 ~
BENTONITE
FLOCCULATION
8192 1600032000
TITER
Fig. 5. R e l a t i o n s h i p b e t w e e n E L I S A values and B F T in 41 patients w i t h other c o n n e c t i v e tissue diseases.
125 ~" 40C
Z
m
UNKNOWN
•
30C
o
~ 9
DIAGNOSIS
° 200
2
{
~<~ ~oc
NEG
32
64
BENTONITE
128
256
512
FLOCCULATION
1024 2048
4096
TITER
Fig. 6. Relationship between ELISA values and BFT in 24 patients with unknown diagnosis.
DISCUSSION The results show t ha t the ELISA assay as p e r f o r m e d under the conditions described can be used to determine the presence and quant i t y of r h e u m a t o i d factor in serum samples. By using a r h e u m a t o i d factor positive control standardized against the WHO reference serum preparation, and constructing a standard curve, one can establish a level o f RF above which < 5 % of normals are positive. At the level o f ~>36 IU/ml, the results of the ELISA assay correspond closely with the bent oni t e flocculation test (overall coefficient o f correlation o f 87%) at roughly the same level of sensitivity. A t t e m p t s to make the test m or e sensitive, by either selecting a lower level for normal or by performing the assay at a 1/25 dilution instead of at 1/51 only increased the false positive rate in bot h controls and patients. U n d o u b t e d l y the bent oni t e flocculation test and the ELISA assay measure antiglobulins with different reactivities and specificities within the heterogeneous response to chronic antigenic stimulation. Conventional methods measure only agglutinating antibodies while the ELISA m e t h o d may also measure non-agglutinating antibodies. This is exemplified by two sera from patients with systemic lupus e r yt hem a t os us t hat were positive only by the ELISA assay. Discordance between the assays m ay further be explained in part by differences in the substrate. The substrate in the bent oni t e flocculation test is h u m an IgG, while it is goat IgG in the ELISA assay. Similar differences have been observed between antibodies directed against hum an IgG and rabbit IgG (Williams and Kunkel, 1963) The results also indicate t hat anti-goat Ig is n o t specific for r h e u m a t o i d arthritis, since it occurs at significant levels in some patients with systemic lupus erythematosus, SjSgren's s y n d r o m e and in a variety of ot her conditions.
126 T h e r e are several a d v a n t a g e s o f t h e E L I S A assay over a g g l u t i n a t i o n reactions. (1) Standardization of reagents: the c o a t e d discs and c o n j u g a t e d anti-IgM have l o n g shelf lives w h i c h allows the use of identical reagents on a d a y - t o - d a y basis. (2) Convenience: t h e use o f i m m u n o g l o b u l i n c o a t e d discs a n d disposable vials m a k e f o r easy handling. T h e C O R D I A ® Disc Washing Device and COP~DIA ® 250 Wash D i s p e n s e r allow t h o r o u g h washing o f t h e discs to decrease n o n - s p e c i f i c b i n d i n g o f p r o t e i n . T h e r e a d i n g o f results is n o t subjective, as in a g g l u t i n a t i o n . T h e R F level o b t a i n e d is a f u n c t i o n of an objective n u m e r i c a l a b s o r b a n c e value m e a s u r e d b y a p h o t o m e t e r . T h e t o t a l t i m e r e q u i r e d to q u a n t i t a t e t h e level o f R F in 100 s a m p l e s is a b o u t 4 h c o m p a r e d t o several d a y s f o r fully titering 1 0 0 positive s a m p l e s b y the B F T m e t h o d . (3) Reproducibility: the m i n i m a l e r r o r o f an a g g l u t i n a t i o n r e a c t i o n using serial 2-fold