mAb epitope on human thyroglobulin
Eur. J. Immunol. 1992. 22: 315-319
315
Mireille Henryo, Eric ZanelliO, Martine Piechaczyk., Bernard Pau. and Yves MalthihryO
A major human thyroglobulin epitope defined with monoclonal antibodies is mainly recognized by human autoantibodies*
Laboratoire de Biochimie MCdicale, INSERM U38, FacultC de MCdecineO, Marseille and Laboratoire d’Immunologie, FacultC de Pharmacie., Montpellier
The antigenic nature of 15 anti-human thyroglobulin (hTg) monoclonal antibody (mAb) epitopes was studied by two different approaches. First, we tested two successive protease-digest products of hTg. Only four mAb from the same cluster of reactivity recognized a low-molecular weight peptide, the other mAb only bound native hTg or high-molecular weight digest fractions. Second, these 15 mAb were used to immunoscreen hTg expression libraries. Only the same four mAb revealed immunoreactive clones corresponding to region 1149-1295 on the hTg primary sequence. After subcloning, this antigenic determinant was reduced to a 102-amino acid peptide (hTg region 1149-1250). The two different methodologies were coherent and complementary, and demonstrated that hTg sequence 1149-1250 is the target for this cluster of four mAb. Moreover, anti-hTg autoantibodies which cross-reacted with these mAb bound the 102-amino acid peptide. This epitope was the one most frequently detected by sera from autoimmune thyroid disease. The data confirm the presence of an immunodominant domain in the central part of the hTg molecule and suggest that this mAb epitope may be a powerful probe for the diagnosis of autoimmune thyroid disorders.
1 Introduction Human thyroglobulin (hTg) is a high-molecular weight glycoprotein which supports the synthesis and storage of thyroid hormones [l].Autoantibodies (aAb) to hTg are classically found with high titers in the sera from patients with Graves’ disease, Hashimoto’s thyroiditis [2] and other thyroid disorders such as thyroid carcinoma [3], but also in healthy patients [4]. By criss-cross inhibition with 15 anti-hTg mouse mAb, Piechaczyk et al. [ S ] identified six clusters of reactivity which reflected the presence of at least six antigenic domains.These authors observed that only the mAb from cluster 11 (mAb 3, 6, 10 and 15) cross-reacted with anti-hTg aAb [6]. In previous report [7] using rabbit polyclonal antibodies we have located seven epitopes and an immunodominant domain in the middle part of the hTg molecule. Each immunoreactive peptide tested with autoimmune sera revealed a heterogeneity of aAb response. We reported here the simultaneous screening with mAb [5] of a double-enzymatic digest of a highly purified hTg, and hTg cDNA expression libraries. We concluded that a large number of conformational or discontinuous epitopes were recognized by these mAb. Only mAb from cluster I1 recognized low-molecular weight digested pep-
[I 95981
*
This work was supported by grants from the Centre National de la Recherche Scientifique (U.A. 178) and the Institut National de la SantC et de la Recherche Medicale (U.38).
Correspondence: Mireille Henry, Laboratoire de Biochimie MCdicale, INSERM U38, Faculte de MCdecine, 27, Boulevard Jean Moulin, F-13385 Marseille cedex 5, France Abbreviations: hTg: Human thyroglobulin aAb: Autoantibodies AITD: Autoimmune thyroid disease AP-mAb: Alkaline phosphatase-labeled mAb pEXT Colony transformed with wild-type plasmid
0 VCH Verlagsgescllschaft mbH, D-6940 Weinheim, 1992
tides and a recombinant epitope expressed in hTg cDNA libraries. The conformation of this determinant seems important for mAb recognition, but it was restored in the fusion protein. The mAb antigenic determinant, exactly localized on hTg peptidic primary sequence, conserved its reactivity for aAb in the sera from patients with autoimmune thyroid disease (AITD).
2 Materials and methods 2.1 Double hTg digestion and gel filtration analysis Fifty milligrams of hTg (purified by C. Marriq from thyroid goiter) in 0.1 M ammonium hydrogen carbonate, pH 7.8 was digested by 1/50 (w/w) staphylococcal V8 protease (Miles Scientific, Naperville, IL), for 23 h at 37°C. Enzymatic reaction was stopped by adding 1 PM PMSF, 6 M urea (final concentration). The digest was resolved by HPLC gel filtration on TSK G 3000 SW column (Pharmacia LKB, Uppsala, Sweden) with PBS as running buffer (0.04 ml/min). Collected fractions corresponding to each peak were pooled and concentrated. Then, endoproteinase LysC (Boehringer Mannheim, Mannheim, FRG) was added in the second peak (P2) according to Jeckel et al. [8] and was allowed to proceed for 8 h at 37°C. The reaction was stopped by addition of aprotinin (Boehringer) for 1 h at 25”C, and urea ( 5 M in final concentration). The hTg fragments were separated by gel filtration on Ultrogel AcA 54 (IBF,Villeneuve-la-Garenne, France) with PBS, 5 M urea as running buffer. Convenient fractions were pooled and dialyzed in PBS. Ultrogel ACA 54 was calibrated with gel filtration LMW calibration kit (Pharmacia LKB). 2.2 Anti-hTg mAb
The mAb were produced and purified as described [ 5 ] . Lyophilized mAb were dissolved in PBS at 1 mg/ml, and used at 1/1000 dilution in immunoscreening and Western OO14-2980/92/O202-0315$3.50+ .25/O
316
M. Henry, E. Zanelli, M. Piechaczyk et al.
blotting, 1150000dilutionin hTg ELISA, and 1/500in fusion protein ELISA.
2.3 Autoimmune sera Sixteen sera from patients with AITD, three with thyroid cancer were obtained. Their titers in anti-hTg antibodies were over 1164000, as determined by passive hemagglutination with commercial kits (Wellcome, Dartfasd, GB). One serum without anti-hTg Ab was tested under the same conditions. All sera were used at 1/20 dilution in PBS, 3% BSA, plus 1% Escherichia coli wild-type bacterial lysate ( E .coli pop2136 transformed with the wild-type pEX plasmid).
2.4 Construction of hTg cDNA expression libraries cDNA fragments corresponding to the complete coding sequence of hTg molecule were inserted in pEX1, pEX2 and pEX3 plasmids (Boehringer) to synthetize the corresponding fusion proteins in the three reading frames as previously described [9, 101.
Eur. J. Immunol. 1992. 22: 315-319
2.7.2 Inhibition of mAb reactivity with different antigens The dilutions of the four different mAb plus an irrelevant one were established to obtain an absorbance between 0.5 and 1.0 at 620nm on competition ELISA. They were preincubated for 90 rnin at 37 "C with the different inhibitor Ag (hTg 50 pglml; fusion protein extracts about 1.5 mg/ml). Each sample (100 pl) was incubated on coated plate for I h at 37 "C. Immunobinding was revealed by adsorbed peroxidase-labeled anti-mouse Ab and 1 mM o-tolidine in acetic acid buffer 0.1 M pH 3.7 plus H202 110 vol. at 1/1500 dilution. Plates were incubated for 30 rnin at 4°C and absorbance measured at 620 nm.
