-
Proc. Natl. Acad. Sci. USAVol. 92, pp. 3541-3545, April
1995Immunology
Autoimmunity in Chagas disease cardiopathy: Biological
relevanceof a cardiac myosin-specific epitope crossreactive to
animmunodominant Trypanosoma cruzi antigenEDIECIO CUNHA_NETO*,
MARCIA DURANTI*, ARTHUR GRUBERt, BIANCA ZINGALESt, IARA DE
MESSIASt,NOEDIR STOLF§, GIovANNI BELLOTTI§, MANOEL E. PATARROYOI,
FuLvIo PILLEGGI§, AND JORGE KALIL*II*Laboratory of Transplantation
Immunology, §Divisions of Surgical and Clinical Cardiology,
Instituto do Coracao, Sao Paulo University Hospital, Faculty
ofMedicine, and tlnstituto de Quimica, Universidade de Sao Paulo,
Sao Paulo, SP, 05403-000, Brazil; tImmunopathology Laboratory,
Hospital de Clinicas,University of Parana, Curitiba, Brazil; and
Instituto de Inmunologia, Hospital San Juan de Dios, Universidad
Nacional de Colombia, Bogota, Colombia
Communicated by Jean Dausset, Centre d'Etude du Polymorphisme
Humain, Paris, France, October 21, 1994
ABSTRACT Heart tissue destruction in chronic Chagasdisease
cardiopathy (CCC) may be caused by autoimmunerecognition of heart
tissue by a mononuclear cell infiltratedecades after Trypanosoma
cruzi infection. Indirect evidencesuggests that there is antigenic
crossreactivity between T.cruzi and heart tissue. As there is
evidence for immunerecognition of cardiac myosin in CCC, we
searched for aputative myosin-crossreactive T. cruzi antigen. T.
cruzi lysateimmunoblots were probed with anti-cardiac myosin
heavychain IgG antibodies (AMA) affinity-purified from CCC
orasymptomatic Chagas disease patient-seropositive sera.
A140/116-kDa doublet was predominantly recognized by AMAfrom CCC
sera. Further, recombinant T. cruzi protein B13-whose native
protein is also a 140- and 116-kDa doubleband-was identified by
crossreactive AMA. Among 28 seratested in a dot-blot assay, AMA
from 100%6 of CCC sera butonly 14% of the asymptomatic Chagas
disease sera recognizedB13 protein (P = 2.3 x 10-6). Sequence
homology to B13protein was found at positions 8-13 and 1442-1447 of
humancardiac myosin heavy chain. Competitive ELISA assays thatused
the correspondent myosin synthetic peptides to inhibitserum
antibody binding to B13 protein identified the heart-specific
AAALDK (1442-1447) sequence of human cardiacmyosin heavy chain and
the homologous AAAGDK B13 se-quence as the respective crossreactive
epitopes. The recogni-tion of a heart-specific T. cruzi
crossreactive epitope, in strongassociation with the presence of
chronic heart lesions, sug-gests the involvement of crossreactivity
between cardiac my-osin and B13 in the pathogenesis of CCC.
Chagas disease (American trypanosomiasis), caused by
theprotozoan Trypanosoma cruzi, is endemic in many countries
ofLatin America, where 16-18 million people may be infected.Chronic
Chagas disease cardiopathy (CCC), a dilated cardiom-yopathy with a
T-cell-rich inflammatory infiltrate that oftenleads to a fatal
course, develops in 25-30% of infectedindividuals 5-30 yr after
primary infection (1). The remaining60-70% chronically T.
cruzi-infected individuals either remainasymptomatic
("indeterminate" patients, ASY) or developdenervation of parietal
smooth muscle in the digestive tract(5-10%).The pathogenesis of
heart lesions in CCC has been the issue
of much debate. Several lines of evidence-including theinability
to consistently find parasites in damaged areas (2) andthe passive
transfer of heart lesions by CD4+ T cells from T.cruzi-infected
mice (3)-indicated that the inflammatory heartlesion could be of an
autoimmune nature (4).According to the antigenic mimicry
hypothesis, lymphocytes
in the heart should recognize and mount delayed-type hyper-
The publication costs of this article were defrayed in part by
page chargepayment. This article must therefore be hereby marked
"advertisement" inaccordance with 18 U.S.C. §1734 solely to
indicate this fact.
