Antibodies against transglutaminases, peptidylarginine deiminase and citrulline in rheumatoid arthritis - New pathways to epitope spreading

Article (PDF Available)inClinical and experimental rheumatology 24(1):12-8 · January 2006with30 Reads
Impact Factor: 2.72 · Source: PubMed
Abstract
The findings of the involvement of tissue transglutaminase (tTg) in the pathogenesis of coeliac disease (CD) have stimulated progress in the field of auto-immune diseases. Another calcium-dependent cysteine enzyme, peptidylarginine deiminase type 4 (PAD4), seems to be involved in the pathogenesis of rheumatoid arthritis (RA). There are obvious similarities between Tgs and PADs. Using enzyme-linked immuno-sorbent assays, we have measured the occurrence of antibodies against guinea pig (gp) and human recombinant (hr) tTg, PAD and citrulline in 59 controls and 184 RA-patients, of whom 71 were treated with methotrexate (mtx). In addition to the expected antibodies against citrulline (62%), sera from the 113 RA-patients without mtx treatment contained significantly increased frequencies of IgG anti-PAD (35%), IgA anti-gp-tTg (34%), IgA anti-hr tTg (20%), IgG anti-gp-tTg (13%) and IgA anti-hr-FXIII (15%) compared to controls. In sera from the mtx-treated RA-patients the expression of antibodies was reduced. In patients not treated with methotrexate there was a statistically significant correlation between, on one hand, IgG anti-PAD and on the other hand, IgG anti-citrulline, IgA anti-gp-tTg, IgA anti-hr-tTg, IgG anti-gp-tTg, IgG anti-hr-tTg, or IgA anti-hr-FXIII. In the mtx-treated group these correlations were less pronounced. In addition to the expected antibodies against citrulline, sera from RA-patients contained antibodies against PAD and against Tgs of at least two kinds, indicating that the specificity for anti-tTg in CD is far from complete. Most of the patients displayed more than one antibody, a possible indication of epitope spreading. MTX-treatment reduced the expression of antibodies.

Figures

Full-text (PDF)

Available from: Bodil Roth
Clinical and Experimental Rheumatology 2006; 24: 12-18.
Antibodies against transglutaminases, peptidylarginine
deiminase and citrulline in rheumatoid arthritis –
new pathways to epitope spreading
E.B. Roth
1, 2
, P. Stenberg
1
, C. Book
2
, K. Sjöberg
2
1
Hospital Pharmacy, Malmö University Hospital, Malmö, Sweden;
2
Department of Medicine,
Lund University, Malmö University Hospital, Malmö, Sweden.
Abstract
Objective
The findings of the involvement of tissue transglutaminase (tTg) in the pathogenesis of coeliac disease (CD) have
stimulated progress in the field of auto-immune diseases. Another calcium-dependent cysteine enzyme, peptidylarginine
deiminase type 4 (PAD4), seems to be involved in the pathogenesis of rheumatoid arthritis (RA). There are obvious
similarities between Tgs and PADs.
Methods
Using enzyme-linked immuno-sorbent assays, we have measured the occurrence of antibodies against guinea pig (gp)
and human recombinant (hr) tTg, PAD and citrulline in 59 controls and 184 RA-patients, of whom 71 were treated with
methotrexate (mtx).
Results
In addition to the expected antibodies against citrulline (62 %), sera from the 113 RA-patients without mtx treatment
contained significantly increased frequencies of IgG anti-PAD (35 %), IgA anti-gp-tTg (34 %), IgA anti-hr tTg (20%),
IgG anti-gp-tTg (13 %) and IgA anti-hr-FXIII (15%) compared to controls. In sera from the mtx-treated RA-patients the
expression of antibodies was reduced. In patients not treated with methotrexate there was a statistically significant cor-
relation between, on one hand, IgG anti-PAD and on the other hand, IgG anti-citrulline, IgA anti-gp-tTg, IgA anti-hr-
tTg, IgG anti-gp-tTg, IgG anti-hr-tTg, or IgA anti-hr-FXIII. In the mtx-treated group these correlations were less pro-
nounced.
Conclusion
In addition to the expected antibodies against citrulline, sera from RA-patients contained antibodies against PAD and
against Tgs of at least two kinds, indicating that the specificity for anti-tTg in CD is far from complete. Most of the
patients displayed more than one antibody, a possible indication of epitope spreading. MTX-treatment reduced the
expression of antibodies.
Key words
Autoimmunity, citrulline, coeliac, epitope spreading, factor XIII, peptidylarginine deiminase, rheumatoid,
transglutaminase, methotrexate.
Page 1
Rheumatoid arthritis and autoantibodies / E.B. Roth et al.
13
E. Bodil Roth, BSci.; Pål Stenberg, PhD,
Assoc. Professor; Christina Book, MD;
Klas Sjöberg, MD, Associate Professor.
This work has been financially supported
by Anna and Edwin Bergers Foundation
and by Ernhold Lundström’s Foundation.
Please address correspondence to:
E. Bodil Roth, Department of Medicine,
Lund University, Malmö University
Hospital, S-205 02 Malmö, Sweden.
