Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
1 156 142
4530 THE SOLID PHASE ANTI-C3 ASSAY
FOR DET~CTION OF IM~IUNE COMPLEXES
This invention relates to immunoche~istry and,
more specifically, to the detection and measurement
of immune complexes ~ICs) as an aid in diagnosis and
treatment of diseases.
Background
Upon exposure to most antigens, an individual responds
by synthesizing specific antibodies that subsequently
may interact with the inciting antigens and unite
noncovalently with them to form immu~e complexes (ICs).
This course of the normal immune response is designed to
eliminate and/or neutralize the antigens, thus benefiting
the host. However, there is no doubt that formation of
ICs under some circumstances is detrimental to the host.
The harmful effects of ICs were first suggested at
the beginning of this century by von Pirquet ~1911),
who proposed that the onset and course of serum sickness
were determined by toxic factors produced by the
interaction of host antibody and antigen in the circulation.
Thereafter, others made similar observations (Longcope,
20 1915; Rich and Gregory, 1943; Hawn and Janeway, 1947). The
definite pathogenic role of the ICs was shown during the
1950's by the work of Germuth (1953), Germuth and
~cKinnon (1957j, Germuth et al (1957), Dixon and associates
(1958, 1961), and Dixon (1963), who confirmed and amplified
25 von Pirquet's original theory using the experimental
rabbit model of "one shot" serum sickness. It was
demonstratedthat the onset of glomerulonephritis and
generalized vasculitis coincided with the appearance of
soluble ICs in rabbits' circulations, a decrease in serum
30 complement (c) activity, and a deposition of ICs in the
sites of injury (Dixon, 1963; Dixon et al, 1958, 1961).
~'
1156~42
In vivo and in ~itro experiments have more recently
clarified many factors involved in IC formation, removal,
and localization as well as the mechanisms of
IC-induced inflammatory reactions. Moreover, ICs are now
5 viewed, apart from their potential phlogogenicity, as
regulators of both cellular and humoral immune responses
by virtue of their capacities to interact with antigen
receptor-bearing lymphocytes and subpopulations of T and
B cells, as well as with unclassified lymphocytes and
10 macrophages having ~c and C receptors. With the
recognition of the immunopathologic consequences of ICs
and the development of new techniques for demonstrating
ICs in tissues and biological fluids, considerable evidence
has accumulated substantiating the primary pathogenic
15 significance of lCs in a variety of animal and human
diseases.
Circulating soluble ICs in man and lower animals are`
responsible for, or associated with, a diverse array of
diseases. These include autoimmune disease, neoplastic
20 diseases, infectious diseases due to bacteria, viruses
and parasites, and other unclassified disorders.
Both exogenous and endogenous antigens can trigger
pathogenic immune responses, resulting in IC disease. ~he
awareness of IC's im~ortant role in many diseases has
25 stimulated de~}elopment of techniques for demonstrating them
in tissues and biological fluids.
The presence of ICs in pathologic tissue specimens
can be inferred from several lines of evidence.
Histologically, the patterns of injury may appear similar
30 to those known to occur in experimental animals in which
IC disease has been induced (Wilson and Dixon, 1976).
Presumed IC deposits in tissues can be identified also by
electron microscopy and by conventional histochemical
techniques. Of the immunohistochemical techniques, the
35 most widely used is immunofluorescence (Wilson and Dixon,
1976), which allows Ig, C components, and in some cases
specific antigens, to be identified. When these materials
ll~S14~
deposit in a granular, discrete pattern, in all
likelihood the individual has an IC disorder. In
addition, if enough of the diseased tissue is available,
deposited ICs can be eluted from it by using low pH buffers
5 ~citrate buffer, ~CI-glycine buffer)(Oldstone, 1975a;
Koffler et al, 1971 l~oodroffe and Wilson, 1977;
Bartolotti, 1977) or chaotropic agents ~RI. KSCN)
~Edgington, 1971; Woodroffe and ~ilson, 1977), etc., after
which the antibodies and sometimes the antigens can be
10 recovered, quantitated, and identified (see Section VI).
Recently, interest has increased in developing
immunologic techniques for demonstrating ICs directly in
bodily fluids.
I Although historically the development of assays for
15 soluble ICs started with physicochemical rather than
im~unologic techniques, the latter are more commonly used
now. It is useful to divide all the available methods
into two main groups; antigen-specific tests, that is,
detection of a specific antigen complexed with antibody
2n or, by far the larger and more readily applicable group,
antigen-nonspecific tests.
* * * * *
TABLE I
ANTIGEN-NONSPECIFIC METHODS FOR
DETECTING CIRCULATION IMMUNE CO~PLEXES
.
1. Physical techniques
Analytical ultracentrifugation
Sucrose density gradient centrifugation
Gel filtration
Ultrafiltration
Electrophoresis
Polyethylene glycol (PEG) precipitation
Cryoprecipitation
1 156142
2. Methods based on the biologic characteristics of
immune complexes
a. Complement techniques
Microcomplement consumption test
Assays based on the interaction of ICs with
purified Clq (Clq precipitation in gels, ..
Clq-PEG test, Clq deviation tests, Clq
solid-phase radioimmunoassays)
Assays of breakdown products of C3 and Cl
The C3 precipitation assay
The conglutinin radioimmunoassay
b. Antiglobulin techiques
Rheumatoid factor tests
Other antiglobulin tests
c. Cellular techniques
The platelet aggregation test
Inhibition of antibody-dependent cell-mediated
to~icity
Intracytoplasmic staining of polymorphonuclear
leukocytes
Release of enzymes from eosinophils and mast
cells
The macrophage inhibition assay
Rosette inhibition.tests
The Raji cell assay
The L1210 murine leukemia cell assay
The human erythrocyte assay
d. Other methods
Binding to staphylococcal protein A
The inventors have summarized studies related to the
formation, localization, and fate of ICs and their
biologic activities, and described in considerable detail
the currently available methods for detecting circulating
35 ICs, the application of these methods to studying
115~ 2
IC-associated human diseases, and the state of present
attempts to isolate and characterize circulating ICs.
~ADVANCES IN IMMUNOLOGY, VO1~ 28, PP. 89-220, THE
BIOLOGY AND DET~CTION OF IM~UNE COMPLEXES1, Argyrios
5 N. Theofilopoulos and Frank J. Dixon). (The works referred
to above are specific~Ally identified at pp. ' 89-220, VQ1.
28, ADV. IN IMMUNOLOGY).
- Disclosure of the Invention
.
10 ABBREVIATIONS:
DEAE - diethylaminoethylcellulose
AGG - aggregated gamma globulin
BBS - borate buffered saline
BSA - bovine serum albumin
EDTA - ethylene diamino tetraacetic acid
HSA - human serum albumin
IC - immune complexes
NHS - normal human serum
PBS - phosphate buffered saline
RF - rheumatoid factor
SDG - sucrose density gradient
SLE - systemic lupus erythematosus
SUMMARY DISCLOSURE
Although the formation of antibodies against antigens
and subsequent formation of immune complexes is meant to
protect the host, in some circumstances the continuity of
this proces~s, or its intensity, may be harmful to the body.
Considerable clinical and laboxatory evidence is now
available indicating that immune complexes play an
important role in several autoimmune, infectious and
neoplastic diseases.
Most of the pathogenic effects of immune complexes are
believed to be due to their ability in activating the
complement system, followed by release of various phlogogenic
mediators from mononuclear cells. Being present and
related to so many pathologic situations, immune complexes
have deserved a great deal of attention and recently
1 1~616l2
numerous techniques have been developed for their detection
in biological fluids in the hope that their measurement
may help in understanding pathogenesis, prognosis and in
monitoring therapy. However, the reagents u~ilized in
these assays may give falsely positive or negative results
due to various interfering substances.
A new assay in which complement fixing immune complexes
are specifically detected. In this new assay an anti-C3
antibody is fixed to a solid matrix, serum samples and
standards are permitted to interact and the amount of
immunoglobulin bound to the solid matrix through the
C3-anti-C3 reaction is measured by a second reaction with a
radio or enzyme-labeled anti-Ig or staphylococcal protein A.
The test is easy to perform, the reagents are easily
preparable and stable. It is reproducible and can detect
6~ of aggregated gammaglobulin per ml of serum. The
reagents utilized and the procedure are outlined below.
Isolation of F(ab')2 anti-C3: Hyperimmune anti-C3
serum is prepared by repeatedly injecting goats with
purified human C3. The IgG fraction of this antiserum
is prepared by ammonium sulfate precipitation and DEAE
chromatography. F~ab')2 fragments are prepared and used
instead of the whole IgG molecule to avoid interference
by endogenous rheumatoid factors in the test samples.
