The Regions on the Light Chain of Botulinum Neurotoxin Type A Recognized by T Cells from Toxin-Treated Cervical Dystonia Patients. The Complete Human T-Cell Recognition Map of the Toxin Molecule.

We have recently mapped the in vitro proliferative responses of T cells from botulinum neurotoxin type A (BoNT/A)-treated cervical dystonia (CD) patients with overlapping peptides encompassing BoNT/A heavy chain (residues 449-1296). In the present study, we determined the recognition profiles, by peripheral blood lymphocytes (PBL) from the same set of patients, of BoNT/A light (L) chain (residues 1-453) by using 32 synthetic overlapping peptides that encompassed the entire L chain. Profiles of the T-cell responses (expressed in stimulation index, SI; Z score based on transformed SI) to the peptides varied among the patients. Samples from 14 patients treated solely with BoNT/A recognized 3-13 (average 7.2) peptides/sample at Z > 3.0 level. Two peptide regions representing residues 113-131 and 225-243 were recognized by around 40% of these patients. Regarding treatment parameters, treatment history with current BOTOX® only group produced significantly lower average T-cell responses to the 32 L-chain peptides compared to treatments with mix of type A including original and current BOTOX®. Influence of other treatment parameters on T-cell recognition of the L-chain peptides was also observed. Results of the submolecular T-cell recognition of the L chain are compared to those of the H chain and the T-cell recognition profile of the entire BoNT/A molecule is discussed. Abbreviations used: BoNT/A, botulinum neurotoxin type A; BoNT/Ai, inactivated BoNT/A; BoNT/B, botulinum neurotoxin type B; CD, cervical dystonia; L chain, the light chain (residues 1-448) of BoNT/A; LNC, lymph node cells; H chain, the heavy chain (residues 449-1296) of BoNT/A; HC, C-terminal domain (residues 855-1296) of H chain; HN, N-terminal domain (residues 449-859) of H chain; MPA, mouse protection assay; SI, stimulation index (SI = cpm of 3H-thymidine incorporated by antigen-stimulated T cells/cpm incorporated by unstimulated cells); TeNT, tetanus neurotoxin; TeNTi, inactivated TeNT.

Influences of HLA DRB1, DQA1 and DQB1 on T-cell recognition of epitopes and of larger regions of the botulinum neurotoxin molecule.

Previously, we have examined the proliferative responses of T-cells from 25 patients and 8 controls to 32 light chain (L1-L32) and 60 heavy chain peptides (N1-N29, C1-C31) representing the entire clostridium botulinum neurotoxin type A (BoNT/A)[OM1-OM3]. In the current work, these T-cell responses were analyzed in the context of the patients HLA-DRB1, DQA1 and DQB1 variation. There were strong associations between the DQA1*01:02 and its derived haplotypes and cumulative T-cell proliferative responses. With or without HLA based differentiation the responses showed marked correlation. Inter-epitope correlation of responses demonstrably associated with particular regions (peptides N1-N29) peaking in the region covered by of N18-29. A second region of higher correlation was observed close to the carboxyl terminal of the heavy chain. Region N15 to N29 was found to have a significantly lower Stimulation Indices when DRB1*01:01-DQA1*01:01-DQB1*05:01 was present. This pattern was also evident in the HLA analysis where DQA1*01:02 associations were collectively most significant in the N1-N13 & C16-C31 region. Responses in these regions correlated well with one another. HLA-specific correlation analysis revealed that DQA1*01:01 bearing haplotype had the strongest inter-epitope correlations despite having a generally negative association with simulation indices. Structural and immunogenic implications of these findings are discussed.

Decreased bone mineral density in experimental myasthenia gravis in C57BL/6 mice.