d i l u t i o n s is 100%, while t h e e r r o r o f the E L I S A assay is generally less t h a n 10% at b o t h high a n d low c o n c e n t r a t i o n s o f R F . T h e low coefficient of v a r i a t i o n m a k e s the use o f s a m p l e d u p l i c a t i o n u n n e c e s s a r y . T h e linearity of the s t a n d a r d curve and o f p a t i e n t s s a m p l e s indicates t h a t t h e assay is q u a n t i t a t i v e . This q u a n t i t a t i v e r e p r o d u c i b l e n a t u r e o f t h e assay allows o n e to d e t e c t small v a r i a t i o n s in R F level a n d c o u l d be useful in evaluating the effects of treatment. SUMMARY T h e E L I S A assay d e s c r i b e d p r o v i d e s a sensitive, r e p r o d u c i b l e , q u a n t i t a t i v e m e a n s o f m e a s u r i n g IgM r h e u m a t o i d f a c t o r in s e r u m a n d m a y p r o v i d e a v e r y c o n v e n i e n t m e a n s o f establishing changes in t h e level of r h e u m a t o i d factor. ACKNOWLEDGEMENT T h e a u t h o r s are i n d e b t e d to Mr. M. A n k e n for e x c e l l e n t t e c h n i c a l assistance. REFERENCES Avrameas, M.S., T. Ternynck and B. Guilbert, 1969, C.R. Acad. Sci. 268, 227. Bozicevitch, J., J.J. Bunim, J. Freund and S.D. Ward, 1958, Proc. Soc. Exp. Biol. Med. 97, 180. Butler, V.P. and J.H. Vaughan, 1964, Proc. Soc. Exp. Biol. Med. 116,585. Cathcart, E.S. and J.B. O'Sullivan, 1965, Arthr. Rheum. 8, 530. Chandler, R.W. and M. Robinson, 1974, Am. J. Clin. Pathol. 61, 47. Engvall, E. and P. Perlmann, 1971, Immunochemistry 8,874. Halbert, S.P. and M. Anken, 1977, J. Infect. Dis. 136, $318. Maiolini, R., B. Ferrua, J.F. Quaranta, A. Pinoteau, L. Euller, G. Ziegler and R. Masseyeff, 1978, J. Immunol. Methods 20, 25. Singer, J.M. and C.M. Plotz, 1956, Am. J. Med. 21,888. Willems, F.T.C. and C.C.M.K. De Kort, 1978, Lancet ii, 994. Williams, R.C. and H.G. Kunkel, 1963, Arthr. Rheum. 6,665. World Health Organization, 1977, Biological Substances. International Standards, Reference Preparations and Reference Reagents (W.H.O., Geneva) p. 24.
115
Q U A N T I T A T I V E D E T E R M I N A T I O N O F R H E U M A T O I D F A C T O R BY AN ENZYME-LABELED IMMUNOASSAY
J. KARSH, S.P. HALBERT, E. KLIMA and A.D. STEINBERG Clinical Immunology Section, National Institute o f Dental Research; The Arthritis Branch, National Institute o f Arthritis, Metabolism and Digestive Diseases, National Institutes o f Health, Bethesda, MD 20205; and the Cordis Laboratories, Miami, FL 33127, U.S.A.
(Received 2 July 1979, accepted 19 September 1979)
The presence and quantity of rheumatoid factor (RF) in human serum were determined by an indirect enzyme-linked immunoabsorbent assay (ELISA). A human rheumatoid factor control serum, standardized against the WHO reference rheumatoid arthritis serum preparation, was used to derive a standard curve in each assay. The results of unknowns were estimated from the standard curve and reported in International Units (IU) per milliliter (ml). The ELISA assay was compared with the bentonite flocculation test. The overall coefficient of correlation between the two assays was 87%; it was 93.3% for patients with rheumatoid arthritis but only 67.5% for patients with undiagnosed conditions. The error of the ELISA assay (coefficient of variation) was generally less than 10% at both high and low concentrations of rheumatoid factor. The quantitative reproducible nature of the assay allows the detection of small variations of rheumatoid factor level and could be useful in the serial evaluation of patients.