E-mail: pal.stenberg@apoteket.se
Received on July 5, 2004; accepted in
revised form on September 1, 2005.
© Copyright C
LINICAL AND EXPERIMEN-
TAL RHEUMATOLOGY 2006.
Introduction
Antibodies against peptide-bound cit-
rulline are considered highly specific
for rheumatoid arthritis (RA) (1). A
family of enzymes, the peptidylargi-
nine deiminases (PADs) (2), catalyzes
the citrullination of arginine-containing
peptides. The ureido group on the cit-
rulline facilitates the unfolding of pro-
teins due to decrease in net charge and
loss of potential ionic bonds (3). Stud-
ies by Suzuki et al. (4) imply that the
PAD4 haplotype is associated with sus-
ceptibility to RA. PAD4 is highly ex-
pressed in haematological cells and in
RA synovial tissues. Deiminated α-
and β-chains of fibrin have been point-
ed out as the major targets of RA-spe-
cific auto-antibodies (5).
In coeliac disease (CD), antibodies
against endomysium (EMA) and glia-
dins are important complements of
intestinal biopsies for establishing di-
agnosis. In 1997, Dieterich et al. (6)
convincingly showed that tissue trans-
glutaminase (tTg) is the major auto-
antigen of EMA. The normal Tg-cat-
alyzed reaction is a transamidation
between protein-bound glutamine and
lysine residues, forming γ-glutamine-ε-
lysine pseudo-peptide bridges between
proteins, for example in the Factor XIII
(FXIII) catalyzed stabilization of fib-
rin. In the absence of a potent nucle-
ophilic second substrate, tTg can cat-
alyze the hydrolysis of specific gluta-
mine residues to glutamic acids in for
example gliadins. This deamidation
strengthens the recognition of gliadin
by HLA-restricted gut-derived T-cells.
For reviews, see (7, 8).
Antibodies against tTg have also been
observed in conditions not directly
connected to CD, such as inflammatory
bowel diseases (9).
There are obvious similarities between
Tgs and PADs. Both these post-transla-
tionally active families of thiol enzymes
have similar molecular weights and are
calcium-dependent, acting in the first
step of the reaction by a nucleophilic
attack on a polarized carbon atom in the
substrate. In both cases, the products
are ammonia and a peptide/protein with
changed electrical charges and conse-
quently with opportunities for new con-
formations. Furthermore, the expres-
sion of the subclasses of the two fami-
lies of enzymes seems to be similar.
Interestingly, the two types of enzymes
might work sequentially. For example,
maximal conversion of arginines to cit-
rullines in a modified form of the hair
follicle trichohyalin makes this structur-
al protein a more efficient substrate for
epidermal Tg (10).
Epitope spreading refers to a situation
where immune responses develop to
new epitopes, distinct from and non-
cross reactive with the primary epitope
causing the disease (11). The mecha-
nisms for this phenomenon are still
under investigation. Interestingly, in
many cases enzymes and their sub-
strates seem to be the auto-immune tar-
gets.
The strategy of the present work was to
apply the progress made in understand-
ing the pathogenesis of CD to the field
of RA. Specifically, we wanted to eval-
uate the occurrence and correlation of
antibodies against citrulline and the
enzymes PAD, tTg and Factor XIII in
sera from patients with RA.
Materials and methods
Patients and controls
All patients attending the Rheumato-
logical Unit of Malmö University Hos-
pital during 1995-2002, diagnosed with
RA according to the criteria by the Am-
erican College of Rheumatology 1987,
were consecutively registered and sys-
tematically monitored. The median
duration of disease at inclusion was 7
months (range 2-12). A total of 184
patients (53 males and 131 females;
mean age (S.D.) 61.0 ± 14.0 years;
range 22-84) were enrolled in the
study. None had been diagnosed with
CD, nor had they any signs or symp-
toms of CD. At the time of sampling,
71 patients had initiated treatment with
methotrexate (MTX). Other treatments
included chloroquine (n = 43), sulfasa-
lazine (n = 13), gold sodium thiomalate
(n = 2) and azathioprine (n =1), while
54 patients received no treatment. Blood
donors (30 males and 29 females; mean
(S.D.) age 42 ± 16.2 years; range 19-
65) served as controls.
All procedures were in accordance
with the Ethical Committee of Lund
University.
Page 2
Rheumatoid arthritis and autoantibodies / E.B. Roth et al.
14
Assays of antibodies
For all determinations of antibodies,
except anti-citrulline, cut-off was cal-
culated as mean + 2 S.D. of controls.
For IgG anti-citrulline the cut-off was
defined according to the manufactur-
ers manual as > 25 units.
EMA were analyzed at the Department
of Clinical Microbiology and Immuno-
logy of Lund University Hospital ac-
cording to standard procedures.
