- Email: [email protected]
Characterization of an HLA-DQ2-specific monoclonal antibody
ELSEVIER
Characterization of an HLA-DQ2-Specific Monoclonal Antibody Influence of Amino Acid Substitutions in DQP PO202 Helge D. Viken, Gunnar Paulsen, Ludvig M. Sollid, Knut E. A. Lundin, Geir E. Tjennfjord, Erik Thorsby, and Gustav Gaudernack
ABSTRACT: The HLA-DQ((Y 1*050 1 ,p lx020 1) and -DQ(a1+050 l,p 1’0202) (i.e., DQ2) heterodimers are probably involved in the pathogenesis of celiac disease and several other HLA-DQ-associated diseases. To obtain a tool for studies of these molecules, a mAb of the IgGl isotype, 2.12.El1, was produced by immunization with purified DQ(a 1*050 1, p 1*020 1) molecules and murine NIH 3T3 cells transfected with both DQA 1*050 11 and DQBl*0202. Panel cell studies with HLA homozygous B-lymphoblastoid cells and HLA-transfected murine cells demonstrated that 2.12.E 11 bound only to cells expressing HLA-DQP lx020 1 or 0202, irrespective of the accompanying DQ 01 chain (i.e., DQol1*0501 or
ABBREVIATIONS amino acid TTCC American Tissue Culture Collection B-LCL B-lymphoblastoid cell line FITC fluorescein isothiocyanate IHWS International Histocompatibility Workshop
DQa1*0201). Another DQZ-specific mAb (XIII 358.4) and the broadly HLA class-II-reactive mAb Tii39 strongly inhibited binding of 2.12.Ell. Epitope mapping employing mutants with single aa substitutions of DQpl*O202 indicated that position 37 may be of some importance for 2.12.Ell binding. A triple mutant (45G+E, 46E+V, 47-Y) failed to bind 2.12.El1, suggesting a crucial role for one or more of these residues in the epitope. However, the expression of the mutant p chain could not be formally proved, as none of the DQ2reactive mAbs recognized this transfectant. Humun Immunoiogy
IIF mAb MF NIH RIA
42, 319-327
(1995)
indirect immunofluorescence monoclonal antibody mean fluorescence National Institutes of Health radioimmunoassay
INTRODUCTION The HLA-DQ2 molecules usually include the DQ(a1*0501,~1*0201) and DQ(cx1X0201,~1*0202) heterodimers 111. The serologic DQ2 specificity is determined chains position
by DQPl”0201 except
for
a single
or DQpl”O202 amino
135 121) and is independent
acid
(identical
p
[aa} residue
in
of the accompany-
From the institute of Transplantation Immunology, The Nationa Hospital and University of Oslo, Oslo, Norway. Address reprint requests to Dr. H. D. Viken, Institute of Transplantation Immunology, The National Hospital, N-002 7 Oslo, Norway. Received(E) July 13, 1994; acceptedOctober G, 1994. Human immunology
42, 319-327 (1995) 6 American Society for Histocompatibility
and Immunogenetics,
1995
ing DQ (x chain. The DQ(a1*0501,~1+0201) (cisencoded) and DQ(a l”O50 1, p 1*0202) (tram-encoded) heterodimers are of special immunobiological interest because of the strong association to diseases such as celiac disease C33. Recent studies from our laboratory demonstrate that CD4+ T cells isolated from the small intestinal mucosa of celiac disease patients recognize gluten in the context of DQ(cr lx050 1, p 1*020 l-2) 141. Gluten-specific, DQ(a l”O50 1, f3l”O20 l-2)-restricted T-cell clones demonstrated heterogeneous sensitivity to introduced aa substitutions in the DQ2 p chain 151. 019%8859/95/$9.50 SSDI 0198-8859(94)00110-C
320
In this paper we describe the production and characterization of the DQ2-specific murine mAb 2.12.E 11 generated by immunization with purified DQ(ol l”O50 1, p 1*020 1) molecules and the murine National Institutes of Health (NIH) 3T3 cell line transfected with both DQAl*OSOll and DQBl”0202. DQ2 molecules have at least five unique sequence motifs, and more than one of these motifs may be recognized by DQ2-specific antibodies [6]. A molecular localization of the 2.12.Ell epitope was attempted using site-directed mutagenesis of codons corresponding to several aa residues specific for the DQP 1*020 l/O202 chains.
MATERIALS AND METHODS Mono&al antibodies (mAbs). The following mAbs were used: XIII 358.4 (DQ2 specific 177, a gift from Dr. C. Mazzilli, Rome, Italy); FN-8 I- 1 (anti-DQ monomorphic 181, a gift from Dr. S. Funderud, Oslo, Norway); Tii35 and Tii39 (both broadly class II reactive 191, obtained from Dr. A. Ziegler, Berlin, Germany and through the 10th International Histocompatibility Workshop {IHWS)); SPV-L3 (anti-DQ monomorphic [lo], a gift from Dr. H. Spits, Amsterdam, The Netherlands); L243 (anti-DR monomorphic) I1 11 and W6/32 (anti-class I monomorphic) {12], both obtained from American Type Culture Collection (ATCC), Rockville, MD, USA; B8.11 (anti-DR monomorphic 1131, a gift from Dr. B. Malissen, Marseille, France); B7/21 (antiDP monomorphic [14], a gift from Dr. F. Bach, Minneapolis, MN, USA); and 9A3 (DQ2 specific, unpublished results) and 12G6 (broad class II reactivity, {15a7, both made in our laboratory. Cell lines, vector-cDNA constructs, and transfection. DQAl*O5011 cDNA from the cell line RAJI El] and DQB 1+0202 cDNA from the cell line no. 9046(BH) [ 11 were gifts from Drs. L. Schenning (Uppsala, Sweden) and J. Lee (New York, NY, USA), respectively. PA317 cells [15b] (ATCC CRL 9078) were transfected with the retroviral vector pLNCL [77 with an insert of DQAl”05011 or DQB l”O202 cDNA, respectively, using Lipofectin (Bethesda Research Laboratories, Gaithersburg, MD, USA) as described [Sl. Retrovirusmediated gene transfer to NIH 3T3 cells (ATCC CRL 1658) was accomplished by growing the cells overnight, first in medium from virus producing PA3 17 cells transfected with pLNCL-DQA1*05011. The cells were then selected in medium containing 1 mg/ml neomycin (Geneticin; Gibco, Paisley, Scotland) and infected with overnight incubation in medium from PA3 17 cells transfected with pLNCL-DQB 1*0202. The NIH 3T3 cell line transfected with both DQAl”O501 and DQB l”O202 was repeatedly selected for DQ2 expression using the DQ2-specific mAb XIII
H. D. Viken et al.
358.4 coupled to sheep anti-mouse immunoglobulin beads (Dynabeads M-450; Dynal, Oslo, Norway) as described [lb]. The beads were detached from the cells using DETACHaBEAD (Dynal) to avoid phagocytosis of beads in culture. The effect of this positive immunoselection was monitored using flow cytometry (see below) with XIII 358.4 and the pan-DQ-reactive mAb FN81-1. The other HLA transfected mouse cell lines and most B-lymphoblastoid cell lines (B-LCLs) with their assigned HLA genotypes are from the 10th and 1 lth IHWSs 117, 181. The B-LCL CD 114 was locally derived from a celiac disease patient who carries DQ(a l”O50 l,p 1*020 1). Site-directed mutagenesis of the DQBl*0202 gene in codons corresponding to aas 30, 37, 45, 46, 47, 57, 71, and 74 as well as transfection of the no. 9037 B-LCL has been described in detail elsewhere {51. The BALB/c strain fusion partner NS-0 is a murine myeloma cell line. Cell culture media used were essentially as described 1197. Immunization, fusion, and culture of hybridomas. Young BALB/c mice received three antigen injections at &week intervals. The first injection was subcutaneous with an emulsion of DQ2 molecules that had been affinity purified from the DR3DQ2 homozygous B-LCL CD 114 using the HLA-DQ monomorphic mAb FN-8 l-l, in 100 ~1 Freund’s complete adjuvant (Difco, Detroit, MI, USA). The subsequent injections were intraperitoneal with 1.0 X 10’ DQAl”05011, DQB1*0202-cotransfected NIH 3T3 cells in RPM1 1640. Three days after the last injection, the spleen cells were fused with NS-0 myeloma cells using polyethylene glycol 2000 (Eastman Kodak, Rochester, NY, USA) essentially as described I197. The myeloma cells were checked for mycoplasma contamination shortly before fusion. Hybridization was postponed if the myeloma cells were less than 95% viable. After the hybridization, cells were distributed in 20 96-well flat-bottomed microtiter plates (Costar, Cambridge, MA, USA) with hypoxanthine-aminoptherine-thymidine (HAT) medium and BALB/c peritoneal macrophages as feeder cells. Hybridomas were cloned by limiting dilution at a density of 0.6 cells/well. mAb containing ascites was purified on protein A or by fast protein liquid cromatography (FPLC) and tested for isotype with immunoglobulin class- and subclass-specific antibodies (Tago, Burlingame, CA; and Southern Biotechnology Associates, Birmingham, AL, USA) in Ouchterlony double diffusion. Antibody reactivity assays. Hybridoma cultures were screened using indirect immunofluorescence (IIF) on B-LCLs in Terasaki microtiter trays (Nunc, Roskilde, Denmark) coated with poly+lysine (50 pg/ml, MW = 180.000; Sigma, St. Louis, MO, USA) in phosphatebuffered saline (PBS) to adhere cells to the plastic {20}.