sensitivity responses toward a tissue-specific heart
componentbearing structural similarities to a given T. cruzi
antigen.Several investigators found autoantibody and
self-reactiveT-cell formation in human and experimental T cruzi
infection(5, 6). Crossreactive autoantibodies were mainly
directedtoward ubiquitous and evolutionarily conserved molecules
(7,8). A recent report (9) demonstrated an epitope sharedbetween a
recombinant flagellar T. cruzi molecule (Fl 160) andan undefined
48-kDa neuronal protein. However, Fl 160-nervecrossreactive
antibodies did not correlate to CCC or chronicdigestive forms of
human Chagas disease. Most T. cruzi-hosttissue crossreactive
antigen systems described seemed to lackclinico-biological
significance, as indicated by a lack of targetorgan specificity of
self-antigen as well as lack of correlationbetween antibody
detection and different clinical forms ofchronic Chagas disease.The
recognition of myosin by CD4+ T cells in a murine
model of CCC (10) revealed a prime candidate target autoan-tigen
in CCC. Myosin heavy chain (myosin HC) is the mostabundant heart
protein-50% of total protein by weight(11)-and is recognized in
many situations of heart-specificautoimmunity, such as rheumatic
heart disease (12) andmurine post-Coxsackie B3 autoimmune
cardiomyopathy (13).The presence of antibody against cardiac myosin
HC is cor-related to the development of chronic inflammatory
cardio-pathy in T cruzi-infected mice (14). Immunization with
cardiacmyosin HC induces an aggressive myocarditis (15, 16).
Thesereports prompted us to look for anti-cardiac myosin HCimmune
reactivity in CCC patients and search for a possiblemyosin
crossreactive T. cruzi antigen.
MATERIALS AND METHODSPatients. CCC patients (severe heart
failure, dilated car-
diomyopathy with a positive serology for T. cruzi, and all
othercauses excluded) in waiting list for cardiac
transplantationwere subjected to transvenous endomyocardial biopsy.
Samplecollection procedures have been cleared by the Committee
ofEthics. Peripheral blood was collected from 24 CCC patients,as
well as from 14 asymptomatic Chagas disease patient
(ASY)individuals (clinically asymptomatic individuals seropositive
toT. cruzi, with a normal electrocardiograph and heart sonog-raphy)
from Instituto do Cora-ao and from Parana UniversityHospital
outpatient clinics, as well as from 10 normal volun-teers. Sera
from Duchenne muscle dystrophy patients wereprovided by Mayana
Zatz, University of Sao Paulo. His-
Abbreviations: CCC, chronic Chagas disease cardiopathy;
ASY,asymptomatic Chagas disease patient; myosin HC, myosin heavy
chain;AMA, anti-cardiac myosin HC IgG antibodies; NR, nonrelated
pep-tide; pB13, 19-mer that encompasses all possible linear
epitopes withinthe 12-mer tandem repeats of B13 protein.IlTo whom
reprint requests should be addressed.
3541
Dow
nloa
ded
by g
uest
on
June
11,
202
1
-
3542 Immunology: Cunha-Neto et al
topathological scoring for myocarditis of biopsy samples wasdone
as described (17).
Antigens. Purified human cardiac myosin (for ELISA), aswell as
myosin-rich 0.3 M KCl soluble fraction of hearthomogenate (for
immunoblots and immunosorbent prepara-tion), was obtained from
normal human heart ventriculartissue of an organ donor as described
(18). After SDS/7.5%PAGE (19), the myosin-rich heart lysate was
electroblotted tonitrocellulose sheets (20). Cardiac myosin
HC-containing ni-trocellulose strips (major 210-kDa protein band)
were used asimmunosorbents. Tissue culture-derived T. cruzi Y
straintrypomastigotes obtained from infected LLC-MK2 cell
linemonolayers (21) were fractionated by SDS/7.5% PAGE in
thepresence of 2-mercaptoethanol, being subsequently trans-ferred
to nitrocellulose sheets (20).Recombinant T. cruzi Antigen. The
recombinant T. cruzi
clone B13 was originally isolated from a Agtll
expressionlibrary, as described (22). The insert of B13 was
subcloned invector pMSgtll, in phase with the lacZ gene. The
expressedfusion protein B13 and ,B-galactosidase were
affinity-purifiedon p-aminophenyl-f3-D-galactopyranoside-agarose
columns.B13 recombinant protein contains 16 tandemly repeated
12-amino acid motifs. A rabbit antiserum to B13 protein wasobtained
as described (22).
