| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Division of Endocrinology, Faculty of Medicine, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
secretion
in recall assays in vitro, but contained nondominant epitopes. All
immunogenic peptides were confirmed as Ak binders because
peptide-specific LNC proliferation was blocked by an
Ak-specific mAb, but not by a control mAb. Peptide-specific
serum IgG was induced only by p2102 and p2596, but these Abs did not
bind to intact mouse Tg. This study reaffirms the predictive value of
Ak-binding motifs in epitope mapping and doubles the number
of known pathogenic T cell determinants in Tg that are now found
scattered throughout the length of this large autoantigen. This
knowledge may contribute toward our understanding of the pathogenesis
of autoimmune thyroiditis. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Evidence supporting the view that Tg is a major autoantigen in clinical or experimental thyroiditis dates back to 1956 (11, 12). The molecule is quite abundant in the thyroid gland, representing 75–80% of the total protein in the thyroid extract (13), and it can be easily isolated by gel filtration chromatography. These features, however, did not aid in the identification of pathogenic T cell epitope sites by biochemical methods, as this was hampered by the large size of the molecule (homodimeric mass = 660 kDa). In addition, previous T cell epitope mapping efforts could not avail themselves of the mouse Tg (mTg) gene sequence information and used either the human (14) or the partial rat (15) Tg sequence data in EAT studies, relying on the high sequence homology among Tg from different species. Despite these difficulties, over the last decade and via a variety of methods, five immunopathogenic Tg peptides have been identified, encompassing at least six distinct T cell epitopes (16).
The use of Tg peptides as model Ags in EAT generated an impetus for studying the immunoregulation of the disease (17). At the same time, the emerged map of the pathogenic Tg epitopes focused attention on new issues that remain unresolved. First, none of the known pathogenic Tg peptides appear to comprise an immunodominant epitope, because these peptides cannot be generated after processing of intact Tg by APCs in vitro (16, 17). The experimental evidence, however, clearly suggests the presence of dominant Ak-binding epitopes(s) within Tg, as supported by the known genetic control of the Tg-mediated EAT by the I-Ak locus (8, 10) and the prevention of EAT by treatment of mice with Ak-specific Abs (18). Second, four of the five known pathogenic peptides are clustered toward the C-terminal end of Tg, raising the question of whether the rest of this large molecule can contain additional EAT-causing epitopes.
In this study, we proceeded to do a systematic search of the complete mTg sequence (19) for the detection of dominant and/or additional EAT-causing T cell epitopes by using an algorithm (20) that searches for Ak-binding motifs within a protein sequence. This algorithm takes into account the physicochemical characteristics and structural properties of amino acids within motifs that are shared among immunogenic Ak-binding peptides. The study was undertaken in the EAT-susceptible strain CBA/J (H-2k), but suggests an approach that can be applied equally well to EAT-susceptible strains of other H-2 haplotypes or to other thyroid autoantigens.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Female CBA/J (H-2k) mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and were used in experiments at 6–8 wk of age. Tg was extracted from thyroids of outbred ICR mice as previously described (21). Briefly, frozen glands (Bioproducts for Science, Indianapolis, IN) were homogenized in phosphate buffer, pH 7.0, and the supernatant was centrifuged three times at 16,000 x g. Tg was obtained from the supernatant by gel filtration using Sepharose CL-4B (Pharmacia, Baie d’Urfé, Quebec, Canada). The fractions of peak II were pooled, concentrated to 3–5 mg of PBS, filter sterilized, and stored at -20°C until use. Tg concentrations are expressed as the molarity of the monomeric form (330 kDa). The Tg peptide (2495–2511) GLINRAKAVKQFEESQG (p2495) was synthesized at the Alberta Peptide Institute (Edmonton, Alberta, Canada), whereas all other Tg peptides were synthesized by Sigma-Genosys (The Woodlands, TX). All peptides were blocked with an acetyl group at the N-terminal and with an amide group at the C-terminal, whereas the thiol group of internal Cys residues was blocked by acetamide. All peptides were used in experiments at >80% purity.
