вход по аккаунту


Relative reactivities of rheumatoid factors in serum and cells. evidence for a selective deficiency in serum rheumatoid factor

код для вставкиСкачать
Investigations of rheumatoid factors by a hemolytic plaque forming cell assay of blood lymphocytes with
sensitized sheep cells have suggested that the rheumatoid
factors released from the cells have higher affinities for
human IgG than do the rheumatoid factors measured in
the patient’s serum. This study reaffirms this observation
and provides evidence that the reported differences are not
artifacts of technique. The findings imply that rheumatoid
factors of higher affinity for IgC than those of the serum
are being released into the tissue fluids by lymphocytes
and locally precipitated. Such rheumatoid factors may
never reach the peripheral blood serum in detectable quantity or may do so only infrequently.
Rheumatoid factors (RF) are anti-IgG antibodies that circulate bound in autoimmune complexes.
From the Division of Rheumatology, Department of Clinical
Research, Research Institute of Scripps Clinic, La Jolla, California.
Supported in part by a grant from the Kroc Foundation, and
NIH Grants RR 00833 and AM 07144.
Dick L. Robbins, M.D.: Assistant Professor of Medicine,
Section of Rheumatology, University of California at Davis; Terry L.
Moore, M.D.: Assistant Professor, Department of Medicine, St. Louis
University, St. Louis, Missouri; Dennis A. Carson, M.D.: Associate,
Department of Clinical Research, Scripps Clinic and Research Foundation, La Jolla, California; John H. Vaughan, M.D.: Head, Division
of Clinical Immunology, Scripps Clinic and Research Foundation, La
Jolla, California.
Address reprint requests to Dr. John H. Vaughan, Scripps
Clinic and Research Foundation, 10666 North Torrey Pines Road, La
Jolla CA 92307.
Submitted for publication February 27, 1978; accepted April
10, 1978.
Arthritis and Rheumatism, Vol. 21, No. 7 (September-October 1978)
Although a great number of studies have been carried
out on serum RF, it is generally recognized that serum
R F is not necessarily representative of R F made elsewhere in the tissues. R F made in the lymph nodes, or in
inflamed tissues, must meet and interact with IgG extravascularly before it finds its way into the blood plasma.
Any R F highly avid for IgG may therefore never reach
the blood plasma in detectable amounts, but can be
trapped in the tissues as deposits of immune complexes.
A means to judge this possibility is now available
in the recently reported (1) hemolytic plaque-forming
cell (PFC) assay for RF. Since this assay uses sheep red
cells (SRC) sensitized with rabbit IgG, it detects the
portion of R F that cross-reacts with rabbit IgG. For
clarity and emphasis this cross-reactive R F will subsequently be referred to as RFR. RFR-PFC can be detected in small but significant numbers among the peripheral blood lymphocytes in a minority of patients
with rheumatoid arthritis (RA), principally those with
very active or severe disease (1).
We have reported that the RFR detected in the
PFC assay appears to have higher affinity for human
IgG than for rabbit IgG, since it is inhibited better with
human monomer IgG than with rabbit monomer IgG.
This is quite different from the RFR in serum, which is
inhibited better by rabbit IgG than by human IgG (2,3).
Our interpretation for this was that the R F revealed in
the PFC assay was of such high avidity for human IgG
that, if made in the tissues, it would reach the peripheral
blood serum only with difficulty, if at all.
82 1
Various uncertainties attended these observations. First, our measurement of the RFR in the PFC
assay had involved hemolytic activity, whereas that of
the RFR in the serum involved agglutinating activity (1).
Secondly, the possibility existed that RFR of high affinity was actually present in the serum, but undetectable
by our assay, because it was still complexed with IgG.
Finally, it was necessary to determine that no artifact
was introduced into the PFC assay by aggregation of the
IgG when it was added as inhibitor to the agarose, or
that aggregates were otherwise determinative of our
findings. The current report brings definitive information to bear on these uncertainties.
Subjects. The sera and cells studied were from patients
with classic or definite RA (ARA criteria), all of whom had
vasculitis and/or active, aggressive joint disease. The patients
were under study in our Clinical Research Center.