HLA-DQ2-Specific
321
Antibody
These trays in inverted position were examined in a fluorescence microscope. IIF panel studies were performed using 10 pg/ml of purified mAb. Samples for flow cytometry with or without inhibiting antibody were analyzed in a FACScan flow cytometer (Becton-Dickinson, Immunocytometry Systems, San Jose, CA, USA) equipped with an air-cooled argon laser tuned at 488 nm. The results are given as linear mean fluorescence (MF) units. For all IIF tests, we used phycoerythrin or fluorescein isothiocyanate (FIT0conjugated second antibodies (Tag0 and Southern Biotechnology Associates) diluted according to the manufacturers’ recommendations. Direct immunofluorescence antibody competition studies were performed essentially as described 12 11, except that the samples were preincubated with blocking antibody for 30 minutes on ice and for 45 minutes on ice with fluoresceinated antibodies. FITC labeling of mAbs was performed using a Quick Tag conjugation kit (Boehringer Mannheim, Indianapolis, IN, USA). Rosetting with sheep-anti-mouse immunoglobulincoated magnetic beads (Dynabeads M-450; Dynal) was done essentially as described [16]. Radioimmunoassays (RIAs) with or without inhibiting antibody were performed in U-bottomed microtiter trays (Costar) with elution of bound antibody with 0.1 M glycin-HCl pH 2.0 before harvesting and counting as described [19]. Protein-A-purified mAbs were labeled with iz51 using IODO-GEN (Pierce, Rockford, IL, USA) {22]. Inhibition of T-cell proliferation. The gluten-reactive, HLA-DQ-restricted T-cell clones 4.66, 5.14, and 1.6 from the small intestinal mucosa of celiac disease patients have been described elsewhere [4, 231. In mAb inhibition assays, irradiated DR3DQ2 + or DR4DQ8 + B-LCLs were preincubated with gluten and mAbs (1:500 dilution of ascites) before addition of the T cells. Cultures were pulsed with f3H)thymidine after 48 hours and harvested 18 hours thereafter. Radioactivity was measured in a liquid scintillation counter (LKB Wallac, Turku, Finland).
TABLE 1
RESULTS Establishment of the mousebybridoma 2.12. E Il. The NIH 3T3DQ(cr. 1*050 1, f31*0202) cell line had relatively low DQ2 expression following transfection and selection in neomycin-containing medium. Before use of these cells in immunization, the fraction of cells with DQ2 expression comparable to B-LCLs in flow cytometry was increased from 17% to 89% using positive selection with the DQ2-specific mAb XIII 358.4 coupled to Dynabeads (results not shown). By screening of hybridoma supernatants for reactivity with the DR3DQ2 homozygous B-LCL no. 9023 (VAVY), 13 positive supernatants were identified. Analyzing a larger panel of B-LCLs with most DQ specificities included, positive reactions with only DQ2positive cell lines were detected in four supernatants. These supernatants did not show reactivity with reactivity with the DR5DQ7 + cell lines, excluding DQa 1*050 1 chain. One of these hybridomas (2.12.E 11) had stable production and gave a high titer of antibody. 2.12. E 11 hybridoma cells were subcloned four times and propagated in mice to produce ascites, and the mAb was purified on protein A before further characterization. 2.12.Ell was of the IgGl isotype and noncytotoxic. The yield of IgG 1 in ascites following the last subcloning was 5- 10 mg/ml. Specificity of mAb 2.12. E I1 defined by transfectants and HLA homozygotisB-LCLs. mAb 2.12.Ell reacted with the NIH 3T3 cell line transfected with both DQA lx050 11 and DQB l”O202 previously used for immunization, and with the 11th IHWS transfectant cell line no. 8204 carrying the DQPl”0202 chain in combination with DQal*0201 (Table 1). There was no reactivity with no. 8205 cells (DQc~1*0201,~1+0302). Thus, the reactivity of mAb 2.12.Ell with HLA transfected murine cells correlated with expression of DQPl*0202 and was independent of the DQa chain. There was no reactivity with NIH 3T3 cells transfected with DQAl”05011 alone or DQBl*0202 alone.
Reactivity of mAb 2.12.Ell with HLA-DQ transfected murine cells demonstrates specificity for DQ2 p chains Ttansfected DQ genes Cell line
NIH NIH NIH No. No.
3T3DQ(alX0501J31*0202) 3T3DQA1”0501 3T3DQB I*0202 82046 8205
L(IIF using 10 pg/ml mAb. b 1lth IHWS identification number.
DQAl
DQB 1
2.12.Ell reactivity”
DQA1*05011 DQAl*05011 -
DQB 1*0202
+
&BP0202 DQB I*0202 DQBl*0302
+ -
DQA 1*020 1 DQA1*020 1
322
H. D. Viken et al.
The specificity of mAb 2.12.E 11 was then analyzed using a panel of HLA homozygous B-LCLs from the 10th and 11th IHWSs (Table 2). Positive reactions were found with all cells carrying DQB lx020 1 or 0202 genes; i.e., cells being DR3DQ2 (e.g., VAVY), DR7DQ2 (PITOUT), or DR9DQ2 (ARBO). Some of the B-LCLs listed were also included in an indirect Dynabeads rosetting assay with 2.12.E 11, giving reactivity patterns identical to those observed by RIA and IIF (results not shown).
ability to inhibit the binding of FITC-labeled 2.12.E 11 to the DR7DQ2 homotygous B-LCL no. 905 1 PITOUT was analyzed (Fig. 1). Unlabeled 2.12.El1, XIII 358.4 (DQ2 specific), and Tii39 (broadly class II reactive) efficiently blocked the binding of FITC-labeled 2.12. E 11. Borderline inhibition was observed with the broadly class-II-reactive mAbs Ti.i35 and 12G6. No inhibition was found with the monomorphic mAbs SPV-L3, L243, B7/21, or W6/32 reactive with DQ, DR, DP, or class I, respectively. mAbs XIII 358.4, Tii39, and i2G6 also demonstrated inhibition of 1251-2. 12.E 11 binding to the DR3DQ2 homozygous B-LCL no. 9023 VAVY in RIA (results not shown). Binding interaction of mAbs 2.12.Ell and XIII 358.4 to DQ2 was investigated in more detail, using flow cytometry with FITC-labeled antibodies. Figure 2 shows complete two-way cross-blocking of 2.12.E 11 and XIII 358.4 with very similar results.
mAb inhibition studies. The DQ-specific mAbs (2.12.E 11, XIII 358.4, SPV-L3) and the broadly classII-reactive mAbs (12G6, Tu35, Tii39) used for the inhibition studies were shown to react with NIH 3T3 cells transfected with both DQAl”05011 and DQBl”0202 (Table 3 and results not shown). To relate the binding site of 2.12.Ell to those of reference HLA mAbs, their
TABLE
2
Specificity of mAb 2.12.Ell. IIF and RIA results using selected B-LCLs, mainly from the 10th and 11th IHWS cell panels
Cell line ID no.” 9023 9018 905 l 9102 9003 9004 9009 9008 902 1 9024 9029 9034 9040 9038 9058 9060 9062 9063 9055 9061 9064 9052 9067 907 1 9066 9070 9073
HLA
2.12.Ell
Name
DR
DQ
DQB lx
IIF
VAVY LOO81785 CD114 PITOUT ARBO KAS116 JESTHOM KASOll DO2089 15 RSH YT2 KT17 WT5 1 SAVC BMl5 BM16 OMW CB6B WDV WT47 HO301 3 1227ABO AMALA DBB BTB OLGA TAB089 LUY KT12
3 (17) 3 (17) 3 (17) 7
2 2 2 2 2
+
2 6 4 4 8 8 8 7 7 6 6 6 6 6 5 7
0201 0201 0201 0202 02 0501 0501 0502 060210603 0402 0401 0302 0302 0302 0301 0301 0603 0603 0603 0604 0609 0503 0301
: 4 6 7 9
0303 0402 0402 0601 0301 0303
9 1
5
1 2 (16) 2 (15) 3 (18) 4 4 4 4 5 5 6 6 6 6 6 6 6 7 8 8 8 8
(11) (12) (13) (13) (13) (13) (13) (14) (14)
9
reactivity
RIA (mean CPM and SD)
+ + + NT NT NT NT NT -
5040 2 NT NT 7161 k 2448 k 293 +
NT, not tested.
Histocompatibility
Workshop
(IHWS)
identification
1065 470 6
295 337 305 229 478
2 19 + 55 2 24 k 18 ” 161 NT 306 f 29 308 +. 5 243 k 36 NT 396 k 42 220 + 32 NT 317 +- 48 314 f. 31 240 2 17 327 * 13 270 k 31 416 It_ 68 333 5 18 286 -I- 39 287 2 12 NT
DQ2 alleles and antigens ate in bold.
G ID no., International
1440
number.
HLA-DQZ-Specific
TABLE
3
323
Antibody
Reactivity of a panel of mAbs with DQ2 wild-type mutant cell lines and a mutine cell line transfected DQA 1*050 11 and DQB 1*0202
and with both
Antibody 2.12.Ell
Cell line
0.3 801.8 635.6 199.1 3.8 534.2 700.9 612.8 609.0 632.2 960.4
9037 (DQal’O501,pl*0301 9037 + DQB lx0202 wild type DQp 1*0202 30-r DQp I#0202 37I+Y DQp1+0202 45-E; 46E--*V; 47P+Y DQp 1*0202 47F+Y DQ@1*0202 57A+D DQP If0202 7 lK--+T DQP lx0202 74A+E DQj3 1’0202 7 lK+T; 74A+E NIH 3T3DQ(a1*0501$1’0202)
XIII 358.4 0.7 572.0 230.1
154.5 3.1 201.5 99.2 435.6 669.7 589.8 388.7
9A3 0.1
123.7 80.2 25.1 0.4 54.0 79.4 60.6 73.2 97.3 215.3
SPV-L3 303.2 553.3 434.0 346.4 470.1 492.3 432.9 450.6 533.9 617.3 280.3
Representative results from one of four experiments given as MF linear units after subtraction of MF for the same cell line with second antibody only. MF values lower than one third of MF with the same mAb and B-LCL no. 9037 + D@ 1’0202 wild type are underlined. a No. 9037 transfected with mutant of DQB1+0202.