Experimental autoimmune myasthenia gravis (EAMG), an animal model of myasthenia gravis (MG), can be induced in C57BL/6 (B6, H-2 b) mice by 2-3 injections with Torpedo californica AChR (tAChR) in complete Freund's adjuvant. Some EAMG mice exhibit weight loss with muscle weakness. The loss in body weight, which is closely associated with bone structure, is particularly evident in EAMG mice with severe muscle weakness. However, the relationship between muscle weakness and bone loss in EAMG has not been studied before. Recent investigations on bone have shed light on association of bone health and immunological states. It is possible that muscle weakness in EAMG developed by anti-tAChR immune responses might accompany bone loss. We determined whether reduced muscle strength associates with decreased bone mineral density (BMD) in EAMG mice. EAMG was induced by two injections at 4-week interval of tAChR and adjuvants in two different age groups. The first tAChR injection was either at age 8 weeks or at 15 weeks. We measured BMD at three skeletal sites, including femur, tibia, and lumbar vertebrae, using dual energy X-ray absorptiometry. Among these bone areas, femur of EAMG mice in both age groups showed a significant decrease in BMD compared to control adjuvant-injected and to non-immunized mice. Reduction in BMD in induced EAMG at a later-age appears to parallel the severity of the disease. The results indicate that anti-tAChR autoimmune response alone can reduce bone density in EAMG mice. BMD reduction was also observed in adjuvant-injected mice in comparison to normal un-injected mice, suggesting that BMD decrease can occur even when muscle activity is normal. Decreased BMD observed in both tAChR-injected and adjuvant-injected mice groups were discussed in relation to innate immunity and bone-related immunology involving activated T cells and tumour necrosis factor-related cytokines that trigger osteoclastogenesis and bone loss.

Injection of inactive Bordetella pertussis and complete Freund's adjuvant with Torpedo californica AChR increases the occurrence of experimental autoimmune myasthenia gravis in C57BL/6 mice.

An animal model of myasthenia gravis (MG), termed experimental autoimmune MG (EAMG), is an important tool for investigations of disease mechanisms and/or methods of treatment for this disease. EAMG can be induced in C57BL/6 (B6, H-2b) mice by 2-3 times injections at 4 weeks intervals with Torpedo californica (t) acetylcholine receptor (AChR) in complete Freund's adjuvant (CFA). However, the protocol especially with a two-injection schedule occasionally produces a poor incidence of EAMG. We have investigated the efficacy of the additional adjuvant, inactive organisms of Bordetella pertussis (iBP), on the induction with a two-injection schedule. In a group immunized with tAChR in CFA + iBP, 76% of mice developed EAMG (average grade in exercise test, 1.02). Whereas, 46% of mice were found EAMG-positive (average grade, 0.73) in a group injected with tAChR/CFA alone. Thus, the combined use of CFA and iBP significantly increased both the occurrence and severity of clinical MG in the immunized mice. This was accompanied by higher antibody (Ab) and T-cell responses to tAChR. The effect on disease occurrence of the iBP use in a three-injection protocol was also described.

Antibody responses to botulinum neurotoxin type A of toxin-treated spastic equinus children with cerebral palsy: A randomized clinical trial comparing two injection schedules.

We have conducted a 26-month-long comparative study involving young patients (2-6years old) with a clinical diagnosis of spastic equinus secondary to cerebral palsy who have been treated with BoNT/A (BOTOX®, Allergan) tri-annually or annually. Serum samples were obtained to determine the presence or absence of blocking antibodies (Abs) by a mouse protection assay (MPA) and levels of anti-BoNT/A Abs by radioimmune assay (RIA). HLA DQ alleles were typed using blood samples to determine the possible association of certain HLA type(s) with the disease or with the Ab status. Blocking Abs were detected in only two out of 18 serum samples of the tri-annual group, but none were found in 20 samples of the annual group. The MPA-positive serum samples gave in RIA significantly higher anti-BoNT/A Ab-binding levels than the MPA-negative samples. On the other hand, when two MPA-positive sample data were excluded, serum samples from tri-annual and annual groups showed similar anti-BoNT/A Ab levels. Linkage of the disorder with a particular HLA DQA1 and DQB1 allele types was not observed due to the small sample size. However, by combining results with other studies on BoNT/A-treated Caucasian patients with cervical dystonia (CD), we found that, among Caucasian patients treated with BoNT/A, DQA1*01:02 and DQB1*06:04 were higher in Ab-positive than in Ab-negative patients. The genetic linkage was on the threshold of corrected significance.

Submolecular recognition regions of the HN domain of the heavy chain of botulinum neurotoxin type A by T cells from toxin-treated cervical dystonia patients.