2.7.3 Direct binding of alkaline phosphatase-labeled mAb (AP-mAb) The different AP-mAb, labeled as described [5], were tested at the same concentration as for hTg. AP-mAb binding was revealed by using 4-nitrophenylphosphate (Sigma Chemical Co., St. Louis, MO) as substrate. After an incubation of 90 rnin at 37"C, the absorbance was measured at 405 nm.
2.5 hTg libraries immunoscreening 2.8 Shorter immunoreactive peptide localization hTg expression libraries were plated by filtration on nitrocellulose disks (BA85; Schleicher and Schull, Dassel, FRG). After replication, fusion protein synsthesis was induced and nitrocellulose disks were treated as described [7]. Disks were incubated for 90 rnin at room temperature first in diluted, absorbed anti-hTg mAb, then incubated in peroxidase-labeled anti-mouse Ab solution (Calbiochem Corporation, La Jolla, CA), and revealed with 4-chloro1-naphthol as chromogen [ 111. Positive recombinant colonies were cloned.The plasmidic DNA were purified on CsCl gradient as in [12], analyzed by restriction mapping and sequenced by the dideoxy-mediated chain termination method (Sequenase; US Biological Corp., Cleveland, OH).
2.6 SDS-PAGE and Western blotting Selected immunoreactive or non-recombinant clones were grown in 100 ml Luria Bertani medium at 32 "C and protein expression was induced 2 h at 42°C. Bacterial extract was prepared as described in [7], dissolved in electrophoresis sample buffer to obtain a final dilution of 1/10 of the bacterial extract and final SDS concentration of 2'30, 15 pl of the sample was separated by 6% SDS-PAGE [13], and transferred to nitrocellulose. Immunodetection was performed as described above.
2.7 ELISA 2.7.1 Coated antigens hTg was coated at 10 pg/ml in carbonate-hydrogen carbonate buffer (50 mM, pH 9.5). Fusion proteins were used as previously described [7]. HPLC fractions were coated at about 5s pg/ml in the same coating buffer as hTg.
Plasmidic DNA of positive recombinant clone was purified, and partial deletions were made in different restriction sites. The resulting fragments were subcloned in pEX1, pEX2 or pEX3 in agreement with the convenient reading frame, and tested for mAb immunoreactivities. In case of positive or negative results, recombinant plasmids were purified and analyzed by restriction mapping to confirm the presence and the correct orientation of cDNA.
3 Results 3.1 Reactivity studies of the different AP-mAb with hTg gel filtration fractions Highly purified hTg was digested with proteaseV8, and hTg fragments were separated by HPLC gel filtration (Fig. 1A). Fraction 2 (P2) was digested by endoproteinase Lys-C, then analyzed on Ultrogel AcA 54 gel filtration column (Fig. 1B). Peptide fractions P1, P2, and P3 collected on HPLC, and fractions 1, 2, and 3 collected on AcA gel filtration of P2 were coated on microtiter plates and tested for their reactivities with the 15 AP-mAb (Fig. 1C). Among the six clusters, only mAb from cluster I1 reacted with the endoproteinase hTg digest fraction which corresponded to collecting tubes 57 to 62 (eluted peptides of about 15 to 10 kDa by calibration). The mAb from clusters V and VI reacted with no digest fraction, and those from clusters I, I11 and IV immunoreacted with all fractions of proteaseV8 hTg digest. No tested mAb reacted with fraction 3 of endoproteinase Lys-C digest which corresponded to hTg fragments smaller than 8 kDa. Fraction 2 of endoproteinase Lys-C digest was purified on mAb 10 affinity column; the eluted peptide reacted with the 3 other mAb of cluster 11. This response was inhibited by hTg and the peptide inhibited hTg mAb binding (data not shown).
Eur. J. Immunol. 1992. 22: 31.5-319 x
A 280 nrn 10
P2
Pi
317
mAb epitope on human thyroglobulin
duced from one of these clones was very well recognized by mAb 3 , 6, 10, 15 in Western blotting (data not shown).
P3
3.3 Search for minimum immunoreactive peptide
0
x
20
10
40
30
50
We established a detailed restriction map. The different selected cDNA were purified and inserted in convenient plasmid. The resulting recombinant cells were screened as above. The strategy to reduce the immunoreactive peptide is summarized on Fig. 2. The minimal selected antigenic peptide was 102 amino acids long. The same reactivity was observed for the four cluster I1 mAb, and corresponds to hTg peptidic sequence 1149-1250. All attempts to reduce the 102-amino acid peptide led to a loss of immunoreactiv-
60
A 280 nrn 10
1
---
....
3 4 4 9 +rmd&nA”
P
20
60
40
H A H
_--
-_
M3-436
- - _-
---
---
3754
M
100 fractions
60
-m_
5’
H E A”
B
H
---
---
HAA”
I
............IL ...............................................................
8244
3’
-
3884
P
M3-305
I
I
(Cl Digest Fractions
AbL4.b 4
5
9
1 1 1 3 14 3
6
10 15 2
8
P1
145
142 137 120 105 98
97
91
96 107 68 85
P2
105
98
102
95
100 92
89
89
73
P3
7
1
12
92
0
0
18
76
70
68
0
0
12
16
15
18
21
13
36
30
32
86
12
15
18
0
0
P2-1
6
8
5
3
9
11
55
47
41
136
6
9
5
0
0
P2-2
0
0
0
Pt-3
0
0
0
-
.1 .
0
0
0 0
0
0
0 0 4 3 1 8 1 6 7 0 0 0 0
0
8
7
8
II
................................................. I..................................... J L ..................., . . , . I I
2
0
0
III
0 0
0
I V V V I
Figure 1 . Reactivity of AP-mAb against hTg fractions obtained by protease digestions. (A) Profile of HPLC gel filtration after hTg digestion by protease V8; pools P1, P2, P3 used for immunoreactivity tests are in boxes. (B) Profile of AcA 54 gel filtration after fraction P2 digestion by endoproteinase Lys-C; pools 1, 2, 3 in boxes correspond to fractions P2-1, P2-2, P2-3 used for immunoreactivity tests. (C) Direct ELISA on the hTg digest fractions with 1.5 AP-mAb. Antigenic clusters I to VI of different mAb are indicated as previously defined (51. Values refer to binding of AP-mAb on coated hTg which was arbitrarily set at 100.
1 L
.......II ....................
J
M3-262
............................I
...................................... 1 ....................................................................
I I
.........J
I
1....1
.............................
1
I
1 I
................II ..............I
Figure2. Summary of strategies used to define the epitope of cluster 11. The upper line indicates the M3-436 region on the hTg primary sequence. Numbering refers to the encoding sequence without the leader peptide. The minimal antigenic M3-305 sequence is spotted; the restriction sites used for the subcloning are indicated as follows: P = Pst I, H = Hae 111, A = Alu I. M = Mlu I, Av = Ava I, B = Bgl I, Hi = Hinf I. Positive constructs are in solid lines, the negative ones in dotted lines.