Determination of PAD
For the analysis of IgG anti-PAD-anti-
bodies, a rabbit skeletal muscle PAD
(Sigma, batch 013K4128) was used as
antigen. Microtitre plates (Maxisorp
Nunc) were coated with 1 µg PAD/
well, freshly prepared with 100 µl of 50
mM Tris-HCl, 150 mM NaCl, pH 7.4
(TBS) and 5.0 mM CaCl
2
. After incu-
bation overnight at 4°C, the plates were
washed three times with washing solu-
tion (TBS with 10 mM EDTA and 0.1%
Tween 20) and blocked for 30 minutes
at room temperature (RT) with block-
ing solution (TBS with 0.5% Tween 20
and 1% human serum albumin (HSA)).
Sera from patients with RA and healthy
blood donors were diluted 200 times in
washing solutions with 0.5% Tween
20. To each well 100 µl diluted serum
was added in duplicate followed by
incubation for one hour at RT. The
washing procedure was repeated as
above and 100 µl of peroxidase-conju-
gated antihuman IgG (DAKO), 1:5000
in washing solution, was added to each
well. After incubation for one hour at
RT the washing procedure was repeat-
ed once more and the plates were
developed according to standard proce-
dure with 1 mg OPD/ml, 0.1 M Na-cit-
rate at pH 4.2 (DAKO), and 0.06%
H
2
O
2
. The enzyme reaction was
stopped after incubation at 30 minutes
in darkness at RT by the addition of
100 µl 0.5 M H
2
SO
4
. Absorbance was
estimated at 490 nm in a microplate-
reader (E max). The intra-assay corre-
lation of variation (CV) was 7.3 % (n =
8) and the inter-assay CV was 10.9 %
(n = 3).
Determination of IgG anti-citrulline
The analysis of anti-CCP was per-
formed in our laboratory with the
Immunoscan RA Mark2 kit (Euro-
Diagnostica) in accordance with the
manufacturers manual. As antigen,
this method utilizes a synthetic, cyclic
citrulline-containing peptide (CCP).
Determination of IgA and IgG
anti-tTg antibodies
For the analysis of anti-tTg, a guinea
pig liver Tg (gp-tTg, Sigma lot
122K7435; 1 µg/well) or a human
recombinant Tg (hr-tTg, N-Zyme lot
0804T0021; 0.5 µg/well) was used.
Freshly prepared solutions of the anti-
gens in TBS containing 5.0 mM CaCl
2
were added to microtitre plates (Cov-
aLink, Nunc). After incubation over-
night at 4°C the plates were washed
three times with washing solution
(TBS with 10 mM EDTA and 0.1%
Tween 20) and blocked for 30 minutes
at RT with blocking solution (TBS with
0.5% Tween 20).
Sera from patients with RA and healthy
blood donors were diluted 400 times
(gp) or 1600 times (hr) with blocking
solution, and 100 µl diluted serum were
added to each well in duplicate fol-
lowed by incubation for one hour at RT.
The washing procedure was repeated
as above and 100 µl of either peroxi-
dase conjugated antihuman IgG
(DAKO), 1:5000, or peroxidase conju-
gated antihuman IgA (DAKO), 1:1000
(gp) or 1:2000 (hr) diluted in washing
solution, were added to each well. After
incubation at RT for one hour the wash-
ing procedure was repeated once more.
The plates were developed according
to standard procedure with 1 mg
OPD/ml, 0.1 M Na-citrate at pH 4.2
(DAKO), and 0.06% H
2
O
2
. After incu-
bation at RT for one hour in darkness,
the enzyme reaction was stopped by
the addition of 100 µl 0.5 M
H
2
SO
4
/well. Absorbance was estimated
at 490 nm in a microplate-reader (E
max).
The intra-assay CV of the guinea pig
based method (n = 8) was 3.0% for IgA
and 8.6% for IgG, while inter-assay CV
was 6.1% for both IgA and IgG. The
intra-assay CV of the method based on
human recombinant tTg (n = 8) was 0.8
% for IgA and 1.3 % for IgG, while
inter-assay CV was 3.1% for IgA and
7.1% for IgG.
Determination of IgA anti-hr-FXIII
a subunit
For the analysis of antibodies against
thrombin-activated FXIII, human re-
combinant FXIII (kindly supplied by
Dr Bruce Carter, Zymogenetics), com-
prising the a subunit, was used as anti-
gen. Microtitre plates (Maxisorp Nunc)
were coated with 1 µg FXIII/well,
freshly prepared with 100 µl of 50 mM
Tris-HCl, 150 mM NaCl, pH 7.4 (TBS)
including 5.0 mM CaCl
2
and 0.2 U hu-
man thrombin (Sigma, lot no. 21K
7602). The procedure was otherwise
the same as for the analysis of antibod-
ies against tTg. Intra-assay CV was
determined as 3.9% (n = 8) and inter-
assay CV was 1.7%.