INTRODUCTION R h e u m a t o i d f a c t o r s are a n t i b o d i e s d i r e c t e d a g a i n s t t h e F c p o r t i o n o f IgG. T h e i r d e t e c t i o n in b i o l o g i c a l f l u i d s has b e c o m e a c o m m o n l a b o r a t o r y p r o c e dure; the information derived regarding both the presence and amount of r h e u m a t o i d f a c t o r is o f t e n u s e d t o s u p p o r t or r e f u t e clinical d i a g n o s i s a n d in the management of individual patients. T h e s o u r c e o f I g G has l i t t l e a p p a r e n t c l i n i c a l s p e c i f i c i t y a n d c o n v e n t i o n a l methods take advantage of the broad reactivity of circulating rheumatoid factors with i m m u n o g l o b u l i n s of m a n y species (Butler and Vaughan, 1964). T h e c o m m o n assays m e a s u r e o n l y IgM r h e u m a t o i d f a c t o r s a n d are b a s e d on the ability of these polyvalent immunoglobulins to agglutinate particles c o a t e d w i t h IgG, such as r e d b l o o d cells ( C a t h c a r t a n d O ' S u l l i v a n , 1 9 6 5 ) , l a t e x ( S i n g e r a n d P l o t z , 1 9 5 6 ) , o r b e n t o n i t e ( B o z i c e v i t c h e t al., 1 9 5 8 ) . T h e s e a s s a y s have c e r t a i n d i s a d v a n t a g e s in r o u t i n e c l i n i c a l use. T h e y are g e n e r a l l y n o t standardized and considerable d a y - t o - d a y variability can occur (Chandler a n d R o b i n s o n , 1 9 7 4 ) . S e m i - q u a n t i t a t i o n is a c h i e v e d b y serial d i l u t i o n o f
116 sera and selection of the last positive reaction. However, the process of serial dilution is time consuming, and the precision of end point titration is poor. In an effort to overcome these problems, several new second generation assays for the measurement of rheumatoid factors have been reported (Maiolini et al., 1978; Willems and De Kort, 1978). They offer the advantages of using standardized reagents and of being quick, reproducible and quantitative. We have developed an indirect enzyme-linked immunosorbent assay (ELISA) to detect and quantitate IgM rheumatoid factor in serum. The substrate is goat immunoglobulin G covalently bound to plastic discs. A human rheumatoid factor control serum pool was standardized against the WHO reference rheumatoid arthritis serum preparation, and was used to derive a standard curve in each assay. The results of single dilutions of the unknowns were estimated from the standard curves and reported in International Units/ml (IU/ml). The accuracy of the m e t h o d was confirmed by multiple replications of individual sera as well as serial dilution of several sera. METHODS AND MATERIALS
Preparation of goat immunoglobulin and linkage to immunosorbent Goat immunoglobulin G (Cohen Fraction II) was obtained from ICN Pharmaceuticals (Cleveland, OH). It was dissolved in 0.1 M sodium carbonate buffer, pH 9.6, and clarified by centrifugation at 9000 rev/min at 4°C. The goat IgG was covalently coated onto small plastic discs (13 mm diameter, 0.3 mm thick) via isothiocyanate groups introduced into their surface (Halbert and Anken, 1977). Exposure at the rate of 7 pg/disc at 2--8°C was permitted for 16--20 h with continuous gentle agitation. After washing the discs with buffered saline, they were rinsed in a bovine serum albumin solution (2 mg/ml in 0.01 sodium phosphate pH 8), drained and lyophilized. In this state, they are stable for at least 14 m o n t h s when stored at 2--8°C in the presence of a desiccant. Preliminary studies were carried out with discs coated with nonaggregated and with heat-aggregated goat immunoglobulin G (heated to 60°C for 10 min), but the results indicated that no significant difference in the ELISA read-out was evident when non-aggregated and aggregated goat IgG were compared.
Preparation and labeling o f goat anti-human IgM Human immunoglobulin M (IgM) was isolated from a patient with WaldenstrSm's macroglobulinemia by precipitation with a m m o n i u m sulfate (1.37 M), followed by Sepharose 6B gel filtration. The purified IgM was shown to be free of detectable IgG by gel diffusion methods, and was used to immunize goats following emulsification with complete Freund's adjuvant. The resulting p o t e n t antiserum was absorbed with a solid-phase im-
117 munoadsorbent prepared from specially selected human cord sera found to contain very low levels of IgM. The cord serum was insolubilized to a gel by glutaraldehyde cross-linking using the m e t h o d of Avrameas et al. (1969). After the antiserum was made monospecific for IgM, it was passed through a column of solid-phase purified IgM, and the antibodies were eluted with glycine buffer at pH 2.5. After neutralization, dialysis and centrifugation, the immunospecifically purified antibodies were lyophilized for storage. The purified anti-IgM was coupled to purified calf intestinal alkaline phosphatase (PL Biochemicals, Milwaukee, WI) with glutaraldehyde by the m e t h o d of Engvall and Perlmann (1971). The stock conjugate was diluted to the appropriate concentration