Synthetic Peptides. Peptides were synthesized with the"teabag"
method (23), HPLC-purified, and sequenced. Table1 lists the
peptides used, including a 19-mer that encompassesall possible
linear epitopes (24) within the 12-mer tandemrepeats of B13 protein
(pB13) (22). In addition, 9-mer B13peptides (S1-6) were designed in
"sliding window" fashion(25) for epitope mapping. Peptides pl-20
and p1439-1453 arederived from the sequence of human cardiac myosin
HC, (3isoform, and contain B13-homologous sequences
(bold-facedtype). The nonrelated peptide (NR) represents residues
22-41of the human T-cell receptor variable-region a-subunit
13chain.
Affinity Purification of Anti-Cardiac Myosin HC Antibod-ies
(AMA). Immunoselection of specific antibodies was doneas described
(26). After adequate blocking, myosin HC im-munosorbents (50 ,ug of
myosin per strip) were incubatedovernight at 4°C with 1.5 ml of a
1:10 dilution of each serumsample. Bound antibodies were eluted
with 0.02 M glycine, pH2.8, neutralized with 1 M Tris, pH 8.6, and
reconstituted toTris-buffered saline/1% nonfat milk. Each AMA
preparationwas tested for specific anti-myosin HC IgG antibody
activity byincubation with human heart lysate
immunoblots.Immunoblots. Immunoblots were done as described
(20).
Briefly, electroblotted nitrocellulose strips were
incubatedovernight at 4°C with either AMA obtained from
individualsera or 1:40 dilutions of unfractionated sera, followed
by goat
Table 1. Synthetic peptides
PeptidesB13-derivedpB13SiS2S3S4S5S6
Human cardiac myosinHC-derivedpl-20p1439-1453
Human T-cell receptorVa chainNR
GDKPSLFGQAAAGDKPSLFGDKPSLFGQ
KPSLFGQAASLFGQAAAGFGQAAAGDK
QAAAGDKPSAAGDKPSLF
MGDSEMAVFGAAAPYLRKSENAAAAALDRKQRNFD
NKSAKQFSLHIMDSQPDGS
Va, variable region a.
anti-human IgG-alkaline phosphatase conjugate (Sigma),
anddeveloped with nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl
phosphate chromogens. T. cruzi strips were alterna-tively incubated
with a rabbit antiserum to B13 recombinantprotein, when a goat
anti-rabbit IgG-alkaline phosphataseconjugate (Sigma) was used.
Recombinant B13 protein andnonrecombinant ,B-galactosidase were
dot-blotted on nitrocel-lulose membranes with a BRL 96-well
manifold (50 ng perdot). AMA of all tested sera were incubated
overnight at 4°Cand processed as described for immunoblots.ELISA
Assays. Anti-myosin antibody assays were done
essentially as described (13). In our system, 1:200 dilutions
ofindividual sera were incubated overnight, at 4°C, in
triplicates,over purified human cardiac myosin as the solid-phase
antigen,followed by goat anti-human IgG-peroxidase conjugate
(Sig-ma) and developed with O-phenylenediamine chromogen.
Forepitope mapping of B13 protein, we adapted competitiveELISA
protocols (13, 14). Briefly, appropriate dilutions of sera(yielding
OD values in the 0.35-0.6 range) were preincubatedwith 1, 10, or
100,uM of each synthetic peptide for 1 hr at 37GC.Each
serum/peptide mixture was then incubated overnight intriplicate
wells of a B13 protein-sensitized polyvinyl chloridemicrotiter
plate (50 ng per well). The reaction was developedas described
above. The percentage inhibition of serum bind-ing to solid-phase
B13 protein by each peptide at 100 ,uMconcentration was normalized
to allow inter-test comparisonsaccording to the equation:
% inhibition = [1 - (OD of peptide
- OD of pB13/OD ofNR - OD of pB13)] x 100.
Statistical Methods. Fisher's exact test was used to
evaluatefrequency differences between different patient groups.
Stu-dent's t test was used to compare means ofOD values in
ELISAtests between different groups of serum. The MACAW program(27)
was used for protein-sequence alignment.