Statistics
The statistical significance of differ-
ences was determined by the χ
2
-test and
2-tailed Student’s t-test. Due to skewed
distribution of the values, logarithmic
transformation was carried out before
analysis. Spearman rank correlation
was used to evaluate correlation. P-val-
ues < 0.05 were considered significant
Results
The levels of antibodies against CCP,
PAD, tTg and FXIII in sera from the
184 patients with RA are presented in
Figure 1. Of the 113 RA patients with-
out MTX-treatment, 40 (35%) were
positive for IgG anti-PAD, 70 (62%)
for IgG anti-CCP, 38 (34%) for IgA
anti-gp-tTg, 15 (13%) for IgG anti-gp-
tTg, 22 (20%) for IgA anti-hr-tTg, 21
(19%) for IgG anti-hr-tTg and 17
(15%) for IgA anti-hr-FXIII. With the
exception of IgG anti-hr-tTg, all anti-
bodies were significantly elevated
compared with the controls (Table I).
The corresponding figures for all the
184 patients, including the 71 MTX-
treated were: 57 (31%) for IgG anti
PAD, 109 (59%) for IgG anti-CCP, 55
(30%) IgA anti-gp-tTg, 22 (12%) for
IgG anti-gp-tTg, 30 (16%) for IgA anti-
hr-tTg, 24 (13%) for IgG anti-hr-tTg
and 22 (12%) for IgA anti-hr-FXIII.
In sera from the 71 MTX-treated RA-
patients only IgG anti-PAD, IgG anti-
CCP and IgA anti-gp-tTg were statisti-
cally elevated (Table I).
When significantly elevated compared
Page 3
Rheumatoid arthritis and autoantibodies / E.B. Roth et al.
15
with the controls, the antibody levels of
anti-tTg were not in any way as high as
those found in true CD-patients (12)
including infants (13). Of the 56 RA-
patients who were positive for IgA
anti-gp-tTg and the 28 RA-patient who
were positive for IgA anti-hr-tTg, none
were positive in the EMA test.
The various antibodies correlated
markedly with each other. Especially,
IgG anti-PAD correlated with all other
antibodies and IgG anti-CCP correlated
with all but IgG anti-hr-tTg and IgA
anti-hr-FXIII. In contrast, IgG anti-gp-
tTg did not correlate with most of the
other transglutaminase antibodies. For
details, see Table IIa.
The effects of treatment with MTX are
illustrated in Figure 2. For all antibod-
ies, patients treated with MTX dis-
played lower levels compared with
patients on other treatment or with no
treatment at all, with significant differ-
ences obtained for anti-PAD, IgA and
IgG anti-hr-tTg, IgG anti-gp-tTg and
IgA anti-hr-FXIII.
Discussion
Antibodies against CCP and PAD
Considering the situation in CD with
antibodies against the substrate (glia-
din) as well as against the correspond-
ing enzyme (tTg), we found it worth-
while investigating the situation in RA
with the enzyme/converted substrate
pair, i.e. PAD and peptide bound citrul-
line. Our results, with a 59% frequency
of IgG anti-CCP in sera from RA-
patients, are in line with an earlier re-
port (1). Moreover, we did not find any
case of antibodies among the blood
donors, further emphasizing the high
specificity of IgG anti-CCP in RA. In-
terestingly, similar to a recent Finnish
study (14), we also observed a signifi-
cantly increased frequency of IgG anti-
PAD among the RA patients. Further-
more, there was a high correlation
between the prevalence of antibodies
against PAD and CCP. Consequently,
the situation in CD, with a combined
immune reaction against an enzyme
(tTg) as well as against a modified sub-
strate (gliadin), is also observed in RA.
Antibodies against transglutaminases
In CD, the occurrence of serum IgA
anti-tTg is regarded as a sensitive labo-
ratory marker especially if the antigen
is exposed to calcium during the coat-
ing in the ELISA procedure (13). In the
present study on RA patients, 30%
were positive for IgA anti-gp-tTg and
15% for IgA anti-hr-tTg. It should be
emphasized that there were no signs or
symptoms of CD among these RA pa-
tients. Furthermore, all patients with
positive IgA anti-gp-tTg and anti-hr-
tTg were EMA negative. IgA anti-hr-
tTg has been observed in other intesti-
nal disorders as well, such as ulcerative
colitis and Crohn’s disease (9). These
patients were all EMA negative. In
both the RA patients and in the patients
with inflammatory bowel disease the
tTg antibody titres were lower than
those found in true CD patients, so it
may be postulated that the tTg epitopes
found in CD could be different from
those found in RA. Compared with our
data based on guinea pig liver Tg, Pi-
carelli et al. (15), using native erythro-
cyte or recombinant human tTg as anti-
gens, found a similar frequency of anti-
bodies in RA. In contrast, Bizzaro et al.
(16) reported a low frequency of anti-
Fig. 1. The levels of antibodies in samples from RA-patients.
: not methotrexate treated, x: methotrexate treated. The horizontal lines indicate cut off levels.
Table I. Frequency (%) of positive antibody levels in the various groups of patients and
controls. Statistical significance compared with the controls is indicated below the frequen-
cies.