3.2 Screening of hTg libraries with hTg mAb of the different clusters of reactivity Between 200 and 500 cells of the three hTg expression libraries were plated. Nitrocellulose disk replicas were tested with one or two mAb of each cluster previously defined [5] (mAb 5 and 9 of cluster I; 6 and 15 of cluster 11; 2 and 8 of cluster 111; 7 cluster of IV; 1 of cluster V; 12 of cluster V I ) . Positive results were obtained only for cluster 11. We revealed 12 identical colonies in pEX3 library with mAb 6 and 15. Plasmidic cDNA from each positive clone was characterized by restriction mapping and sequence analysis. The peptide encoded by these cDNA was 144 amino acids long, and stemmed from M3 clone [14]. They correspond to fragment Pst-Pst of 436 nucleotides in position 3449-3884 on hTg encoding cDNA and 1149-1295 on the peptidic chain (M3-436). The fusion protein pro-
Figure 3. SDS-PAGE of fusion proteins and wild-type pEX (A) and corresponding Western blotting (B, C, D, E). Samples were electrophoresed as follows: lane 1, pEX T; lane 2, M3-305; lane 3, M3-327; lane 4, M3-262. Position of wild-type cro-lac Z protein is indicated by an arrow “cro-lac Z”. (A) SDS-PAGE stained with Coomassie blue. (B-E) Immunorevelation with respectively mAb 3, mAb 6, mAb 10 and mAb 15. Lane 1 and 4 are negative controls.
318
Eur. J. Immunol. 1992.22: 315-319
M. Henry, E. Zanelli, M. Piechaczyk et al.
ity. Positive clones M3-327 (Pstl-Smal, 327 bp) and M3305 (Pstl-Hinfl, 305 bp), the largest negative clone M3-262 ( A M - A M , 262 bp), and the wild-type clone (pEX T) were expressed in proteins and tested in Western blotting (Fig. 3). The results obtained through the bacterial screening were confirmed.
3.4 Identity between the antigenic determinant recognized on purified hTg and that of bacterial fusion protein We tested mAb 3, 6, 10 and 15 in direct ELISA with peroxidase revelation to obtain A620 between 0.4 and 1.0. Competition tests were performed as described above. hTg mAb reactivities were tested in competition on coated hTg, M3-305, pEXTagainst hTg, M3-305 and pEXTas inhibitor Ag. Results are reported on Table 1. The inhibition of binding of the four mAb was > 60% in all cases, i.e. it was the same antigenic determinant which bound cluster I1 mAb on purified hTg and on fusion protein M3-305.
3.5 Western blotting on mAb antigenic determinant with 19 pathologic sera Fusion proteins M3-305 was separated by electrophoresis, transferred, and tested with the 16 AITD sera, 3 cancer sera, and 1normal serum (Fig. 4).We obtained 13 positive
Figure 4 . Immunorevelation of fusion proteins M3-305 with 19 human pathological sera and one healthy.The first strip was stained with amido schwarz. Pathologies are indicated as follows: A: AITD, C: thyroid cancer, N: normal.The arrow shows the position of fusion proteins.
results out of 16 AITD patient sera and 1positive result out of 3 cancer sera.Thus 74% of the tested sera recognized the M3-305 epitope.
4 Discussion We have tried to localize epitopes recognized by 15 different anti-hTg mAb on peptides produced by protease digestion of hTg or expressed in a prokaryotic system. The recombinant bacteria E.coli p o p 2136 allowed synthesis of hybrid peptides (cro-lac Z foreign protein) in high quantities, which can be directly tested as well in Western blotting as in ELISA. By the first approach, all mAb except those of clusters V and VI recognized all fractions of proteaseV8 digest. These peptidic fractions, however, were highly complex and heterogeneous, and the different epitopes could not be isolated without disrupting the structure of the antigenic determinants. The digest with endoproteinase Lys-C of the second protease V8 fraction revealed that only cluster I1 mAb reacted with fractions 1 and 2, and that no mAb reacted with the fraction 3, which contains peptides < 8 kDa. These results agree with those of Male et al. [15] who found that Ab conserved their reactivity on hTg peptides produced by protease V8 digestion [16]. Conversely, they differ from those of Khono et al. [17] and Fukuma et al. [18] who reported no response in ELISA with reduced or degraded hTg. Except for cluster I1 mAb, all others reacted against epitopes which required a native structure of the molecule (mAb 1 and 12) or a peptide large enough to maintain conformational structure (mAb from cluster I, 111, and IV). Since they are probably discontinuous, these types of antigenic determinants could not be screened easily in prokaryotic expression libraries. We detected positive clones only with mAb of cluster 11, showing that the epitope they recognized was less discontinuous. The same M3-436 clone corresponding to hTg sequence (1149-1295) was revealed both by the rabbit anti-hTg Ab previously described [9], and by mAb from cluster 11, but the restricted determinant was different.The epitope defined by rabbit anti-hTg Ab was reduced to 38 amino acids (1258-1295), whereas the antigenic peptide selected by the mAb could not be reduced to fewer than 102 amino acids (1149-1250). In the peptidic sequence analysis on Fig. 5, we noted that this peptide contained ten cysteines and ten prolines; thus, a secondary structure including a disulfide bridge may be necessary to restore the peptidic conformation of the determinant and allow a detection by
+
t h e m A b of cluster 11. Competition tests attested t h a t the
same determinant site was detected by the mAb on hTg and on the hybrid peptide, proving that a peptidic conformation close to that of native hTg can be restored in the foreign
Table 1. Inhibition of mAb cluster I1 binding on hTg or M3.305 by hTg or M3.305
mAb
3 6 10 15
/Oa)
/M3.305b)
/oa)
M3.305 /M3.305b)
0.79') 0.47 0.61 0.73
0.16(80%)d) O.l6(66%) 0.18(70%) 0.16(78%)
0.40 0.40 0.37 0.35
0.10(75%) 0.10(75%) 0.09(76%) 0.10(71%)
hTg
hTg
M3.305
M3.305 /hTgb)
0.07(82%) 0.06(85%) 0.04(89%) 0.05(86%)
a) Direct binding on coated hTg or M3.305. b) Competition binding : upper line, coated Ag; outer line, competitor Ag. C) Absorbance at 620 nm. d, Percentage of inhibition as compared to the direct binding.
Eur. J. Immunol. 1992. 22: 315-319 3445 1148 34 75 1158 3505 1168
3535 1178
mAb epitope on human thyroglobulin
319
TGC-AGG-GCA-GAG-GAT-GGG-GGC-TTT-TCC-CCA CYB-arg-ala-glu-asp-gly-gly-phe-ser-pro
M3-305 peptide in Western blotting, probably because the epitope presentation is better than on total hTg molecule. In contrast with other observations [23], epitopes of human GTG-CAA-TGT-GAC-CAG-GCC-CAG-GGC-AGC-TGC v a l - g l n - ~ y ~ - a s p - q l h - a l a - g l n - g l y - s e r - ~ ~anti-hTg ~ aAb can be defined in prokaryotic expression system. In a recent report, Dietrich et al. [24] described the TGG-TGT-GTC-ATG-GAC-AGC-GGA-GAA-GAG-GTG importance of the hTg region recognized by the cluster I1 trp-CYS-val-met-asp-ser-gly-glu-qlu-val mAb. Further investigations are now necessary to deterCCT-GGG-ACG-CGC-GTG-ACC-GGG-GGC-CAG-CCC mine if the M3-305 hTg epitope could be a marker of the pro-gly-thr-arg-val-thr-gly-qly-gln-pro thyroid disorder.