in a buffer at pH 8.0 containing 0.15 M NaC1, 0.0025 M Tris, 0.2% bovine serum albumin, 0.001 M magnesium chloride, 0.03 M sodium phosphate, 10 /~g/ml of purified human IgG, and 0.02% sodium azide. The sterile reagent is stable for at least 13 months at 2--8 ° C. The enzyme-labeled anti-IgM was tested for specificity by exposing it to discs coated with purified human IgG, IgM or normal goat serum, then assaying the discs for enzyme activity. Based on the absorbance of the final substrate solution with the IgM discs (33.5), the IgG discs (0.15), and the goat serum discs (0.08), it was estimated that the enzyme-labeled a n t i b o d y was 99.8% specific for IgM. The rheumatoid factor positive control represented a pool from 12 sera containing elevated levels of rheumatoid factor. This was diluted appropriately in normal sera negative for rheumatoid factor. Its potency in IU/ml was established by assaying it by the ELISA technic described below, while using the World Health Organization (WHO) 1st reference preparation (1970) of human rheumatoid arthritis serum as the primary standard (WHO Biological Substances, International Standards, 1977). This has been arbitrarily assigned a p o t e n c y of 100 IU/ampoule by the WHO. The lyophilized powder of 1 ampoule was dissolved in 1 ml of water, and serial 2-fold dilutions of this were made in the diluent for specimens, to represent 100, 50, 25 and 12.5 IU/ml. Using the logarithms of the observed absorbance readings and the corresponding IU/ml values for the WHO standard, a regression line was established in the m e m o r y of a hand-held calculator (Texas Instruments TIo55). From the straight line obtained, the IU/ml p o t e n c y of the ELISA positive control was established based on its absorbance value. It was found to have 96 IU/ml and was positive in the bentonite floculation test at a titer of 1/256.
Buffers and solutions Diluent for specimens. The test sample is diluted in 6% bovine serum albumin in phosphate-buffered saline (pH = 7.2, 0.15 M ) c o n t a i n i n g 0.05% Tween-20 and 0.1% sodium azide. Washing buffer. During the assay, the plastic discs were washed with 0.15 M sodium chloride buffered to pH 7.2--8.0 with 0.002 M Tris. When the absorbance at 405 nm was greater than the accurate readability of the
118 p h o t o m e t e r (> 1.5), dilutions of the substrate reaction solution were made in the washing buffer. Carbonate buffer. The a m o u n t of alkaline phosphatase activity per disc were d e t e r m i n e d using P- ni t r ophe nyl phosphate as substrate at a dilution of 1 mg/ml in a buf f er containing 0.016 M sodium bicarbonate, 0.012 M sodium carbonate, and 0.001 M magnesium chloride, pH 9.8 -+ 0.1.
Technique of ELISA assay To establish the standard curve in each assay, the r h e u m a t o i d factor positive control p o ol was diluted 1 : 2, 1 : 4, and 1 : 8 with the diluent for specimens. Ten pl of the undiluted c ontrol pool and each dilution were added to 500 pl of the diluent for specimens in f l a t - b o t t o m e d vials. Similarly, 10 pl o f each u n k n o w n test specimen was added to 500 pl of the diluent (1 : 51 final dilution for each sample). Using the CORDIA ® Disc Transfer Device (Cordis, Miami, FL), one plastic disc coated with goat IgG was added to each vial, and the vials were incubated for 45 min at 37°C in a shaking water bath. The vials were then transferred to an ice-water bath and each disc was washed 5 times with 2.5 ml of wash solution using the CORDIA ® Disc Washing Device and the CORDIA ® 250 Wash Dispenser. Each wash cycle was 1.3 sec. After washing the discs, 500 pl of the alkaline phosphate-labeled anti-IgM were added to each vial, and t h e y were again incubated for 45 min at 37°C in a shaking water bath. The vials were then immersed in an ice-water bath, and each disc was washed 5 times. The washed discs were then transferred to clean unused vials with the CORDIA ® Disc Transfer Device, and 1000 pl of cold freshly prepared p - n i t r o p h e n y l p h o s p h ate substrate solution were added to each vial. After 30 min incubation at 37°C in a shaking water bath, the chromogenic reaction was terminated by the addition of 2 drops of 3 M sodium h y d r o x i d e . The absorbance o f each sample at 405 n m was read against a distilled water blank. When the absorbance of a specimen was greater than 1.5, appropriate dilutions were made and the true absorbance determined.
In terpretation of results The straight line standard curve was obtained by plotting the logarithms o f the absorbance values o f each dilution of the r h e u m a t o i d factor positive control against the logarithms of its p o t e n c y in IU/ml. Although this could be accomplished by the use of log-log graph paper, it is much more conveniently and precisely done by establishing the regression line. The coefficient o f correlation of the points in the standard curve was usually 0.995 or greater. The r h e u m a t o i d factor level in IU/ml of each u n k n o w n was established f r o m the standard curve.