This is done by pepsin digestion in O.lM acetate buffer,
pH 4.1, for 18 hours at 37~C, with an enzyme: substrate
ratio of 2:100. The reaction is stopped by neutralization
with 1. OM Tris and the F(ab')2 fragments are isolated by
gel filtration through Sephadex G-150. Elution is carried
out with borate buffered saline (0~2M borate, 0.075 M
NaCl, p~ 7.5 - BBS). The F(ab')2 peak is collected, pooled
and concentrated if necessary. Sodium azide is added as
preservative or it can be stored frozen. The best
concentration to be used in the assay (see below) is
determined by running standard curves with different
concentrations of F(ab')2. We have been using
B * Trademark
11~6~2
concentrations of 1-2 mg/ml but this is expected to vary
depending on the amount and avidity of the antibodyO
.. . . .. . . .. .. ..
Radioiodinated or en~yme-linked I~ anti-human IgG:
Goat anti-human IgG serum is used. The IgG fraction is
S prepared as described for the anti-C3 anti-serum~ but
the whole IgG molecule is used instead of the Ftab')2
fragments. It is radiolabeled with 125I by the Chloramine
T me~hod to give a specific activity of 0.3-0.5 ~Ci/~g
and diluted in 1~ human serum albumin in BBS (HSA-BBS),
- 10 to a final concentration of 100 ~g/ml. Alternatively,
we have utilized enzyme-linked instead of radiolabeled
antibody anti-human IgG. The anti-human IgG is linked
to horseradish peroxidase (HRPO) at a ratio of 8 mg of
IgG per 4 mg of enzyme following the procedure described
15 by Wilson and Nakane (53).*
Standards and samples: Aggregated human gamn~globulin
~AHG) is prepared by heating human Cohn Fraction II
dissolved in phosphate buffered saline (O.OlM at 3000
rpm for 15 min., in order to remove large aggregates,
20 this preparation is stored in 1% HSA-BBS at a concentration
of 1 mg/ml, at -70C, remaining active for several months.
For the standard curve, serial dilutions of AHG in HSA-BBS,
containing 20~g to 78 ng are added to 25~1 of normal
human serum, freshly collected or s~ored at -70C (as
25 source of complement) and incubated at 37C for 15 min.
After this period, H5A-BBS containing O.OlM EDTA (HSA-BBS-EDTA)
is added, to give a final dilution of 1/21, and 100 ~1 are
transferred to wells of flexible microtiter plates for
reaction with the anti-C3 antibody. The standard curve
30 so pr~pared corresponds to concentrations of 800 to 3.0
~g AHG/ml of serum. The serum samples are prepared by
*See references at the end.
.
115~ 1~2
simple dilution ~1/21) in the same diluent buffer and 100
~1 are added per well. Blood is collected in tubes containing
EDTA (final concentration of 0.01M) and plasma is
collected by centrifugation. Serum can be used but should
5 not be frozen and thawed many times so as to avoid
complement activation.
.
- ~ssay method: Flexible microtiter plates (Dynatech
. . .
Laboratories Inc., Alexandria, VA) are coated with
F(ab')2, 100 ~1 per well, dissolved in BBS, pH 7.5.
10 Plates are left overnight at 4~C, or 2-3 hours at 37C
with the same binding efficiency. F(ab')2 is then
aspirated (it can be reused at least 2 times for coating
new plates), 200 ~1 of 1% HSA in BBS, pH 7.5 are added
per well, to cover unreacted sites, and the plates are
let standing at room temperature for 1 hour. In the
meantime, standards and samples are prepared. The
HSA-BBS is aspirated off, the plate is washed once with
BBS, samples and standards are added (100 ~1 per well),
and the plate is incubated at 37C for 2 hours. After
5 washes with BBS, 10 ~g of 1 I goat anti-human IgG
is added, dissolved in 1~ HSA-BBS, 100 ~1 per well.
The plate is again incubated at 37C for 2 hours.
After this final incubation, the radioactive antibody
is aspirated off, the plate is washed 5 times with BBS
and dried up. Finally the wells are cut and counted in
a gammacounter.
When utilizing enzyme linked antibody, after ~he
final washes 200 ~1 of ABTS ~2,2'-azino-di (3-ethyl-
benzthiazoline sulphonic acid -6) ammonium salt] 1 mg/ml
of 0.1M citrate - phosphate buffer, pH 4.0 and 0.003~
H2O2, are added per well. After color development for
10-15 min. at room temperature, the reaction is stopped
by addition of 50 ~1 of 37 mM NaCN. The content of each
well is diluted to 1.0 ml and the OD read at 415 nm.
;1 4 2
C11 cal_application: We have applied the assay
to various human sera and good concordance was obtained with
the Raji cell (2) and conglutinin (3) assays previously
described, as far as positive and negative results. High
levels of circulating immune complexes were found in
systemic lupus erythematosus, rheumatoid arthritis,
Sjogrenls and Reiter's syndromes, subacute bacterial
endocarditis and acute glomerulonephritis.
* * * ~ *
TABLE II
ANTI-HUMAN C3 ASSAY APPLIED TO SEVERAL HUMAN
PAT~O~OGIC CONDITIONS FOR THE DETECTION OF
CIRCULATING IMMUNE COMPLEXES
t ~g AHG/ml
N (mean + SEM) % Positivity*
- Normal Controls 42 9.9 + 2.53
Rheumatoid Arthritis 22 234.7 + 57.88 68.0
Systemic Lupus
Erythematosus 20 270.8 + 117.73 37.0*
Sjogren's Syndrome 22 356.3 + 113.64 ~2.0
Reiter's Syndrome 25 406.2 ~ 93.95 52.0
Subacute Bacterial
Endocarditis 20 183.1 + 38.32 45.0
25 Acute Glomerulonephritis 20 432.8 + 164.38 30.0
* Values considered positive above 43 ~g AHG/ml (mean +
SD for normal population = 9.9 + 16.41)
** Includes patients with active and inactive SLE
* * * * *
Moreover, we have applied the assay to the analysis of sera
from mice with autoimmune diseases and rabbits with chronic
serum sickness by utilizing species specific reagents, and
shown the applicability of the assay to animal models of
diseases.
1 1~6 142
Thus, the anti-C3 assay has several advantages over
other assays~ t is specific and detects all
complement-fixing immune complexes, independently of
their ability to activate the classical or alternative
- 5 pathways; (2) the reagents are stable and easy to prepare;
(3) by using as a final reagent antisera against various
isotypes and against suspected antigens, both the nature
of the antibody and antigen involved in the immune
complexes can be determined; (4) by using appropriate
10 antisera to C3 and IgG it can be applied to different
animal species; (5) as a solid phase immunoassay the
principle of enzyme lirked immunoassays (ELIS~) can be
easily applied, so precluding the use of radioisotypes
in the assay.
Brief Descr_~ion of the Drawings
Figure 1 is a graph depicting binding of AGG in the
presence of fresh NHS and EDTA-NHS to anti-C3.
Figure 2 is a graph depicting levels of ICs in sera
of normal and SLE-prone mice by the anti-C3 assay. BALB/c,
20 C3H/St and MRL/l mice were females or males, NZBxW females
and B~SB males. All mice were 5-6 months of age, except
the NZB~ which were 8 months old.
Figure 3 is a graph depicting anti~C3 binding
activity in fractions obtained after SDG centrifugation
25 of pathologic (solid lines) and normal human sera (dotted
lines). The position of l9S and 7S markers is indicated
by the arrows. a) ~elty's syndrome; ~) SLE with
proliferative nephritis; c) Sjagren's syndrome; d) acute
glomerulonephritis; e) rheumatoid arthritis; f) SLE
30 without renal disease; g) Reiter's syndrome; h) chronic
granulocytic leukemia in blastic crisis. Three other sera
analyzed (SLE withou~ renal disease~ Reiter's syndrome,
rheumatoid arthritis) showed anti-C3 acti~ity e~clusively
in the 7-8S fractions whereas another serum (Sjogren's
35 syndrome)showed activity in intermediate (8-19S) fractions.
i 1 4 2
Figure ~ is a graph depicting conglutinin binding
activity in fractions obtained after SDG centrifugation
of pathologic (solid lines) and normal human sera (dotted
lines). The position of 19~ and 7S markers is indicated
5 by the arrows. Same sera as in Fig. 3.