3565 1188
GCC-TGT-GAG-AGC-CCG-CGG-TGT-CCG-CTG-CCA a 1a - e l y S - g l u - s e r - p r o - a r g - C y S - p r O - l e u - p r o
3595 1198
TTC-AAC-GCG-TCG-GAG-GTG-GTT-GGT-GGA-ACA phe-asn-ala-ser-glu-val-val-gly-gly-thr
3625 1208
ATC-CTG-TGT-GAG-ACA-ATC-TCG-GGC-CCC-ACA ile-leu-CyS-glu-thr-ile-ser-gly-pro-thr
3655 1218
GGC-TCT-GCC-ATG-CAG-CAG-TGC-CAA-TTG-CTG gly-ser-ala-met-gln-gln-CYS-gln-leu-leu
3685 1228
TGC-CGC-CAG-GGC-TCC-TGG-AGC-GTG-TTT-CCA Cy5-arg-qln-qly-ser-trp-ser-val-phe-pro
3705 1238
CCA-GGG-CCA-TTG-ATA-TGT-AGC-CTG-GAG-AGC pro-g 1 y-pro- l e u - i 1 e - C Y B - s e r - le u - g l u - s e r
3735 1248
GGA-CGC-TGG g l y -a rg -t r p
Figures. Nucleotidic and peptidic sequence of mAb cluster I1 epitope. Numbering corresponds to encoding sequence without leader peptide.
peptide synthetized by prokaryotic cells [20]. The immunoreactivity of the four mAb cannot be dissociated on fusion proteins. It appears that the domain defined in prokaryotic system is the antigenic support of cluster I1 defined by criss-cross experiments on native hTg [6]. Using 26 and 38 anti-hTg mAb, we defined 8 and 10 clusters of reactivity respectively [19,20], and established that the four cluster I1 mAb [5], named K, B, F, J in [20], and the 5 mAb of cluster ILV produced and characterized by Ruf et al. [21], named 5, 6, 10, 11, 8 in [20], constituted an unique entity. These findings indicate that the 5 cluster IV mAb would recognize either the M3-305 epitope as did the 4 cluster I1 mAb, or a contiguous one. We have reported that aAb from AITD presented a high heterogeneity of Ab response [7], but the antigenic sites recognized were mostly located on an immunodominant region corresponding to the cDNA nucleotidic sequence (3294-4680). The epitope of mAb from cluster I1 was also encoded in this cDNA region (3449-3754). We concluded that this corresponded to a long highly antigenic domain from amino acids 1097 to 1560 on the hTg peptide chain. Studying the cross-reactivity of these 15 mAb with anti-hTg aAb in the serum from subjects with or without AITD [6], Piechaczyk et al. [6] suggested that pathological anti-hTg Ab recognized mainly cluster 11, and Bouanani et al. that natural anti-hTg mainly cluster V [22]. We herein demonstrate that the exact localization of the antigenic determinant of cluster I1 mAb is made possible by screening hTg cDNA expression libraries. Conversely, it was not possible to determine the antigenic determinant of cluster V mAb probably because of the highly conformational nature of the epitope. Using 19 pathological sera from AITD and thyroid cancer with high anti-hTg titer, we obtained 14 positive results on the M3-305 peptide. Some sera which did not present crossreactivity with cluster I1 mAb on ELISA inhibition test with hTg as coated antigen (data not shown), reacted with
We thank C. Marriq for purified h Q and very much appreciated the expert technical assistance of B. Clzarvet. Received May 23, 1991; in revised form September 20, 1991.
5 References 1 Lissitsky, S. in De Groot, L. J. (Ed.), Endocrinologj vol. 1,W. B. Saunders Compagny, Philadelphia 1989, p. 512. 2 Weetman, A. P. and McGregor. A. M., Endocrine Rev. 1984. 5: 309. 3 Rochman, H . , DeGroot, L. J., Rieger, C. H. L.,Varnavides, L. A., Refetoff, S., Joung, J. I. andHoye, K., CancerRex 1975. 35: 2692. 4 Guilbert, B., Dighiero, G. and Avrameas, S., J. Immunol. 1982. 128: 2779. 5 Piechaczyk, M., Chardes,T., Cot, M. C., Pau, B. and Bastide, J. M., Hybridoma 1985. 4: 361. 6 Piechaczyk, M., Bouanani, M., Salhi, S. L., Baldet, L., Bastide, M . , Pau, B. and Bastide, J. M., Clin. Immunol. Immunopathol. 1987. 45: 114. 7 Henry, M., Malthiery,Y., Zanelli, E . and Charvct, B., J. Imrnunol. 1990. 145: 3692. 8 Jekel, €! A.,Weijer,W. J. and Beintema, J. J., Anal. Biochem. 1983. 134: 347. 9 Stanley, K. K . and Luzio, J. l?, EMBO J. 1984. 3: 1429. 10 Haymerle, H . , Hertz, J., Bressan, G. M., Frank, R. and Stanley, K. K., Nucleic Acids Kes. 1986. 14: 8615. 11 Hawkes, R . , Anal. Biochem. 1982. 123: 143. 12 Maniatis, T., Fritsch, E . F. and Sambrook, J., Moleculur cloning: a luborutory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor 1982. 13 Towbin, H . , Staehelin,T. andGordon, J., Proc. Natl. Acad. Sci. USA 1979. 76: 4350. 14 Malthiery, Y. and Lissitzky, S., Eur. J. Biochem. 1987. 165: 491. 15 Male, D. K., Champion, B. R., Pryce, G . , Matthens, H. and Shepherd, P., Immunology 1985. 54: 954. 16 Kohno, Y., Naito, N., Hiyama, Y., Shimojo, N., Suzuki, N., Tarutani, O., Niimi, H . , Nakajima, H . and Hosoya,T., J. Clin. Endocrinol. Metab. 1988. 67: 899. 17 Fukuma, N., McLachlan, S. M., Rapoport, B., Goodacre, J., Middleton, S. L., Phillips, D. I . W., Pegg, C. A. S. and Rees Smith. B., Clin. Exp. Immunol. 1990. 82: 275. 18 DeMacedo Brigido, M., Sabbaga, J. and Renzo Brentani, R., Immunnl. Lett. 1990. 24: 191. 19 Henry. M., Piechaczyk, M., Durand-Gorde. J. M., Pau, B. and Costagliola, S., Abstracts of 16th Annual Meeting of the European Thyroid Associntion, Masson, Paris, 1987, p. 130. 20 Ruf, J., Henry, M., DeMicco. C. and Carayon, l?, Workshop Proceeding Schilddruse 1Y85, Georg Thiemc Verlag, Stuttgart, New York 1987, p. 21. 21 Ruf, J., Carayon, P. and Lissitsky, S., Eur. J. Immunol. 1985. 15: 268. 22 Bouanani, M., Piechaczyk, M., Pau, B. and Bastide, M., .I. Immunol. 1989. 143: 1129. 23 Dong, G., Ludgate, M. and Vassart, G., J. Endocrinol. 1989. 122: 169. 24 Dietrich, G . , Piechaczyk, M., Pau, B. and Kazatchkine, M. D., Eur. .I. Immunol. 1991. 21: 811.