Reactivity of 2.12. El 1 and other mAbs with cell lines expressing mutant HLA-DQ. For a more detailed intramolecular mapping of the epitope recognized by 2.12.E 11, the reactivities of this mAb and the DQ2-specific mAbs XIII 358.4 and 9A3 with SPV-L3 as a control mAb were tested on a panel of cell lines transfected with mutated FIGURE 1 Ability of a panel of unlabeled mAbs to inhibit binding of FITC-labeled mAb 2.12.E 11 to the DQ2 homozygous B-LCL no. 905 1 PITOUT. The results are given as linear flow cytometry MF units.
No inhibitor
DQB 1+0202 (Table 3; see also Fig. 3 for a topographical view of the substituted aa residues). The mutations involved aas 30, 37, 45, 46, 47, 71, and 74, most of which are unique for DQ~1*0201/0202 compared to other DQ /3 chains and have been postulated binding sites for XIII 358.4 Lb]. The DQ2-specific mAbs 2.12.El1, XIII 358.4, and 9A3 did not bind to the untransfected DRSDQ7 + cell line no. 9037, but reacted strongly with the same cell line following transfection with the DQBl”0202 gene. Cells with DQP1+0202 substituted in all three aa positions 45, 46, and 47 demonstrated no binding of 2.12.El1, XIII 358.4, or 9A3, and cells transfected
2.12.Ell
FIGURE 2 Competition for binding to the DQ2 homozygous B-LCL no. 9051 PITOUT between FITC-labeled and unlabeled mAbs 2.12.Ell and XIII 358.4 analyzed by flow cytometry: 0, 2.12.Ell self-blocking; l , XIII 358.4 selfblocking; n, 2.12.Ell blocking of XIII 358.4 binding; and A, XIII 358.4 blocking of 2.12.Ell binding. MF in the absence of inhibiting antibody: 160 (2.12.El l-FITC) and 188 (XIII 358.4-FITC).
XIII 366.4 antLDQ2 21 1206 broad class II TX% broad class II TX!9 broad class II I
% inhibition 100
I243 anti-DR 67/21 anti-DP
50
W6/32 anti-classI I
0
I
I
30
I
I
60
I
I
90
MF
1
I 120
0
1.56
comp50ing rn*tZ&/rnl)
0 200
400
324
H. D. Viken et al.
FIGURE 3 Model of the structure of an HLA class II molecule. The figure is modified from the reported structure of DR(a,Bl*OlOl) 1251. Large black dots with aa number and aa substitutions (e.g., 3OS-Y) indicate the DQB1+0202 aa residues that were substituted in this study. Other aa differences between DQB 1*0202 and DQB l”O30 1 are given as /arge open circleswith aa number only.
p chain
with DQBl”0202 mutated in the codon for aa 37 showed reduced binding with these three mAbs. Binding of mAb XIII 358.4 to cells with substitution in DQPl”0202 aa residue 57 was also decreased. Single substitutions of aa 30, 47, 7 1, or 74 as well as both aas 7 1 and 74 had small or no effects compared to wild-type DQB l”O202. DQBl gene sequencing of the mutant DQB lx0202 cell lines, including the 45, 46, 47 triple mutant, demonstrated that transfection of the mutant DQBl”0202 genes had taken place. Following transfection with the triple mutant DQB l”O202 gene, total DQ expression of
1.6
FIGURE 4 mAb 2.12.Ell inhibits DQZ-restricted, antigenspecific T-cell clones. Bars show T-cell uptake of 13H]thymidine measured as mean CPM of triplicate cultures. T-cell clones 4.66 and 5.14 are gluten specific, HLA-DQ(a1*0501$1*0201-2) restricted; 1.6 is gluten specific DQ8 restricted.
anti-DQ 2.12.Ell c
B7/2lanti-DPm
_ , _ . c 0
2
4
6
6
10
to the other cell lines ex3).
lnhibition of TLC activation by 2.12. El 1. The mAb 2.12, E 11 has potential importance for characterization of T-cell HLA restriction elements. Figure 4 shows that restimulation of the gluten-specific, DQ(cz 1*050 1, p l*O201-2)-restricted T-cell clones 4.66 and 5.14 was inhibited by 2.12.Ell and the pan-DQ-reactive mAb SPV-L3, whereas W6/32 (class I monomorphic), B8.11 (pan-DR reactive), and B7/21 (pan-DP reactive) had no
5.14
4.66
SW-L3
this cell line was comparable pressing mutant DQ2 (Table
I
-
0
10
_
-
20
-
-
30
Cpmxld
3
HLA-DQ2-Specific
1
325
Antibody
10
20
30
40
effect. As a control, Fig. 4 also shows that the glutenspecific, HLA-DQ&restricted clone 1.6 was inhibited by SPV-L3 but not by 2.12.Ell.
50
60
70
80
90
100
FIGURE 5 Comparison of the aa sequences of the first domain of DQ p chains. The aa residues that were substituted in this study are boxed. The standard one-letter aa code is used. All sequences are aligned with the DQp1*0501 sequence at the top, and &_rbe~indicate identity to this sequence. Asterisks in the sequences indicate lack of sequence data.
DISCUSSION We describe here the production and characterization of a mAb, 2.12.El1, which reacts with all transfectants and B-LCLs expressing DQP 1*020 1 or 0202. Cells with triple substitutions of DQPl”0202 aa 45, 46, and 47 demonstrated no binding by 2.12.E 11 or the two DQ2specific mAbs XIII 358.4 and 9A3. Substitution of aa 37 resulted in reduced binding of all three mAbs. MAb XIII 358.4 also bound weakly to cells with substitution of DQB lx0202 aa residue 57. Single substitutions of aa 30, 47, 7 1, or 74 as well as both 7 1 and 74 gave small or no effects. The results obtained indicated mAb 2.12.E 11 to be specific for the B chains of HLA-DQ2 (DQP 1*0201 or 0202). First, 2.12.Ell reacts with DR3DQ2+, DR7DQ2 + , and DR9DQ2 + cell lines having only the DQ2 p chains in common. This mAb does not react with DR7DQ9 + cells, expressing the same DR but a different DQ molecule compared to a positively reacting cell line (Table 2), nor does it bind to DR5DQ7+ cehs, expressing the same DQ (Y chain but different DQ p chain compared to DR3DQ2+ cells. Second, binding of 2.12.Ell is inhibited by the DQ2-specific mAb XIII 358.4 (Fig. 1). Third, restimulation of HLA-DQ2restricted, antigen-specific T-cell clones is inhibited by 2.12.Ell (Fig. 4). The specificity for DQ2 p chains was confirmed using both murine and human transfected cell lines expressing DQBl”0202 (Tables 1 and 3). As DQ2 p chains have severa unique aa residues compared to other DQ p chains (Fig. 5), it is difficult to deduce a “best fit” hypothetical epitope for DQ2-specific mAbs on the basis of sequence data. The observed crossblocking between mAbs 2.12.Ell and XIII 358.4 indicates that these mAbs have spatially related epitopes, but does not contribute new information about the intramolecular location of the 2.12.E 11 epitope, as the
binding site of XIII 358.4 has not been described previously. mAb Tii35 (weak inhibition of 2.12.Ell binding) also has an unknown epitope. Tii39, which completely inhibited binding of 2.12.El1, has a linear epitope detected by peptide epitope scanning corresponding to aa 59-67 on the a-helix of HLA class II B chains 1241. Topographically (Fig. 3) this epitope is situated between two possible 2.12.Ell epitopes, aa 46-47 and aa 52-56 I-6, 251. Inhibition could thus be caused by steric hindrance of access of 2.12.Ell to both sites. Competition has been shown previously to occur between DQ3-specific mAbs (postulated epitope DQBl aa 53-56) 161 and a DQl,4,8,9-specific mAb (postulated epitope DQBl aa 45-47) f6] in binding to DQS molecules, and also between DQ3-specific mAbs and a DQ7-specific mAb (postulated epitope DQBl aa 45-47) I61 in binding to DQ7 molecules 12 1). Thus, the epitope of mAb Tii39 is probabIy in the vicinity of the 2.12.Ell epitope, but does not need to be identical with this epitope to explain the observed inhibition. In this study we also describe binding of Tii35 to the NIH 3T3 cell line transfected with both DQAl”05011 and DQBl”0202. Tii35 also inhibits the restimulation of DQ2_restricted, antigen-specific T-cell clones (H. A. Gjertsen, unpublished results). This, to our knowledge, previously unpublished reactivity of Tii35 with DQ2, adds to the characterization of this reference antibody. In an attempt to map the epitopes of 2.12.E 11 and other DQ2-specific mAbs in detail, we performed sitedirected mutagenesis in the codons corresponding to DQBl*0202 aas 30, 37, 45, 46, 47, 57, 7 1, and 74. For each DQPl”0202 aa substitution included in this study (Fig. 5), the aa was changed to the aa found in
326
the same position in DQpl”O30 1. The use of a human B-LCL as recipient of wild-type and mutated DQB lx0202 gave a panel of cell lines that could also be used as antigen-presenting cells for T-cell clones. The transfectant carrying three aa substitutions (aa 45, 46, and 47) gave no detectable binding of mAbs 2.12.El1, XIII 358.4, or 9A3, while a substitution in aa 47 alone did not drastically decrease binding of these antibodies, indicating that aa 45 and 46 may be crucial for binding of DQ2-specific mAbs. DNA sequence analysis has confirmed that the 45, 46, 47 triple mutant cell line has been transfected with the DQBl”0202 gene with correct mutations. We were unable to show whether mutant DQBl”0202 is expressed in the cell membrane of this cell line (see also Paulsen et al. {5]>, as all DQ2 mAbs were nonreactive. We therefore cannot discriminate between the two possibilities: lack of expression or destruction of the epitope. Cells having substitution in aa 37 also demonstrated reduced binding of 2.12.El1, XIII 358.4, and 9A3 compared to wild-type DQ2, and mAb XIII 358.4 bound relatively weakly to cells with substitution in aa 57. The finding that substitutions in several different parts of the molecule affect binding of mAbs 2.12. E 11, XIII 358.4, and 9A3 could possibly be due to allosteric effects. The 45-E and 46E*V substitutions both change the charge of the aa side chain. These combined changes could possibly alter the conformation of distant epitopes. In analogy with the previously observed salt bridge between DR@ l*OlOl 57 Asp and DR cx 76 Arg {25], the DQP1’0202 57 Ala-Asp substitution may possibly lead to salt bridge formation with DQ cx 79 Arg. Thus, substitutions in aa 57 (and aa 37, on the edge of the peptide-binding groove) could maybe exert their effects through altered peptide-binding repertoire. That DQ2-specific mAbs have varying sensitivity to aa substitutions in the DQ2 p chain is perhaps not surprising due to the several unique sequence motifs of DQ2 molecules. Gluten-specific, DQ((~1*0501,~1*0201-2)restricted T-cell clones also demonstrated heterogeneous response patterns towards antigen presented by DQPl”0202 mutant cells 151. In conclusion, we described the production and general characterization of the DQ2-specific mouse mAb 2.12. E 11, as well as the attempts to localize the epitopes of DQ2-specific mAbs using aa substitutions in DQpl”O202. mAb 2.12.Ell has been useful in a number of applications for functional studies of the diseaseassociated DQ(a1~0501,~1*0201) and DQ(c~l”0501, @1#0202) molecules, including studies of T-cell clone restriction elements (present report) and immunohistochemistry 1261, and may form the basis for a diagnostic test on biopsy specimens from celiac disease patients. 2.12.E 11 is also efficient in affinity purification of