Figure 5 is a graph depicting anti-C3 binding
activity in fractions obtained after SDG centrifugation
of a Felty's syndrome serum under neutral (pH 7.4, panel
A) and acidic ~pH 2.8, panel B) conditions. Note that
10 under acidic conditions (panel B) there is a shift of
the anti-C3 activity to the lighter fractions of the
gradient.
Figure 6 is a graph depicting SDG analysis of
5I-BSA- 31I-anti-BSA complexes made at equivalence and
15 subsequently solubilized with fresh NHS (C source). Note
that a substantial amount of the solubilized antibody
(31%) migrates in the 7-8S position.
Figure 7 is microtiter plate prepared according
to this invention.
Figure 8 is an enlarged cross-section of one well
of the plate of Figure 7 showing the active layer in
greatly magnified thickness.
Best Mode for Carryi~ Out the Invention
INTRODUCTION
Immune complexes (ICs)l unquestionably play a
pathogenic role in many infectious, autoimmune and
neoplastic diseases (reviewed in 1); therefore numerous
antigen-specific and antigen-nonspecific assays have been
developed to detect such complexes (reviewed in 1).
30 Antigen-nonspecific techniques have the wider applicability
because in so many immunologic diseases ~he etiologic
agents are unknown. These antigen-nonspecific ~ethods
measure certain physicochemical characteristics of
antigen-antibody complexes as well as their interactions
35 with cellular Fc and C3 receptors and with certain serum
-- --
ll~S142
12
components such as rheumatoid factors (RF), complement
~c~ and conglutinin. Although the use of these
techniques in human and animal diseases has provided
much useful information in regard to immunopathology,
5 prognosis and follow-up of immunologic disorders, several
drawbacks limit their usefulness. First, some of the
purification procedures are technically difficult or
overly time-consuming. Second, and more importantly,
many of these assays detect interfering non-IC substances
10 which may lead to false positive or negati~e results.
We present here a simple, sensitive and reproducible
solid phase radioimmunoassay for the detection of C-fixing
ICs. Our strategy was to interact matrix-fixed F(ab')2
i anti-C3 with C-fixing ICs after which the bound IG was
15 detected with radiolabeled anti-Ig antibodies. By using
this procedure C-fixing ICs were detected in humans with
a variety of diseases, mice with genetic autoimmune disorders
and rabbits with serum sickness. When the IC-positive
human sera were tested further to characterize the anti-C3
20 binding ICs, some sera had heavy (>19S) and intermediate
(8-19S) size complexes; other samples contained
predominantly light (7-8S) anti-C3 reactive materials that
seemed to be IgG associated with small C3 fragments.
MATERIALS AND MET~ODS
Monomeric a~d Aggregated Gamma Globulins Monomeric
IgG was purified from human, mouse and rabbit Cohn Fraction
II (Miles Laboratories, Inc., Elkhart, Illinois; Hyland
Laboratories, Costa Mesa, California) by DEAE chromatography
and rendered free of aggregates by ultracentrifugation at
30 150,000 g for 90 min. Aggregated gamma globulins (AGG)
were prepared from the respective Cohn Fraction II by heating
a solution in phosphate buffered saline (PBS) at 63C for
30 min. After centrifugation at 3,000 rpm for 15 min. to
remove large insoluble aggregates, the preparations were
35 stored in small aliquots at -70C and used throughout the
study.
ll~S142
13
Antisera: Hyperimmune goat anti-human C3 serum was
a generous gift from Dr. H.J. Muller-Eberhard (Scripps
Clinic and Research ~oundation). This antis~ru~was
monospecific as shown by double immunodiffusion
5 against human serum and isolated human C3. The antiserum
reacted with native human C3 as well as its fragments C3b, -
C3c and C3d and was devoid of any activity against human
immunoglobulins. All other antisera used in these studies
were produced in goats and purchas~d from Cappel Laboratories
10 (Cochranville, PA). Goat anti-mouse C3 and goat anti-rabbit
C3 each formed a single line with the respective sera
during double immunodiffusion. F(ab')2 fragments of
anti-human, anti-mouse and anti-rabbit C3 were prepared by
a modification of the method of Nisonoff et al. (2).
15 Brie~ly, IgG fractions of these antisera were submitted
to pepsin digestion in 0.1 M acetate buffer, pH 4.1, with
an enzyme to substrate ratio of 2:100 (w/w) for 18 hours
at 37C. After neutralization with 1.0 M Tris, F(ab')2
was separated from the other fragments by gel filtration
20on Sephadex G-150 with borate buffered saline (BBS) as the
eluting buffer. The preparations were free of intact IgG
and Fc fragments as shown by polyacrylamide gel
electrophoresis and by double immunodiffusion.
Radioiodination: IgG fractions of goat anti-human,
25 anti-mouse or anti-rabbit IgG, F(ab)2 anti-human C3, human
IgG, bovine serum albumin (BSA) and anti-BSA were labeled
with 125I or 131I as described by McConahey and Dixon (3)
to a specific activity of 0.3-0.5 ~Ci/~g of protein.
Radiolabeled yoat IgG anti-human and anti-mouse IgG were
30 diluted in 1~ human serum albumin (HSA)-BBS or 1% BSA-BBS,
pH 7.5, respectively, and radiolabeled anti-rabbit IgG
was diluted in BBS without carrier protein. All the
radiolabeled proteins were stored at 100 ~g/ml.
1156142
14
Test Sera: Sera from heàlthy individuals (n=42) and
from patients with SL~ (n=24), rheumatoid arthritis (n=22),
Sjogrenls (n=22) and Reiter's (n=25) syndromes, infectious
endocarditis (n~20), acute glomerulonephritis (n=20)
5 and malignancies (n=51) were analyzed. Aliquots of these
sera were stored at -70C and thawed once. Plasmas from
some of these patients were obtained by collecting the
blood in 0.01 M EDTA (Ethylenediamine tetra-acetic acid;
final concentration). Serum samples of immunologically
10 normal mice (C3H~St, BALB/c) and SLE-prone mice (r~RL/l,
BXSB, NZBxW) were also tested. The number, age and sex
of the mice examined are given in the Results section.
The immunopathologic characteristics of the SLE-prone
r mice have been detailed (4). Sera from 6 normal rabbits
15 and 5 rabbits with chronic serum sickness resulting from
injections of BSA were kindly provided by Drs. C.B. Wilson
and T.J. Neale (Scripps Clinic and Research Foundation).
The sera from the rabbits with serum sickness were drawn
before and iO, 30 and 60 min. after the daily injection
20 of BSA (5)-
~uantitation of IgG and C3: IgG in serum samples and
gradient fractions and C3 in the normal human serum used
as C source were quantitated by radial immunodiffusion (6).
Detection of ICs: The solid phase anti-C3
25 radioimmunoassa~ for circulating ICs was performed as
follows: ~ells of U-shaped flexible microtiter plates
(Dynatech Laboratories, Inc., Alexandria, VA, Catal. No.
1-220-24) were coated with 100 ~1 of F(ab')2 anti-C3 in
BBS (1 mg/ml), pH 7.5, and incubated at 4C overnight.
30 The material was then aspirated and 200 ~1 of 1~ HSA-BSS,
p~I 7.5, was added to each well to cover unreacted sites.
After standing for 1 hr. at room temperature, the wells
were washed once with BBS, pH 7.5. To these wells we
added 100 ~1 of serum samples diluted 1/21 in 1% HSA-BBS-
35 0.01 M ED~A and standards containing various amounts of
AGG (3 to 800 ~g AGG/ml~ in normal human serum (C source~
11~6142
The various amounts of AGG in fresh ser~m were incubated
for 15 mun. at 37C before being diluted to 1:~1 wi h the
diluent buffer. For murine sera, BSA was used instead of
HSA in the diluent buffer whereas ~or rabbit sera the
5 carrier protein was omitted. After incubation for 2 hours
at 37C, the wells were washed 5 times with BBS, and
100 ~1 (10 ~g) of radioiodinated goat IgG fraction of
anti-IgG (human, mouse or rabbit) were added. After
further incubation (2 hrs., 37C), the plates were again
10 washed 5 times with BBS and dried, and then each well
was cut free and counted. The uptake in each well was
referred to the standard curve and results were expressed
as ~g AGG equivalent per milliliter of serum~ For
comparison levels of IC-like materials in test sera were
15 also determined by the Raji (7) and the conglutinin (8)
assays as described.