Eur. J. Immunol. 1992. 22: 315-319
315
Mireille Henryo, Eric ZanelliO, Martine Piechaczyk., Bernard Pau. and Yves MalthihryO
A major human thyroglobulin epitope defined with monoclonal antibodies is mainly recognized by human autoantibodies*
Laboratoire de Biochimie MCdicale, INSERM U38, FacultC de MCdecineO, Marseille and Laboratoire d’Immunologie, FacultC de Pharmacie., Montpellier
The antigenic nature of 15 anti-human thyroglobulin (hTg) monoclonal antibody (mAb) epitopes was studied by two different approaches. First, we tested two successive protease-digest products of hTg. Only four mAb from the same cluster of reactivity recognized a low-molecular weight peptide, the other mAb only bound native hTg or high-molecular weight digest fractions. Second, these 15 mAb were used to immunoscreen hTg expression libraries. Only the same four mAb revealed immunoreactive clones corresponding to region 1149-1295 on the hTg primary sequence. After subcloning, this antigenic determinant was reduced to a 102-amino acid peptide (hTg region 1149-1250). The two different methodologies were coherent and complementary, and demonstrated that hTg sequence 1149-1250 is the target for this cluster of four mAb. Moreover, anti-hTg autoantibodies which cross-reacted with these mAb bound the 102-amino acid peptide. This epitope was the one most frequently detected by sera from autoimmune thyroid disease. The data confirm the presence of an immunodominant domain in the central part of the hTg molecule and suggest that this mAb epitope may be a powerful probe for the diagnosis of autoimmune thyroid disorders.
1 Introduction Human thyroglobulin (hTg) is a high-molecular weight glycoprotein which supports the synthesis and storage of thyroid hormones [l].Autoantibodies (aAb) to hTg are classically found with high titers in the sera from patients with Graves’ disease, Hashimoto’s thyroiditis [2] and other thyroid disorders such as thyroid carcinoma [3], but also in healthy patients [4]. By criss-cross inhibition with 15 anti-hTg mouse mAb, Piechaczyk et al. [ S ] identified six clusters of reactivity which reflected the presence of at least six antigenic domains.These authors observed that only the mAb from cluster 11 (mAb 3, 6, 10 and 15) cross-reacted with anti-hTg aAb [6]. In previous report [7] using rabbit polyclonal antibodies we have located seven epitopes and an immunodominant domain in the middle part of the hTg molecule. Each immunoreactive peptide tested with autoimmune sera revealed a heterogeneity of aAb response. We reported here the simultaneous screening with mAb [5] of a double-enzymatic digest of a highly purified hTg, and hTg cDNA expression libraries. We concluded that a large number of conformational or discontinuous epitopes were recognized by these mAb. Only mAb from cluster I1 recognized low-molecular weight digested pep-
[I 95981
*
This work was supported by grants from the Centre National de la Recherche Scientifique (U.A. 178) and the Institut National de la SantC et de la Recherche Medicale (U.38).
Correspondence: Mireille Henry, Laboratoire de Biochimie MCdicale, INSERM U38, Faculte de MCdecine, 27, Boulevard Jean Moulin, F-13385 Marseille cedex 5, France Abbreviations: hTg: Human thyroglobulin aAb: Autoantibodies AITD: Autoimmune thyroid disease AP-mAb: Alkaline phosphatase-labeled mAb pEXT Colony transformed with wild-type plasmid
0 VCH Verlagsgescllschaft mbH, D-6940 Weinheim, 1992
tides and a recombinant epitope expressed in hTg cDNA libraries. The conformation of this determinant seems important for mAb recognition, but it was restored in the fusion protein. The mAb antigenic determinant, exactly localized on hTg peptidic primary sequence, conserved its reactivity for aAb in the sera from patients with autoimmune thyroid disease (AITD).
2 Materials and methods 2.1 Double hTg digestion and gel filtration analysis Fifty milligrams of hTg (purified by C. Marriq from thyroid goiter) in 0.1 M ammonium hydrogen carbonate, pH 7.8 was digested by 1/50 (w/w) staphylococcal V8 protease (Miles Scientific, Naperville, IL), for 23 h at 37°C. Enzymatic reaction was stopped by adding 1 PM PMSF, 6 M urea (final concentration). The digest was resolved by HPLC gel filtration on TSK G 3000 SW column (Pharmacia LKB, Uppsala, Sweden) with PBS as running buffer (0.04 ml/min). Collected fractions corresponding to each peak were pooled and concentrated. Then, endoproteinase LysC (Boehringer Mannheim, Mannheim, FRG) was added in the second peak (P2) according to Jeckel et al. [8] and was allowed to proceed for 8 h at 37°C. The reaction was stopped by addition of aprotinin (Boehringer) for 1 h at 25”C, and urea ( 5 M in final concentration). The hTg fragments were separated by gel filtration on Ultrogel AcA 54 (IBF,Villeneuve-la-Garenne, France) with PBS, 5 M urea as running buffer. Convenient fractions were pooled and dialyzed in PBS. Ultrogel ACA 54 was calibrated with gel filtration LMW calibration kit (Pharmacia LKB). 2.2 Anti-hTg mAb
The mAb were produced and purified as described [ 5 ] . Lyophilized mAb were dissolved in PBS at 1 mg/ml, and used at 1/1000 dilution in immunoscreening and Western OO14-2980/92/O202-0315$3.50+ .25/O
316
M. Henry, E. Zanelli, M. Piechaczyk et al.
blotting, 1150000dilutionin hTg ELISA, and 1/500in fusion protein ELISA.
2.3 Autoimmune sera Sixteen sera from patients with AITD, three with thyroid cancer were obtained. Their titers in anti-hTg antibodies were over 1164000, as determined by passive hemagglutination with commercial kits (Wellcome, Dartfasd, GB). One serum without anti-hTg Ab was tested under the same conditions. All sera were used at 1/20 dilution in PBS, 3% BSA, plus 1% Escherichia coli wild-type bacterial lysate ( E .coli pop2136 transformed with the wild-type pEX plasmid).
2.4 Construction of hTg cDNA expression libraries cDNA fragments corresponding to the complete coding sequence of hTg molecule were inserted in pEX1, pEX2 and pEX3 plasmids (Boehringer) to synthetize the corresponding fusion proteins in the three reading frames as previously described [9, 101.
Eur. J. Immunol. 1992. 22: 315-319
2.7.2 Inhibition of mAb reactivity with different antigens The dilutions of the four different mAb plus an irrelevant one were established to obtain an absorbance between 0.5 and 1.0 at 620nm on competition ELISA. They were preincubated for 90 rnin at 37 "C with the different inhibitor Ag (hTg 50 pglml; fusion protein extracts about 1.5 mg/ml). Each sample (100 pl) was incubated on coated plate for I h at 37 "C. Immunobinding was revealed by adsorbed peroxidase-labeled anti-mouse Ab and 1 mM o-tolidine in acetic acid buffer 0.1 M pH 3.7 plus H202 110 vol. at 1/1500 dilution. Plates were incubated for 30 rnin at 4°C and absorbance measured at 620 nm.