We have recently reported the submolecular T-cell recognition profile of the C-terminal half (HC, residues 855-1296) of the heavy (H) chain of botulinum neurotoxin type A (BoNT/A) with peripheral blood lymphocytes (PBL) from 25 BoNT-treated cervical dystonia (CD) patients. In the current study, we describe the mapping of the T-cell responses of the patients to the N-terminal half (HN, residues 449-859) of the heavy chain by using 29 synthetic overlapping peptides encompassing the entire HN domain of BoNT/A. The profiles of the T-cell responses to the peptides varied among the patients. Samples from 14 patients treated solely with BoNT/A recognized 1-9 (average 3.7) peptides/sample at Z>3.0 level. Three peptide regions representing residues 631-649, 659-677 and 743-761 were frequently recognized by 29-64% of the patients. In patients with positive anti-BoNT/A antibody responses the overall positive T cell responses to the HN peptides were significantly increased compared to antibody-negative patients. Influence of treatment parameters on the T-cell recognition of the HN peptides was also observed. The results were compared with those of previously identified HC region.

Submolecular recognition of the C-terminal domain of the heavy chain of botulinum neurotoxin type A by T cells from toxin-treated cervical dystonia patients.

We determined the T-cell proliferative responses of the peripheral blood lymphocytes (PBL) from 25 botulinum neurotoxin (BoNT)-treated patients to 31 overlapping synthetic peptides encompassing the C-terminal half (residues 855-1296) of BoNT/A heavy chain. Responses of PBL to HC peptides varied among patients. Samples from 14 patients treated solely with BoNT/A recognized 2-13 (average 6.4) peptides/sample at Z>3.0 level. Six peptide regions representing residues 855-873, 1023-1041, 1051-1069, 1093-1111, 1135-1153 and 1247-1265 were frequently recognized by 36-57% of these PBLs. Influence of treatment parameters on T-cell recognition of the peptides was also investigated.

Regions recognized on the light chain of botulinum neurotoxin type A by T lymphocytes of SJL and BALB/c mice primed with inactivated toxin.

Lymph node cells (LNC) from SJL (H-2(s)) and BALB/c (H-2(d)) mice primed once with inactivated botulinum neurotoxin type A (BoNT/A) were examined for their T-cell responses to each of 32 synthetic overlapping peptides (19 residues each, L1-L32) that encompass the entire L chain (residues 1-448) of BoNT/A. LNC of SJL gave strong responses to 6 regions on, L2 (residues 15-23), L10/11/12 (127-173), L19 (253-271) and L21 (281-299), and moderate to weak responses to L9 (113-131), L14/15 (183-215) and L27 (365-383). In BALB/c, LNC gave a substantial T-cell response only against peptide L12 (residues 155-173), and responded very weakly to 9 other peptides. The results were compared with the recognition profiles determined previously in these two strains after multiple BoNT/A injections. Overall responses to the L-chain peptides of T cells in later profiles were found to be somewhat weakened in SJL and stayed essentially at a similar level in BALB/c, although responses to BoNT/A increased. In SJL, response to L10 (127-145) remained the highest in the later profile. Strong responses against L12 (155-173) observed in both strains at early stage were reduced to an insignificant level. Cross-reactivity to tetanus neurotoxin by BoNT/A-specific T cells was observed in SJL but not in BALB/c. Design of an effective synthetic peptide vaccine will require incorporation of both T cell- and Ab-recognition elements of the BoNT molecule. Significance and possible implications of these results on BoNT/A-specific T-cell responses of BoNT-treated patients are discussed.

T-cell recognition of acetylcholine receptor provides a reliable means for monitoring autoimmunity to acetylcholine receptor in antibody-negative myasthenia gravis patients.

Myasthenia gravis (MG) is an autoimmune disease usually associated with autoantibodies (auto-Abs) against nicotinic acetylcholine receptor (AChR). Some MG patients appear negative for anti-AChR Abs (seronegative), and a fraction of these have auto-Abs against muscle-specific kinase. The remaining patients, although displaying MG symptoms, show no detectable auto-Abs. We describe here a possible association of a rare human leukocyte antigen (HLA)-DQ type and AChR Ab-negative MG. We also found that the majority of seronegative patients exhibit an anti-AChR autoimmune T lymphocyte response. We investigated the existence of AChR-reactive T cells in peripheral blood lymphocytes from seronegative patients by their proliferative responses against a mixture of 18 overlapping synthetic peptides encompassing the extracellular part of human AChR α-chain. Of the 10 samples, eight exhibited positive T-cell proliferative responses against the peptide mixtures. The proliferative assay was equally efficient using a mixture of eight peptides frequently recognized by MG T cells. This T-cell proliferative assay should provide a reliable method for monitoring seronegative MG patients.

Antibody and T cell recognition of the light chain of botulinum neurotoxin A in two high-responder mouse strains.