T cell activation assays
Mice were immunized s.c. under ether anesthesia with 100 nmol of peptide or 100 µg of Tg in 100 µl of 1:1 PBS/CFA (with Mycobacterium butyricum; Difco Laboratories, Detroit, MI) emulsion. Nine days later, the inguinal, brachial, and axillary lymph nodes were collected aseptically and single cell suspensions were prepared in DMEM supplemented with 10% FBS (Cansera, Ontario, CA), 20 mM HEPES buffer, 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin (all from Life Technologies, Rockville, MD), and 5 x 10-5 M 2-ME (Sigma-Aldrich, St. Louis, MO). After centrifugation and washing, cells (4 x 105 cells/200 µl/well) were cultured in the presence or absence of Ag in flat-bottom 96-well plates and incubated for 4 days at 37°C in a 10% CO2, 90% air-humidified incubator. Eighteen hours before harvesting, 1 µCi of [3H]thymidine (6.7 Ci/mmol; DuPont, Mississauga, ON, Canada) was added to each well in 25 µl of complete medium. The cells were harvested using a Harvester 96 Mach III M (Tomtec, Hamden, CT), and incorporated radioactivity was measured using the TopCount NXT microplate counter (Canberra Packard Canada, Mississauga, ON, Canada). Stimulation index (S.I.) is defined as follows: (cpm in the presence of peptide)/(cpm in the absence of peptide). mAbs were purified by affinity chromatography on protein G-Sepharose 4 Fast Flow columns (Pharmacia) from culture supernatants of the hybridomas 10-3.6.2 (IgG2a) reactive with I-Ak (22) and H16-L10-4R5 (IgG2a) specific for the influenza A nucleoprotein (23) purchased from the American Type Culture Collection (Manassas, VA). Inhibition of proliferation was performed at 10 µg/ml final concentration of the blocking mAb, a dose that has been previously shown to be nontoxic and effective in the blockade of T cell hybridoma clones (24). The data are expressed as follows: % inhibition = [1 - (cpm in the presence of mAb)/(cpm in the absence of mAb)] x 100.
Detection of cytokines and peptide-specific IgG by ELISA
Cytokine production was determined in culture supernatants
harvested after 48-h stimulation of lymph node cells (LNCs) with Ag
(8–20 µM). Detection of IL-2, IL-4, IL-10, and IFN- was performed
by sandwich ELISA based on noncompeting pairs of capture and detection
(biotinylated) mAbs as follows: IL-2, JES6-1A12, and JES6-5H4; IFN-,
R4-6A2, and XMG1.2 (BD PharMingen, San Diego, CA); IL-4, 11B11
(American Type Culture Collection), and DVD-6-24G2 (BD PharMingen).
IL-10 was detected via the use of affinity-purified polyclonal rabbit
Ab 500-P60 and 500-P60Bt (PeproTech, Rocky Hill, NJ). Alkaline
phosphatase-conjugated streptavidin was purchased from Sigma-Aldrich.
Standard curves were generated for each individual cytokine using known
amounts of murine rIL-2 and rIFN- (BD PharMingen) or rIL-4 and
rIL-10 (PeproTech). The detection limits were 4 pg/ml for IL-2 and
IFN-, 10 pg/ml for IL-4, and 17 pg/ml for IL-10. The presence of
peptide-specific IgG in pooled sera was determined by ELISA as
previously described (21), using an alkaline
phosphatase-conjugated goat anti-mouse IgG (Sigma-Aldrich) as the
second Ab. Light absorption of the p-nitrophenolate product
at 405 nm was measured using a Vmax plate reader (Molecular Devices,
Sunnyvale, CA).
Induction and histological assessment of EAT
Mice were challenged s.c. with 100 nmol of each peptide in CFA emulsion and were boosted 21 days later with 50 nmol of peptide in IFA. EAT was assessed 35 days after the initial challenge. Adoptive transfer of thyroiditis was performed as previously described (25). Briefly, LNCs from peptide-primed donor mice were cultured for 72 h in the presence of 20 µM of the respective peptide. The cells were then harvested, and after washing three times, 2 x 107 cells in 200 µl of PBS were injected i.p. into syngeneic recipients (six mice per group). EAT was assessed 14 days post-transfer. Fixation, embedding, and sectioning of thyroids were performed as previously described (25). Histological sections were stained with H&E, and the mononuclear cell infiltration index (I.I.) was scored as follows: 0, no infiltration; 1, interstitial accumulation of cells between two or three follicles; 2, one or two foci of cells at least the size of one follicle; 3, extensive infiltration 10–40% of total area; 4, extensive infiltration 40–80% of total area; and 5, extensive infiltration >80% of total area.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The complete mTg sequence (19) was scanned for the
presence of two Ak-binding motifs, a heptamer
motif A and a pentamer motif B (Table I
),
by using the algorithm described by Altuvia et al. (20).