2.7.3 Direct binding of alkaline phosphatase-labeled mAb (AP-mAb) The different AP-mAb, labeled as described [5], were tested at the same concentration as for hTg. AP-mAb binding was revealed by using 4-nitrophenylphosphate (Sigma Chemical Co., St. Louis, MO) as substrate. After an incubation of 90 rnin at 37"C, the absorbance was measured at 405 nm.
2.5 hTg libraries immunoscreening 2.8 Shorter immunoreactive peptide localization hTg expression libraries were plated by filtration on nitrocellulose disks (BA85; Schleicher and Schull, Dassel, FRG). After replication, fusion protein synsthesis was induced and nitrocellulose disks were treated as described [7]. Disks were incubated for 90 rnin at room temperature first in diluted, absorbed anti-hTg mAb, then incubated in peroxidase-labeled anti-mouse Ab solution (Calbiochem Corporation, La Jolla, CA), and revealed with 4-chloro1-naphthol as chromogen [ 111. Positive recombinant colonies were cloned.The plasmidic DNA were purified on CsCl gradient as in [12], analyzed by restriction mapping and sequenced by the dideoxy-mediated chain termination method (Sequenase; US Biological Corp., Cleveland, OH).
2.6 SDS-PAGE and Western blotting Selected immunoreactive or non-recombinant clones were grown in 100 ml Luria Bertani medium at 32 "C and protein expression was induced 2 h at 42°C. Bacterial extract was prepared as described in [7], dissolved in electrophoresis sample buffer to obtain a final dilution of 1/10 of the bacterial extract and final SDS concentration of 2'30, 15 pl of the sample was separated by 6% SDS-PAGE [13], and transferred to nitrocellulose. Immunodetection was performed as described above.
2.7 ELISA 2.7.1 Coated antigens hTg was coated at 10 pg/ml in carbonate-hydrogen carbonate buffer (50 mM, pH 9.5). Fusion proteins were used as previously described [7]. HPLC fractions were coated at about 5s pg/ml in the same coating buffer as hTg.
Plasmidic DNA of positive recombinant clone was purified, and partial deletions were made in different restriction sites. The resulting fragments were subcloned in pEX1, pEX2 or pEX3 in agreement with the convenient reading frame, and tested for mAb immunoreactivities. In case of positive or negative results, recombinant plasmids were purified and analyzed by restriction mapping to confirm the presence and the correct orientation of cDNA.
3 Results 3.1 Reactivity studies of the different AP-mAb with hTg gel filtration fractions Highly purified hTg was digested with proteaseV8, and hTg fragments were separated by HPLC gel filtration (Fig. 1A). Fraction 2 (P2) was digested by endoproteinase Lys-C, then analyzed on Ultrogel AcA 54 gel filtration column (Fig. 1B). Peptide fractions P1, P2, and P3 collected on HPLC, and fractions 1, 2, and 3 collected on AcA gel filtration of P2 were coated on microtiter plates and tested for their reactivities with the 15 AP-mAb (Fig. 1C). Among the six clusters, only mAb from cluster I1 reacted with the endoproteinase hTg digest fraction which corresponded to collecting tubes 57 to 62 (eluted peptides of about 15 to 10 kDa by calibration). The mAb from clusters V and VI reacted with no digest fraction, and those from clusters I, I11 and IV immunoreacted with all fractions of proteaseV8 hTg digest. No tested mAb reacted with fraction 3 of endoproteinase Lys-C digest which corresponded to hTg fragments smaller than 8 kDa. Fraction 2 of endoproteinase Lys-C digest was purified on mAb 10 affinity column; the eluted peptide reacted with the 3 other mAb of cluster 11. This response was inhibited by hTg and the peptide inhibited hTg mAb binding (data not shown).
Eur. J. Immunol. 1992. 22: 31.5-319 x
A 280 nrn 10
P2
Pi
317
mAb epitope on human thyroglobulin
duced from one of these clones was very well recognized by mAb 3 , 6, 10, 15 in Western blotting (data not shown).
P3
3.3 Search for minimum immunoreactive peptide
0
x
20
10
40
30
50
We established a detailed restriction map. The different selected cDNA were purified and inserted in convenient plasmid. The resulting recombinant cells were screened as above. The strategy to reduce the immunoreactive peptide is summarized on Fig. 2. The minimal selected antigenic peptide was 102 amino acids long. The same reactivity was observed for the four cluster I1 mAb, and corresponds to hTg peptidic sequence 1149-1250. All attempts to reduce the 102-amino acid peptide led to a loss of immunoreactiv-
60
A 280 nrn 10
1
---
....
3 4 4 9 +rmd&nA”
P
20
60
40
H A H
_--
-_
M3-436
- - _-
---
---
3754
M
100 fractions
60
-m_
5’
H E A”
B
H
---
---
HAA”
I
............IL ...............................................................
8244
3’
-
3884
P
M3-305
I
I
(Cl Digest Fractions
AbL4.b 4
5
9
1 1 1 3 14 3
6
10 15 2
8
P1
145
142 137 120 105 98
97
91
96 107 68 85
P2
105
98
102
95
100 92
89
89
73
P3
7
1
12
92
0
0
18
76
70
68
0
0
12
16
15
18
21
13
36
30
32
86
12
15
18
0
0
P2-1
6
8
5
3
9
11
55
47
41
136
6
9
5
0
0
P2-2
0
0
0
Pt-3
0
0
0
-
.1 .
0
0
0 0
0
0
0 0 4 3 1 8 1 6 7 0 0 0 0
0
8
7
8
II
................................................. I..................................... J L ..................., . . , . I I
2
0
0
III
0 0
0
I V V V I
Figure 1 . Reactivity of AP-mAb against hTg fractions obtained by protease digestions. (A) Profile of HPLC gel filtration after hTg digestion by protease V8; pools P1, P2, P3 used for immunoreactivity tests are in boxes. (B) Profile of AcA 54 gel filtration after fraction P2 digestion by endoproteinase Lys-C; pools 1, 2, 3 in boxes correspond to fractions P2-1, P2-2, P2-3 used for immunoreactivity tests. (C) Direct ELISA on the hTg digest fractions with 1.5 AP-mAb. Antigenic clusters I to VI of different mAb are indicated as previously defined (51. Values refer to binding of AP-mAb on coated hTg which was arbitrarily set at 100.
1 L
.......II ....................
J
M3-262
............................I
...................................... 1 ....................................................................
I I
.........J
I
1....1
.............................
1
I
1 I
................II ..............I
Figure2. Summary of strategies used to define the epitope of cluster 11. The upper line indicates the M3-436 region on the hTg primary sequence. Numbering refers to the encoding sequence without the leader peptide. The minimal antigenic M3-305 sequence is spotted; the restriction sites used for the subcloning are indicated as follows: P = Pst I, H = Hae 111, A = Alu I. M = Mlu I, Av = Ava I, B = Bgl I, Hi = Hinf I. Positive constructs are in solid lines, the negative ones in dotted lines.