We investigated in two inbred mouse strains the submolecular recognition of botulinum neurotoxin type A (BoNT/A) by Abs (B cells) and by T lymphocytes. For mapping, we employed a set of overlapping synthetic peptides that encompassed the entire light (L) chain of BoNT/A. After 3 BoNT/A toxoid injections, BALB/c T cells responded in vitro to challenge by peptides L18 (residues 239-257), L23 (309-327), L27 (365-383), L29 (393-411), or L31 (421-439) and more weakly to peptides L3 (29-47), L9/L10 (113-145), L15 (197-215), L17 (225-243), or L26 (351-369). The other peptides stimulated little or no T cell responses. SJL mice mounted, after 3 BoNT/A injections, stronger T cell responses that were medium-to-strong to peptides L2/L3 (15-47), L10/L11 (127-159), L19 (253-271), or L23 (309-327) and low to peptides L17 (225-243), L21 (281-299), L27 (365-383), or L30/L31 (407-439). After 3 BoNT/A injections, BALB/c and SJL antisera protected mice against lethal BoNT/A doses, but displayed restricted epitope profiles compared to outbred (ICR) mice Abs. BALB/c Abs displayed medium-to-high binding to peptides L4/L5 (43-75), L10/L11 (127-159), L18 (239-257) or L27 (365-383). SJL Abs were high to peptides L4/L5 (43-75), L14 (183-201), L16 (211-229), or L18/L19 (239-271), and medium to peptides L10 (127-145), L11 (141-159), L12 (155-173) or L29 (393-411). The other peptides had little or no binding. Responses to each T cell or Ab epitope were under separate genetic control. T and B (antibody) cell recognition regions may coincide, but there were also regions recognized only by Abs or by T cells.

Human T-cell responses to botulinum neurotoxin. Responses in vitro of lymphocytes from patients with cervical dystonia and/or other movement disorders treated with BoNT/A or BoNT/B.

We have previously reported that botulinum neurotoxin type A (BoNT/A)-specific T-cell responses occur in a majority of patients treated with botulinum neurotoxins (BoNT). In this study, we first determined if T-cell responses against BoNT/A and tetanus toxin (TeNT) differ between cervical dystonia (CD) patients and other movement disorder cases. Secondly, we have examined in CD cases the treatment parameters that may have an effect on the T-cell responses against BoNT/A. We found that T-cell responses to BoNT/A were significantly higher in patients with CD than in those with other movement disorders. An increase in TeNT T-cell response in CD was observed when compared to un-treated controls. CD patients who were injected with BoNT/B mounted higher responses to BoNT/A than patients treated with BoNT/A only. Frequent injections (more than 2.1/year) were associated with a significantly higher T-cell response to BoNT/A in CD. T cell responses to BoNT/A did not differ between CD patients who had clinically responsive and non-responsive status at the time of enrollment.

Human T-cell responses to botulinum neurotoxin: proliferative responses in vitro of lymphocytes from botulinum neurotoxin A-treated movement disorder patients.

We determined the T-cell responses against botulinum neurotoxin type A (BoNT/A) and tetanus toxin (TeNT) of peripheral blood lymphocytes from 95 BoNT-treated patients and 63 non-treated control subjects. The patient group included 80 cervical dystonia and 15 other movement disorder cases. Positive T-cell responses to BoNT/A were detected in 70% of the treated patients, and in only 3% of controls. T-cell responses of BoNT-treated patients against BoNT/A did not differ between patients who were clinically responsive and those who had become non-responsive to the treatment. BoNT-treated patients gave significantly higher in vitro T-cell responses to TeNT than did the controls.

Association of HLA Class II alleles and haplotypes with cervical dystonia: HLA DR13-DQ6 (DQB1*0604) homozygotes are at greatly increased risk of cervical dystonia in Caucasian Americans.

An unanticipated discovery was made while examining genetics of the immune response in patients treated with botulinum neurotoxin (BoNT), which included cervical dystonia (CD) patients. Initial examination of HLA DQA1:DQB1 frequencies revealed an unexpectedly high number of DQA1*0102:DQB1*0604 homozygotes (hz) in the CD patients. We typed the BoNT-treated CD Caucasian subset for HLA-DRB1, DQA1, and DQB1 and succeeded in typing HLA-DRB1, -DQA1, and -DQB1 for 75 of the patients. Two statistical methods found the DQB1 locus associated with CD and one method found a probable association of DQB1*0604. Examination of the allele and haplotype pairing indicated that DQB1*0604 hz comprised most to all of the positive association. Other than this genotype, one other allele, DQB1*0504 contributes to the association of the DQB1 locus. These findings indicate a probable infectious and/or autoimmune component in some CD patients. However, longer distance associations within an extended and conserved DQB1*0604 bearing haplotype leave a possibility that a locus proximal to DQB1 might be involved.