All patients Mtx-treated Not mtx-treated Blood donors
n = 184 n = 71 n = 113 n = 59
IgG anti-PAD 31 25 35 3.4
p < 0.0001 p < 0.0001 p < 0.0001
IgG anti-CCP 59 55 62 0.0
p < 0.0001 p < 0.0001 p < 0.0001
IgA anti-gp-tTg 30 25 34 0.0
p < 0.0001 p < 0.0001 p < 0.0001
IgG anti-gp-tTg 12 9.9 13 3.4
p = 0.023 p = 0.065 p = 0.011
IgA anti-hr-tTg 16 9.9 20 6.9
p = 0.080 p = 0.549 p = 0.026
IgG anti-hr-tTg 13 7.0 19 6.1
p = 0.204 p = 0.843 p = 0.085 (n=49)
IgA anti-hr-FXIII 12 7.0 15 5.1
p = 0.081 p = 0.573 p = 0.020
Page 4
Rheumatoid arthritis and autoantibodies / E.B. Roth et al.
16
tTg in RA. Riente et al. also reported a
low frequency of anti-tTg in RA both
for anti-bovine and anti-human tTg
antibodies (17).
Differences in the three-dimensional
structure in bovine and guinea pig Tg
may explain the varying results obtain-
ed for the two species, while the results
for human tTg are more intriguing. The
negative result obtained by Riente et al.
was based on a commercial kit that has
been used in several studies with good
accuracy as regards CD. The varying
results in RA may be due to differences
in the tTg preparation or the technique
used and/or a reaction towards other
epitope(s). Some minor differences be-
tween the Eurospital kit and our meth-
od are most likely. Changes in, for ex-
ample, the calcium concentration have
great impact on the results for tTg anti-
body titres. Since the composition of
the kit is not completely official it is
not possible to fully explain the differ-
ences observed. Obviously, treatment
with MTX might also contribute to this
complex situation.
Thrombin-activated FXIIIa and tTg
have similar characteristics and possi-
bly also similar physiological func-
tions. In an early preliminary study on
RA synovial fluids, we detected Tg
activity, sometimes originating from
tTg and sometimes from Factor XIII,
seemingly in an irreproducible pattern
(18). The explanation was given many
years later when we found that human
monocytes express FXIIIa, but during
the transformation of the cell, the ex-
Table IIa. The correlation (Spearman rank with 95% confidence interval) between antibodies in sera from the 113 RA-patients without
methotrexate treatment.
IgG anti-PAD IgG anti-CCP IgA anti-gp- tTg IgG anti-gp- tTg IgA anti-hr- tTg IgG anti-hr- tTg
IgG anti-CCP 0.40 _____ _____ _____ _____ _____
(0.23; 0.55)
p < 0.0001
IgA anti-gp-tTg 0.39 0.24 _____ _____ _____ _____
(0.22; 0.54) (0.05; 0.41)
p < 0.0001 p = 0.01
IgG anti-gp-tTg 0.44 0.23 0.39 _____ _____ _____
(0.28; 0.58) (0.04; 0.40) (0.21; 0.54)
p < 0.0001 p = 0.02 p < 0.0001
IgA anti-hr-tTg 0.25 0.23 0.56 0.13 _____ _____
(0.06; 0.42) (0.04; 0.40) (0.42; 0.68) (-0.07; 0.31)
p = 0.01 p = 0.02 p < 0.0001 p = 0.18
IgG anti-hr-tTg 0.20 0.03 0.09 0.13 0.28 _____
(0.01; 0.37) (-0.16; 0.22) (-0.11; 0.27) (-0.06; 0.31) (0.09; 0.45)
p = 0.04 p = 0.75 p = 0.37 p = 0.17 p = 0.01
IgA anti-hr-FXIII 0.36 0.18 0.63 0.15 0.52 0.15
(0.18; 0.52) (-0.01; 0.36) (0.50; 0.73) (-0.04; 0.33) (0.36; 0.65) (-0.04; 0.03)
p < 0.0001 p = 0.05 p < 0.0001 p = 0.11 p < 0.0001 p = 0.10
Table IIb. The correlation (Spearman rank with 95% confidence interval) between antibodies in sera from the 71 RA-patients with
methotrexate treatment.
IgG anti-PAD IgG anti-CCP IgA anti-gp- tTg IgG anti-gp- tTg IgA anti-hr- tTg IgG anti-hr- tTg
IgG anti-CCP 0.38 _____ _____ _____ _____ _____
(0.15;0.56)
p = 0.001
IgA anti-gp-tTg 0.12 0.09 _____ _____ _____ _____
(-0.12;0.35) (-0.15;0.33)
p = 0.32 p = 0.44
IgG anti-gp-tTg 0.20 0.24 0.32 _____ _____ _____
(-0.04;0.42) (0.00;0.46) (0.09;0.52)
p = 0.096 p = 0.04 p = 0.006
IgA anti-hr-tTg 0.15 0.29 0.35 0.04 _____ _____
(-0.09;0.38) (0.0.4;0.49) (0.12;0.54) (-0.20;0.28)
p = 0.21 p = 0.02 p = 0.003 p = 0.74
IgG anti-hr-tTg 0.29 -0.02 0.02 0.24 0.37 _____
(0.053;0.50) (-0.26;0.22) (-0.22;0.26) (0.00;0.46) (0.15;0.56)
p = 0.01 p = 0.86 p = 0.89 p = 0.042 p = 0.001
IgA anti-hr-FXIII 0.09 -0.01 0.34 -0.22 0.45 0.07
(-0.16;0.32) (-0.25;0.23) (0.11;0.54) (-0.44;0.02) (0.24;0.62) (-0.17;0.30)
p = 0.46 p = 0.94 p = 0.004 p = 0.062 P < 0.0001 P = 0.56
Page 5
Rheumatoid arthritis and autoantibodies / E.B. Roth et al.