3.2 Screening of hTg libraries with hTg mAb of the different clusters of reactivity Between 200 and 500 cells of the three hTg expression libraries were plated. Nitrocellulose disk replicas were tested with one or two mAb of each cluster previously defined [5] (mAb 5 and 9 of cluster I; 6 and 15 of cluster 11; 2 and 8 of cluster 111; 7 cluster of IV; 1 of cluster V; 12 of cluster V I ) . Positive results were obtained only for cluster 11. We revealed 12 identical colonies in pEX3 library with mAb 6 and 15. Plasmidic cDNA from each positive clone was characterized by restriction mapping and sequence analysis. The peptide encoded by these cDNA was 144 amino acids long, and stemmed from M3 clone [14]. They correspond to fragment Pst-Pst of 436 nucleotides in position 3449-3884 on hTg encoding cDNA and 1149-1295 on the peptidic chain (M3-436). The fusion protein pro-
Figure 3. SDS-PAGE of fusion proteins and wild-type pEX (A) and corresponding Western blotting (B, C, D, E). Samples were electrophoresed as follows: lane 1, pEX T; lane 2, M3-305; lane 3, M3-327; lane 4, M3-262. Position of wild-type cro-lac Z protein is indicated by an arrow “cro-lac Z”. (A) SDS-PAGE stained with Coomassie blue. (B-E) Immunorevelation with respectively mAb 3, mAb 6, mAb 10 and mAb 15. Lane 1 and 4 are negative controls.
318
Eur. J. Immunol. 1992.22: 315-319
M. Henry, E. Zanelli, M. Piechaczyk et al.
ity. Positive clones M3-327 (Pstl-Smal, 327 bp) and M3305 (Pstl-Hinfl, 305 bp), the largest negative clone M3-262 ( A M - A M , 262 bp), and the wild-type clone (pEX T) were expressed in proteins and tested in Western blotting (Fig. 3). The results obtained through the bacterial screening were confirmed.
3.4 Identity between the antigenic determinant recognized on purified hTg and that of bacterial fusion protein We tested mAb 3, 6, 10 and 15 in direct ELISA with peroxidase revelation to obtain A620 between 0.4 and 1.0. Competition tests were performed as described above. hTg mAb reactivities were tested in competition on coated hTg, M3-305, pEXTagainst hTg, M3-305 and pEXTas inhibitor Ag. Results are reported on Table 1. The inhibition of binding of the four mAb was > 60% in all cases, i.e. it was the same antigenic determinant which bound cluster I1 mAb on purified hTg and on fusion protein M3-305.
3.5 Western blotting on mAb antigenic determinant with 19 pathologic sera Fusion proteins M3-305 was separated by electrophoresis, transferred, and tested with the 16 AITD sera, 3 cancer sera, and 1normal serum (Fig. 4).We obtained 13 positive
Figure 4 . Immunorevelation of fusion proteins M3-305 with 19 human pathological sera and one healthy.The first strip was stained with amido schwarz. Pathologies are indicated as follows: A: AITD, C: thyroid cancer, N: normal.The arrow shows the position of fusion proteins.
results out of 16 AITD patient sera and 1positive result out of 3 cancer sera.Thus 74% of the tested sera recognized the M3-305 epitope.
4 Discussion We have tried to localize epitopes recognized by 15 different anti-hTg mAb on peptides produced by protease digestion of hTg or expressed in a prokaryotic system. The recombinant bacteria E.coli p o p 2136 allowed synthesis of hybrid peptides (cro-lac Z foreign protein) in high quantities, which can be directly tested as well in Western blotting as in ELISA. By the first approach, all mAb except those of clusters V and VI recognized all fractions of proteaseV8 digest. These peptidic fractions, however, were highly complex and heterogeneous, and the different epitopes could not be isolated without disrupting the structure of the antigenic determinants. The digest with endoproteinase Lys-C of the second protease V8 fraction revealed that only cluster I1 mAb reacted with fractions 1 and 2, and that no mAb reacted with the fraction 3, which contains peptides < 8 kDa. These results agree with those of Male et al. [15] who found that Ab conserved their reactivity on hTg peptides produced by protease V8 digestion [16]. Conversely, they differ from those of Khono et al. [17] and Fukuma et al. [18] who reported no response in ELISA with reduced or degraded hTg. Except for cluster I1 mAb, all others reacted against epitopes which required a native structure of the molecule (mAb 1 and 12) or a peptide large enough to maintain conformational structure (mAb from cluster I, 111, and IV). Since they are probably discontinuous, these types of antigenic determinants could not be screened easily in prokaryotic expression libraries. We detected positive clones only with mAb of cluster 11, showing that the epitope they recognized was less discontinuous. The same M3-436 clone corresponding to hTg sequence (1149-1295) was revealed both by the rabbit anti-hTg Ab previously described [9], and by mAb from cluster 11, but the restricted determinant was different.The epitope defined by rabbit anti-hTg Ab was reduced to 38 amino acids (1258-1295), whereas the antigenic peptide selected by the mAb could not be reduced to fewer than 102 amino acids (1149-1250). In the peptidic sequence analysis on Fig. 5, we noted that this peptide contained ten cysteines and ten prolines; thus, a secondary structure including a disulfide bridge may be necessary to restore the peptidic conformation of the determinant and allow a detection by
+
t h e m A b of cluster 11. Competition tests attested t h a t the
same determinant site was detected by the mAb on hTg and on the hybrid peptide, proving that a peptidic conformation close to that of native hTg can be restored in the foreign
Table 1. Inhibition of mAb cluster I1 binding on hTg or M3.305 by hTg or M3.305
mAb
3 6 10 15
/Oa)
/M3.305b)
/oa)
M3.305 /M3.305b)
0.79') 0.47 0.61 0.73
0.16(80%)d) O.l6(66%) 0.18(70%) 0.16(78%)
0.40 0.40 0.37 0.35
0.10(75%) 0.10(75%) 0.09(76%) 0.10(71%)
hTg
hTg
M3.305
M3.305 /hTgb)
0.07(82%) 0.06(85%) 0.04(89%) 0.05(86%)
a) Direct binding on coated hTg or M3.305. b) Competition binding : upper line, coated Ag; outer line, competitor Ag. C) Absorbance at 620 nm. d, Percentage of inhibition as compared to the direct binding.
Eur. J. Immunol. 1992. 22: 315-319 3445 1148 34 75 1158 3505 1168
3535 1178
mAb epitope on human thyroglobulin
319
TGC-AGG-GCA-GAG-GAT-GGG-GGC-TTT-TCC-CCA CYB-arg-ala-glu-asp-gly-gly-phe-ser-pro
M3-305 peptide in Western blotting, probably because the epitope presentation is better than on total hTg molecule. In contrast with other observations [23], epitopes of human GTG-CAA-TGT-GAC-CAG-GCC-CAG-GGC-AGC-TGC v a l - g l n - ~ y ~ - a s p - q l h - a l a - g l n - g l y - s e r - ~ ~anti-hTg ~ aAb can be defined in prokaryotic expression system. In a recent report, Dietrich et al. [24] described the TGG-TGT-GTC-ATG-GAC-AGC-GGA-GAA-GAG-GTG importance of the hTg region recognized by the cluster I1 trp-CYS-val-met-asp-ser-gly-glu-qlu-val mAb. Further investigations are now necessary to deterCCT-GGG-ACG-CGC-GTG-ACC-GGG-GGC-CAG-CCC mine if the M3-305 hTg epitope could be a marker of the pro-gly-thr-arg-val-thr-gly-qly-gln-pro thyroid disorder.