This computerized method was developed following a compilation of an
extended database of helper T cell sites and takes into account
physical-chemical and structural properties of peptides (dictated by
the primary amino acid sequence) that may be responsible for binding to
MHC class II Ags. Sixty-nine and 47 sites containing motif A or B,
respectively, were identified (data not shown). To maximize the chances
for success, we focused our attention to thirteen peptides encompassing
completely overlapping motifs A and B flanked by 4 aa (Table I). Of
these, peptides (224–238) and (228–242) identified an overlapping
motif-rich site, thus prompting the synthesis of a single peptide
(226–239) containing this region. The peptides (824–838) with three
proline residues flanking the motifs, and (837–851) with a proline
residue inside a motif, were not considered for further study because
of concerns that Pro may drastically affect the secondary
structure. Peptides (2490–2504) and (2543–2557) were shown previously
to contain EAT-causing T cell epitopes (21, 26, 33). With
these considerations in mind, we proceeded to synthesize the eight
motif A and B-containing peptides (p110, p226, p306, p1579, p1826,
p2026, p2102, and p2596; Table I) and examine their immunopathogenic
properties.
|
To examine the immunogenicity of the Tg peptides, CBA/J mice were
challenged s.c. at two sites on the back with 100 nmol of each peptide.
Nine days later, the inguinal, axillary, and brachial LNCs were
collected and cultured in the presence of varying concentrations of the
immunizing peptide. Five of eight Tg peptides, p306, p1579, p1826,
p2102, and p2596 (Fig. 1A),
induced significant and specific LNC proliferation because there was no
detectable response against the control peptide p2495. This reactivity
profile correlated well with the capacity of the same peptides to
elicit IL-2 (Fig. 1B) and/or IFN- (Fig. 1C)
release from such activated LNCs. IL-4 and IL-10 were undetectable in
all culture supernatants (data not shown), indicating that these
peptides—with the exception of p2102, which elicited IL-2 but not
IFN- release—activated Th1 cells. In all cultures, peptide-specific
proliferation was significantly blocked by an
I-Ak-specific mAb, but not by a control
nucleoprotein-specific mAb (Fig. 2), strongly suggesting that recognition
of these five immunogenic peptides occurred in the context of
Ak molecules. These results confirmed the
predictive value of the algorithm and the identification of five new Tg
peptides encompassing Ak-restricted T cell
epitopes.
|
|
To examine whether any of the five immunogenic peptides contain
dominant T cell determinants, CBA/J mice were s.c. primed with 100 µg
of intact Tg in CFA as above, and 9 days later the draining LNCs were
cultured in the presence of Tg or free peptide. As shown in Fig. 3, Tg-primed LNCs responded strongly to
Tg in vitro but failed to respond to equimolar (0.2–0.9 µM range)
concentrations of free peptide. This lack of responsiveness was
observed even at higher concentrations of free peptide (up to 75 µM),
thus excluding the possibility that a narrow range of peptide
concentration might influence the results. Conversely, LNCs primed in
vivo with any of the five immunogenic peptides did not respond to
intact Tg in recall assays in vitro (data not shown). These results do
not support the view that these immunogenic peptides encompass dominant
T cell epitopes.
|
Subsequently, CBA/J mice (six per group) were immunized with each
of the five immunogenic peptides above in CFA and were boosted, 3 wk
later, with the same peptide in IFA. Five weeks from the initial
challenge, thyroid glands were removed for histological examination of
EAT development, and the I.I. was scored as described in
Materials and Methods and representatively shown in Fig. 4. With the exception of p2102, which did
not elicit detectable pathology, all other peptides induced mild and
variable EAT, with mean I.I. varying from 0.2 (p1826) to 1.2 (p306)
(Table II). In contrast, adoptive
transfer of peptide-primed LNCs into syngeneic CBA/J hosts increased
both the incidence and the severity (I.I. 0.7–2.2) of EAT in all
groups, including the one challenged with p2102 (Table II), confirming
that all five immunogenic peptides have thyroiditogenic potential.