RESULTSPrevalence of Anti-Cardiac Myosin IgG Antibodies in
Pa-
tients' and Normal Sera. All 45 sera from CCC, ASY, andnormal
groups were reactive when tested in a purified humancardiac ELISA.
The distribution of reactivities was compara-ble, and mean ODs
(horizontal bars) in each group were notsignificantly different by
Student's t test (Fig. 1). In addition,all sera recognized the
major 210-kDa band-comigratingwith myosin HC-in human heart lysate
immunoblots (datanot shown).
1.8
1.60
1.4]-1.2-
a 1-00.8-
0.6-
0.4
0.2iIn
00
0
i
0
IS
0
I1
Ba
Normal ASY CCC(n=10) (n=11) (n=24)
FIG. 1. ELISA with purified human cardiac myosin as the
solid-phase antigen incubated with 1:200 dilution of sera from
normal, ASY,and CCC individuals. A 1:200 dilution was found to
yield OD valuesin the most informative range. Horizontal lines
represent mean values.
u YIV
Proc. Natl. Acad Sci USA 92 (1995)
Dow
nloa
ded
by g
uest
on
June
11,
202
1
-
Proc. Natl. Acad. Sci. USA 92 (1995) 3543
A B
a b
kDa
210
140116
97
68
FIG. 2. (A) Immunoblot with T. cruzi trypomastigote lysate
incu-bated with AMA purified from a CCC serum (strip a) and a
rabbitantiserum to B13 recombinant protein (strip b); molecular
massmarkers (kDa) are shown at left. Arrowheads indicate
140/116-kDabands. (B) Dot-blots with 50 ng of B13 recombinant
protein incubatedwith AMA immunoselected from 14 CCC sera (strip a)
and 14 ASYsera (strip b). The positions where protein and antibody
samples wereapplied are indicated with dashes.
Identification of a Cardiac Myosin HC Crossreactive T.cruzi
Antigen. Trypomastigote lysate immunoblots wereprobed with AMA
individually purified from 23 CCC and 11ASY sera. A doublet of
bands of 140 and 116 kDa (Fig. 2A,strip a) was recognized by AMA
from 14 of 23 (61%) CCCpatients, in contrast to 1 of 11 (9%) ASYs.
The difference incrossreactive recognition among AMA from CCC and
ASYsera is statistically significant (P = 5.1 X 10-3 , Fisher's
exacttest). AMA from normal individuals did not recognize
thedoublet of T. cruzi bands. Furthermore, AMA from
chagasicpatients or normal subjects do not bind to LLC-MK2 cell
lineantigens in immunoblots (data not shown).
Fig. 2A, strip b, shows the same 140/116-kDa doubletrecognized
by a rabbit antiserum to the recombinant T. cruziprotein B13 (22).
We then investigated whether AMA fromchagasic patients recognized
B13 immobilized in dot blots. Fig.2B indicates that AMA from 14 of
14 CCC sera (100%) and2 of 14 ASY sera (14%) give a positive
reaction. The differencein binding to B13 between AMA from CCC and
ASY sera ishighly significant (P = 2.3 x 10-6, Fisher's exact
test). Cross-reactivity in CCC AMA was not secondary to higher
levels ofantimyosin or anti-B13 antibody levels, as they were
compa-rable in both CCC and ASY unfractionated sera groups,
asassessed by ELISA tests (data not shown). AMA from alimited
number of ASY individuals with myocarditis at en-domyocardial
biopsy also recognized B13 in dot-blots (data notshown).
,B-Galactosidase, the fusion protein support, was notrecognized by
AMA from any of the sera.Mapping of the Immunodominant and Cardiac
Myosin
HC-Crossreactive B13 Epitopes. To identify regions of pri-mary
sequence homology between B13 and cardiac myosinthat could harbor a
crossreactive epitope, the sequence of theB13 repetitive motif was
aligned (27) to the f3 isoform ofhuman cardiac myosin HC, the
prevalent isoform in humanventricle (28). Two hexapeptide regions
of 80% homologywere disclosed, at positions 8-13 (VFGAAA vs.