Titration and ~bsorption of RF: Titers of IgM RF
in serum samples were determined by the latex
agglutination test (9). To determine whether IgM RF
20 could cause false positive results in the assay and
whether C-fixing ICs detected in sera of patients with
rheumatoid arthritis or Sjogren's syndrome and in sera of
MRL/l muce contained IgM RF, selected sera were absorbed
on IgG-Sepharos~. 10 ~1 of serum were incubated with 200
25 ~1 of human or mouse IgG-Sepharose.containing 3 mg of the
respective IgG per milliliter of packed gel. As a control,
portions of the same sera were absorbed on BSA-Sepharose*
After incubation for 18 hrs. at 4C, the Sepharose was
removed by centrifugation, and the supernatants were
30 tested for ICs with the anti-C3 assay and for IgM RF
activity by latex agglutination.
Sucrose Density Gradient (SDG) Fractionation: To
determine the size of the materials that were detected
in the anti-C3 assay we sampled sera from 6 healthy donors
* Trademar~
~3
1 1~6142
16
and 12 IC-positive patients (one with SLE and
proliferative glomerulonephritis; two with SLE but no clear
renal involvement, two with Reiter's syndrome, two with
Sjogren's syndrome, one with acute glomerulonephritis,
5 one with seropositive and one with seronegative
rheumatoid arthritis, one with Felty~s syndrome and one
with chronic granulocytic leukemia in blastic crisis).
Sera were diluted 1:2 in PBS, and 300 ~1 portions were
placed on the top of 4.5 ml tubes containing 10-40%
10 sucrose gradients in PBS. Sera were centrifuged at
190,000 g for 15 hrs. using a SW60.1 Beckman rotor.
Fifteen fractions were collected from the bottom of the
tube and analyzed for levels of IgG by radial immunodiffusion
f and for positivity in the anti-C3 and conglutinin
15 radioimmunoassays. Because the IgG content of fractions
from the different sera varied after SDG centrifugation,
binding activity in the anti-C3 assay was expressed as ng
of radiolabeled anti-human IgG bound per mg of IgGxml 1
in a given fraction. To verify that IgG bound to
20 anti-C3 during the solid phase assay was indeed associated
wi~h C3 fragments, precipi~ation and inhibition
experiments were performed. For precipitation, test
fractions were treated with anti-C3 in antibody excess
(2 hrs., 37C), and supernatants obtained after
25 centrifugation (3,000 rpm, 20 min.) were retested in
the solid phase anti-C3 assay; binding activity was
compared to that of unmanipulated fractions. For
inhibition, test fractions were incubated (30 min. at
37C and overnight at 4C) with isolated conglutinin
30 (205 ~g/ml final concentration), and the inhibition of
binding was assessed by the solid phase anti-C3 method~
Solubilization of-ICs: To evaluate whether C can
_
solubilize IC and free antibody from antigen, we reacted
affinity purified, 131I-rabbit IgG anti-BSA with
1 1~6142
~ -BSA at equivalence for 30 min. at 37C and
overnight at 4C. Precipitates containing 15 ~g of
I-IgG anti-BSA and 2.2 ~g of I BSA were mixed
with 0.5 ml fresh normal human serum as the solubilizing
5 agent at 37C for 1 hr. (10). ~he serum was then
subjected to SDG analysis as described abo~e. Fractions
corresponding to the 7S peak were pooled and reacted
with goat anti-human C3 serum, anti-human transferrin
or anti-sSA all in antibody excess, or with normal
10 goat serum. Precipitates obtained after centrifugation
were washed twice in cold P~S and counted for radioactivity.
Percentages and moles of antigen and antibody present in
each fraction of the SDG ultracentrifuged serum and in
immunoprecipitates of the fractions around the 7S position
15 were calculated. Corrections for 131I isotype decay
were carried out when indicated.
RESULTS
SOLID PHASE ANTI-C3 RADIOIMMUNOASSAY:
Bindinq of F(ab'32 Anti-C3 to Microtiter Plates: The
... . _ .
20 amount of F(ab')2 anti-human C3 bound to microtiter
plates was measured at different concentrations (0.25
to 8 mg/ml), acidities (pE 6.5 to 9.5), intervals (2 to
18 hours) and temperatures (4-20-37C) of incubation. At
neutral pH and after overnight-incubation at 4C, the
25 amount of bound anti-C3 increased with increasing
concentrations as expected. The binding was better at pH
7.5 than at lower or higher acidities, and appro~imately
the same after incubation at 37C for 2 hours, room
te~perature for 5 hours or 4C for 16 hours. The
30 concentration of 1.0 mg/ml was chosen since higher
concentrations did not improve the assay's sensitivity nor
the slope of the standard curve with serum containing AGG.
For practical reasons incubation lasted overnight at 4C
to coat the plates with anti~C3.
1~614~
18
Binding of AGG to Anti-C3 Coated Plates: As shown
in Figure l, the binding of AGG to the anti-C3 coated
wells depended on activation of C and binding was
proportional to the amount of AGG added to normal serum.
5 A sensitivity of 6 ~g AGG/ml serum was consistently
obtained. Data not shown indicated that, at a
concentration of 400 ~g 125I labeled AGG/ml serum,
approximately llO ~g precipitated with anti-C3 serum in
antibody excess and that at concentrations about 800 ~g of
lO AGG/ml of serum the amount of AGG precipitated by anti-C3
remained constant t 200 ~g). These results were obtained
with the normal human serum used throughout the study as
tlle source of C for the standard curve and containing
1.45 mg/ml of C3. Apparently, no additional C was available
15 for binding above 800 ~g AGG/ml serum, which we therefore
adopted as the upper point of the standard curve.
Plates coated with F(ab')2 anti-human C3 and stored
at 4C consistently produced standard curves of binding
activity for at least three months, although for the
20 present study plates were prepared just the day before the
assay. ~he same F(ab')2 anti-C3 could be used at least
three times for coating new plates with no apparent loss
of activity.
Reproducibility of the Assay and Effect of Freezing
25 and ~hawing, EDTA, ~lonomeric Ig and other Substances: No
significant effect on IC le~els was observed after 5 cycles
of freezing (-70C) and thawing (22C) of three serum
samples containing 200 ~g AGG/ml. Similarly, freezing a~d
thawing of sera from five normal volunteers did not convert
- 30 the test results from negative to positive. When mul-tiple
samples (6-10) of 3 positive sera were tested within the
same day, the variability of results did not exceed 5%.
I~en 5 IC-positive human sera were tested consecutively for
3 days, using the same anti-C3 coated plates and radiolabeled
35 anti-IgG antibody, the variability was 9.5%. Adding ED~A
to a serum containing AGG and previously interacted with C
not only did not prevent the binding of AGG -to the anti-C3
11~61d52
19
but actually increased the binding of AGG + C 10~ above
that in EDTA-free samples. High levels of monomeric
IgG seemed to interfere with the results, ~ut the
difference in binding between monomeric IgG and AGG
5 in serum was more than 2 log units. Thus, 10 mg of
isolated human mo~o~ieric IgG added together with 0.01 M
EDTA to a sample of normal human serum containing 12.5
mg/ml of endogenous IgG yielded upta~e values in the
anti-C3 assay that resembled uptake by the same normal
10 serum to which 50 ~g AGG was added.
Removal of IgM RF activity from 5 IC-positive sera
from rheumatoid patients and 5 IC-positive sera of MRL/l
mice by usin~ Sepharose-IgG had no effect on activity
in the anti-C3 assay. Nor did heparin (200 ~g/ml),
l 15 lipopolysaccharide (LPS, 2 ~g/ml) or DNA (2-50 ~g/ml)
convert normal humanisera to positive.
DETECTION OF ICs IN SERUM: Forty-two human serum
. ~ .. _ .. . . . ..
samples obtained from healthy volunteers were analyzed
with the anti-C3 assay and the mean value + SE~ was
20 9.9 + 2.5 ~g AGG eq/ml. In comparison, high levels (above
2 SD of normals) and incidences of ICs were found in sera
of patients with SLEb rheumatoid arthritis, Sjogren's and
Reiter's syndromes, infectious endocarditis and acute
glomerulonephritis.
* * * *
1 4 ~
TABLE III
I~CIDENCE AND LEVELS OF ICs IN PA~IOLOGIC
SERA TESTED BY THE ANTI-C3 ASSAY
_ _ . , . _ .. .. _ . _ . . .
~g AGG/ml
CASES (mean + SEM~ % POSITIVE*
. . .