Intrathyroidal homing of mononuclear cells was specific because
analogous infiltration was not observed in liver or kidney samples of
these mice (data not shown). Peptide pathogenicity did not correlate
with the presence of peptide-specific IgG in pooled sera of CBA/J mice
with EAT, because only p2102 and p2596 elicited IgG responses (Fig. 5). In addition, peptide-reactive IgG did
not appear to bind to intact Tg (Fig. 5), suggesting either that p2102
and p2596 are not expressed on the surface of Tg or that they adopt in
free form a conformation different from the one they assume within the
intact Tg molecule.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), which is known to
contain three families of homologous domains with cysteine-rich motifs
(14, 28). Interestingly, p306 is localized within the I.4
domain with type 1 homology, whereas p1579 as well as p2102 are
encompassed, respectively, in the IIIa.1 and IIIa.3 domains that
express type 3 homology (Table III). The
fifth pathogenic peptide, p2596, maps in the vicinity of the
(2549–2560) epitope, which encompasses a thyroxine molecule at
position 2553 (26).
|
|
). Thus, it is
difficult to predict from such data how many additional
Ak-binding peptides are harbored within Tg, and
it is likely that alternative approaches might still identify other
EAT-causing determinants. Interestingly, none of the five pathogenic Tg
peptides described herein overlap with seven putative
I-Ek-binding peptides uniquely expressed in
murine Tg, as identified by the EpiMer algorithm (31). In
contrast, p306 and p2102 are being indicated as sites that contain one
of five Ek-binding motifs according to the
algorithm of Altuvia et al. (20). The current data also bring into focus the fact that none of the seven known Ak-restricted pathogenic Tg peptides—five from the present study, plus the (2499–2507) and (2549–2560) epitopes (24, 26)—can be classified as immunodominant because they cannot be generated after processing of intact mTg in vivo and/or in vitro. This is a rather paradoxical finding in view of the fact that Tg-induced EAT is under control of the I-Ak locus (8, 10), implying the presence of dominant Ak-restricted T cell epitopes in this large autoantigen. The following explanations can be proposed to account for this apparent discrepancy. First, one could maintain that immunodominant T cell epitopes in Tg exist but remain unidentified because 1) the sheer size of Tg prevents their easy detection, 2) such epitopes do not contain hormonogenic sites but are normally iodinated, and 3) detection is precluded by limitations of algorithm-based approaches in epitope mapping (here and in Refs. 21 and 32) or inherent restrictions in using cloned Tg-reactive T cell hybridomas for screening the antigenicity of overlapping Tg peptides, a method followed by Champion et al. (26) for the discovery of the pathogenic (2549–2560) sequence. Second, the theoretical possibility exists that all Ak-restricted Tg peptides interact with MHC with similar affinity and activate T cell precursors of low frequency, thus not allowing a clear-cut hierarchy of immunodominance to emerge. In this case, Ak-controlled susceptibility to EAT or reactivity to mTg in vitro would be detected as a result of additive or synergistic effects of Tg peptide-specific T cell clones, which nevertheless would remain individually undetectable due to their low frequency. Third, and most likely, the conventional criteria for the definition of immunodominance might not apply to Tg. This is an explanation we have previously elaborated on (16), as it is conceivable that in vitro processing of this large autoantigen does not normally generate enough of any given epitope to activate peptide-specific lymphoid cells. For example, optimal processing of 100 µg/ml (150 nM) of intact dimeric Tg would not generate >0.6 µg/ml (300 nM) of any given 2-kDa peptide.
For most Tg peptides tested in this study, the adoptive transfer
protocol led to EAT with higher incidence and severity than that
induced after direct challenge of hosts with peptide in CFA. These data
might be explained on the basis of a peptide dose constraint on direct
EAT induction (100 nmol of peptide may be a relatively low dose given
the nondominant nature of these Ags), vis-à-vis the capacity of
preformed peptide-specific effector T cells to home selectively to the
thyroid after the adoptive transfer protocol. In analogy with the
peptide (2549–2560) previously identified by Roitt and colleagues
(33), the pathogenicity of p2102 was shown only by
adoptive transfer of p2102-primed LNCs to syngeneic naive CBA mice,
confirming the view (16) that the thyroiditogenicity of
candidate peptides should not be examined only by direct challenge of
host mice with the respective epitope. We cannot explain why peptide
p2102 does not elicit detectable levels of IFN- in culture. However,
its potential to elicit mild EAT via the adoptive transfer protocol
indicates that p2102 activates a very low number of Th1 cells and/or
other types of effector cells that can infiltrate the thyroid. Indeed,
previous studies have shown that mice lacking the IFN- receptor gene
are able to develop disease upon challenge with Tg (34),
and adoptive transfer of Tg-activated T-cells along with an
IFN--specific Ab into naive hosts results in the induction of
granulomatous EAT (35).