17
pression of FXIII is down-regulated
and substituted completely by tTg in
the mature macrophage (19). Thus, al-
though at a lower frequency, it is not
surprising that we also observed IgA
anti-FXIII in sera from RA-patients.
This relationship is further emphasized
by the high correlation between the
occurrence of IgA against tTg and
FXIII (Table IIa).
The effects of methotrexate
In general, patients treated with MTX
had lower levels of antibodies than
patients on no – or other – drugs. This
result is in line with a recent prospec-
tive study that reported reduced levels
of anti-tTg after treatment with MTX
(15). In the case of MTX-treated pa-
tients, the correlations between the
antibody levels were less clear, proba-
bly due to a general depression of anti-
body expression. Even though no im-
munosuppression, measured by a de-
crease in the total number of circulating
leukocytes, occurred in a group of pa-
tients with SLE, their B-cells and dif-
ferent types of autoantibodies (ANA
and anti-dsDNA among others) de-
creased after more than 10 weeks of
MTX-treatment (20).
Epitope spreading
For the patients not being treated with
MTX, there was a clear correlation
between the levels of most antibodies.
Compare the situation in CD, where
antibodies against gliadins appear
hand-in-hand with auto-antibodies
against tTg. This intriguing situation
may be explained by the formation of
an immunogen complex between the
substrate, gliadin, and the enzyme, tTg.
The nature of this complex is still under
investigation. In vitro in the presence of
calcium and reducing agents, highly
purified tTg forms high molecular
weight structures with retained enzyme
activity. Moreover, in vitro a substrate
can be covalently incorporated into tTg
via γ-glutamine-ε-lysine bonds. Indeed,
a recent report showed that a complex
between tTg and gliadin could be iden-
tified ex vivo (21). In 2003, we pro-
posed another character of the com-
plex, namely a thioester of the Mi-
chael’s intermediate in the enzyme
reaction (12). In the absence of a suit-
able nucleophilic substrate such as a
primary amine, this thioester might be
fairly long-lived. Recently, Flecken-
stein et al. (22) were indirectly able to
detect a thioester between human tTg
and some synthetic, glutamine-contain-
ing gliadin peptides. They also found
peptides incorporated into the enzyme
via glutamyl-lysine pseudopeptide
bridges. The reactions were performed
in vitro in the presence of calcium and
at pH 7.4. Obviously, at a lower pH such
as in inflamed tissues, the fraction of
the thioester might have been even
higher. Indeed, this result strengthens
our theory of a thioester being the anti-
gen in CD (12).
Our present findings of antibodies in
RA against the enzyme PAD and
against the deiminated substrate, pep-
tide-bound citrulline, resemble the situ-
ation with CD. It is probable that the
PAD reaction occurs in three steps:
1. formation of the Michael’s complex;
2. the formation of an amidino-enzyme
intermediate with liberation of am-
monia;
3. hydrolysis of the amidino-enzyme.
However, in contrast to the situation
with tTg/gliadin, there are no obvious
reasons to believe that peptide-bound
arginine or citrulline residues will be
incorporated into PAD. These facts
favour the explanation of the covalent
enzyme-substrate intermediate being
the antigen.
Fig. 2. A comparison between the relative levels of antibodies in samples from RA-patients with (mtx) and without (not mtx) methotrexate treatment.
Page 6
Rheumatoid arthritis and autoantibodies / E.B. Roth et al.
18
If an enzyme such as tTg or PAD forms
a temporary, covalent intermediate
with a protein substrate previously mo-
dified by the enzyme, this complex
might be antigenic, resulting in anti-
bodies against both the enzyme and the
modified substrate. This might explain
the phenomenon called epitope (deter-
minant) spreading. In such a situation,
accidental activation of the enzyme
plays the key role. Recently, we des-
cribed that zinc at physiological levels
inhibits the binding of CD antibodies to
tTg (12). Indeed, zinc might be the
physiological moderator of Tg activity.
Unfortunately, with most enzymes in-
cluding PAD, very little is known about
the physiological moderation of en-
zyme activity.
Acknowledgements
We are grateful to Jan-Åke Nilsson for
statistical advice and Peter Baston for
linguistic assistance.
References
1. VA N VENROOIJ WJ, HAZES JM, VISSER H:
Anticitrullinated protein/peptide antibody
and its role in the diagnosis and prognosis of
early rheumatoid arthritis. Neth J Med 2002;
60: 383-8.