Systemic Lupus
Erythematosus 24 168.2 + 35.0 75
Rheumatoid Arthritis 22 234.7 + 57.8 73
10 Sjagren's Syndrome 22 356.3 + 113.6 82
Reiter's Syndrome 25 406.2 + 93.9 56
Infectious Endocarditis 20 183.1 + 38.3 45
Acute glomerulonephritis 20 432.8 + 164.3 30
Malignancies:
I5 Breast Cancer 3869.2 + 11.8 21
Lung Cancer 1366.2 + 16.6 31
Normal Controls 429.9 + 2.5
* Sera with activity higher than 2SD above the mean of
20 normal controls ~9.9 + 16.4) were considered positive. The
mean values of ICs were derived from sera recorded as
positive. .
* * * * *
The incidence of positivity in the anti-C3 assay generally
25 lay intermediately between that in the other two assays
(~aji and conglutinin assays), with the Raji cell assay
giving more positive results.
* * * * *
11~614~
E~
H
~ ~ Z
E~
~; V Z __ ~_ _
I O ~ ~ ~ ~ U~ ~
E~ H O ~) ~ I` 1` ~r 1-1
'¢p:j Z R~
u~ ~ o o r~
V
m o 1 ~ ~ co n ~9 co 1`
E~ C ) H K -- --
: Z El
H ~:1 0 0 Z ~ N 1S~ D C~
,¢ Z: t.) 1¢
æ
m H
1~ H ~ n o û~ ~1 o
' m ~1 ~ o 'I ~
U~ IcS H
E'l E~
3 U~ U~ _ ~ _ _
~ U~ O ~ ~ o ~D ~ CO
H ,Y Z ~ 00 co ~ 1` oo t` ~C
K S H -- -- ~
m~ o ~:~ ~ ~ ~ ,t ,1 ~ ~ ~ ~
Z
t.1 H lc
Z ~ ~ ~ ~ ~ ~ ~ t)
~; U~ O C~ 1 ~ ~ ~ h
O U~H -- `~
o u~ æ ,
o o~~ 'I ~ ~ ~ a~
O Z
o ~ ~ ~ O. U~
H Z g ~ `I
~:1 H
Z l¢ Ul rl
H Z Ul
H ra
H H rl ~ a) C)
E I ~I tl ~ O h
O O
Z ~ O ~
O
k rl
oa) o o u~
i ~ h ~U
h
~:1 ~ h
= O ~ Z
H ~ 0 r) 0
U~ K H ~
O Lt ) O , .
. ~
11~6142
In general, sera positive in the anti-C3 assay were also
positive in the Ra~i cell assay. Seventy percent of all
sera tested by both assays gave concordant positive or
negative results. No correlation was found between titers
5 of RF and levels of IC-like materials as detected by the
anti-C3 assay, with samples from seropositive patients
having values and incidences (42-500 ~g/ml; 75% positive)
as high as of those from seronegative patients ~44-470~g
AHG/ml; 67% positive). In all three murine strains
10 (NZBxW, BXSB, MRL/l) known to develop spontaneous autoimmune
syndromes, in contrast to normal strains, IC-like
materials were apparent with the anti-C3 assay (Figure Z),
and MRL/l mice had the highest values. Sera of rabbits
r with chronic serum sickness also contained high levels
15 (> 800 ~g AGG/ml) of circulating IC-like materials that
bound to anti-rabbit C3. Unexpectantly, no clear
differences were observed in incidences and levels of IC
binding among serum samples collected before and after the
daily injection of the antigen BSA (data not sho~).
-CHARACTERIZATION OF IC-LIKE MATERIALS IN HUMAN SERA
DETECTED BY THE ANTI-C3 ASSAY: Twelve human serum samples
showing strong positivity in the anti-C3 assay were
subjected to SDG ultracentrifugation. The fractions
obtained were individually analyzed for activity in the
25 anti-C3 and conglutinin assays and for IgG levels by
radial immunodiffusion. The results obtained with these
sera are depicted in Figures 3 and 4. As one sees,
basically two types of patterns emerged. Some of the
pathologic sera exhibited reactivity with anti-C3 and
30 conglutinin in fractions heavier than the 7S marker whereas
others exhibited predominant reactivity in fractions around
the 7-8S position.
When a serum from a patient with Felty's syndrome
containing heavy anti-C3 reacting materials was ultracentrifuged
35 again in similar gradients made in acidic buffer (Glycine-
HCl 0.1M, pH 2.8), the reactivity with anti-3 shifted to
1 4 2
the upper fractions of the gradient (Figure 5). Although
only a small number of samples was analyzed, it is notable
that sera of patients with more generalized diseases like
lupus proliferative nephritis, Felty's syndrome,
5 Sjogren's syndrome and acute poststreptococcal
glomerulonephritis (Figures 3 and 4, patterns a,b,c and
d) contained heavy reactive IgG, i.e., 8 to l9S and
occasionally beyond l9S. In contrast, related to more
localized chronic inflammatory conditions like SLE
10 without renal involvement, rheumatoid arthritis without
extraarticular manifestations and Reiter's syndrome,
the peak of reactivity consistently occupied the 7-8S
position (Figures 3 and 4, patterns e,f,g,h). Since
material from the 7-8S region that reacted with anti-C3
15 and conglutinin could represent interference by
monomeric IgG, all values for binding were compared to the
respective IgG concentrations. Even when calculated as
binding per mg of IgG X ml 1, binding was still much
higher in the fraction derived from these pathologic sera
(46-90 ng 5I anti-IgG per mg IgG X ml ) than in those
derived from the sera of 6 normal humans (3.4 ~ 0.54 ng
5I anti-IgG per mg IgG X ml ).
To confirm that complexes of IgG and C3 were
responsible for reactivity at the 7-8S position, we then
25 pooled corresponding fractions separately from each of
three-sera that exhibited such activity and treated the
pools with goat F(ab')2 anti-human C3, conglutinin or
simply buffer (HSA-BBS). After centrifugation (3,000
rpm, 20 min.) to remove immunoprecipitates, the pools
30 were resubmitted to analysis by the anti-C3 assay. As
shown in Table V, anti-Ig binding decreased significantly
in all three samples after incubation with anti-C3 or
conglutinin as compared to controls treated with HSA-BBS.
1 4 2
24
* * * * *
TABLE V
ANTI-C3 BINDI~G ACTIVI~Y IN 7S FRACTIONS OF
CERTAIN SE~A FOLLOWING TREAT~NT WITH BUFFER,
FLUID PHASE ANTI-C3 OR CONGLUTININ
S
.......... . .. ...... . .. . . . . .
TREATED WITH
~UFFER ANTI-C3 CONGLUTININ
Systemic Lupus Erythematosus 108* 0~100~)** 14(87%)
10 Rheumatoid Arthritis 83 0(100%) 13(84~)
SjOgrenlS Syndrome 33 14(58%) 12(64~)
* Values represent ng of 125I-IgG anti-human Ig bound to
F(ab')2 anti-C3 coated plate.
15 ** Numbers in parenthesis express % reduction in comparison
with control (buffer values).
* * * * *
Immunoconglutinins are classically described as
autoantibodies directed against C3 and its fragments (11).
20 Therefore, we investigated whether the anti-C3 reacting
light material found in some sera represented IgG
immunoconglutinin or immunoconglutinin-C3 fragment complex.
~or this purpose, 7-8S fractions from three such sera were
incubated (30 min., 4C) with C3-Sepharose or BSA-Sepharose
25 (control) and checked again in the anti-C3 assay. No
removal of the anti-C3 activity in these light fractions
was observed after such treatments, strongly suggesting
that the reactivity was not due to IgG immunoconglutinin.
Finally, fractions corresponding to 8-19S and 7-8S
30 position yeilded by SDG centrifugation-of 4 positive sera
were mixed with 2.5~ polyethylene glycol (12). In the
1 1561d~2
resul-ting supernatants more than 95% of the heavy
fractions' binding activity was gone but, as expected,
the 7-8S fractions lost less than 5% of this activity.
.. .... .. ....... ... .....
IN VITRO SOLUBILI~ATION OF ICs: As indicated above,
5 the 7-8S materials detected in some samples of pathologic
sera by the anti-C3 and conglutinin assays were apparently
composed of IgG associated with C3 fragments. Since
solubilization of antigen bound antibody molecules by the
C system (10) could produce such complexes, we directly
10 examined their generation after solubilizing BSA-anti-BSA
complexes with human C. Radiolabeled antigen-antibody
complexes were formed at equivalence, and incubaced with
fresh human serum at 37~C for 60 min. and then at 4C for
2 hours. This treatment solubilized 68.5~ of the
15 precipitated complexes. When the serum containing these
solubilized complexes was subjected to ultracentrifugation
in SDG, 31.1~ of the antibody molecules were found in the
7S region (mean of two experiments); 10.8~ of the antigen
(BSA) was located in the 4-5S region, and less than 1%
20 appeared in the 7S region (~igure 6). The remaining
portions of antigen and antibody were distributed in the
heavier fractions of the gradient. Consequently, upon
solubilization by C a large portion of cornplexed antibody
molecules are freed of antigen.