B13FGQAAA) and 1442-1447 (AAALDK vs. B13 AAAGDK),respectively, of
cardiac myosin HC. Synthetic peptides span-ning the B13-homologous
sequences in cardiac myosin HC(pl-20 and p1439-1453) were incubated
with CCC and ASYsera in competitive ELISA experiments with B13 as
thesolid-phase antigen. Fig. 3A shows that p1439-1453, but not
the pl-20 peptide, competes for binding of the
anti-B13antibodies in serum of a CCC patient in a
dose-dependentfashion. On the other hand, neither myosin peptide
competesfor binding of the antibodies in a ASY serum (Fig.
3B).Apart from the myosin-crossreactive epitope, the B13 pro-
tein repeat motif is known to contain an immunodominant T.cruzi
epitope, recognized by unfractionated sera from 97% ofT.
cruzi-infected individuals irrespective of disease manifesta-tion
(22, 29). To map this immunodominant epitope, synthetic9-mer
peptides derived from B13 were also tested. Fig. 5indicates that
only the 9-mer S4 peptide, apart from the 19-merpB13 peptide,
completely inhibits the binding of both CCCserum (Fig. 4A) and ASY
serum (Fig. 4B) to B13 recombinantprotein. The p1439-1453 myosin
peptide, along with pB13 andS4 B13-derived peptides, consistently
inhibits the binding ofCCC anti-B13 antibodies in six CCC sera in
B13 ELISA (Fig.5). Three other ASY sera were tested, corroborating
that onlyB13-derived peptides pB13 and S4 inhibit ASY serum
antibodybinding to B13 (data not shown). At low dilutions
(1:40-1:160),sera from 10 normal controls and 10 cardiopathic
Duchennemuscular dystrophy patients displayed OD readings in
therange of 0.2-0.3. Such OD readings in B13 ELISA could notbe
inhibited by preincubation with peptides p1439-1453 andpB13, thus
demonstrating that Duchenne muscle dystrophypatient sera and normal
sera are devoid of specific anti-B13antibodies (data not
shown).
DISCUSSIONIn this study, we identify the myosin crossreactive T.
cruziantigen B13, and we have mapped the crossreactive epitope toa
cardiac myosin-specific 6-amino acid region. The almost
0.40.35
0.3
g 0.2n +
0.1
0.05
0
0.6 _
0.5
n A
0 n 'AL
0.2
0.1
A
p10NR
bpl-20
1 J,M 10,lM
F- .;E-
\pl439-1453
lOOulLM
B
NR
------- -~ pl-20
p1439 -1453
pB1310,lM 100 ,uM
FIG. 3. Competitive ELISA for B13 protein. Different
concentra-tions of synthetic peptides derived from human cardiac
myosin HC ,3isoform (pl-20 and p1439-1453), B13 protein (pB13), and
humanT-cell receptor variable-region a chain (NR) were preincubated
withserum from CCC patient WA (A) or ASY CA (B). Each mixture
wasadded to wells of a microtiter plate sensitized with B13
protein. pB13and NR stand as positive and negative competition
controls, respec-tively. Resulting OD values are displayed for each
peptide concentra-tion.
_
0.25 -1- -- " 17'N-,
34
IU.15
jk_
l l~i
u.-4
.. II
Immunology: Cunha-Neto et at
.f:
fl
31.
)0-
U.J,
Dow
nloa
ded
by g
uest
on
June
11,
202
1
-
3544 Immunology: Cunha-Neto et at
n A-
0.41"0.35-0.3-
0 0.25-0 0.2-
0.15-
0.1As AC
A
oS1-- S2
--cS3-4.-54
-S6-x- pB13--- NR
K`>
NR
_ x pB13Ul-
1 pLM 10 AM
1 ,M 10 ,uM
100 JIM
pB13
100 ,uM
FIG. 4. Competitive ELISA for B13 protein. Different
concentra-tions of synthetic peptides (S1-S6 and pB13, compare with
Table 1)covering B13 linear epitopes, as well as peptides derived
from T-cellreceptor variable-region a chain (NR), were preincubated
with serumfrom CCC patient WA (A) or from ASY CA (B). Assays were
doneas described in the Fig. 3 legend.