The present findings support the notion that Tg encompasses many noniodinated T cell epitopes, which can be cryptic but pathogenic under conditions that allow their generation and presentation in professional APCs. In that regard, we have previously shown that processing of highly iodinated Tg (36) or Tg-Ab complexes (37) in APCs allows selective presentation of pathogenic but cryptic Tg peptides. These mechanisms would promote epitope spreading (38, 39) and rapid emergence of thyroiditogenic T cells because it is unlikely that peripheral tolerance would have been pre-established against cryptic Tg determinants (39). Some epitopes may even promote expansion of autoreactive T cells via molecular mimicry with foreign pathogens, as has been shown previously with the pathogenic Tg (2695–2706) determinant that exhibits high homology with the (368–381) peptide of murine adenovirus type 1 sequence (25). In contrast, an immunoregulatory role for peptide-reactive IgG Abs is quite unlikely for the Tg sequences described herein, because only p2102 and p2596 elicited IgG responses, and these Abs did not bind to intact Tg.
In conclusion, the present study has established the immunopathogenic potential of five novel Tg peptides in CBA/J mice out of a list of eight candidate sequences that contain Ak-binding motifs. The overall importance of all sequences in mouse EAT can be further evaluated in future studies using H-2k strains of diverse non-H-2 backgrounds and/or strains of other H-2 haplotypes. Work with previously identified pathogenic Tg epitopes, such as the 9-mer (2496–2504), which binds to nonisotypic I-Ek and I-As determinants (24), and the 11-mer (2549–2559), which induces EAT in CBA/J but not in DBA/1 (H-2q) or SJL (H-2s) mice (40), suggests that the outcome of such studies is impossible to predict. However, our findings demonstrate a promising approach for the identification of immunopathogenic Tg peptides with MHC class II-binding properties in EAT that can be analogously applied in the search of Tg epitopes relevant to human thyroiditis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. George Carayanniotis, Faculty of Medicine, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada, A1B 3V6. E-mail address: gcarayan{at}mun.ca
3 Abbreviations used in this paper: EAT, experimental autoimmune thyroiditis; Tg, thyroglobulin; mTg, mouse Tg; S.I., stimulation index; LNC, lymph node cell; I.I., infiltration index.
Received for publication June 19, 2002. Accepted for publication August 30, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
receptor gene. Eur. J. Immunol. 28:201.[Medline]
antibody. J. Immunol. 149:2219.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  H. P. Ng and A. W. C. Kung Induction of Autoimmune Thyroiditis and Hypothyroidism by Immunization of Immunoactive T Cell Epitope of Thyroid Peroxidase Endocrinology, June 1, 2006; 147(6): 3085 - 3092. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. S. Li and G. Carayanniotis Iodination of Tyrosyls in Thyroglobulin Generates Neoantigenic Determinants That Cause Thyroiditis J. Immunol., April 1, 2006; 176(7): 4479 - 4483. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Verginis, H. S. Li, and G. Carayanniotis Tolerogenic Semimature Dendritic Cells Suppress Experimental Autoimmune Thyroiditis by Activation of Thyroglobulin-Specific CD4+CD25+ T Cells J. Immunol., June 1, 2005; 174(11): 7433 - 7439. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. D. Dai, P. Eliades, K. A. Carayanniotis, D. J. McCormick, Y.-c. M. Kong, V. Magafa, P. Cordopatis, P. Lymberi, and G. Carayanniotis Thyroxine-Binding Antibodies Inhibit T Cell Recognition of a Pathogenic Thyroglobulin Epitope J. Immunol., March 1, 2005; 174(5): 3105 - 3110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. W. Purcell and J. J. Gorman Immunoproteomics: Mass Spectrometry-based Methods to Study the Targets of the Immune Response Mol. Cell. Proteomics, March 1, 2004; 3(3): 193 - 208. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
,
Peptides identified in the present report; numbers denote amino acid
coordinates within the mTg molecule (
, no asterisk: Peptides identified in
previous studies with their amino acid coordinates as originally
published and in agreement with the mTg sequence according to Ref.