H. D. Viken et al.
DQ(o 1*050 1,p 1*020 1) heterodimers for use in in vitro peptide-binding assays [27]. Further studies are planned for detailed mapping of the 2.12. E 11 epitope, including single substitutions of DQPl”0201 aas 28, 45, 46, 52, and 55. ACKNOWLEDGMENTS
We thank Drs. Fritz Bach, Steinar Funderud, Bernhard Malissen, Christina Mazzilli, Hergen Spits, and Andreas Ziegler for the generous gift of mAbs and Bente H. Johansen for the generous gift of affinity-purified DQ2 molecules. We are grateful to Drs. Lena Schenning and Janet Lee for the gift of DQAl*OSOll and DQB1*0202 cDNA, respectively. This work was supported in part by the Research Council of Norway (L.M.S.), Pronova (G.P., K.E.A.L.), the Norwegian Cancer Society (G.G.), Proprietaer Christopher gverland og hustru Ingeborg @verlands legat, and Aktieselskabet Freia Chocolade Fabriks Medicinske Fond. G.E.T. was supported by a research fellowship from the Norwegian Cancer Society, which was made possible thtough donations from Egil A. B&hen, Sigval Betgesen d.y. og husttu Nankis Almennyttige Stiftelse, and Betgliot og Sigurd Skaugens fond til bekjempelse av kreft. REFERENCES 1. Bodmet JG, Marsh SGE, Albert ED, Bodmet WF, Dupont B, Erlich HA, Mach B, Mayt WR, Parham P, Sazasuki T, Schreuder GMT, Strominger JL, Svejgaard A, Terasaki PI: Nomenclature for factors of the HLA system, 1994. Tissue Antigens 44:1, 1994. 2. Hall MA, Lanchbury JS, Lee JS, Welsh KI, Ciclitira PJ: HLA-DQ2 second-domain polymorphisms may explain increased trans-associated risk in celiac disease and dermatitis herpetiformis. Hum Immunol 38:284, 1993. 3. Sollid LM, Matkussen G, Ek J, Gjetde H, Vattdal F, Thorsby E: Evidence for a primary association of celiac disease to a particular HLA-DQ d/p hetetodimet. J Exp Med 169:345, 1989. 4. Lundin KEA, Scott H, Hansen T, Paulsen G, Halstensen TS, Fausa 0, Thorsby E, Sollid LM: Gliadin specific, HLA-DQ(al*0501,P 1*0201) restricted T cells isolated from the small intestinal mucosa of celiac disease patients. J Exp Med 178:187, 1993. 5. Paulsen G, Lundin KEA, Gjertsen HA, Hansen T, Sollid LM, Thorsby E: HLA-DQ2-restricted T-cell recognition of gluten-derived peptides in celiac disease: influence of amino acid substitutions in the membrane distal domain of DQ~1*0201. Hum Immunol 42, 145-153, 1995. 6. Marsh SGE, Bodmer JG: HLA-DR and -DQ epitopes and monoclonal antibody specificity. Immunol Today 10: 305, 1989. 7. Kwok WW, Schwarz D, Nepom BS, Hock RA, Thurtle PS, Nepom GT: HLA-DQ molecules form a-p heterodimers of mixed allotype. J Immunol 141:3123, 1988. 8. Holte H, Blomhoff HK, Beiske K, Fundetud S, Torjesen P, Gaudernack G, Stokke T, Smeland EB: Intracellular
HLA-DQ2-Specific
327
Antibody
events associated with inhibition of B cell activation by monoclonal antibodies to HLA class II antigens. Eur J Immunol 19: 122 1, 1989.
clonal antibody component. In Tsuji K, Aizawa M, Sazasuki T (eds): HLA 199 1, vol 1. Oxford, Oxford University Press, 1992.
9. Ziegler A, Heinig J, Miiller C, Gijtz H, Thinnes FP, Uchanska-Ziegler B, Wernet P: Analysis by sequential immunoprecipitation of the specificities of the monocional antibodies Tii22, 34, 35, 36, 37, 39, 43, 58 and YD1/63.HLK directed against human HLA class II antigens. Immunobiology 17 1:77, 1986.
18. Kimura A, Dong RP, Harada H, Sazasuki T: DNA typing of HLA class II genes in B-Iymphoblastoid cell lines homozygous for HLA. Tissue Antigens 40:5, 1992.
10. Spits H, T4 + and products complex
Ijssel H, Thompson A, de Vries JE: Human T8 + cytotoxic T lymphocyte clones directed at of different class II major histocompatibility loci. J Immunol 131:678, 1983.
19. Viken HD, Gaudernack G, Thorsby E: Characterization of a monoclonal antibody recognizing a polymorphic epitope mainly on HLA-DPw2 and DPw4 molecules. Tissue Antigens 34:250, 1989. 20.
Kolstad A, Hansen T, Hannestad K: A human-human hybridoma antibody (TrB12) defining subgroups of HLADQwl and -DQw3. Hum Immunol 20:219, 1987.
mole1980.
21.
12. Barnstable CJ, Bodmer WF, Brown G, Galfre G, Milstein C, Williams AF, Ziegler A: Production of monoclonal antibodies to group A erythrocytes, HLA and other cell surface antigens: new tools for genetic analysis. Cell 14:9, 1978.
Radka SF, Nelson KA, Johnston JV: HLA-DQw3-related determinants: analysis of subunit and spatial relationships. Hum Immunol 25:225, 1989.
22.
Markwell MAK, Fox CF: Surface specific iodination of membrane proteins of viruses and eucaryotic cells using 1,3,4,6-tetrachloro-3a,bol-diphenylglycoluril. Biochemistry 17:4807, 1978.
11. Lampson LA, Levy R: Two populations cules on a human B cell line. J Immunol
of Ia-like 125:293,
13. Rebai N, Malissen B, Pierres M, Accolla RS, Corte G, Mawas C: Distinct HLA-DR epitopes and distinct families of HLA-DR molecules defined by 15 monoclonal antibodies (mAb) either anti-DR or allo-anti-Iak crossreacting with human DR molecule. I. Cross-inhibition studies of mAb cell surface fixation and differential binding of mAb to detergent-solubilized HLA molecules immobilized to a solid phase by a first mAb. Eur J Immunol 13: 106, 1983. IS: Monoclonal an14. Royston I, Omary MB, Trowsbridge tibodies to a human T-cell antigen and Ia-like antigen in the characterization of lymphoid leukemia. Transplant Proc 8:761, 1981. 15a. Viken HD, Thorsby E, Gaudernack G: Characterization and epitope mapping of four HLA class II reactive mouse monoclonal antibodies using transfected L cells and human cells transfected with mutants of DQB 1*0302. Tissue Antigens 1995 (in press). 15b. Miller AD, Buttimore C: Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production. Mol Cell Biol 6:2895, 1986. 16. Gaudernack phenotyping
G, Lundin KEA: Rapid of cells. J Immunogenet
immunomagnetic 16: 169, 1989,
17. Inoko H, Bodmer JG, Heyes JM, Drover S, Trowsdale J, Marshall WH: Joint report on the transfectant/mono-
23. Lundin KEA, Scott H, Fausa 0, Thorsby E, Sollid LM: T cells from the small intestinal mucosa of a DR4,DQ7/ DR4,DQ8 celiac disease patient preferentially recognize gliadin when presented by DQS. Hum Immunol4 1:285, 1994. 24.
Riley GS, Worthington J, Dyer PA, Oilier WER: Epitope scanning of the HLA-DRB 1 allele sequences with monoclonal antibodies. In Tsuji K, Aizawa M, Sasazuki T (eds): HLA 199 1, ~012. Oxford, Oxford University Press, 1992.
25
Brown JH, Jardetzky TS, Gorga JC, Stern LJ, Urban RG, Strominger JL, Wiley DC: Three-dimensional structure of the human class II histocompatibility antigen HLA-DRl. Nature 364:33, 1993.
26
Halstensen TS, Scott H, Fausa 0, Brandttaeg P: Gluten stimulation of coeliac mucosa in vitro induces activation (CD25) of lamina propria CD4+ T cells and macrophages but no crypt-cell hyperplasia. Stand J Immunol 38:58 1, 1993.
27
Johansen BH, Buus S, Vartdal F, Viken H, Eriksen JA, Thorsby E, Sollid LM: Binding of peptides to HLA-DQ molecules: peptide binding properties of the diseaseassociated HLA-DQ(a lx050 1, p 1’020 1) molecule. Int Immunol 6:453, 1994.