Inhibition experiments Hu man IgG was dissolved in 0.2 M NaC1 (0.1 M Tris, pH 8.3) and heat aggregated at 63°C for 10 min. The supernatant was ultracentrifuged and
119 soluble ag~egates separated on a Sepharose 6B column. Goat IgG was aggregated by heating at 63°C for 20 min. Aggregated and non-aggregated human IgG and goat IgG were diluted in normal human serum devoid of rheumatoid factor activity. Ten pl of each aggregate at various concentrations and 10 pl of serum containing rheumatoid factor were each added to 500 gl of the incubation solution and allowed to react at 37°C for 30 min. The ELISA assay was then performed as described. Bentonite flocculation test The bentonite flocculation test was performed according to the m e t h o d of Bozicevitch et ah (1958). The substrate for the test is human IgG and serum specimens giving agglutination titers/> 1/32 are considered positive. Samples tested A total of 360 serum samples from separate persons were tested by both the bentonite flocculation test and the ELISA assay. All sera tested had been stored frozen at --20°C for less than 2 years. A standard curve of the reference RF positive control was established for each assay, which always included a negative control. The samples consisted of 162 normal controls, 60 from patients with rheumatoid arthritis, 66 from patients with systemic lupus erythematosus, 41 from patients with other diseases (23 Sj5gren's syndrome, 8 scleroderma, 5 mixed connective tissue disease, 2 osteoarthritis and 1 each from polymyositis, psoriatic arthritis, and thalassemia arthritis), as well as 24 from patients with undiagnosed conditions. RESULTS Standard curve Two-fold dilutions of the reference standard were made in the diluent for specimens. Fig. 1 shows a typical standard curve on log-log paper. By plotting the absorbance against concentration, a straight line is obtained between 12 and 96 IU/ml. Specificity o f the assay Three lines of evidence support the specificity of the assay for measuring rheumatoid factor. (1) Fig. I shows that the standard curves obtained either with dilutions of the positive control rheumatoid factor serum pool or with dilutions of a representative rheumatoid factor serum sample are parallel. This has been repeated with 5 other samples with identical results (data not shown). The straight line relationship holds over an extended range of dilutions. (2) With two exceptions, all sera from patients negative by the BFT were negative by the ELISA assay. These samples were from two patients with systemic lupus erythematosus and may represent examples of non-agglutinating rheumatoid factors.
120 • RHEUMATOID FACTOR POSITIVECONTROL96 TU/ML 0 PATIENTRNEUMATI~DFACTOR(BFT 1 32,000 )
10.C
td < CO -CO <
PATIENT DILUTIONS UND I U / M L , P O S I T I V E CONTROL 9 6
112 48
118 1116 I~32
1':4 24
1164-
12
Fig. 1. Straight-line relationship b e t w e e n absorbance values for r h e u m a t o i d f a c t o r E L I S A assay and: the I U / m l values o f the positive control standard curve; serial dilutions of a serum f r o m a patient w i t h a high titer o f r h e u m a t o i d factor. Bars represent standard errors o f the m e a n for each point.
Two samples from patients with WaldenstrSm's macroglobulinema were tested. The IgM paraprotein in each case was devoid of RF activity by the BTF and, despite serum IgM levels of 10 and 50 mg/ml, were also negative by the ELISA method. Elevated levels of IgM thus did not produce false positive results by non-specific adherence to the discs. (3) Fig. 2 illustrates the ability of aggregated human and goat immunoglobulins to inhibit the binding of RF to the discs. Unaggregated human IgG showed no inhibition while unaggregated goat immunoglobulin provided a 20% decrease in absorbance at a final concentration in the assayed sample of HUMAN IgG • GOAT IgG
8L
D
,< ~3 d3 < < u) uJ
1
46 AGGREGATED
,~.2 IgG
3&.,~
4.8
CONC. ( p G / M L )
Fig. 2. I n h i b i t i o n o f b i n d i n g o f r h e u m a t o i d f a c t o r t o the discs by pre-incubation o f one test sample w i t h aggregated h u m a n or goat i m m u n o g l o b u l i n G.