2. VOSSENAAR ER, ZENDMAN AJ, VA N VEN-
ROOIJ WJ, PRUIJN GJ
: PAD, a growing fam-
ily of citrullinating enzymes: genes, features
and involvement in disease. Bioessays 2003;
25: 1106-18.
3.
TARCSA E, MAREKOV LN, MEI G, MELINO
G, LEE SC, STEINERT PM
: Protein unfolding
by peptidylarginine deiminase. Substrate
specificity and structural relationships of the
natural substrates trichohyalin and filaggrin.
J Biol Chem 1996; 271: 30709-16.
4. SUZUKI A, YAMADA R, CHANG X et al.:
Functional haplotypes of PADI4, encoding
citrullinating enzyme peptidylarginine deim-
inase 4, are associated with rheumatoid arth-
ritis. Nat Genet 2003; 34: 395-402.
5.
MASSON-BESSIERE C, SEBBAG M, GIRBAL-
NEUHAUSER E
et al.: The major synovial
targets of the rheumatoid arthritis-specific
antifilaggrin autoantibodies are deiminated
forms of the alpha- and beta-chains of fibrin.
J Immunol 2001; 166: 4177-84.
6. DIETERICH W, EHNIS T, BAUER M et al.:
Identification of tissue transglutaminase as
the autoantigen of celiac disease. Nat Med
1997; 3: 797-801.
7. MOLBERG O, MCADAM SN, SOLLID LM:
Role of tissue transglutaminase in celiac dis-
ease. J Pediatr Gastroenterol Nutr 2000; 30:
232-40.
8.
DIETERICH W, ESSLINGER B, SCHUPPAN D:
Pathomechanisms in celiac disease. Int Arch
Allergy Immunol 2003; 132: 98-108.
9.
FARRACE MG, PICARELLI A, DI TOLA M et
al.: Presence of anti-”tissue” transglutami-
nase antibodies in inflammatory intestinal
diseases: an apoptosis-associated event? Cell
Death Differ 2001; 8: 767-70.
10. TARCSA E, MAREKOV LN, ANDREOLI J et
al.: The fate of trichohyalin. Sequential post-
translational modifications by peptidyl-argi-
nine deiminase and transglutaminases. J Biol
Chem 1997; 272: 27893-901.
11. LEHMANN PV, FORSTHUBER T, MILLER A,
SERCARZ EE
: Spreading of T-cell autoim-
munity to cryptic determinants of an autoan-
tigen. Nature 1992; 358: 155-7.
12. ROTH EB, SJOBERG K, STENBERG P: Bio-
chemical and immuno-pathological aspects
of tissue transglutaminase in coeliac disease.
Autoimmunity 2003; 36: 221-6.
13. AGARDH D, ROTH B, LERNMARK A, STEN-
BERG P
: Calcium activation of tissue trans-
glutaminase in radioligand binding and en-
zyme-linked autoantibody immunoassays in
childhood celiac disease. Clinica Chimica
Acta 2005; 358: 95-103.
14. NISSINEN R, PAIMELA L, JULKUNEN H et
al.: Peptidylarginine deiminase, the arginine
to citrulline converting enzyme, is frequently
recognized by sera of patients with rheuma-
toid arthritis, systemic lupus erythematosus
and primary Sjögren syndrome. Scand J
Rheumatol 2003; 32: 337-42.
15.
PICARELLI A, DI TOLA M, SABBATELLA L
et al.: Anti-tissue transglutaminase antibod-
ies in arthritic patients: a disease-specific
finding? Clin Chem 2003; 49: 2091-4.
16. BIZZARO N, VILLALTA D, TONUTTI E, TAM-
POIA M, BASSETTI D, TOZZOLI R
: Associa-
tion of celiac disease with connective tissue di-
seases and autoimmune diseases of the diges-
tive tract. Autoimmun Rev 2003; 2: 358-63.
17.
RIENTE L, CHIMENTI D, PRATESI F et al.:
Antibodies to tissue transglutaminase and
Saccharomyces cerevisiae in ankylosing
spondylitis and psoriatic arthritis. J Rheuma-
tol 2004; 31: 920-4.
18. WOLLHEIM FA, STENBERG, P, MARSAL, L:
Characterisation of transglutaminases in
plasma and synovial fluid from patients with
arthritis. 15th International Congress of
Rheumatology, Paris, 1981.
19. SEIVING B, OHLSSON K, LINDER C, STEN-
BERG P
: Transglutaminase differentiation
during maturation of human blood mono-
cytes to macrophages. Eur J Haematol 1991;
46: 263-71.
20. BOHM I
: Decrease of B-cells and autoanti-
bodies after low-dose methotrexate. Biomed
Pharmacother 2003; 57: 278-81.
21.
CICCOCIOPPO R, DI SABATINO A, ARA C et
al.: Gliadin and tissue transglutaminase com-
plexes in normal and coeliac duodenal mu-
cosa. Clin Exp Immunol 2003; 134: 516-24.
22. FLECKENSTEIN B, QIAO SW, LARSEN MR,
JUNG G, ROEPSTORFF P, SOLLID LM
: Mol-
ecular characterization of covalent complex-
es between tissue transglutaminase and
gliadin peptides. J Biol Chem 2004; 279:
17607-16.