Characterization of an HLA-DQ2-Specific Monoclonal Antibody Influence of Amino Acid Substitutions in DQP PO202 Helge D. Viken, Gunnar Paulsen, Ludvig M. Sollid, Knut E. A. Lundin, Geir E. Tjennfjord, Erik Thorsby, and Gustav Gaudernack
ABSTRACT: The HLA-DQ((Y 1*050 1 ,p lx020 1) and -DQ(a1+050 l,p 1’0202) (i.e., DQ2) heterodimers are probably involved in the pathogenesis of celiac disease and several other HLA-DQ-associated diseases. To obtain a tool for studies of these molecules, a mAb of the IgGl isotype, 2.12.El1, was produced by immunization with purified DQ(a 1*050 1, p 1*020 1) molecules and murine NIH 3T3 cells transfected with both DQA 1*050 11 and DQBl*0202. Panel cell studies with HLA homozygous B-lymphoblastoid cells and HLA-transfected murine cells demonstrated that 2.12.E 11 bound only to cells expressing HLA-DQP lx020 1 or 0202, irrespective of the accompanying DQ 01 chain (i.e., DQol1*0501 or
ABBREVIATIONS amino acid TTCC American Tissue Culture Collection B-LCL B-lymphoblastoid cell line FITC fluorescein isothiocyanate IHWS International Histocompatibility Workshop
DQa1*0201). Another DQZ-specific mAb (XIII 358.4) and the broadly HLA class-II-reactive mAb Tii39 strongly inhibited binding of 2.12.Ell. Epitope mapping employing mutants with single aa substitutions of DQpl*O202 indicated that position 37 may be of some importance for 2.12.Ell binding. A triple mutant (45G+E, 46E+V, 47-Y) failed to bind 2.12.El1, suggesting a crucial role for one or more of these residues in the epitope. However, the expression of the mutant p chain could not be formally proved, as none of the DQ2reactive mAbs recognized this transfectant. Humun Immunoiogy
IIF mAb MF NIH RIA
42, 319-327
(1995)
indirect immunofluorescence monoclonal antibody mean fluorescence National Institutes of Health radioimmunoassay
INTRODUCTION The HLA-DQ2 molecules usually include the DQ(a1*0501,~1*0201) and DQ(cx1X0201,~1*0202) heterodimers 111. The serologic DQ2 specificity is determined chains position
by DQPl”0201 except
for
a single
or DQpl”O202 amino
135 121) and is independent
acid
(identical
p
[aa} residue
in
of the accompany-
From the institute of Transplantation Immunology, The Nationa Hospital and University of Oslo, Oslo, Norway. Address reprint requests to Dr. H. D. Viken, Institute of Transplantation Immunology, The National Hospital, N-002 7 Oslo, Norway. Received(E) July 13, 1994; acceptedOctober G, 1994. Human immunology
42, 319-327 (1995) 6 American Society for Histocompatibility
and Immunogenetics,
1995
ing DQ (x chain. The DQ(a1*0501,~1+0201) (cisencoded) and DQ(a l”O50 1, p 1*0202) (tram-encoded) heterodimers are of special immunobiological interest because of the strong association to diseases such as celiac disease C33. Recent studies from our laboratory demonstrate that CD4+ T cells isolated from the small intestinal mucosa of celiac disease patients recognize gluten in the context of DQ(cr lx050 1, p 1*020 l-2) 141. Gluten-specific, DQ(a l”O50 1, f3l”O20 l-2)-restricted T-cell clones demonstrated heterogeneous sensitivity to introduced aa substitutions in the DQ2 p chain 151. 019%8859/95/$9.50 SSDI 0198-8859(94)00110-C
320
In this paper we describe the production and characterization of the DQ2-specific murine mAb 2.12.E 11 generated by immunization with purified DQ(ol l”O50 1, p 1*020 1) molecules and the murine National Institutes of Health (NIH) 3T3 cell line transfected with both DQAl*OSOll and DQBl”0202. DQ2 molecules have at least five unique sequence motifs, and more than one of these motifs may be recognized by DQ2-specific antibodies [6]. A molecular localization of the 2.12.Ell epitope was attempted using site-directed mutagenesis of codons corresponding to several aa residues specific for the DQP 1*020 l/O202 chains.
MATERIALS AND METHODS Mono&al antibodies (mAbs). The following mAbs were used: XIII 358.4 (DQ2 specific 177, a gift from Dr. C. Mazzilli, Rome, Italy); FN-8 I- 1 (anti-DQ monomorphic 181, a gift from Dr. S. Funderud, Oslo, Norway); Tii35 and Tii39 (both broadly class II reactive 191, obtained from Dr. A. Ziegler, Berlin, Germany and through the 10th International Histocompatibility Workshop {IHWS)); SPV-L3 (anti-DQ monomorphic [lo], a gift from Dr. H. Spits, Amsterdam, The Netherlands); L243 (anti-DR monomorphic) I1 11 and W6/32 (anti-class I monomorphic) {12], both obtained from American Type Culture Collection (ATCC), Rockville, MD, USA; B8.11 (anti-DR monomorphic 1131, a gift from Dr. B. Malissen, Marseille, France); B7/21 (antiDP monomorphic [14], a gift from Dr. F. Bach, Minneapolis, MN, USA); and 9A3 (DQ2 specific, unpublished results) and 12G6 (broad class II reactivity, {15a7, both made in our laboratory. Cell lines, vector-cDNA constructs, and transfection. DQAl*O5011 cDNA from the cell line RAJI El] and DQB 1+0202 cDNA from the cell line no. 9046(BH) [ 11 were gifts from Drs. L. Schenning (Uppsala, Sweden) and J. Lee (New York, NY, USA), respectively. PA317 cells [15b] (ATCC CRL 9078) were transfected with the retroviral vector pLNCL [77 with an insert of DQAl”05011 or DQB l”O202 cDNA, respectively, using Lipofectin (Bethesda Research Laboratories, Gaithersburg, MD, USA) as described [Sl. Retrovirusmediated gene transfer to NIH 3T3 cells (ATCC CRL 1658) was accomplished by growing the cells overnight, first in medium from virus producing PA3 17 cells transfected with pLNCL-DQA1*05011. The cells were then selected in medium containing 1 mg/ml neomycin (Geneticin; Gibco, Paisley, Scotland) and infected with overnight incubation in medium from PA3 17 cells transfected with pLNCL-DQB 1*0202. The NIH 3T3 cell line transfected with both DQAl”O501 and DQB l”O202 was repeatedly selected for DQ2 expression using the DQ2-specific mAb XIII
H. D. Viken et al.
358.4 coupled to sheep anti-mouse immunoglobulin beads (Dynabeads M-450; Dynal, Oslo, Norway) as described [lb]. The beads were detached from the cells using DETACHaBEAD (Dynal) to avoid phagocytosis of beads in culture. The effect of this positive immunoselection was monitored using flow cytometry (see below) with XIII 358.4 and the pan-DQ-reactive mAb FN81-1. The other HLA transfected mouse cell lines and most B-lymphoblastoid cell lines (B-LCLs) with their assigned HLA genotypes are from the 10th and 1 lth IHWSs 117, 181. The B-LCL CD 114 was locally derived from a celiac disease patient who carries DQ(a l”O50 l,p 1*020 1). Site-directed mutagenesis of the DQBl*0202 gene in codons corresponding to aas 30, 37, 45, 46, 47, 57, 71, and 74 as well as transfection of the no. 9037 B-LCL has been described in detail elsewhere {51. The BALB/c strain fusion partner NS-0 is a murine myeloma cell line. Cell culture media used were essentially as described 1197. Immunization, fusion, and culture of hybridomas. Young BALB/c mice received three antigen injections at &week intervals. The first injection was subcutaneous with an emulsion of DQ2 molecules that had been affinity purified from the DR3DQ2 homozygous B-LCL CD 114 using the HLA-DQ monomorphic mAb FN-8 l-l, in 100 ~1 Freund’s complete adjuvant (Difco, Detroit, MI, USA). The subsequent injections were intraperitoneal with 1.0 X 10’ DQAl”05011, DQB1*0202-cotransfected NIH 3T3 cells in RPM1 1640. Three days after the last injection, the spleen cells were fused with NS-0 myeloma cells using polyethylene glycol 2000 (Eastman Kodak, Rochester, NY, USA) essentially as described I197. The myeloma cells were checked for mycoplasma contamination shortly before fusion. Hybridization was postponed if the myeloma cells were less than 95% viable. After the hybridization, cells were distributed in 20 96-well flat-bottomed microtiter plates (Costar, Cambridge, MA, USA) with hypoxanthine-aminoptherine-thymidine (HAT) medium and BALB/c peritoneal macrophages as feeder cells. Hybridomas were cloned by limiting dilution at a density of 0.6 cells/well. mAb containing ascites was purified on protein A or by fast protein liquid cromatography (FPLC) and tested for isotype with immunoglobulin class- and subclass-specific antibodies (Tago, Burlingame, CA; and Southern Biotechnology Associates, Birmingham, AL, USA) in Ouchterlony double diffusion. Antibody reactivity assays. Hybridoma cultures were screened using indirect immunofluorescence (IIF) on B-LCLs in Terasaki microtiter trays (Nunc, Roskilde, Denmark) coated with poly+lysine (50 pg/ml, MW = 180.000; Sigma, St. Louis, MO, USA) in phosphatebuffered saline (PBS) to adhere cells to the plastic {20}.