Preparation of Lymphocytes (1). Lymphocytes were
isolated from these patients’ peripheral blood by differential
centrifugation of Ficoll-Hypaque. For their handling thereafter, the conditions were modified somewhat from previous
methods (I). Specifically, after isolation the lymphocytes were
washed at 4°C three times in full tissue culture medium RPMI
1640 (Grand Island Biological Co., Grand Island, New York)
containing 10% fetal bovine serum (FBS), resuspended in undiluted FBS, and added in 1 to 5 volume ratios to agarose
containing RPMI 1640. Inclusion of full tissue culture media
256 1024
l/Serum Dilution
Figure 1. Hemolysis of sheep cells sensitized with reduced and alkylated,
acid treated rabbit IgG antibody by sera containing rheumatoidfaciors.
in the agarose phase of the assay enhanced both the numbers
and size of the RF,-PFC seen.
Sensitized Sheep Red Cells. Rabbit IgG hemolysin
(anti-SRC) (BBL, Cockeysville, Maryland) was separated by
DE52 or Sephadex (3-200 (Pharmacia Fine Chemicals, Inc.,
Piscataway, New Jersey) gel filtration. The IgG hemolysin was
reduced and alkylated (R/A) as previously described (4). The
R/A rabbit IgG hemolysin was then dialyzed for 1-3 hours
against a pH 3.0 glycine buffer to produce conditions favorable
to the formation of hybrid molecules ( 5 ) . It was then returned
to neutrality by dialysis against phosphate buffered saline
(PBS) pH 7.3. The optical density, direct agglutinating titer,
and Coombs’ titer of the hemolysin before acid treatment were
2.680, 1 : 1024, and 1 : 1024, and after treatment 1.004, I :32,
and I :512, respectively. Thus SRC could be sensitized with
much larger quantities of rabbit IgG than previously used
without encountering interfering agglutination.
Source of Complement. Lyophilized guinea pig serum
(Hyland Laboratories, Los Angeles, California) was reconstituted with the diluent provided by the distributor.
PFC Assay for RF Production. Five-tenths milliliter of
0.5% agarose in RPMI 1640 at 44°C (agarose A-37 Induboise
R, L’industrie Biologique, Francaise, C.A., Gennevilliers,
France) was mixed with 0.1 ml of lymphocyte suspension
containing 3-7 X 108 cells in FBS and 0.05 ml of a 6.7%
suspension of SRC sensitized with acid-treated R/A rabbit
IgG hemolysin. The additions to the agarose were made in the
following order: 0.05 ml of 6.7% SRC suspension, 0.05 ml of
the inhibiting IgG, and 0.1 ml lymphocytes. The contents were
promptly mixed and poured onto precoated microscope slides.
The slides were then incubated at 37°C for 90 minutes in a
moist chamber, immersed in guinea pig serum at a 1 : 10 dilution, and incubated for an additional 90 minutes at 37°C;
finally the RF-PFC were enumerated.
Preparation of IgG. In the initial studies, human and
rabbit Cohn Fraction I1 were used. In later studies the Fraction I1 was passed through DE52 at 0.01 M phosphate, pH 8.0,
or IgG was isolated from individual normal human or rabbit
sera by DE52. In no instance were differences in the behavior of the IgG discerned, based upon these differences in
method of isolation.
Preparation of Monomeric IgC (6). The human and
rabbit IgG preparations were ultracentrifuged at 35,000 rpm
for 60 minutes at 4°C. After this the upper one third of the
solution was collected. The OD 280 of an aliquot was read to
determine protein concentration.
Preparation of Aggregated IgC (7). The human and
rabbit IgG (20 mg/ml) were aggregated at 63°C for 30 minutes
or until the solution became turbid (whichever came first). The
suspension was then centrifuged at 2,000 rpm for 5 minutes at
4°C to sediment the gross aggregates. The supernatant was
ultracentrifuged at 40,000 rpm for 90 minutes at 4°C. and the
pellet was recovered and diluted to 2.0 ml in PBS. The protein
concentration of the redissolved pellet was determined in a 0.2
ml portion mixed with 1.8 ml of 0.1 N NaOH and read at OD
Hemolytic Assay of Serum RF (4). To tubes containing
0.1 ml of serial dilutions of RA serum in buffer was added 0.1
ml of 1% SRC sensitized with acid treated R/A rabbit IgG.
After incubation at 37°C for 60 minutes and at 4°C for 90
minutes, 1.0 ml of cold veronal buffered saline (GVB) was
added to all tubes, and the tubes were spun at 2,000 rpm for 5
minutes at 4°C. The supernatants were discarded, and to the
cells was added 0.4 ml of 1:40 guinea pig complement previously absorbed with SRC. The tubes were then incubated a t
37OC for 60 minutes with gentle vortexing every 10 minutes,
after which the hemolysis was stopped with 1.5 ml of cold PBS,
and the supernatants were analyzed at OD 412. Zero and 100%
controls were included in all assays.