In subsequent experiments, fractions corresponding
to the 7S peak were pooled and portions submitted to
treatment with goat anti-C3, anti-transferrin or normal
goat serum. mhe results clearly demonstrated that the
antibody was selectively precipitated only by anti-C3
30 serum (41~), not by the controls, anti-transferrin (2.8%),
normal goat serum (3%) or by anti-BSA (7%). These
experiments ~urther confirm that a considerable amount of
the solu~ilized antibody was not bound to the antigen
but carried C3 determinants.
DISCUSSION
Despite the many antigen nonspecific assays currently
available to detect ICs or like materials in biologic
fluids, criticism has been voiced for most ~13), and the
5 drawbacks of these assays are considerable (1?. However,
because ICs appear to be involved in the immunopathology of
many diseases and exert profound effects on humoral and
cellular immune responses ~1,14), the development of new,
more specific techniques or IC-detection is, in our opinion,
justifiable. We have described here the use of F(ab')2
anti-C3 antibody in a solid phase radioimmunoassay for the
detection and partial characterization of C-fixing ICs in
human and animal sera. The reagents used in this highly
sensitive anti-C3 assay are simple to prepare or commercially
- 15 available. Results not shown demonstrated that the principle
of enzyme immunoassay can be easily applied to preclude the
use of radioisotopes as well as assay by binding to
staphylococcal protein A. The assay, like all other
non-antigen specific IC assays, does not distinguish between
20 nonspecifically aggregated Ig and antigen-complexed Ig.
However, we believe that most nonspecific aggregates are
formed in vitro during blood clotting and serum preparation.
Therefore, the addition of FDTA during bloo~ collection
would prevent C-fixation to such aggregates wi'thout
25 affecting the detection of ICs formed ln vivo and containing
C in the anti-C3 assay. Moreover, the assay is not
affected by non-IC substances such as polyanions,
endotoxins, nucleic acid, R~s and anti-lymphocyte
antibodies that are known to interfere with many of the
30 existing IC detection techniques (1). Our use of
F(ab')2 anti-C3, instead of whole IgG anti-C3 outlined in
a soluble anti-C3 assay by o'~hers (15~, eliminates false
positivity due to the presence of RFs, unless the RFs
represent part of C-fixing ICs. The additional
1 4 2
advantage over the Clq assays is that the anti-C3 solid
phase assay detects ICs that fix C activated not only
via the classical pathway but also via the alternative
pathway~ It is known that IgM and IgG antibodies
5 activate the classical C pathway (16,17) whereas those
of the IgA and IgE class activate the alternative C
pathway (18,19~. Unlike the Clq deviation and Clq and
cause false positive results, the use of anti-Ig antibod~
as the final reagent in our solid phase anti-C3 assay
10 permits detection of Ig-containing substances only.
Although we have some data on the size of ICs
detected by the anti-C3 assay (see below), the required
ratio of antigen to antibody and amount of C fixation
for an IC to become detectable are unknown. There is
15 little agreement in the literature about the molecular
composition or minimum size of IgG-type ICs that can
activate C; some investigators advocate the doublet
Ab2Agl (20), others claim no less than a tetramer of antibody
(21), whereas still others have shown activation of the
20 classical pathway of C by a single IgG molecule binding
a monovalent hapten t22). In general, it would appear
that fixation of C by soluble ICs made with IgG antibodies
requires a lattice structure containing more than two or
three antibody molecules ~23).
When we applied the anti-C3 assay to the detection
of C-fixing ICs in sera, we found a high incidence of such
ICs in patients with SLE, rheumatoid arthritis, Felty's,
Reiter's and Sjogren's syndromes and acute glomerulonephritis
and to a lesser degree in neoplasia. Previously, IC-like
30 materials were detected in sera of SIE patients by a
variety of procedures (24,25), and IC determinations were
often found useful in following the course of disease
in patients with SLE ( 26,27). Although glomerular elution
studies indicated a role for DNA-anti-DNA complexes in the
35 pathogenesis of SIE nephritis (28), direct evidence of
1156142
28
circulating DNA-anti-DNA complexes has been controversial
(29,30). However, mice with spontaneous SLE-like syndromes
clearly form a variety of complexes including DNA-anti-DNA,
RF-Ig and retroviral gp70-anti-gp70 (31~. According to
5 many assays, ICs are present in persons with Felty's (32),
Reiter's (33) and Sjogren's (34) syndromes, but again the
nature of the materials detected has not yet been defined.
ICs have also been found in patients with infectious
endocarditis (~5) and the value of IC-determinations in
10 the prognosis and management of such patients is clear
(35,36). C-fixing ICs or like materials have also been
observed in the course of leukemia (37) and a variety of
other malignancies; in fact a role has been proposed for
;~ such complexes in tumor immunity, and IC measurement may
15 have substantial prognostic value in such patients
(reviewed in 1,38). In malignancies the nature of complexes
detected often remains undefined, despite some evidence
that such patients sometimes have specific tumor
antigen-antibody complexes in their sera and kidney
20 deposits lreviewed in 1,38). We (39) and o~ers (~0 42)
are using a variety of physicochemical and immunochemical
procedures as well as absorption to substances that bind
ICs to characterize antigens in the ICs detected by the
various techniques. Anti-C3 bound to insoluble matrices
25 should be helpful in such attempts.
Sera in which ICs or like materials have been detected
are being evaluated in several laboratories to deter~ine
the size of the reactive matexials. In most of these
studies the materials that react with substrates which
30 detect ICs have been found to be mainly heavy (>19S) or
intermediate (8-19S~ in size in agreement with the concept
that complexes are composed of multiple antigen-antibody
molecules and sometimes C components. ~Iowever, low
molecular weight (7-8S) materials, reactive with Clq and
1 4 2
29
P~aji cells have also been detected in some sera of
patients with SLE (25,43), Lyme arthritis (44~, IgA
deficiency (45~ and other disorders. In using SDG
ultracentrifugation to analyze materials reactive in the
5 anti-C3 assay, we found heavy and/or intermediate size
complexes in some sera. However, in sera from several
patients the F(abl)2 anti-C3 reactivity was primarily,
if not exclusively, confined to the 7-8S fractions.
Not only F(abl)2 anti-C3 but also conglutinin bound the
10 7-8S fractions, and binding was specific since it was
inhibited by anti-C3 or conglutinin in the fluid phase.
Particularly patients with more generalized disease were
those with intermediate and heavy complexes in their sera,
whereas patients with less severe disease had
r. 15 predominantly the 7-8S component. This 7-8S material,
which contained IgG that bound to F(ab')2 anti-C3 and
conglutinin, may well represent the antibody moiety of
ICs along with some small C3 fragments generated after
C-induced IC solubilization. Solubilization of immune
20 precipitates mediated by C in fresh serum has been
described by Miller and Nussenzweig (10). The alternative
C pathway mediates the solubilization, since factor B,
factor D, properdin, C3 and Mg~+ must be present, and ~he
apparent sequence is insertion of C3b into the immune
25 precipitates to form properdin and factor B dependent
convertase (46). ~he classical C pathway alone is
incapa~le of solubilizing ICs (47), possibly because the
assembled classical C3 convertase (C~), in contrast to
the properdin- and factor B-dependent C3 convertase
30 (C3b,PrB), has very short half-life (48,49) and therefore
binds too few C3b molecules to disrupt the antigen-antibo2y
lattice. Nevertheless, activation of the classical C
pathway greatly enhances the solubilizing activity of the
alternative pathway (47). Solubilization does not seem
35 to result from rupture of all the antigen-antibody
bonds considering that almost all the antigen remains
ammonium s~h]fate-precipitable (10) and is, therefore,
comple~ed after solubilization. In the present
experiments, double labeling of antigen and antibody
left little doubt that most of the antigen actually
remained complexed; however, many antibody molecules
~ 30%) were freed during the C-induced solubilization
process. The presence of free antibody molecules is
possible due to the fact that certain antigens, as tha-t
used here, at equivalence bind several antibody molecules.
Whether solubilization of ICs and generation of free
10 antibody molecules takes place ln ViVO remains to be
demonstrated, but our results suggest this is the case.
