Autoimmunity in Scleroderma
489
DNA at the N-terminus, and forming a
homodimer at the C-terminus.
Epitope mapping of CENP-B were con-
ducted to deduce the important anti-
genic regions for autoimmunity reactions
against human centromere. Bacterial re-
combinant proteins expressed from a set
of truncated CENP-B cDNAs were used
for immunoblotting analysis. From the
C-terminus to the N-terminus, four rep-
resentative epitopes have been shown,
epitopes I to IV. Epitopes III and IV
locate in the N-terminal 73 residues cor-
responding to the N-terminal half of the
DNA binding domain. They are overlap-
ping since the N-terminal boundary of
epitope III has not yet been precisely
determined. Interestingly, the inhibitory
effect of the patients’ sera containing an-
tibodies against epitopes III and/or IV
on the DNA binding activity was shown
in vitro
, while no inhibitory effect was
observed with ACA-positive sera without
reactivities to these epitopes. Epitope III
plus IV (1–73) reacts with
75% of the
ACA-positive sera. In contrast, almost all
ACA sera reacted with the region span-
ning amino acids 535 to 599, the epitope
I, which corresponds to the region re-
quired for its homodimerization. In order
to restrict a minimal region of epitope I,
further studies for epitope mapping were
performed but no shorter reactive region
in epitope I was identi±ed. The epitope
I, a major epitope of CENP-B, is used as
a substrate of ELISA for ACA for clini-
cal use.
CENP-B is a highly conserved protein
through species. The similarity of the
amino acid sequences of CENP-B between
mouse and human are 96%. Amino acid
sequences of the epitopes I, III, and IV
are identical between the two species. In
contrast, the epitope II (438–492), which
is only reactive to the minority of ACA,
is called a ‘‘human speci±c’’ epitope be-
cause the antibody to this region did not
react to mouse centromere. Actually, this
epitope has lower homology to amino
acid sequences for CENP-B between hu-
man and mouse. This fact supported an
idea that the autoantibodies were derived
from stimulation of the autoantigen it-
self, consistent with the ‘‘antigen driven
theory.’’ In addition, the antigenic pleural
sites in a single molecule also supported
this theory. For ordinary antibodies’ and
antigens’ interaction, 15 to 24 residues of
the antigen participate in antibody bind-
ing. As for the epitopes I and II, we
could not limit the region below the 30 to
40 amino acid length. These results sup-
por
tedtheideatha
tsomeconforma
t
iona
l
structures in the epitopes were required
for the interaction between the antibodies
and antigens.
In order to con±rm the importance
of protein conformation for antigenic-
ity, we tried a DNA immunoprecipitation
method for quantitative assays of ACA.
A
32
P radio-labeled 59 bp oligonucleotide
probe, including the 17 bp motif of the
CENP-B box, bound with cellular puri-
±ed CENP-B proteins was immunopre-
cipitated and the radioactivity was used
for the estimation of anti-CENP-B/DNA
complex–antibody activity. Even though
no ACAs reacted to only DNA molecules,
a few sera without reactivities to CENP-B
in immunoblotting methods did immuno-
precipitate the CENP-B/DNA complex.
Such a discrepancy of results between the
DNA immunoprecipitation method and
the immunoblotting as well as that rela-
tively long stretches of amino acids were
required for epitope mapping led us to an
idea that the conformational structure was
critical for the epitope targeted by some
autoantibodies.
previous page 489 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 491 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off