exclusive presence of cardiac myosin HC-B13
140/116-kDacrossreactive antibodies in CCC sera (Fig. 2) suggests
thatcardiac myosin HC-B13 crossreactive recognition may
bebiologically relevant to the pathogenesis of CCC.The universal
reactivity of sera from CCC, ASY, and normal
controls against human cardiac myosin both in immunoblotsand in
ELISA (Fig. 1) indicates the existence of naturalantimyosin
antibodies. Although "natural" anti-cardiac myo-sin antibodies have
already been reported in mice (30) andhumans (31), our data
establish their finding at high preva-lence. Our results show a
qualitative difference between"cardiopathy-related" anti-myosin
antibodies in CCC sera and"natural" anti-myosin antibodies in ASY
sera (Fig. 2), match-ing the findings in experimental murine
infection (14). Thepresence of crossreactive antibodies in sera
from ASYs withsubclinical Chagas myocarditis may indicate that such
anti-bodies are associated to Chagas myocarditis per se rather
thanto extensive heart damage. It is tempting to speculate
whetherthe two ASYs whose sera displayed crossreactive antibodies
in
human cardiachuman cardiac ahuman adult skeleletalrabbit smooth
musclehuman nonmuscle
w
m.2a._2._
Sl S2 S3 S4 S5 S6 pl-20p1439-1453
FIG. 5. Epitope mapping of the B13 repeat motif.
Competitiveinhibition of the binding of anti-B13 antibodies in B13
ELISA bysynthetic peptides at 100 ,uM concentration. The percentage
ofinhibition of serum antibody-binding solid-phase B13 protein
wascalculated according to the equation in Materials and Methods.
Inhi-bition values for each peptide were averaged from six CCC sera
afternormalization.
Fig. 2B, strip b, could also present subclinical Chagas
myo-carditis.The presence of two distinct epitopes in the B13
repeat
motifs was suspected in face of the findings of
restrictedrecognition of B13 by AMA from CCC sera (Fig. 2B)
andunrestricted recognition of an "immunodominant" epitope
byunfractionated serum in B13 ELISA serodiagnosis tests (22,29).
Epitope mapping with competitive inhibition of B13ELISA (Figs. 3-5)
suggested the existence of two distinctspecificities of B13-binding
antibodies: (i) antibodies inhibitedby B13 peptides pB13 or S4,
found in unfractionated sera fromall T. cruzi-infected individuals
tested and possibly responsiblefor the binding of B13 protein in
serodiagnosis tests, recog-nizing the immunodominant epitope
contained in B13 peptideS4; and (ii) antibodies inhibited by the
cardiac myosin peptidep1439-1453, found mainly in CCC sera and
representing theantibodies that recognize B13 protein and the
140/116-kDadoublet in AMA preparations, indicating the
sequencesAAALDK in myosin and AAAGDK in B13 as the
respectivecrossreactive epitopes. B13 peptide S5 (QAAAGDKPS)
wasunable to inhibit antibody binding to B13 in CCC sera
(Figs.3-5), which may indicate that residues neighboring the
cross-reactive hexapeptide also play a role in antibody
reactivity.The characterization of the p1439-1453 myosin
peptide
containing the AAALDK hexapeptide homologous to B13AAAGDK as the
crossreactive epitope established a structuralbasis for myosin-B13
crossreactivity. The AAALDK 1442-1447 myosin sequence is strictly
conserved only among cardiacmyosin HC isoforms (Fig. 6). Thus, the
B13-crossreactivemyosin epitope recognized by serum antibodies of
CCC pa-tients is cardiac myosin HC-specific. The fact that the
cross-reactive epitope in B13, AAAGDK, is embedded into
theimmunodominant epitope FGQAAAGDK (S4 peptide) ispuzzling, as ASY
sera recognize the latter and ignore theformer. This finding
suggests that production of cardiac myosin
1428IEDLMVDVERSNAAAAALDDEKQRNFDKILAE1458
V . . .. I .. . .T . . .C .. .. . .. T . .. .. .. .
LD. .V. .LDNQRQLVSN.E ... KK. .L...
LD. .L. .LDHQRQS.CN.E .. .KK. .QL ...
FIG. 6. Alignment of the 1428-1458 region of human cardiac
myosin HC containing the B13-crossreactive epitope with other
tissue-specificmyosin HC isoforms. Genenbank peptide locus codes
are as follows: human cardiac ,B myosin HC, HUMBMYH7CD_1; human
cardiac a myosinHC, HUMCAMHC 1; human adult skeletal myosin HC,
HSMYH1R 1; rabbit uterus smooth muscle myosin HC, S68021 1; and
human nonmusclemyosin HC, HUMMYONM_1.