3565 1188
GCC-TGT-GAG-AGC-CCG-CGG-TGT-CCG-CTG-CCA a 1a - e l y S - g l u - s e r - p r o - a r g - C y S - p r O - l e u - p r o
3595 1198
TTC-AAC-GCG-TCG-GAG-GTG-GTT-GGT-GGA-ACA phe-asn-ala-ser-glu-val-val-gly-gly-thr
3625 1208
ATC-CTG-TGT-GAG-ACA-ATC-TCG-GGC-CCC-ACA ile-leu-CyS-glu-thr-ile-ser-gly-pro-thr
3655 1218
GGC-TCT-GCC-ATG-CAG-CAG-TGC-CAA-TTG-CTG gly-ser-ala-met-gln-gln-CYS-gln-leu-leu
3685 1228
TGC-CGC-CAG-GGC-TCC-TGG-AGC-GTG-TTT-CCA Cy5-arg-qln-qly-ser-trp-ser-val-phe-pro
3705 1238
CCA-GGG-CCA-TTG-ATA-TGT-AGC-CTG-GAG-AGC pro-g 1 y-pro- l e u - i 1 e - C Y B - s e r - le u - g l u - s e r
3735 1248
GGA-CGC-TGG g l y -a rg -t r p
Figures. Nucleotidic and peptidic sequence of mAb cluster I1 epitope. Numbering corresponds to encoding sequence without leader peptide.
peptide synthetized by prokaryotic cells [20]. The immunoreactivity of the four mAb cannot be dissociated on fusion proteins. It appears that the domain defined in prokaryotic system is the antigenic support of cluster I1 defined by criss-cross experiments on native hTg [6]. Using 26 and 38 anti-hTg mAb, we defined 8 and 10 clusters of reactivity respectively [19,20], and established that the four cluster I1 mAb [5], named K, B, F, J in [20], and the 5 mAb of cluster ILV produced and characterized by Ruf et al. [21], named 5, 6, 10, 11, 8 in [20], constituted an unique entity. These findings indicate that the 5 cluster IV mAb would recognize either the M3-305 epitope as did the 4 cluster I1 mAb, or a contiguous one. We have reported that aAb from AITD presented a high heterogeneity of Ab response [7], but the antigenic sites recognized were mostly located on an immunodominant region corresponding to the cDNA nucleotidic sequence (3294-4680). The epitope of mAb from cluster I1 was also encoded in this cDNA region (3449-3754). We concluded that this corresponded to a long highly antigenic domain from amino acids 1097 to 1560 on the hTg peptide chain. Studying the cross-reactivity of these 15 mAb with anti-hTg aAb in the serum from subjects with or without AITD [6], Piechaczyk et al. [6] suggested that pathological anti-hTg Ab recognized mainly cluster 11, and Bouanani et al. that natural anti-hTg mainly cluster V [22]. We herein demonstrate that the exact localization of the antigenic determinant of cluster I1 mAb is made possible by screening hTg cDNA expression libraries. Conversely, it was not possible to determine the antigenic determinant of cluster V mAb probably because of the highly conformational nature of the epitope. Using 19 pathological sera from AITD and thyroid cancer with high anti-hTg titer, we obtained 14 positive results on the M3-305 peptide. Some sera which did not present crossreactivity with cluster I1 mAb on ELISA inhibition test with hTg as coated antigen (data not shown), reacted with
We thank C. Marriq for purified h Q and very much appreciated the expert technical assistance of B. Clzarvet. Received May 23, 1991; in revised form September 20, 1991.
5 References 1 Lissitsky, S. in De Groot, L. J. (Ed.), Endocrinologj vol. 1,W. B. Saunders Compagny, Philadelphia 1989, p. 512. 2 Weetman, A. P. and McGregor. A. M., Endocrine Rev. 1984. 5: 309. 3 Rochman, H . , DeGroot, L. J., Rieger, C. H. L.,Varnavides, L. A., Refetoff, S., Joung, J. I. andHoye, K., CancerRex 1975. 35: 2692. 4 Guilbert, B., Dighiero, G. and Avrameas, S., J. Immunol. 1982. 128: 2779. 5 Piechaczyk, M., Chardes,T., Cot, M. C., Pau, B. and Bastide, J. M., Hybridoma 1985. 4: 361. 6 Piechaczyk, M., Bouanani, M., Salhi, S. L., Baldet, L., Bastide, M . , Pau, B. and Bastide, J. M., Clin. Immunol. Immunopathol. 1987. 45: 114. 7 Henry, M., Malthiery,Y., Zanelli, E . and Charvct, B., J. Imrnunol. 1990. 145: 3692. 8 Jekel, €! A.,Weijer,W. J. and Beintema, J. J., Anal. Biochem. 1983. 134: 347. 9 Stanley, K. K . and Luzio, J. l?, EMBO J. 1984. 3: 1429. 10 Haymerle, H . , Hertz, J., Bressan, G. M., Frank, R. and Stanley, K. K., Nucleic Acids Kes. 1986. 14: 8615. 11 Hawkes, R . , Anal. Biochem. 1982. 123: 143. 12 Maniatis, T., Fritsch, E . F. and Sambrook, J., Moleculur cloning: a luborutory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor 1982. 13 Towbin, H . , Staehelin,T. andGordon, J., Proc. Natl. Acad. Sci. USA 1979. 76: 4350. 14 Malthiery, Y. and Lissitzky, S., Eur. J. Biochem. 1987. 165: 491. 15 Male, D. K., Champion, B. R., Pryce, G . , Matthens, H. and Shepherd, P., Immunology 1985. 54: 954. 16 Kohno, Y., Naito, N., Hiyama, Y., Shimojo, N., Suzuki, N., Tarutani, O., Niimi, H . , Nakajima, H . and Hosoya,T., J. Clin. Endocrinol. Metab. 1988. 67: 899. 17 Fukuma, N., McLachlan, S. M., Rapoport, B., Goodacre, J., Middleton, S. L., Phillips, D. I . W., Pegg, C. A. S. and Rees Smith. B., Clin. Exp. Immunol. 1990. 82: 275. 18 DeMacedo Brigido, M., Sabbaga, J. and Renzo Brentani, R., Immunnl. Lett. 1990. 24: 191. 19 Henry. M., Piechaczyk, M., Durand-Gorde. J. M., Pau, B. and Costagliola, S., Abstracts of 16th Annual Meeting of the European Thyroid Associntion, Masson, Paris, 1987, p. 130. 20 Ruf, J., Henry, M., DeMicco. C. and Carayon, l?, Workshop Proceeding Schilddruse 1Y85, Georg Thiemc Verlag, Stuttgart, New York 1987, p. 21. 21 Ruf, J., Carayon, P. and Lissitsky, S., Eur. J. Immunol. 1985. 15: 268. 22 Bouanani, M., Piechaczyk, M., Pau, B. and Bastide, M., .I. Immunol. 1989. 143: 1129. 23 Dong, G., Ludgate, M. and Vassart, G., J. Endocrinol. 1989. 122: 169. 24 Dietrich, G . , Piechaczyk, M., Pau, B. and Kazatchkine, M. D., Eur. .I. Immunol. 1991. 21: 811.