HLA-DQ2-Specific
321
Antibody
These trays in inverted position were examined in a fluorescence microscope. IIF panel studies were performed using 10 pg/ml of purified mAb. Samples for flow cytometry with or without inhibiting antibody were analyzed in a FACScan flow cytometer (Becton-Dickinson, Immunocytometry Systems, San Jose, CA, USA) equipped with an air-cooled argon laser tuned at 488 nm. The results are given as linear mean fluorescence (MF) units. For all IIF tests, we used phycoerythrin or fluorescein isothiocyanate (FIT0conjugated second antibodies (Tag0 and Southern Biotechnology Associates) diluted according to the manufacturers’ recommendations. Direct immunofluorescence antibody competition studies were performed essentially as described 12 11, except that the samples were preincubated with blocking antibody for 30 minutes on ice and for 45 minutes on ice with fluoresceinated antibodies. FITC labeling of mAbs was performed using a Quick Tag conjugation kit (Boehringer Mannheim, Indianapolis, IN, USA). Rosetting with sheep-anti-mouse immunoglobulincoated magnetic beads (Dynabeads M-450; Dynal) was done essentially as described [16]. Radioimmunoassays (RIAs) with or without inhibiting antibody were performed in U-bottomed microtiter trays (Costar) with elution of bound antibody with 0.1 M glycin-HCl pH 2.0 before harvesting and counting as described [19]. Protein-A-purified mAbs were labeled with iz51 using IODO-GEN (Pierce, Rockford, IL, USA) {22]. Inhibition of T-cell proliferation. The gluten-reactive, HLA-DQ-restricted T-cell clones 4.66, 5.14, and 1.6 from the small intestinal mucosa of celiac disease patients have been described elsewhere [4, 231. In mAb inhibition assays, irradiated DR3DQ2 + or DR4DQ8 + B-LCLs were preincubated with gluten and mAbs (1:500 dilution of ascites) before addition of the T cells. Cultures were pulsed with f3H)thymidine after 48 hours and harvested 18 hours thereafter. Radioactivity was measured in a liquid scintillation counter (LKB Wallac, Turku, Finland).
TABLE 1
RESULTS Establishment of the mousebybridoma 2.12. E Il. The NIH 3T3DQ(cr. 1*050 1, f31*0202) cell line had relatively low DQ2 expression following transfection and selection in neomycin-containing medium. Before use of these cells in immunization, the fraction of cells with DQ2 expression comparable to B-LCLs in flow cytometry was increased from 17% to 89% using positive selection with the DQ2-specific mAb XIII 358.4 coupled to Dynabeads (results not shown). By screening of hybridoma supernatants for reactivity with the DR3DQ2 homozygous B-LCL no. 9023 (VAVY), 13 positive supernatants were identified. Analyzing a larger panel of B-LCLs with most DQ specificities included, positive reactions with only DQ2positive cell lines were detected in four supernatants. These supernatants did not show reactivity with reactivity with the DR5DQ7 + cell lines, excluding DQa 1*050 1 chain. One of these hybridomas (2.12.E 11) had stable production and gave a high titer of antibody. 2.12. E 11 hybridoma cells were subcloned four times and propagated in mice to produce ascites, and the mAb was purified on protein A before further characterization. 2.12.Ell was of the IgGl isotype and noncytotoxic. The yield of IgG 1 in ascites following the last subcloning was 5- 10 mg/ml. Specificity of mAb 2.12. E I1 defined by transfectants and HLA homozygotisB-LCLs. mAb 2.12.Ell reacted with the NIH 3T3 cell line transfected with both DQA lx050 11 and DQB l”O202 previously used for immunization, and with the 11th IHWS transfectant cell line no. 8204 carrying the DQPl”0202 chain in combination with DQal*0201 (Table 1). There was no reactivity with no. 8205 cells (DQc~1*0201,~1+0302). Thus, the reactivity of mAb 2.12.Ell with HLA transfected murine cells correlated with expression of DQPl*0202 and was independent of the DQa chain. There was no reactivity with NIH 3T3 cells transfected with DQAl”05011 alone or DQBl*0202 alone.
Reactivity of mAb 2.12.Ell with HLA-DQ transfected murine cells demonstrates specificity for DQ2 p chains Ttansfected DQ genes Cell line
NIH NIH NIH No. No.
3T3DQ(alX0501J31*0202) 3T3DQA1”0501 3T3DQB I*0202 82046 8205
L(IIF using 10 pg/ml mAb. b 1lth IHWS identification number.
DQAl
DQB 1
2.12.Ell reactivity”
DQA1*05011 DQAl*05011 -
DQB 1*0202
+
&BP0202 DQB I*0202 DQBl*0302
+ -
DQA 1*020 1 DQA1*020 1
322
H. D. Viken et al.
The specificity of mAb 2.12.E 11 was then analyzed using a panel of HLA homozygous B-LCLs from the 10th and 11th IHWSs (Table 2). Positive reactions were found with all cells carrying DQB lx020 1 or 0202 genes; i.e., cells being DR3DQ2 (e.g., VAVY), DR7DQ2 (PITOUT), or DR9DQ2 (ARBO). Some of the B-LCLs listed were also included in an indirect Dynabeads rosetting assay with 2.12.E 11, giving reactivity patterns identical to those observed by RIA and IIF (results not shown).
ability to inhibit the binding of FITC-labeled 2.12.E 11 to the DR7DQ2 homotygous B-LCL no. 905 1 PITOUT was analyzed (Fig. 1). Unlabeled 2.12.El1, XIII 358.4 (DQ2 specific), and Tii39 (broadly class II reactive) efficiently blocked the binding of FITC-labeled 2.12. E 11. Borderline inhibition was observed with the broadly class-II-reactive mAbs Ti.i35 and 12G6. No inhibition was found with the monomorphic mAbs SPV-L3, L243, B7/21, or W6/32 reactive with DQ, DR, DP, or class I, respectively. mAbs XIII 358.4, Tii39, and i2G6 also demonstrated inhibition of 1251-2. 12.E 11 binding to the DR3DQ2 homozygous B-LCL no. 9023 VAVY in RIA (results not shown). Binding interaction of mAbs 2.12.Ell and XIII 358.4 to DQ2 was investigated in more detail, using flow cytometry with FITC-labeled antibodies. Figure 2 shows complete two-way cross-blocking of 2.12.E 11 and XIII 358.4 with very similar results.
mAb inhibition studies. The DQ-specific mAbs (2.12.E 11, XIII 358.4, SPV-L3) and the broadly classII-reactive mAbs (12G6, Tu35, Tii39) used for the inhibition studies were shown to react with NIH 3T3 cells transfected with both DQAl”05011 and DQBl”0202 (Table 3 and results not shown). To relate the binding site of 2.12.Ell to those of reference HLA mAbs, their
TABLE
2
Specificity of mAb 2.12.Ell. IIF and RIA results using selected B-LCLs, mainly from the 10th and 11th IHWS cell panels
Cell line ID no.” 9023 9018 905 l 9102 9003 9004 9009 9008 902 1 9024 9029 9034 9040 9038 9058 9060 9062 9063 9055 9061 9064 9052 9067 907 1 9066 9070 9073
HLA
2.12.Ell
Name
DR
DQ
DQB lx
IIF
VAVY LOO81785 CD114 PITOUT ARBO KAS116 JESTHOM KASOll DO2089 15 RSH YT2 KT17 WT5 1 SAVC BMl5 BM16 OMW CB6B WDV WT47 HO301 3 1227ABO AMALA DBB BTB OLGA TAB089 LUY KT12
3 (17) 3 (17) 3 (17) 7
2 2 2 2 2
+
2 6 4 4 8 8 8 7 7 6 6 6 6 6 5 7
0201 0201 0201 0202 02 0501 0501 0502 060210603 0402 0401 0302 0302 0302 0301 0301 0603 0603 0603 0604 0609 0503 0301
: 4 6 7 9
0303 0402 0402 0601 0301 0303
9 1
5
1 2 (16) 2 (15) 3 (18) 4 4 4 4 5 5 6 6 6 6 6 6 6 7 8 8 8 8
(11) (12) (13) (13) (13) (13) (13) (14) (14)
9
reactivity
RIA (mean CPM and SD)
+ + + NT NT NT NT NT -
5040 2 NT NT 7161 k 2448 k 293 +
NT, not tested.
Histocompatibility
Workshop
(IHWS)
identification
1065 470 6
295 337 305 229 478
2 19 + 55 2 24 k 18 ” 161 NT 306 f 29 308 +. 5 243 k 36 NT 396 k 42 220 + 32 NT 317 +- 48 314 f. 31 240 2 17 327 * 13 270 k 31 416 It_ 68 333 5 18 286 -I- 39 287 2 12 NT
DQ2 alleles and antigens ate in bold.
G ID no., International
1440
number.
HLA-DQZ-Specific
TABLE
3
323
Antibody
Reactivity of a panel of mAbs with DQ2 wild-type mutant cell lines and a mutine cell line transfected DQA 1*050 11 and DQB 1*0202
and with both
Antibody 2.12.Ell
Cell line
0.3 801.8 635.6 199.1 3.8 534.2 700.9 612.8 609.0 632.2 960.4
9037 (DQal’O501,pl*0301 9037 + DQB lx0202 wild type DQp 1*0202 30-r DQp I#0202 37I+Y DQp1+0202 45-E; 46E--*V; 47P+Y DQp 1*0202 47F+Y DQ@1*0202 57A+D DQP If0202 7 lK--+T DQP lx0202 74A+E DQj3 1’0202 7 lK+T; 74A+E NIH 3T3DQ(a1*0501$1’0202)
XIII 358.4 0.7 572.0 230.1
154.5 3.1 201.5 99.2 435.6 669.7 589.8 388.7
9A3 0.1
123.7 80.2 25.1 0.4 54.0 79.4 60.6 73.2 97.3 215.3
SPV-L3 303.2 553.3 434.0 346.4 470.1 492.3 432.9 450.6 533.9 617.3 280.3
Representative results from one of four experiments given as MF linear units after subtraction of MF for the same cell line with second antibody only. MF values lower than one third of MF with the same mAb and B-LCL no. 9037 + D@ 1’0202 wild type are underlined. a No. 9037 transfected with mutant of DQB1+0202.