0
V.-4J
39E-m- ----- \o~-- Z:Z_ ]
.,. 4i.!-
., I .-1
I
Proc. Natl. Acad Sci. USA 92 (1995)
Dow
nloa
ded
by g
uest
on
June
11,
202
1
-
Proc. Natl. Acad Sci. USA 92 (1995) 3545
HC-B13 crossreactive antibodies in CCC may be secondaryto
selective recognition of the crossreactive sequencesAAAGDK/AAALDK,
under immune response gene control.Thus, it can be hypothesized
that crossreactive recognition-and the attending heart damage of
CCC-would only occur inT. cruzi-infected individuals endowed, for
instance, with per-missive HLA alleles. Family clustering (32) and
HLA associ-ation (33, 34) studies support a possible role for
immunoge-netic components in the differential susceptibility of
the20-30% of T. cruzi-infected individuals who develop CCC.
Theabsence of crossreactive antibodies in cardiopathic
Duchennemuscular dystrophy patients' sera (data not shown)
suggeststhat such antibodies may not be formed as a consequence
ofheart damage alone. However, it cannot be formally discardedthat
heart damage in a T. cruzi-infected individual might raisesuch
crossreactive antibodies.
In spite of the 100% association with CCC patients,
directparticipation of the myosin-B13 crossreactive antibody in
CCCpathogenesis is not likely. The bulk of evidence supports
theparticipation of T cells rather than antibody in heart damage(2,
3, 5, 10, 35). The association of cardiac myosin
HC-B13crossreactive antibodies with CCC may be only signaling
thepresence of "helper" T cells of similar specificity, these
onespotentially involved in heart tissue damage. The
identificationof a defined heart-specific epitope (AAALDK) of a
majorheart protein (cardiac myosin HC) that is antigenically
mim-icked by a secondary epitope (AAAGDK) of an immunodom-inant T.
cruzi antigen (B13) is in itself a corollary of the theoryof
autoimmune pathogenesis of CCC. The demonstration thatcrossreactive
antibody recognition is mainly restricted to CCCpatients defines
human cardiac myosin HC and T. cruzi proteinB13 as an antigen pair
of potential relevance to the pathogen-esis of heart lesions in
human CCC. Specific treatment withanti-T cruzi drugs may not abort
the progression of CCCamong T. cruzi-infected individuals (36).
Thus, alternativeantigen-specific immunosuppression approaches
blocking thedeleterious autoimmune response without interfering
withantiparasite immunity may help control heart damage inChagas
disease. Further, our findings may allow the search foreffective
subunit vaccines devoid of components of T. cruzithat bear the
pathogenic heart-crossreactive determinants.
We thank Drs. A. C. Pereira-Barreto and B. Ianni for
allowingaccess to patients. We are indebted to Dr. A. C. Goldberg
for criticallyreviewing the manuscript. This work has been
supported by grantsfrom the Brazilian National Research Council
(CNPq) and the SpecialProgram for Research and Training in Tropical
Diseases/WorldHealth Organization.
1. Rassi, A., Tranchesi, J. & Tranchesi, B. (1982) in
DoenqasInfecciosas e Parasitdrias, ed. Veronesi, R. (Guanabara
Koogan,Rio de Janeiro, Brazil), 7th Ed., pp. 675-712.
2. Torres, C. M. (1930) An. 5a Reun. Soc. Argent. Pat. Reg.
Norte 2,902-916.
3. Ribeiro-dos-Santos, R., Rossi, M.-A., Laus, J. L., Santana
Silva,J., Savino, W. & Menguel, J. (1992) J. Exp. Med. 175,
29-39.
4. Andrade, Z. (1958) Rev. Goiana Med. 4, 103-119.5. Teixeira,
A., Teixeira, M. L. & Santos-Buch, C. (1975) Am. J.
Pathol. 80, 163-180.6. Cossio, P. M., Diez, C., Szarfman, A.,
Kreutzer, E., Candiolo, B.
& Arana, R. M. (1974) Circulation 49, 13-21.7. Levin, M. J.,
Mesri, E., Benarous, R., Levitus, G., Schijman, A.,
Torres, H. & Segura, E. (1989) Am. J. Trop. Med. Hyg.
41,530-538.
8. Kerner, N., Liegeard, P., Levin, M. J. & Joskowicz, M. H.
(1991)Exp. Parasitol. 73, 451-459.