121 154 pg/ml. R h e u m a t o i d factors f r om 6 patients with r h e u m a t o i d arthritis were subjected to inhibition experiments with aggregated human IgG. At a final c o n c e n t r a t i o n of 19.2 pg/ml absorbance decreased a mean of 46% (range 27--63%) while at a final c o n c e n t r a t i o n of 9.6 pg/ml absorbance decreased a mean of 23% (1--41%). At lower concentrations little or no inhibition occurred. The degree of inhibition was i n d e p e n d e n t of the initial c o n c e n t r a t i o n of r heum a t oi d f act or (absorbance) and implied an equilibrium between the immunoglobulin b o u n d to the discs and the aggregates in solution.
Reproducibility The intra-assay precision was estimated by establising the coefficient of variation for quadruplicates of samples at several dilutions. The results are summarized in Table 1. The mean coefficient of variation was 8% and was generally equal at b o t h higher and lower concentrations. After logarithmic transformation, the very high coefficients of correlation in the straight-line relationship between absorbance and IU/ml of these regression lines is evident (0.999, 0.999, 0.998). Comparison between the B F T and ELISA assay Normals. One hundr e d and sixty-two sera from normal controls were studied. The mean ELISA value (IU) was 13.0-+ 11.4 (S.D.). The ELISA value o f ~>36 IU (2 S.D. above the mean) was t herefore taken to represent the presence o f r h e u m a t o i d factor in the serum. Of the 162 normal sera, 4 (2.5%) exhibited positive B F T titers of 1/32, 1/64, 1/64, and 1/512. Three of the 4 B F T positive sera were ELISA negative, while the serum with a BFT titer o f 1/32 was ELISA positive at 42 IU. A total of 7 sera from the normal controls were ELISA positive (4.3%) at values ranging from 38 to 84 IU. Patients. A total of 191 sera from individual patients were tested for RF. The overall coefficient of correlation between the two assays was 87.0% (Table 2). For patients with r h e u m a t o i d arthritis it was 93.3%, for patients with systemic lupus e r yt hem a t os us it was 86.4%, for patients with ot her diseases it was 90.0%, and for patients with undiagnosed conditions it was only 67.5%. The relationships between B FT titer and RF concent rat i on determined by the ELISA m e t h o d for the above groups of patients are shown in Figs. 3, 4, 5 and 6. Several points stand out: (1) B F T titers of 1/32 are almost always negative by ELISA (8/9), while titers greater than 1/ 2048 are always positive. Titers between 1/64 and 1 / 1 0 2 4 are most oft en positive (77%; 49/ 64) with the majority of discordant results occurring in patients with systemic lupus e r y t h e m a t o s u s and u n k n o w n conditions. (2) For any specific BFT titer there are considerable differences in the level of RF as established by the ELISA assay. (3) For patients with r h e u m a t o i d arthritis and ot her diseases, increasing B F T titer is generally associated with higher levels in the ELISA assay. In contrast, patients with systemic lupus e r y t h e m a t o s u s
: 2 : 4 : 8 : 16 : 32 : 64 : 128
.
.
. .
.
.
96 48 24 12
. .
0.999
.
.
. .
71 62 21 11
8 12 7 6
0.999
526 253 132 66 33 16 8.2 4.1
5000 3400 2025 1325 830 490 330 225
Absorb. x 1000 (mean) 240 85 25 25 25 22 8.7 6.3
S.E.M.
9.6 5.0 2.4 3.8 6.0 8.9
Coeff. var. (%)
0.998
467 234 117 58 29 15 7 3.6
IU/ml a
4300 2700 2100 1200 770 470 310 210
Absorb. x 1000 (mean)
Serum patient 2
340 410 85 41 14 17 4.1 2.5
S.E.M.
16 30 8 7 4 7 3 2
Coeff. vat. (%)
a The I U / m l value for these specimens in this e x p e r i m e n t was d e t e r m i n e d by averaging the values f o u n d for each dilution t h r o u g h 1 : 32.
. .
.
.
1700 1000 590 380
Coeff. var. (%)
IU/ml a
S.E.M.
IU/ml a
Absorb. × 1000 (mean)
Serum p a t i e n t 1
ELISA R F positive c o n t r o l
Coefficient o f correlation (r)
0 1 1 1 1 1 1 1
Serum dilution
In addition, the coefficient o f correlation b e t w e e n the I U / m l value and the E L I S A a b s o r b a n c e for each s p e c i m e n was d e t e r m i n e d after logarithmic t r a n s f o r m a t i o n .