Reactivity of 2.12. El 1 and other mAbs with cell lines expressing mutant HLA-DQ. For a more detailed intramolecular mapping of the epitope recognized by 2.12.E 11, the reactivities of this mAb and the DQ2-specific mAbs XIII 358.4 and 9A3 with SPV-L3 as a control mAb were tested on a panel of cell lines transfected with mutated FIGURE 1 Ability of a panel of unlabeled mAbs to inhibit binding of FITC-labeled mAb 2.12.E 11 to the DQ2 homozygous B-LCL no. 905 1 PITOUT. The results are given as linear flow cytometry MF units.
No inhibitor
DQB 1+0202 (Table 3; see also Fig. 3 for a topographical view of the substituted aa residues). The mutations involved aas 30, 37, 45, 46, 47, 71, and 74, most of which are unique for DQ~1*0201/0202 compared to other DQ /3 chains and have been postulated binding sites for XIII 358.4 Lb]. The DQ2-specific mAbs 2.12.El1, XIII 358.4, and 9A3 did not bind to the untransfected DRSDQ7 + cell line no. 9037, but reacted strongly with the same cell line following transfection with the DQBl”0202 gene. Cells with DQP1+0202 substituted in all three aa positions 45, 46, and 47 demonstrated no binding of 2.12.El1, XIII 358.4, or 9A3, and cells transfected
2.12.Ell
FIGURE 2 Competition for binding to the DQ2 homozygous B-LCL no. 9051 PITOUT between FITC-labeled and unlabeled mAbs 2.12.Ell and XIII 358.4 analyzed by flow cytometry: 0, 2.12.Ell self-blocking; l , XIII 358.4 selfblocking; n, 2.12.Ell blocking of XIII 358.4 binding; and A, XIII 358.4 blocking of 2.12.Ell binding. MF in the absence of inhibiting antibody: 160 (2.12.El l-FITC) and 188 (XIII 358.4-FITC).
XIII 366.4 antLDQ2 21 1206 broad class II TX% broad class II TX!9 broad class II I
% inhibition 100
I243 anti-DR 67/21 anti-DP
50
W6/32 anti-classI I
0
I
I
30
I
I
60
I
I
90
MF
1
I 120
0
1.56
comp50ing rn*tZ&/rnl)
0 200
400
324
H. D. Viken et al.
FIGURE 3 Model of the structure of an HLA class II molecule. The figure is modified from the reported structure of DR(a,Bl*OlOl) 1251. Large black dots with aa number and aa substitutions (e.g., 3OS-Y) indicate the DQB1+0202 aa residues that were substituted in this study. Other aa differences between DQB 1*0202 and DQB l”O30 1 are given as /arge open circleswith aa number only.
p chain
with DQBl”0202 mutated in the codon for aa 37 showed reduced binding with these three mAbs. Binding of mAb XIII 358.4 to cells with substitution in DQPl”0202 aa residue 57 was also decreased. Single substitutions of aa 30, 47, 7 1, or 74 as well as both aas 7 1 and 74 had small or no effects compared to wild-type DQB l”O202. DQBl gene sequencing of the mutant DQB lx0202 cell lines, including the 45, 46, 47 triple mutant, demonstrated that transfection of the mutant DQBl”0202 genes had taken place. Following transfection with the triple mutant DQB l”O202 gene, total DQ expression of
1.6
FIGURE 4 mAb 2.12.Ell inhibits DQZ-restricted, antigenspecific T-cell clones. Bars show T-cell uptake of 13H]thymidine measured as mean CPM of triplicate cultures. T-cell clones 4.66 and 5.14 are gluten specific, HLA-DQ(a1*0501$1*0201-2) restricted; 1.6 is gluten specific DQ8 restricted.
anti-DQ 2.12.Ell c
B7/2lanti-DPm
_ , _ . c 0
2
4
6
6
10
to the other cell lines ex3).
lnhibition of TLC activation by 2.12. El 1. The mAb 2.12, E 11 has potential importance for characterization of T-cell HLA restriction elements. Figure 4 shows that restimulation of the gluten-specific, DQ(cz 1*050 1, p l*O201-2)-restricted T-cell clones 4.66 and 5.14 was inhibited by 2.12.Ell and the pan-DQ-reactive mAb SPV-L3, whereas W6/32 (class I monomorphic), B8.11 (pan-DR reactive), and B7/21 (pan-DP reactive) had no
5.14
4.66
SW-L3
this cell line was comparable pressing mutant DQ2 (Table
I
-
0
10
_
-
20
-
-
30
Cpmxld
3
HLA-DQ2-Specific
1
325
Antibody
10
20
30
40
effect. As a control, Fig. 4 also shows that the glutenspecific, HLA-DQ&restricted clone 1.6 was inhibited by SPV-L3 but not by 2.12.Ell.
50
60
70
80
90
100
FIGURE 5 Comparison of the aa sequences of the first domain of DQ p chains. The aa residues that were substituted in this study are boxed. The standard one-letter aa code is used. All sequences are aligned with the DQp1*0501 sequence at the top, and &_rbe~indicate identity to this sequence. Asterisks in the sequences indicate lack of sequence data.
DISCUSSION We describe here the production and characterization of a mAb, 2.12.El1, which reacts with all transfectants and B-LCLs expressing DQP 1*020 1 or 0202. Cells with triple substitutions of DQPl”0202 aa 45, 46, and 47 demonstrated no binding by 2.12.E 11 or the two DQ2specific mAbs XIII 358.4 and 9A3. Substitution of aa 37 resulted in reduced binding of all three mAbs. MAb XIII 358.4 also bound weakly to cells with substitution of DQB lx0202 aa residue 57. Single substitutions of aa 30, 47, 7 1, or 74 as well as both 7 1 and 74 gave small or no effects. The results obtained indicated mAb 2.12.E 11 to be specific for the B chains of HLA-DQ2 (DQP 1*0201 or 0202). First, 2.12.Ell reacts with DR3DQ2+, DR7DQ2 + , and DR9DQ2 + cell lines having only the DQ2 p chains in common. This mAb does not react with DR7DQ9 + cells, expressing the same DR but a different DQ molecule compared to a positively reacting cell line (Table 2), nor does it bind to DR5DQ7+ cehs, expressing the same DQ (Y chain but different DQ p chain compared to DR3DQ2+ cells. Second, binding of 2.12.Ell is inhibited by the DQ2-specific mAb XIII 358.4 (Fig. 1). Third, restimulation of HLA-DQ2restricted, antigen-specific T-cell clones is inhibited by 2.12.Ell (Fig. 4). The specificity for DQ2 p chains was confirmed using both murine and human transfected cell lines expressing DQBl”0202 (Tables 1 and 3). As DQ2 p chains have severa unique aa residues compared to other DQ p chains (Fig. 5), it is difficult to deduce a “best fit” hypothetical epitope for DQ2-specific mAbs on the basis of sequence data. The observed crossblocking between mAbs 2.12.Ell and XIII 358.4 indicates that these mAbs have spatially related epitopes, but does not contribute new information about the intramolecular location of the 2.12.E 11 epitope, as the
binding site of XIII 358.4 has not been described previously. mAb Tii35 (weak inhibition of 2.12.Ell binding) also has an unknown epitope. Tii39, which completely inhibited binding of 2.12.El1, has a linear epitope detected by peptide epitope scanning corresponding to aa 59-67 on the a-helix of HLA class II B chains 1241. Topographically (Fig. 3) this epitope is situated between two possible 2.12.Ell epitopes, aa 46-47 and aa 52-56 I-6, 251. Inhibition could thus be caused by steric hindrance of access of 2.12.Ell to both sites. Competition has been shown previously to occur between DQ3-specific mAbs (postulated epitope DQBl aa 53-56) 161 and a DQl,4,8,9-specific mAb (postulated epitope DQBl aa 45-47) f6] in binding to DQS molecules, and also between DQ3-specific mAbs and a DQ7-specific mAb (postulated epitope DQBl aa 45-47) I61 in binding to DQ7 molecules 12 1). Thus, the epitope of mAb Tii39 is probabIy in the vicinity of the 2.12.Ell epitope, but does not need to be identical with this epitope to explain the observed inhibition. In this study we also describe binding of Tii35 to the NIH 3T3 cell line transfected with both DQAl”05011 and DQBl”0202. Tii35 also inhibits the restimulation of DQ2_restricted, antigen-specific T-cell clones (H. A. Gjertsen, unpublished results). This, to our knowledge, previously unpublished reactivity of Tii35 with DQ2, adds to the characterization of this reference antibody. In an attempt to map the epitopes of 2.12.E 11 and other DQ2-specific mAbs in detail, we performed sitedirected mutagenesis in the codons corresponding to DQBl*0202 aas 30, 37, 45, 46, 47, 57, 7 1, and 74. For each DQPl”0202 aa substitution included in this study (Fig. 5), the aa was changed to the aa found in
326
the same position in DQpl”O30 1. The use of a human B-LCL as recipient of wild-type and mutated DQB lx0202 gave a panel of cell lines that could also be used as antigen-presenting cells for T-cell clones. The transfectant carrying three aa substitutions (aa 45, 46, and 47) gave no detectable binding of mAbs 2.12.El1, XIII 358.4, or 9A3, while a substitution in aa 47 alone did not drastically decrease binding of these antibodies, indicating that aa 45 and 46 may be crucial for binding of DQ2-specific mAbs. DNA sequence analysis has confirmed that the 45, 46, 47 triple mutant cell line has been transfected with the DQBl”0202 gene with correct mutations. We were unable to show whether mutant DQBl”0202 is expressed in the cell membrane of this cell line (see also Paulsen et al. {5]>, as all DQ2 mAbs were nonreactive. We therefore cannot discriminate between the two possibilities: lack of expression or destruction of the epitope. Cells having substitution in aa 37 also demonstrated reduced binding of 2.12.El1, XIII 358.4, and 9A3 compared to wild-type DQ2, and mAb XIII 358.4 bound relatively weakly to cells with substitution in aa 57. The finding that substitutions in several different parts of the molecule affect binding of mAbs 2.12. E 11, XIII 358.4, and 9A3 could possibly be due to allosteric effects. The 45-E and 46E*V substitutions both change the charge of the aa side chain. These combined changes could possibly alter the conformation of distant epitopes. In analogy with the previously observed salt bridge between DR@ l*OlOl 57 Asp and DR cx 76 Arg {25], the DQP1’0202 57 Ala-Asp substitution may possibly lead to salt bridge formation with DQ cx 79 Arg. Thus, substitutions in aa 57 (and aa 37, on the edge of the peptide-binding groove) could maybe exert their effects through altered peptide-binding repertoire. That DQ2-specific mAbs have varying sensitivity to aa substitutions in the DQ2 p chain is perhaps not surprising due to the several unique sequence motifs of DQ2 molecules. Gluten-specific, DQ((~1*0501,~1*0201-2)restricted T-cell clones also demonstrated heterogeneous response patterns towards antigen presented by DQPl”0202 mutant cells 151. In conclusion, we described the production and general characterization of the DQ2-specific mouse mAb 2.12. E 11, as well as the attempts to localize the epitopes of DQ2-specific mAbs using aa substitutions in DQpl”O202. mAb 2.12.Ell has been useful in a number of applications for functional studies of the diseaseassociated DQ(a1~0501,~1*0201) and DQ(c~l”0501, @1#0202) molecules, including studies of T-cell clone restriction elements (present report) and immunohistochemistry 1261, and may form the basis for a diagnostic test on biopsy specimens from celiac disease patients. 2.12.E 11 is also efficient in affinity purification of