Targeting the antigen-binding site of HLA-restricting alleles in treatment of autoimmune disease.

In autoimmune disease, production of disease-causing auto-antibodies (Abs) depends on autoreactive T cells that recognize the epitopes of the pathogenic antigen in the context of MHC class II molecules. It is possible that selective inhibition of an antigen-presenting function of disease-associated MHC alleles could lead to suppression of the disease. Myasthenia gravis (MG) is a disabling neuromuscular disease in which autoimmune responses against acetylcholine receptor (AChR), especially against the alpha chain of AChR, cause a postsynaptic defect. HLA linkage of MG has been thus far best detailed for DQB1. Recently, we have shown that certain DQ haplotypes are associated with presentation of AChR alpha-chain peptides in MG. In a mouse model for MG, which can be induced in disease-susceptible C57BL/6 (B6, H-2b) mice by injection with Torpedo AChR, region 62-76 of I-Ab beta chain is involved in the disease mechanism. Monoclonal Abs (mAbs) against synthetic peptide I-Abetab62-76, which localizes at the rim of the antigen-binding site of I-Ab, inhibited in vitro proliferation of disease-associated T cells. Passive transfer of these mAbs as well as vaccination with this peptide strongly suppressed occurrence of clinical MG in B6 mice. In both cases, Ab and T-cell responses against AChR, especially those related to disease pathogenesis, also decreased. mAbs against peptides from the ridge of the antigen-binding region of the correlate DQB1 sequences inhibited in vitro the proliferation of AChR-specific T cells from MG patients. The results indicated that the function of disease-associated MHC alleles may be blocked by directly and selectively targeting the antigen-presenting region on these MHC molecules. The strategy could provide an effective means for immunointervention in other autoimmune and allergic responses.

Subtle differences in HLA DQ haplotype-associated presentation of AChR alpha-chain peptides may suffice to mediate myasthenia gravis.

The HLA DQA1 and DQB1 alleles were determined on a set of 24 myasthenia gravis patients that had previously been examined for their T-cell proliferative responses to the 18 overlapping peptides representing the extracellular domain of hAChR alpha-chain. Patient responses according to assumed cis or trans haplotypes were significantly higher in most cases relative to normal controls. Comparisons of in vitro peptide-stimulated T-cell responses of patient pairs which had DQA1:DQB1 in common displayed responses in tighter distribution relative to comparisons in which patient pairs did not share the same DQA1:DQB1 haplotype. Similar haplotypes, such as DQA1*0102:DQB1*0602 and DQA1*0102:DQB1*0604, tended to exhibit similar responses and were grouped according to this similarity. Modified F-test and Student's T-test analyses on DQ isoform bearing groups revealed that high responses to peptide alpha34-49 were associated with A1*0102:B1*0602/0604, A1*0301:B1*0302 and A1*0401/0303:B1*0301. Peptide alpha146-162 showed higher responses in A1*0301:B1*0302 group and moderate responses in A1*0401/0303:B1*0301 groups. Differences in the age of disease onset relative to DQ haplotypes were also observed. Groups of A1*0301:B1*0302, A1*0501:B1*0201 and A1*0102:B1*0604 showed earlier ages of disease onset relative to those of A1*0102:B1*0602 or A1*0505:B1*0301.

Use of alum and inactive Bordetella pertussis for generation of antibodies against synthetic peptides in mice.

We have investigated the efficacy of the combined use of Alum and inactive Bordetella pertussis (iBP) adjuvants for eliciting anti-peptide antibodies. ICR mice were immunized four times at 3-week intervals with each of 7 free (i.e., not conjugated to any carrier) synthetic peptides of 15-17 amino acid residues in Alum + iBP, in the commonly used adjuvant protocols (CFA; CFA (initial) followed by IFA), or in CFA + iBP. Serum samples after 3 and 4 injections were tested by RIA. Use of Alum + iBP greatly increased the production of antibodies for most of the peptides. The results have important implications for human vaccine formulation involving peptides.

Myasthenia gravis induced in mice by immunization with the recombinant extracellular domain of rat muscle-specific kinase (MuSK).