R E P R O D U C I B I L I T Y OF E L I S A F O R R H E U M A T O I D F A C T O R , E X P R E S S E D AS S T A N D A R D E R R O R OF THE MEAN (S.E.M.) AND C O E F F I C I E N T OF V A R I A T I O N (%) OF EACH A S S A Y P E R F O R M E D IN Q U A D R U P L I C A T E
TABLE 1
123 TABLE 2 C O M P A R I S O N OF R E S U L T S F O R THE D E T E C T I O N OF R H E U M A T O I D F A C T O R U S I N G THE B E N T O N I T E F L O C C U L A T I O N T E S T A N D E L I S A A total o f 1 6 6 / 1 9 1 sera (87%) gave c o m p a r a b l e results (positive or negative by b o t h assays). RA, r h e u m a t o i d arthritis; SLE, s y s t e m i c lupus e r y t h e m a t o s u s . B e n t o n i t e f l o c c u l a t i o n test
Total
4-
SLE
Other
Unknown
RA
53 16 S u b t o t a l 96
22
5
0 2 Subtotal
4. 7 S u b t o t a l 23
4
RA
Elisa + Elisa - -
Total
8
SLE
Other
Total
0
0
15
11
2
3 41 S u b t o t a l 70
119
Unknown
72
98 93 191
RHEUMATOID ARTHRITIS
91210
800 °
3 E
70C
i-t v
60C
o
5oc
D~ Ld T n-
40C
°•
•
°
°
°
~ 3oo w
200
• •
° °
0
•
100
• °°~ NEG 32
S
•
°
64 128 2 5 6 512 1024 2048 4 0 9 6 e19216000 32000 BENTONtTE FLOCCULATION T I T E R
Fig. 3. R e l a t i o n s h i p b e t w e e n E L I S A values and B F T in 60 p a t i e n t s with r h e u m a t o i d arthritis.
124 (900)
E SYSTEMIC LUPUS ERYTHEb4ATOSUS 300
2OC
|
~oc
NEG
32
64
128
256
512 1024 2048
32000
BENTONITE FLOCCULATION TITER
Fig. 4. Relationship between ELISA values and BFT in 66 patients with systemic lupus erythematosus.
a n d u n d i a g n o s e d c o n d i t i o n s s h o w n o a p p a r e n t relationship. T h e o n l y p a t i e n t s w i t h a negative B F T a n d a positive E L I S A were 2 p a t i e n t s w i t h systemic lupus erythematosus.
600
OTHER
CONNECTIVE
TISSUE DISEASES
50C H
n--
400
2 h
c~
30C
<
IE D
t.d "1-
200
< .-I Ld
100
NEG
32
~4
t
"
128
256 512 1024 204B 4 0 ~
BENTONITE
FLOCCULATION
8192 1600032000
TITER
Fig. 5. R e l a t i o n s h i p b e t w e e n E L I S A values and B F T in 41 patients w i t h other c o n n e c t i v e tissue diseases.
125 ~" 40C
Z
m
UNKNOWN
•
30C
o
~ 9
DIAGNOSIS
° 200
2
{
~<~ ~oc
NEG
32
64
BENTONITE
128
256
512
FLOCCULATION
1024 2048
4096
TITER
Fig. 6. Relationship between ELISA values and BFT in 24 patients with unknown diagnosis.
DISCUSSION The results show t ha t the ELISA assay as p e r f o r m e d under the conditions described can be used to determine the presence and quant i t y of r h e u m a t o i d factor in serum samples. By using a r h e u m a t o i d factor positive control standardized against the WHO reference serum preparation, and constructing a standard curve, one can establish a level o f RF above which < 5 % of normals are positive. At the level o f ~>36 IU/ml, the results of the ELISA assay correspond closely with the bent oni t e flocculation test (overall coefficient o f correlation o f 87%) at roughly the same level of sensitivity. A t t e m p t s to make the test m or e sensitive, by either selecting a lower level for normal or by performing the assay at a 1/25 dilution instead of at 1/51 only increased the false positive rate in bot h controls and patients. U n d o u b t e d l y the bent oni t e flocculation test and the ELISA assay measure antiglobulins with different reactivities and specificities within the heterogeneous response to chronic antigenic stimulation. Conventional methods measure only agglutinating antibodies while the ELISA m e t h o d may also measure non-agglutinating antibodies. This is exemplified by two sera from patients with systemic lupus e r yt hem a t os us t hat were positive only by the ELISA assay. Discordance between the assays m ay further be explained in part by differences in the substrate. The substrate in the bent oni t e flocculation test is h u m an IgG, while it is goat IgG in the ELISA assay. Similar differences have been observed between antibodies directed against hum an IgG and rabbit IgG (Williams and Kunkel, 1963) The results also indicate t hat anti-goat Ig is n o t specific for r h e u m a t o i d arthritis, since it occurs at significant levels in some patients with systemic lupus erythematosus, SjSgren's s y n d r o m e and in a variety of ot her conditions.