9. Van Voorhis, W., Schelekewy, L. & Trong, H. (1991) Proc.
Natl.Acad. Sci. USA 88, 5993-5997.
10. Rizzo, L. V., Cunha-Neto, E. & Teixeira, A. (1989)
Infect.Immun. 57, 2640-2644.
11. Harrington, W. F. & Rodgers, M. E. (1984)Annu. Rev.
Biochem.53, 35-73.
12. Dale, J. B. & Beachey, E. H. (1986) J. Exp. Med. 164,
1785-1790.13. Neu, N., Beisel, K., Traytsman, M., Rose, N. &
Craig, S. (1987)
J. Immunol. 138, 2488-2492.14. Tibbetts, R. S., McCormick, T.
S., Rowland, E. C., Miller, S. D.
& Engman, D. M. (1994) J. Immunol. 152, 1493-1499.15. Neu,
N., Rose, N., Beisel, K., Herskowitz, A., Gurri-Glass, G. &
Craig, S. (1987) J. Immunol. 139, 3630-3636.16. Smith, S. &
Allen, P. (1991) J. Immunol. 147, 2141-2147.17. Higuchi, M. L.,
Floriano de Morais, C., Pereira Barreto, A. C.,
Lopes, E. A., Stolf, N., Bellotti, G. & Pileggi, F. (1987)
Clin.Cardiol. 10, 665-669.
18. Margossian, S. & Lowey, S. (1982) Methods Enzymol. 85,
55-71.19. Laemmli, U. K. (1970) Nature .(London) 227, 680-685.20.
Towbin, H., Staehelin, L. & Gordon, J. (1979) Proc. Natl.
Acad.
Sci. USA 76, 4350-4354.21. Zingales, B., Andrews, N. W.,
Kuwajima, V. Y. & Colli, W.
(1982) Mol. Biochem. Parasitol. 6, 111-124.22. Gruber, A. &
Zingales, B. (1993) Exp. Parasitol. 76, 1-12.23. Houghten, R. A.
(1985)Proc. Natl. Acad. Sci. USA 82, 5131-5135.24. Kabat, E. A.
(1970) Ann. N.Y Acad. Sci. 169, 43-54.25. Geysen, H. M., Rodda, S.
J., Mason, T., Tribbick, G. & Schoofs,
P. (1987) J. Immunol. Methods 102, 259-274.26. Tovey, E. &
Baldo, B. (1989) in Protein Blotting, eds. Baldo, B.,
Tovey, E. & St. Leonards, N. (Karger, Basel), pp. 43-64.27.
Schuler, G. D., Altschul, S. F. & Lipman, D. J. (1991)
Proteins
Struct. Funct. Genet. 9, 180-190.28. Gorza, L., Mercadier,
J.-J., Schwartz, K., Thornell, L. E., Sartore,
S. & Schiaffino, S. (1984) Circ. Res. 54, 694-702.29.
Tropical Disease Research/World Health Organization (1990)
Mem. Inst. Oswaldo Cruz 85, 487-543.30. Berneman, A., Ternynck,
T. & Avrameas, S. (1992) Eur. J.
Immunol. 22, 625-633.31. Cunningham, M. W., McCormack, J. M.,
Talaber, L. R., Harley,
J. B., Ayoub, E. M., Muneer, R. S., Chun, L. T. & Reddy, L.
V.(1988) J. Immunol. 14, 2760-2766.
32. Zicker, F., Smith, P. G., Almeida Netto, J. C., Oliveira, R.
M. &Zicker, E. M. S. (1990) Am. J. Trop. Med. Hyg. 43,
498-505.
33. Layrisse, Z., Acquatella, H., Piccone, C., Catalioti, C.
& Gomez,J. R. (1988) Hum. Immunol. 23, 116 (abstr.).
34. Ascurra, M., Pistilli, N., Otha, N. & Rosner, J. (1990)
Mem. Inst.Oswaldo Cruz. 85, Suppl. 1, 78 (abstr.).
35. Khoury, E. L., Diez, C., Cossio, P. M. & Arana, R. M.
(1983)Clin. Immunol. Immunopathol. 27, 283-288.
36. Teixeira, A., Cunha-Neto, E., Rizzo, L. & Silva, R.
(1990) J.Infect. Dis. 162, 1420.
Immunology: Cunha-Neto et at
Dow
nloa
ded
by g
uest
on
June
11,
202
1