UNLABELLED: Myasthenia gravis (MG) is mostly caused by anti-acetylcholine receptor (AChR) auto-antibodies (Abs). Such Abs are undetectable in 10-15% of MG patients, but many have anti-muscle-specific kinase (MuSK) Abs. We injected recombinant rat-MuSK extracellular domain in H-2(a), H-2(b), H-2(bm12) and H-2(d) mice. Certain strains exhibited exercise-induced fatigue, tremors, weight loss, and some died after 2-3 injections. Compound muscle action potentials showed decrement with low-frequency repetitive nerve stimulation. Miniature endplate potentials decreased, suggesting lower numbers of endplates functional AChRs. Myasthenic sera inhibited agrin-induced AChR aggregation in C2C12 myotubes. CONCLUSION: Anti-MuSK Abs induce MG, which might also result from blocking the agrin-signaling pathway.

Vaccination with a MHC class II peptide in Alum and inactive pertussis strongly ameliorates clinical MG in C57BL/6 mice.

We have investigated the efficacy of immunization against peptides from predisposing MHC class II molecules in human-compatible adjuvants for ameliorating experimental autoimmune myasthenia gravis (EAMG). C57BL/6 mice were immunized three times with the peptide I-Abetab62-76 in Alum+killed pertussis organisms (PT) prior to two injections with tAChR. The treatment greatly reduced the occurrence and severity of clinical MG relative to controls that received saline/Alum+PT or none. It also reduced antibody and T-cell responses against tAChR. The results have important implications for the possible immunotherapy of MG by targeting disease-associated MHC.

Suppression by mAbs against DQB1 peptides of in vitro proliferation of AChR-specific T cells from myasthenia gravis patients.

It has been indicated that multiple genes, including HLA genes, are collectively involved in the susceptibility to myasthenia gravis (MG). DQB1 alleles represent one of those associated with MG. We have prepared B-cell hybridomas that produce mAbs against peptides corresponding to the tip of the MHC antigen-binding cavity (region 70-90) of alleles DQB1*02, *03, *05 and *06. The mAbs bound to DQ molecules isolated from cells. In the assays using peripheral blood lymphocytes (PBL) from patients with MG, the mAbs against peptides of the correlate HLA DQ sequences inhibited the in vitro proliferation of acetylcholine receptor (AChR)-specific T cells. The results indicate that the function of disease-related MHC alleles may be blocked by directly and selectively targeting the antigen-presenting region on these MHC molecules. The results also suggest that DQ molecules are one of those involved in the restriction of autoimmune anti-AChR responses in MG. The strategy could provide an effective means for immunointervention in MG. It may also potentially be adapted for down-regulation of undesirable immune responses such as in other autoimmune diseases, allergic reactions, or clinical conditions where immune responses to a therapeutic protein develop.

Responses in vitro of peripheral blood lymphocytes from patients with myasthenia gravis to stimulation with human acetylcholine receptor alpha-chain peptides: analysis in relation to age, thymic abnormality, and ethnicity.

Peripheral blood lymphocytes (PBLs) were isolated from 24 patients with myasthenia gravis of three ethnic groups (Caucasian, African American, and Hispanic) and ten healthy individuals. We determined the in vitro proliferative responses of the PBL samples to each of 18 overlapping synthetic peptides corresponding to the entire main extracellular domain (residues 1-210) of the alpha-subunit of human acetylcholine receptor. The profiles of the T-cell responses (expressed in stimulation index [SI]) to the peptides varied among the 24 patient samples. There was a significant difference in the overall patient responses relative to controls toward 17 of 18 peptides. T cells from the patients gave responses greater than control mean SI + 4 standard deviation (Z(SI) > 4) to 2 approximately 9 peptides/sample. Six peptides, alpha 23-38, alpha 34-49, alpha 78-93, alpha 122-138, alpha 146-162, and alpha 182-198, were recognized with Z > 4 level by 42% to 58% of the patients' PBLs. The grouped patient responses, divided according to age, thymic diagnosis, or ethnicity, were compared with controls and with each other. Significant differences were observed between early- and late-onset cases in recognition of residues alpha 34-49 (p = 0.015) and alpha 78-93 (p = 0.053), and in recognition of residues alpha 12-27, alpha 56-71, alpha 134-150, and alpha 146-162 (0.0072 < p < 0.064) when two ethnic groups were compared with each other.