Page 7
    • "In vivo, differences in ACPA levels [73], fine specificity or epitope spreading [218, 219], avidity220221222, isotype usage [223], and glycosylation [224] may be associated with differences in the potential to activate effector mechanisms, thereby influencing their biological potency [220]. Epitope spreading is often a hallmark of progressive B-cell responses and was described for ACPA and was associated with an increase and/or shift in antigen recognition during the course of an autoimmune response [218, 225, 226]. Fine mapping analysis of preclinical sera compared to early and established RA showed subtle difference in either the identity or the numbers of epitope detected between the different phases of the disease [43, 215, 216, 227, 228]. "
    [Show abstract] [Hide abstract] ABSTRACT: Autoantibodies have been associated with human pathologies for a long time, particularly with autoimmune diseases (AIDs). Rheumatoid factor (RF) is known since the late 1930s to be associated with rheumatoid arthritis (RA). The discovery of anticitrullinated protein antibodies in the last century has changed this and other posttranslational modifications (PTM) relevant to RA have since been described. Such PTM introduce neoepitopes in proteins that can generate novel autoantibody specificities. The recent recognition of these novel specificities in RA provides a unique opportunity to understand human B-cell development in vivo. In this paper, we will review the three of the main classes of PTMs already associated with RA: citrullination, carbamylation, and oxidation. With the advancement of research methodologies it should be expected that other autoantibodies against PTM proteins could be discovered in patients with autoimmune diseases. Many of such autoantibodies may provide significant biomarker potential.
    Full-text · Article · Mar 2014
    • "To confirm the validity of protein array detection, we used the same proteins in ELISAs to screen more sera from patients and controls. Among these proteins, only one, PAD4, the enzyme that converts arginine into citrulline , was already known to be a target for RA autoantibo- dies45678. Autoantibodies to PAD4 are more frequent in established than in early RA. In a previous study, we have identified peptide targets of anti-PAD4 autoantibodies [8]. "
    [Show abstract] [Hide abstract] ABSTRACT: Rheumatoid arthritis (RA) is a chronic inflammatory joint disease causing articular cartilage and bone destruction. Since irreversible joint destruction can be prevented by intervention at the early stages of disease, early diagnosis of RA is important. In this study, we identified new autoantibodies in the sera of patients with early (less than one year) RA. We screened the sera of 20 RA patients with a disease duration of less than one year, 19 RA patients with a disease duration of more than five years and 23 controls on 8268 human protein arrays. We confirmed the validity of protein array detection by ELISA assays. We then performed epitope mapping with overlapping 15mers to analyze RA sera reactivity. WIBG (within BGCN homolog (Drosophila)), GABARAPL2 (GABA(A) receptor associated protein like 2) and ZNF706 (zinc finger protein 706) proteins are preferentially recognized by autoantibodies from early RA patients. Of interest, autoantibodies to WIBG are very specific for early RA. Indeed, 33% of early RA patients' sera recognize WIBG versus 5% of RA patients with disease duration more than 5 years and 2% of controls. We identified three linear peptides on WIBG GABARAPL2 and ZNF706 that are preferentially recognized by sera of early RA patients. We identified new autoantibodies associated with RA with disease duration less than one year. These autoantibodies could be used as diagnosis markers in RA patients.
    Full-text · Article · Jul 2013
    • "PAD4 is not only involved in the generation of citrullinated epitopes; it is in itself a target for RA specific autoantibodies . Autoantibodies to PAD4 have already been described in RA141516171819. We observed that 29% of RA patients have anti PAD4 autoantibodies versus 0% of AS patients and 3% of healthy individuals (p b 10 − 7 , by chi-square test). BRAF encodes a 766 amino acid serine-threonine kinase involved in the transduction of mitogenic signals from the cell membrane to the nucleus [20,21]. "
    [Show abstract] [Hide abstract] ABSTRACT: Rheumatoid arthritis (RA) is a chronic autoimmune disease that causes cartilage and bone destruction. The mechanisms leading to RA are unknown. There is currently no reliable cure for RA. Early treatment can reduce inflammation, joint damage and bone destruction. Thus, early diagnosis of RA is critical. However, definitive diagnosis of RA can be difficult. Immunologic tests that can be performed for the diagnosis of RA include detection of anti citrullinated protein antibodies (ACPAs). However, one third of RA patients have no ACPAs. To identify new autoantibodies in RA, we used the sera of RA patients to screen protein arrays containing 8000 human proteins. We found and validated two major autoantigens: PAD4 (peptidyl arginine deiminase 4) and BRAF (v raf murine sarcoma viral oncogene homolog B1) catalytic domain. We identified peptide targets of anti PAD4 and BRAF autoantibodies. We observed that anti PAD4 are inhibitory whereas anti BRAF stimulate BRAF activity. Anti PAD4 and anti BRAF antibodies may be used to diagnose RA, particularly in the absence of anti citrullinated protein antibodies.
    Full-text · Article · Sep 2012
Show more

Recommended publications

Discover more