For the inhibition assays, dilutions of the RA serum
that had been predetermined t o give partial hemolysis were
used. IgG was serially diluted and mixed in 0.1 mi volumes
with 0.1 mi of the predetermined dilutions of the RA serum.
The procedure was thereafter performed as above for the
hemolytic assay.
Specificities of Hemolytic Serum R F R and of R F R PFC by Inhibition Analyses. The results of hemolytic
assays of RFR are shown in Figure 1. Some RA sera
exhibit prozones, as illustrated, which are presumably
due to inhibition of the RFR by either autologous serum
IgG or endogenous immune complexes present in effectively inhibitory concentrations at the lower serum dilutions. All inhibition studies subsequently performed
were carried out at serum dilutions beyond the prozones
to reduce the likelihood of confusion from their effects.
Four of the five patients illustrated in Figure 1
provided both serum and lymphocyte specimens for
study. The RFR-PFC of these four were markedly inhibited by human monomer IgG (Figure 2). Relative to the
rabbit IgG, the human IgG inhibited the RFR-PFC even
more strikingly than previously described in studies of
other patients (1). However, in all cases the human IgG
failed to inhibit hemolysis by the serum RFR, although
the rabbit IgG inhibited very well. Thus the difference in
serologic specificity of the RFR detected in the PFC
assay, as compared to RFR in the serum, was the same
as that noted previously (1) when agglutination was
used to assess the serum RFR.
Assessment for Aggregates in the Agarose. To test
whether aggregates of the human IgG developed during
the PFC assay, thus allowing the human IgG to inhibit
the RF,-PFC better simply as an artifact, 12SI-1abeled
monomer human IgG was added to agarose at 44OC in
RPMI 1640 and 10% FBS, as in the RFR-PFC assay.
One half milliliter of the agarose solution was immediately delivered to a small column made from a Pasteur
Serum RFR
R.C. (641
V.M. (1281
E.B. ( 3 2 1
c . 1 . 132)
' \
\ \
A E.B.
E.B. 132)
R.C. 164)
C.A. 1321
V.Y. 1128)
Rabbit IgG
A c.1.
P g IgG/ml
Human IgG
Figure 2. Comparison of human and rabbit monomer IgG as inhibitors of hemolytic rheumatoid factor activity in the sera
(serum RFR)or the lymphocytes (RF,-PFC) of RA patients. The numbers in parentheses specify the dilutions of the sera used.
The numbers of RF,-PFC per slide in the controls varied from 12 to 27, and all slides were prepared in triplicate.
inhibits RFR-PFC because of artifactual aggregation.
A Search for “Hidden” RFR in Serum. We then
examined the possibility that RF,with greater reactivity
for human IgG than for rabbit IgG exists in R A sera
but is not easily demonstrated because of being bound in
RFR-IgG complexes. Euglobulin precipitates were made
by dialyzing 0.25-0.5 ml RA sera against 100 ml of 0.001
M phosphate at pH 7.5 for 20-24 hours. The precipitates obtained were dissolved in 0.25 ml of a 0.05 M
glycine-HCI buffer at pH 3.2 and immediately delivered
to the top of a 10-40% sucrose density gradient in PBS
at pH 8.0. The tubes were centrifuged for 18 hours at
35,000 rpm. In the 7s fractions, IgG was demonstrable
by immunodiffusion analysis, but no RFR was detected
in it by slide latex test. The 19s fractions contained all
the detectable RFR and no detectable IgG.
Figure 3 illustrates the hemolysis-inhibition
curves of four sera and the 19s fractions prepared from
them. RFR was distinctly more inhibited by rabbit IgG
than by human IgG in all cases. The slopes of the
inhibition curves described with the 19s fractions did
not look significantly different from those with serum.
Thus evidence of “hidden” RFR in the serum was not
found by this procedure.