126 T h e r e are several a d v a n t a g e s o f t h e E L I S A assay over a g g l u t i n a t i o n reactions. (1) Standardization of reagents: the c o a t e d discs and c o n j u g a t e d anti-IgM have l o n g shelf lives w h i c h allows the use of identical reagents on a d a y - t o - d a y basis. (2) Convenience: t h e use o f i m m u n o g l o b u l i n c o a t e d discs a n d disposable vials m a k e f o r easy handling. T h e C O R D I A ® Disc Washing Device and COP~DIA ® 250 Wash D i s p e n s e r allow t h o r o u g h washing o f t h e discs to decrease n o n - s p e c i f i c b i n d i n g o f p r o t e i n . T h e r e a d i n g o f results is n o t subjective, as in a g g l u t i n a t i o n . T h e R F level o b t a i n e d is a f u n c t i o n of an objective n u m e r i c a l a b s o r b a n c e value m e a s u r e d b y a p h o t o m e t e r . T h e t o t a l t i m e r e q u i r e d to q u a n t i t a t e t h e level o f R F in 100 s a m p l e s is a b o u t 4 h c o m p a r e d t o several d a y s f o r fully titering 1 0 0 positive s a m p l e s b y the B F T m e t h o d . (3) Reproducibility: the m i n i m a l e r r o r o f an a g g l u t i n a t i o n r e a c t i o n using serial 2-fold d i l u t i o n s is 100%, while t h e e r r o r o f the E L I S A assay is generally less t h a n 10% at b o t h high a n d low c o n c e n t r a t i o n s o f R F . T h e low coefficient of v a r i a t i o n m a k e s the use o f s a m p l e d u p l i c a t i o n u n n e c e s s a r y . T h e linearity of the s t a n d a r d curve and o f p a t i e n t s s a m p l e s indicates t h a t t h e assay is q u a n t i t a t i v e . This q u a n t i t a t i v e r e p r o d u c i b l e n a t u r e o f t h e assay allows o n e to d e t e c t small v a r i a t i o n s in R F level a n d c o u l d be useful in evaluating the effects of treatment. SUMMARY T h e E L I S A assay d e s c r i b e d p r o v i d e s a sensitive, r e p r o d u c i b l e , q u a n t i t a t i v e m e a n s o f m e a s u r i n g IgM r h e u m a t o i d f a c t o r in s e r u m a n d m a y p r o v i d e a v e r y c o n v e n i e n t m e a n s o f establishing changes in t h e level of r h e u m a t o i d factor. ACKNOWLEDGEMENT T h e a u t h o r s are i n d e b t e d to Mr. M. A n k e n for e x c e l l e n t t e c h n i c a l assistance. REFERENCES Avrameas, M.S., T. Ternynck and B. Guilbert, 1969, C.R. Acad. Sci. 268, 227. Bozicevitch, J., J.J. Bunim, J. Freund and S.D. Ward, 1958, Proc. Soc. Exp. Biol. Med. 97, 180. Butler, V.P. and J.H. Vaughan, 1964, Proc. Soc. Exp. Biol. Med. 116,585. Cathcart, E.S. and J.B. O'Sullivan, 1965, Arthr. Rheum. 8, 530. Chandler, R.W. and M. Robinson, 1974, Am. J. Clin. Pathol. 61, 47. Engvall, E. and P. Perlmann, 1971, Immunochemistry 8,874. Halbert, S.P. and M. Anken, 1977, J. Infect. Dis. 136, $318. Maiolini, R., B. Ferrua, J.F. Quaranta, A. Pinoteau, L. Euller, G. Ziegler and R. Masseyeff, 1978, J. Immunol. Methods 20, 25. Singer, J.M. and C.M. Plotz, 1956, Am. J. Med. 21,888. Willems, F.T.C. and C.C.M.K. De Kort, 1978, Lancet ii, 994. Williams, R.C. and H.G. Kunkel, 1963, Arthr. Rheum. 6,665. World Health Organization, 1977, Biological Substances. International Standards, Reference Preparations and Reference Reagents (W.H.O., Geneva) p. 24.