On the antibody molecules, bound C3 may be degraded by
serum enzymes such as BlH and C3b inactivator (reviewed
in 1) leaving attached small fragments (i.e. C3d) that
15 are the light-weight molecules we found reactive with
anti-C3 and conglutinin. We are currently analyzing these
low molecular weight reactants for binding substances
like Clq, RF and cell receptors and subsequent biologic
activity, if any. Conceivably, their light weight may
23 preclude deposition in tissues and so "IC disease" is
circumvented. Interestingly enough, ~iller et al (50)
observed that IC solubilized from mouse lymphocytes by C
lost their ability to bind to fresh lymphocytes. These
authors also observed that in older NZB~7 mice with
25 advanced disease there was a reduced C-mediated IC
solubilization (51). Others (52) suggested that the
alternative C pathway may be important in resolving ICs
deposited in the kidney during experimental acute serum
nephritis. Our indirect fin~ings suggest that patients
30 with generalized processes may become deficient in C-induced
solubilization of ICs. The cause could be rapid consumption
of C resulting from a high rate of IC formation, inability
to assem'ole the C3b,P,B convertase or high affinity of the
antigen-antibody bonds.
~Industrial Application
The inventive concept which has been described
has application in immunological diagnostics and
treatment of various disorders and diseases resulting
from or associated witll the presence of immune complexes.
While the invention can be carried out in many ways,
the examples given in the preceding discussion are presently
the best mode of carrying out the invention. The foregoing
must not be considered as limiting in any way, except
10 as to the principle of the assay.
It is convenient to prepare, either for future use iTI
the same laboratory or for sale and distribution to
laboratories, hospitals, pre-coated substrates for use
in carrying out the assay. A very simplified substrate
15 is shown, merely as exemplary, in Figures 7 and 8. The
substrate 100 comprises a sheet of suitable material
with one or more areas which are suitable for being coated
with a reagent specific for C3. Eight such areas, in
this instance, U-shaped wells, are indicated at 101 through
~0 108. It is to be clearly understood, however, that no
criticality is attached either to the size or shape of the
substrate or the areas, a flat substrate with U-shaped
wells being merely a simple and expedient shape which is
readily available. The substrate may, of course, comprise
25 as many areas to be coated as may be desired.
Figure 8 depicts, for illustrative purposes only,
one area which has been prepared for use in the C3 assay
of this invention. Well 101, for example~ is coated, as
described hereinbefore or in any other convenient manner,
30 with F(ab')2 anti-C3 in a layer illustratively indicated
at 111. It will be immediately apparen~ that the thickness
of the substrate 110 is infinitely thicker than the F(ab')2
anti-C3 layer, rather in the dimension or ratio in the figure.
In the exemplary embodiment, the substrate is a polystyrene
1~56142
flexible micro-titer pla~e ol the type which is conventionally
used in agglutin~tion and com~arable tests and ~lhicn is
readily available on the marke-t. The size, sha e and
com~osition oS the substra~e is, ho~iever, absolutel~
5 inmaterial so long as no a~verse reac~ion occurs and the
substrate is in a configuration ~hich is sui-table for use
in ~ radioihmur.oassay) or ELISA, or other, detection
or measuring technigues.
It is ~i~h~n the inteni and sco?e OL- ~his inven-~ion
10 to provide, ei~er fo in-laboratory use oi for sale
to others, an assay ki-t which, conveniently will incluae
a substrate having at least one measurable area coated
wi,h Y(~b')2 anti-C3. (By ~easura~12 area, we mean an
area t~7hioh can be subjected to ~easurement or deteciion
15 ky P~I~, E~lS~ or o-ther counting, observation or measuring
technique.) In addition, the ~it may irclude radio- or
enzyme- lin~ed IgG anti-human IgG, or generally equivalent,
de~ectable reagent. (By detectable reagent, we mean a
reagent which ~ill couple ~ith F(ab')2 anti-C3 and which
20 can be detected qualitatively a-t least and preferably
quantitatively. RIA and ELISA reagents are typical of such
detectabla reag2nts.) A preferred detectable reagen~
for the kit e~odiment of this invention is radiolabeled
staphyloccal protein A, ~-7hich is available in sta~le,
25: consisten~ qualily LOrm and ~Jhich is easily detected as a ; `-
r..eans ol measuring, by RIA ~.ethods, the quantity o~ C3- ~ Ç
~ound to the co_ted s~strate. The ~:it may also includ~
othe.r reagents, iL desired, to facilitate the carrying out
of the assay of ~liS invention.
The in-~ention resides in the disco-~er~ and in the
seleotion of the co~ting for the substrate, the resulting
substrate, and the overall assay, and not in particular
r~agents, materials, or conditions and the claims which are
appended hereto are to be so understood.
i 1 4 2
REFERENCES
1. Theofilopoulos, A.N. and F.J. Dixon. 1979. The
biology and detection of immune complexes. Aav.
Immunol. 28, 89-220.
2. Nisonoff, A., F.C. Wissler, L.N. Lipman and D.L.
Woernley. 1960. Separation of univalent fragments
from the bivalent rabbit antibody molecule by
reduction of disulfide bonds. Arch. Biochem. Biophys.
89:230~244.
3. McConahey, P.J. and F.J. Dixon. 1966. A method of
trace iodination of proteins for immunologic
studies. Int. Arch. Allery 29-185-189.
4. Andrews, B.S., R.A. Eisenberg, A.N. Theofilopoulos,
S. Izui, C.B. Wilson, P.J. McConahey, E.D. Murphy,
; 15 J.B. Roths and F.J. Dixon. 1978. Spontaneous
murine lupus-like syndromes. Clinical and
immunopathological manifestations in several
strains. J. Exp. Med. 148:1198-1215.
5. Wilson, C.B. and F.J. Dixon. 1971. Quantitation
of acute and chronic serum sickness in the rabbit.
J. Exp. Med. 134:7S-18S.
6. Mancini, G., A.O. Carbonara and J~F. ~eremans.
1965. Immunochemical quantitation of antigens
by single radial immunodiffusion. Immunochem.
2:235-25~.
7. Theofilopoulos, A.N., C.B. Wilson and F.J. Dixon.
1976. The Raji cell radioimmune assay for
detecting immune complexes in human sera. J. Clin.
Invest 57:169-182.
8. Eisenberg, R.A.~ A.N. Theofilopoulos and F.J. Dixon.
1977. Use of bovine conglutinin for the assay of
immune complexes. J. Immunol. 118:1428-1434.
9. Singer, J.M. and C.M. Plotz. 1956. The latex
fixation test. I. Application to the serologic
diagnosis of rheumatold arthritis. Amer. J. Med.
21:888-892.
1 15~ 1~2
3~1
10. Miller, G.W. and V. Nussenzweiz. 1975. A new
complement function: solubilization of antigen-
antibody aggregates. Proc. Natl. Acad. Sci. USA
72:418-422.
11. Lachman, P.J. 1967. Conglutinin and immunocon-
glutinins. Adv. Immunol. 6:479-527.
12. Creighton, W.D., P.H. Lambert and P.A. Miescher.
1973. Detection of antibodies and soluble antigen-
antibody complexes by precipitation with
polyethylene glycol. J. Immunol. 111:1219-1227.
13. Schur, P.H. 1978. Immune complex assays: the
state of the art. N. Eng. J. Med. 298:161-162.
14. Haakenstad, A.O. and M. r~annik. The biology of
immune complex. In: Autoimmunity. Talal, N.
(ed.). Academic Press, New York, p. 278.
r. 15. Jacobs, R. and M. Reichlin. 1979. A new
quantitative assay for immune complexes. Fed.
Proc. 38:1359. (Abs.).
16. M~ller-Eberhard, H.J. 1975. Complement. Ann.
Rev. ~iochem. 44:697-724.
17. Spiegelberg, H.L. 1974. Biological activities
of immunoglobulins of different classes and
subclasses. Adv. Immunol. 19:259-294.
18. Ishizaka, T., C.M. Sian and K. Ishizaka. 1972.
Complement fixation by aggregated IgE through
alternate pathway. J. Immunol. 108:848-851.
19. Spiegelberg, H.L. and O. Gotze. 1972. Conversion of
C3 proactivator and activation of the alternative
pathway of complement activation by different
classes and subclasses of human immunoglobulins.
Fed. Proc. 31:655 (Abs.).
20. Cohen, S. 1968. The requirement for the association
of two ad~acent rabbit rG-antibody molecules in the
fixation of complement by immune complexes. J.
Immunol. 100:409-413.
21. Hyslop, N.E., R.R. Dourmashkin, N.M. Green and R.R.
Porter. 1970. The fixation of complement and the
,1
1156142
aetivated first component (Cl) of complement by
complexes formed between antibody and divalent
hapten. J. Exp. Med. 131:783-802.