Table 1. Recovery of Monomer ‘“I-IgG from Agarose in Glass Wool
cpm ( X 109)
IgG incorporated*
Eluate Tail (20-50)
Tail (51-170)
Tail ( I 71-235)
Residue in agarose-wool
Total recovered cpm
* Sample mixed with agarose at 44°C and allowed to harden in glass
PBS containing 10% FBS.
are indicated in parentheses.
wool in a column. Column eluted with
t Tube numbers
pipette plugged with glass wool, and the agarose was
allowed to harden. The labeled protein was then eluted
from the column. All of it could be recovered from the
column with neutral buffer containing 10%FBS and the
protein remained 7s in size when examined at various
positions in the elution profile by sucrose density gradient ultracentrifugation. The data from one of three experiments yielding similar results are given in Table 1.
From those results, it seems unlikely that human IgG
Serum RFR
19s RFR
IgG pg/ml
Human IgG
200 400
IgG pg/ml
Rabbit IgG
Figure 3. Comparison of human and rabbit monomer IgG as inhibitors of hemolytic rheumatoid factor
activity in assays of whole sera of 19s fractions from the sera.
Effects of Heat Aggregation of the Human IgG on
Inhibition. To provide some measure of what should be
expected if our preparations of monomer human IgG
contained traces of aggregates, human IgG was aggregated by heating to 63"C, and the aggregates were concentrated according to Dickler et al. (7). The aggregated
IgG inhibited the RFR well in all three systems studied
(Figure 4). About 1/30-1/100 of the concentrations of
aggregated IgG was needed for the same degrees of
inhibition given by the monomer IgG, and sucrose density gradient studies revealed this to be attributable to
aggregates in the 2 19s zone. Repeatedly, when preparations of monomer lSII-IgGhave been tested on sucrose
density gradients, no evidence of such aggregates has
been found with conditions that would have detected
less than 1%, if present.
These inhibition effects could not be attributed to
anticomplementarity of the IgG preparations used. In
the serum and 19s hemolytic assays, the added IgG was
washed out of the system before the addition of complement. In the PFC assay, we believe the anticomplementarity of the IgG was effectively neutralized
by the high concentration of complement used in the
assay (1/ 10 guinea pig serum), since no anticomplementarity was detected with 30 pg/ml of aggregate IgG in
dilutions of guinea pig serum less than 1/200.
A classic observation has been that the activity of
serum RF, is inhibited better with rabbit IgG than with
human IgG. In this report it is established that in a PFC
assay of RA blood lymphocytes (RFR-PFC) an RFR is
released for which human IgG is more inhibitory than
rabbit IgG. RFR assessed at a site of its origin, i.e. the
lymphocyte, is therefore different from RFR obtained
from the serum of the same patient and assayed by
hemolytic capacity. Our interpretation is that lymphocytic cells produce an RFR that reacts avidly with human
IgG, but that this RFR fails to reach the blood serum, or
is immediately cleared from it.
Our previous findings suggesting this fact involved different techniques (agglutinating versus hemolytic) in assaying for the two RFR. In the present study it
is shown that the difference in the RFR is not attributable to the difference in techniques, nor to technical
inadequacies in the inhibiting monomer IgG reagents,
such as the presence of small but significnat amounts of
contaminating aggregates. The findings also did not result from the development of aggregates of IgG during
the PFC assay.
In most cases the complete inhibition of RFRPFC was observed with as little as 30 pg/ml (2 X lo-'
19s RFR
Serum RFR
-. Patient
Monomer IgG
Aggregated IgG
Human IgG pg/ml
Figure 4. Inhibition of hemolytic rheumatoid factor by monomer and aggregated human IgG.
M ) of added monomer IgG. This implies that the RFR
has an association constant for the IgG in the range of
lo’ L/M, which is two orders of magnitude greater than
association constants described for serum R F (8-1 I ) .
Whether there is multivalency in the RFR-PFC inhibition reactions which makes this comparison invalid is
not known, but observations by others (8,9,11) suggest
otherwise and therefore encourage the comparison.
Ashman and Metzger (11) and Normansell (9) noted
that intact IgM R F reacts with monovalent antigens or
haptens with the same association constants as do the
Fab,, fragments isolated from the IgM, so the pentavalency of the RFR does not in itself invalidate the
comparison, as long as the IgG with which it reacts
behaves univalently. Stone and Metzger (8) found that
Fcy behaved as a univalent antigen with a monoclonal
R F (Lay), and Normansell (9) and others (12,13) have
indicated that the whole IgG molecule also behaves
univalently with RF. Since each plaque in the RFR-PFC
assay represents RFR from an individual cell (and thus a
monoclonal RFR), and since the IgG appears to have
remained in monomer form in the agar, monovalency
may in fact be the true state of the reaction in the agar,
and the above comparison of the relative avidities of
RFR-PFC and serum RFR may therefore be valid.