H. D. Viken et al.
DQ(o 1*050 1,p 1*020 1) heterodimers for use in in vitro peptide-binding assays [27]. Further studies are planned for detailed mapping of the 2.12. E 11 epitope, including single substitutions of DQPl”0201 aas 28, 45, 46, 52, and 55. ACKNOWLEDGMENTS
We thank Drs. Fritz Bach, Steinar Funderud, Bernhard Malissen, Christina Mazzilli, Hergen Spits, and Andreas Ziegler for the generous gift of mAbs and Bente H. Johansen for the generous gift of affinity-purified DQ2 molecules. We are grateful to Drs. Lena Schenning and Janet Lee for the gift of DQAl*OSOll and DQB1*0202 cDNA, respectively. This work was supported in part by the Research Council of Norway (L.M.S.), Pronova (G.P., K.E.A.L.), the Norwegian Cancer Society (G.G.), Proprietaer Christopher gverland og hustru Ingeborg @verlands legat, and Aktieselskabet Freia Chocolade Fabriks Medicinske Fond. G.E.T. was supported by a research fellowship from the Norwegian Cancer Society, which was made possible thtough donations from Egil A. B&hen, Sigval Betgesen d.y. og husttu Nankis Almennyttige Stiftelse, and Betgliot og Sigurd Skaugens fond til bekjempelse av kreft. REFERENCES 1. Bodmet JG, Marsh SGE, Albert ED, Bodmet WF, Dupont B, Erlich HA, Mach B, Mayt WR, Parham P, Sazasuki T, Schreuder GMT, Strominger JL, Svejgaard A, Terasaki PI: Nomenclature for factors of the HLA system, 1994. Tissue Antigens 44:1, 1994. 2. Hall MA, Lanchbury JS, Lee JS, Welsh KI, Ciclitira PJ: HLA-DQ2 second-domain polymorphisms may explain increased trans-associated risk in celiac disease and dermatitis herpetiformis. Hum Immunol 38:284, 1993. 3. Sollid LM, Matkussen G, Ek J, Gjetde H, Vattdal F, Thorsby E: Evidence for a primary association of celiac disease to a particular HLA-DQ d/p hetetodimet. J Exp Med 169:345, 1989. 4. Lundin KEA, Scott H, Hansen T, Paulsen G, Halstensen TS, Fausa 0, Thorsby E, Sollid LM: Gliadin specific, HLA-DQ(al*0501,P 1*0201) restricted T cells isolated from the small intestinal mucosa of celiac disease patients. J Exp Med 178:187, 1993. 5. Paulsen G, Lundin KEA, Gjertsen HA, Hansen T, Sollid LM, Thorsby E: HLA-DQ2-restricted T-cell recognition of gluten-derived peptides in celiac disease: influence of amino acid substitutions in the membrane distal domain of DQ~1*0201. Hum Immunol 42, 145-153, 1995. 6. Marsh SGE, Bodmer JG: HLA-DR and -DQ epitopes and monoclonal antibody specificity. Immunol Today 10: 305, 1989. 7. Kwok WW, Schwarz D, Nepom BS, Hock RA, Thurtle PS, Nepom GT: HLA-DQ molecules form a-p heterodimers of mixed allotype. J Immunol 141:3123, 1988. 8. Holte H, Blomhoff HK, Beiske K, Fundetud S, Torjesen P, Gaudernack G, Stokke T, Smeland EB: Intracellular
HLA-DQ2-Specific
327
Antibody
events associated with inhibition of B cell activation by monoclonal antibodies to HLA class II antigens. Eur J Immunol 19: 122 1, 1989.
clonal antibody component. In Tsuji K, Aizawa M, Sazasuki T (eds): HLA 199 1, vol 1. Oxford, Oxford University Press, 1992.
9. Ziegler A, Heinig J, Miiller C, Gijtz H, Thinnes FP, Uchanska-Ziegler B, Wernet P: Analysis by sequential immunoprecipitation of the specificities of the monocional antibodies Tii22, 34, 35, 36, 37, 39, 43, 58 and YD1/63.HLK directed against human HLA class II antigens. Immunobiology 17 1:77, 1986.
18. Kimura A, Dong RP, Harada H, Sazasuki T: DNA typing of HLA class II genes in B-Iymphoblastoid cell lines homozygous for HLA. Tissue Antigens 40:5, 1992.
10. Spits H, T4 + and products complex
Ijssel H, Thompson A, de Vries JE: Human T8 + cytotoxic T lymphocyte clones directed at of different class II major histocompatibility loci. J Immunol 131:678, 1983.
19. Viken HD, Gaudernack G, Thorsby E: Characterization of a monoclonal antibody recognizing a polymorphic epitope mainly on HLA-DPw2 and DPw4 molecules. Tissue Antigens 34:250, 1989. 20.
Kolstad A, Hansen T, Hannestad K: A human-human hybridoma antibody (TrB12) defining subgroups of HLADQwl and -DQw3. Hum Immunol 20:219, 1987.
mole1980.
21.
12. Barnstable CJ, Bodmer WF, Brown G, Galfre G, Milstein C, Williams AF, Ziegler A: Production of monoclonal antibodies to group A erythrocytes, HLA and other cell surface antigens: new tools for genetic analysis. Cell 14:9, 1978.
Radka SF, Nelson KA, Johnston JV: HLA-DQw3-related determinants: analysis of subunit and spatial relationships. Hum Immunol 25:225, 1989.
22.
Markwell MAK, Fox CF: Surface specific iodination of membrane proteins of viruses and eucaryotic cells using 1,3,4,6-tetrachloro-3a,bol-diphenylglycoluril. Biochemistry 17:4807, 1978.
11. Lampson LA, Levy R: Two populations cules on a human B cell line. J Immunol
of Ia-like 125:293,
13. Rebai N, Malissen B, Pierres M, Accolla RS, Corte G, Mawas C: Distinct HLA-DR epitopes and distinct families of HLA-DR molecules defined by 15 monoclonal antibodies (mAb) either anti-DR or allo-anti-Iak crossreacting with human DR molecule. I. Cross-inhibition studies of mAb cell surface fixation and differential binding of mAb to detergent-solubilized HLA molecules immobilized to a solid phase by a first mAb. Eur J Immunol 13: 106, 1983. IS: Monoclonal an14. Royston I, Omary MB, Trowsbridge tibodies to a human T-cell antigen and Ia-like antigen in the characterization of lymphoid leukemia. Transplant Proc 8:761, 1981. 15a. Viken HD, Thorsby E, Gaudernack G: Characterization and epitope mapping of four HLA class II reactive mouse monoclonal antibodies using transfected L cells and human cells transfected with mutants of DQB 1*0302. Tissue Antigens 1995 (in press). 15b. Miller AD, Buttimore C: Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production. Mol Cell Biol 6:2895, 1986. 16. Gaudernack phenotyping
G, Lundin KEA: Rapid of cells. J Immunogenet
immunomagnetic 16: 169, 1989,
17. Inoko H, Bodmer JG, Heyes JM, Drover S, Trowsdale J, Marshall WH: Joint report on the transfectant/mono-
23. Lundin KEA, Scott H, Fausa 0, Thorsby E, Sollid LM: T cells from the small intestinal mucosa of a DR4,DQ7/ DR4,DQ8 celiac disease patient preferentially recognize gliadin when presented by DQS. Hum Immunol4 1:285, 1994. 24.
Riley GS, Worthington J, Dyer PA, Oilier WER: Epitope scanning of the HLA-DRB 1 allele sequences with monoclonal antibodies. In Tsuji K, Aizawa M, Sasazuki T (eds): HLA 199 1, ~012. Oxford, Oxford University Press, 1992.
25
Brown JH, Jardetzky TS, Gorga JC, Stern LJ, Urban RG, Strominger JL, Wiley DC: Three-dimensional structure of the human class II histocompatibility antigen HLA-DRl. Nature 364:33, 1993.
26
Halstensen TS, Scott H, Fausa 0, Brandttaeg P: Gluten stimulation of coeliac mucosa in vitro induces activation (CD25) of lamina propria CD4+ T cells and macrophages but no crypt-cell hyperplasia. Stand J Immunol 38:58 1, 1993.
27
Johansen BH, Buus S, Vartdal F, Viken H, Eriksen JA, Thorsby E, Sollid LM: Binding of peptides to HLA-DQ molecules: peptide binding properties of the diseaseassociated HLA-DQ(a lx050 1, p 1’020 1) molecule. Int Immunol 6:453, 1994.