22. Goers, J.W., R.J. Ziceardi, V.N. Sehumaker and
M.M. Glovsky. 1977. ~he mechanism of activation
of the first component of eomplement by a univalent
hapten-IgG antibody complex. J. Immunol. 118:2182-
2191.
23. Mannik, M., W.P. Arend, A.P. Hall and B.C.
Gilliland. 1971. Studies on antigen-antibody
complexes. I. Elimination of soluble complexes
from rabbit circulation. J. Exp. Med. 133 713-739.
24. Zubler, R.H., G. Lange, P.H. Lambert and P.A.
Miescher. 1976. Deteetion of immune complexes
i 15 in unheated sera by a modified 125I-Clq binding
test. Effect of heating on the binding of Clq by
immune eomplexes and application of the test to
systemic lupus erythematosus. J. Immunol. 116:232-
235.
25. Gabriel Jr., A. and V. Agnello. 1977. Detection
of immune eomplexes. The use of radioimmunoassays
with Clq and monoelonal rheumatoid factor. J.
Clin. Invest. 59:990-1001.
26. Levinsky, R.J., J.S. Cameron and J.F. Soothill.
1977. Serum immune eomplexes and disease activity
in lupus nephritis. Laneet 1:564-567.
27. Pussell, B.A., C.M. Loekwood, D.M. Seott, A.J.
Pinehing and D.K. Peters. 1978. Value of
immune-complex assays in diagnosis and
management. Laneet 2:359-363.
28. Koffler, D., V. Agnello and H.G. Kunkel. 1974.
Polynucleotide immune complex in serum and
glomeruli o~ patients with systemic lupus
erythematosus. Amer. J. Pathol. 74:109-124.
29. Izui, S., P.H. Lambert and P.A. Mieseher. 1977.
Failure to deteet eirculating DNA-anti-DNA
1 1.56 1~2
complexes by four radioimmunological methods in
patients with systemic lupus erythematosus.
Clin. Exp. Immunol. 30:384-392.
30. Bruneau, C. and J. Benveniste. 1979. Circulating
DNA-anti-DNA complexes in systemic lupus
erythematosus. Detection and characterization
by ultra-centrifugation. J. Clin. Invest. 64:
191-198.
31. Theofilopoulos, A.N., P.J. McConahey, S. Izui,
R.A. Eisenberg, A.B. Pereira and W.D. Creighton.
1980. A comparative immunologic analysis of
several murine strains with autoimmune manifesta-
tions. Clin. Immunol. Immunopath., in press.
32. Andreis, M., E.R. Hurd, J. LoSpalluto and M. Ziff.
1 15 1978. Comparison of the presence of immune complexes
r in Felty's syndrome and rheumatoid arthritis.
Arthritis Rheum. 21:310-315.
33. Rosenbaum, J.T., A.N. Theofilopoulos, H.O. McDevitt,
A.B. Pereira, D. Carson and A. Calin. 1980.
Presence of circulating immune complexes in
Reiter's syndrome and ankylosing spondylitis.
Lancet, submitted.
34. Lawley, T.J., H.M. Moutsopoulos, S.I. Katz, A.N.
Theofilopoulos, T.M. Chused and M.M. Frank. 1979.
Demonstration of circulating immune complexes in
Sjogren's syndrome. J. Immunol. 123:1382-1387.
35. Bayer, A.S., A.N. Theofilopoulos, R.A. Eisenberg,
F.J. Dixon and L.B. Guze. 1976. Circulating
immune complexes in infective endocarditis. N.
Engl. J. Med. 29S:1500-1505.
36. Bayer, A.S., A.N. Theofilopoulos, F.J. Dixon and
L.B. Guze. 1979. Use of circulating immune
complex levels in the sero-differentiation of
endocarditic and non-endocarditic septicemias.
Amer. J. Med. 66:58-62.
37. Carpentier, N.A., G.T. Lange, D.M. Fiere, G.J.
Fournie, P.H. Lambert and P.A. ~5iescher. 1977.
1 1~6142
Clinical relevance of circulating immune complexes
in human leukemia. J. Clin. Invest. 60:874-884.
38. Theoilopoulos, A.N. and F.J. Dixon. 1979. Immune
complexes associated with neoplasia. In:
Immunodiagnosis of Cancer. Herberman, R. (ed.).
M. Dekker/ Inc., New York.
39. Theofilopoulos, A.N., R.A. Eisenberg and F.J. Dixon.
1978. Isolation of circulating immune complexes
using Raji cells. J. Clin. Invest. 61:1570-1581.
40. Chenais, F., G. Virella, C.C. Patrick and H.H.
Fudenberg. 1977. Isolation of soluble immune
complexes by affinity chromatography using
staphylococcal protein A-sepharose as substrate.
J. Immunol. Meth. 18:183-192.
41. Casali, P. and P.H. Lambert. 1979. Purification
of soluble immune complexes from serum using
polymethylmetacrylate beads coated with conglutinin
or Clq. Clin. Exp. Immunol. 37:295-309.
42. Natali, P.G., L. Walker and M.A. Pellegrino. 1980.
Isolation o~ soluble immune complexes from serum
using protein A-bearing Staphylococcus aureus
bacteria: Separation of the antigen from immune
complex and production of antisera. Clin. Immunol.
and Immunopath. 15:76-87.
25 43. Robinson, ~.F., J.L. Roberts, J. Verrier Jones
and E.J. Lewis. 1979. Circulating immune complex
in patients with lupus and membranous glomerulo-
nephritis. Clin. Immunol. Immunopathol. 14:348-360.
44. Hardin, J.A., L.C. Walker, A.C. Steere, T.C.
Trumble, K.S.K. Tung, R.C. Williams, Jr., S. Ruddy
and S.E. Malawista. 1979. Circulating immune
complexes in Lyme arthritis. J. Clin. Invest.
63:468-477.
45. Kwitko, A.O., P.~. McKenzie, D.J.C. Shearman, A.A.
Gormly and A.J. Woodroffe. 1979. Circulating
immune complexes in IgA deficiency. Clin. Exp.
Immunol. 38:45-51.
1 156142
38
46. Takahashi, M., B.F. Tack and V. Nussenzweig. 1977.
Requirements for the solubilization of immune
aggregates by complement. Assembly of a factor
B-dependent C3-convertase on the immune complexes.
J. Exp. Med. 145:86-100.
47. Takahashi, M., S. Takahashi, V. Brade and V.
Nussenzweig. 1978. Requirements for the
solubilization of immune aggregates by complement.
The role of the classical pathway. J. Clin.
Invest. 62:349-358.
48. Fearon, D.T. and K.F. Austen. 1975. Properdin:
binding to C3b and stabilization of the C3b-
dependent C3 convertase. J. Exp. Med. 142:856-863.
49. Schreiber, R.D., R.G. Medicus, O. Gotze and H.J.
M~ller-Eberhard. 1975. Properdin and nephritic
factor-dependent C3 convertases: requirement of
native C3 for enzyme formation and the function
of bound C3b as properdin receptor. J. Exp. Med.
142:760-772.
20 50. Miller, G.W., P.~l. Saluk and V. Nussenzweig. 1973.
Complement-dependent release of immune complexes
from the lymphocyte membrane. J. Exp. Med. 138:
495-507.
51. Miller, G.W., A.D. Steinberg, I. Green and V.
Nussenzweig. 1975. Complement-dependent alterations
in the handling of immune complexes by NZB/W mice.
J. Immunol. 114:1166-1170.
52. Bartolotti, S.R. and D.Ko Peters. 1978. Delayed
removal of renal bound antigen in decomplemented
rabbits with acute serum sickness. Clin. Exp.
Immunol. 32:199-206.
53. Wilson, M.B. and P.K. Nakane: (1978). Recent
developments in the periodate method of conjugating
horseradish peroxidase (HRPO) to antibodies in
immunofluorescence and related staining techniques.
Edited by W. Knapp, K. Hombar and G. Wick,
11~6:142
39
Elsevier/North ~olland Biomedical Press, pg. 215.
54. Theofilopoulos, A.N., C.B. Wilson and F.J. Dixon
(1976). The Raji cell radioi~mune assay for
detecting immune complexes in human sera. J.
Clin. Invest. 57:169.
55. Jacobs, R. and Reichlin, M. The new Quantitative
Assay for Immune Complexes to be presented at the
Federation of American Societies for Experimental
Biology, March 1, 1980, Abstract, Volume 38, p.
1359, No. 5977.
- 15