Under any circumstances the ability of monomer
IgG at concentrations as low as 30 pg/ml to bind effectively with RFR indicates that such binding can take
place even in the low concentrations of IgG that exist in
the interstitial fluids. This obviously may have pathogenetic significance.
We considered the possibility that the high affinity RFR actually was present in the serum, but was not
easily demonstrated there because it was hidden in undissociated complexes with IgG. Allen and Kunkel (14)
used acid dissociated 19s serum fractions and clearly
demonstrated examples of such “hidden” R F reactive
with native IgG. They then postulated a higher binding
affinity for the hidden R F than that exhibited for other
serum RF. Our findings go one step further. Unlike
Allen and Kunkel, we did not find high affinity RFR in
acid dissociated IgM fractions of RA sera (Figure 3).
The logical deduction is that the RFR detected in our
RFR-PFC assay (and which must be made in the extravascular tissues) is of such high affinity that it reacts
locally with autologous IgG, is precipitated in the tissues, and thereby is denied access to the serum. In the
few patients who have RFR-PFC in their peripheral
blood, a small amount of RFR obviously must be released directly into the plasma from the circulating cells.
This RFR is probably promptly swept out of the circulation by the reticuloendothelial system, deposited onto
circulating leukocytes or platelets, or precipitated in the
blood vessel walls. It is noteworthy in this last respect
that we have observed ( 1 ) a significant correlation between presence of RFR-PFC in the peripheral blood and
the presence of necrotizing vasculitis in rheumatoid arthritis patients.
Our studies have been limited to IgM-RF, since
we have studied only the 19s fraction from serum and
have dealt with only direct plaques in the RFR-PFC
assay. We expect, however, that our findings with IgMR F would be equally applicable to IgG-RF.
We very much appreciate the excellent technical assistance of Ms Jean Rose and the aid in manuscript preparation
by Ms Anna Milne.
I . Vaughan JH, Chihara T , Moore TL, et al: Rheumatoid
factor-producing cells detected by direct hemolytic plaque
assay. J Clin Invest 58:933-941, 1976
2. Butler VP Jr, Vaughan JH: Hemagglutination by rheumatoid factor of cells coated with animal gamma globulin.
Proc SOCExp Biol Med 116:585-593, 1964
3. Butler V P Jr, Vaughan JH: The reaction of rheumatoid
factor with animal gamma-globulins: quantitative considerations. Immunol 8:144-159, 1965
4. Tanimoto K, Cooper N R , Johnson JS, et al: Complement
fixation by rheumatoid factor. J Clin Invest 55:437-445,
5. Nisonoff A , Palmer J L Hybridization of half molecules of
rabbit gamma globulin. Science 143:376-379, 1964
6. Spiegelberg HL, Weigle WO: The production of antisera
t o human gamma G subclasses in rabbits using immunologic unresponsiveness. J Immunol 101:377-380, 1968
7. Dickler H G , Kunkel HG: Interaction of aggregated y
globulin with B-lymphocytes. J Exp Med 136191-196,
8. Stone MJ, Metzger H: Binding properties of a Waldenstrijm macroglobulin antibody. J Biol Chem 24359775984, 1968
9. Normansell DE: Anti-y-globulins in rheumatoid arthritis
sera. I. Studies o n the 22s complex. Immunochem 7:787797, 1970
10. Eisenberg R:The specificity and polyvalency of binding of
a monoclonal rheumatoid factor. Immunochem 13:355359, 1976
11. Ashman RF, Metzger H: A Waldenstrom macroglobulin
which binds nitrophenyl ligands. J Biol Chem 244:34053414, 1969
12. Chavin SI,Franklin EC: Studies on antigen-binding activity of macroglobulin antibody subunits and their enzymatic fragments. J Biol Chem 244:1345-1352, 1969
13. Schrohenloher RE, Barry CB: Ultracentrifugal studies of
the reaction of IgM anti-gamma-globulin subunits with
human IgG. J Immunol 100:1006-1013, 1968
14. Allen JC, Kunkel HG: Hidden rheumatoid factors with
specificity for native y globulins. Arthritis Rheum 9:758768, 1966
Без категории
Размер файла
489 Кб
factors, reactivities, evidence, selective, deficiency, relative, serum, rheumatoid, cells
Пожаловаться на содержимое документа