Alimentary Pharmacology and Therapeutics
Dissecting the T-cell response to hordeins in coeliac disease can develop barley with reduced immunotoxicity G. J. Tanner*, C. A. Howitt*, R. I. Forrester , P. M. Campbellà, J. A. Tye-Din§,– & R. P. Anderson§,–
*CSIRO Food Futures National Research Flagship, Canberra, ACT, Australia. CSIRO Plant Industry, Canberra, ACT, Australia. à CSIRO Ecosystem Sciences, Canberra, ACT, Australia. § The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Vic. 3052, Australia. – Department of Gastroenterology, The Royal Melbourne Hospital, Grattan St., Parkville, Vic. 3050, Australia.
Correspondence to: Dr G. J. Tanner, CSIRO Food Futures National Research Flagship, GPO Box 1600, Canberra, ACT, 2601, Australia. E-mail:
[email protected]
Publication data Submitted 8 June 2010 First decision 30 June 2010 Resubmitted 16 August 2010 Accepted 19 August 2010 EV Pub Online 15 September 2010
SUMMARY Background Wheat, rye and barley prolamins are toxic to patients with coeliac disease. Barley is diploid with pure inbred cultivars available, and is attractive for genetic approaches to detoxification. Aim To identify barley hordein fractions which activated the interferon-c (IFNc) secreting peripheral blood T-cells from coeliac volunteers, and compare immunotoxicity of hordeins from experimental barley lines. Methods To reactivate a T-cell response to hordein, volunteers underwent a 3-day oral barley challenge. Peripheral blood mononuclear cells (PBMC) were isolated from twenty-one HLA DQ2+ patients with confirmed coeliac disease. IFN-c ELISpot assays enumerated T-cells activated by purified prolamins and positive controls. Results Hordein-specific T-cells were induced by oral barley challenge. All prolamin fractions were immunotoxic, but D- and C-hordeins were most active. Barley lines lacking B- and C-hordeins had a 5-fold reduced hordeincontent, and immunotoxicity of hordein extracts were reduced 20-fold compared with wild-type barley. Conclusions In vivo oral barley challenge offers a convenient and rapid approach to test the immunotoxicity of small amounts of purified hordeins using fresh T-cells from patients in high throughput overnight assays. Barley lines that did not accumulate B- and C-hordeins were viable, yet had substantially reduced immunotoxicity. Creation of hordein-free barley may therefore be possible. Aliment Pharmacol Ther 2010; 32: 1184–1191
1184
ª 2010 Blackwell Publishing Ltd doi:10.1111/j.1365-2036.2010.04452.x
Barley hordeins are immunotoxic in coeliac disease INTRODUCTION Wheat, rye and barley cannot be included in the diet of patients with coeliac disease (CD).1–3 We recently showed that most of the toxicity of gluten and the related grain storage proteins from wheat (gliadins and glutenins), rye (secalins) and barley (hordeins) can be attributed to T cells recognizing three critical peptides.4 Two of the three peptides critical to gluten’s toxicity are present in barley. We set out to determine whether barley toxicity might be reduced by removing two dominant protein families, the B- and C-hordeins. In contrast to the genetic complexity of modern bread-wheat, which has three genomes, barley has a simpler, diploid genome.5 Pure inbred barley lines are available, which do not accumulate members of the B-hordeins (Risø 56) and the C-hordeins (Risø 1508).6 Together, these hordein families account for over 90% of the hordeins in barley; peptides derived from these protein families account for much of the toxicity of barley in the 80% of patients with coeliac disease who are HLA DQ2+ without other HLA susceptibility genes.5, 7 We took advantage of hordein-reactive T cells present in the blood of volunteers with coeliac disease 6-days after commencing oral barley challenge4 to screen hordein fractions from standard barley strains and hybrids of Risø 56 and Risø 1508, to identify ‘Low Gluten Barley’ lines with reduced toxicity in coeliac disease. METHODS Patients and T cell studies The study was approved by the Melbourne Health Human Research Ethics Committee. Thirty-one volunteers were recruited from the Coeliac Society of Victoria; membership of the organization requires a medical doctor’s letter indicating a proven need for gluten free diet (GFD). Subjects were required to be aged between 18 and 70 years and to have followed a strict GFD for at least 8 weeks. All participants had biopsy-proven CD conforming to ESPGAN diagnostic criteria8 and were included if they possessed HLA DQA1*05 and DQB1*02 (encoding HLA DQ2), but did not possess either DQA1*03 or DQB1*0302 (alleles encoding HLA DQ8). Leucocyte-derived DNA from volunteers was genotyped using a panel of sequence-specific primers to determine HLA DQA and DQB alleles (Victorian Transplantation Immunology Service, Parkville, Australia). Compliance with GFD was confirmed by a self-administered questionnaire and transglutaminase-IgA (tTG-IgA) (INOVA Diagnostics, San Diego, CA, USA) serology being 20 SFU ⁄ well). Cereal sources of prolamins. Prolamins were isolated from a variety of grains. Wild-type barley (cv Himalaya, Sloop, Bomi, and Carlsberg II), barley single hordeinnulls (Risø 56, and Risø 1508), Low Gluten Barley lines (below), maize (cv W22), oats (cv Jumbo) and wheat (cv Chara) were inspected to exclude contamination, ground to a flour in the order from least toxic to most toxic and the machine was dissembled and void spaces and surfaces were cleaned between samples. Risø 56 and Risø 1508 barley seeds deficient in B- or C-hordein protein families, respectively, due to single, recessive mutations10 were obtained from the Nordic Gene Bank and crossed to yield five hordein double-null F2 barley seeds (J1, G1, 9RE, 4BH, and 5RB) that did not accumulate significant B- or C- hordein. F4 families of seeds (Low Gluten Barley) were produced from these F2 seeds. Prolamin extraction. Prolamins in aqueous washed wholemeal flour (10 gm) were dissolved in 50% (v ⁄ v) propan-2-ol, 1% (w ⁄ v) DTT (IPA ⁄ DTT), and precipitated with two volumes of propan-2-ol at )20 C. The precipitated prolamins were dissolved in 8 M urea, 1% DTT, 25 mM TEA (pH 6), and purified by FPLC on a 4 mL Resource RPC column (GE Healthcare Australia Pty. Ltd., Sydney, NSW, Australia) eluted with a 30 mL linear gradient (2 mL ⁄ min) from 3% to 60% acetonitrile in 1% (v ⁄ v) trifluoroacetic acid. Hordeins from barley cv Himalaya were fractionated with a shallower stepwise 1185
G. J. Tanner et al.
M
#1
#2
#3
#4
#5
#6
(b) 3
100 kDa-
6
4
7
50 kDa40 kDa30 kDa-
8
5 20 kDa-
Sloop
CII
Bomi
K8
R56
R1508
9RE
A280
20 kDa-
5RB
%B
50
30 kDa-
4BH
0.4
40 kDa-
J1
100
50 kDa-
G1
Protein analysis Lyophylized hordeins were dissolved in 6M urea, 2% (w ⁄ v) SDS, 1% (w ⁄ v) DTT at room temperature and analysed by SDS-PAGE (Figure 2). One millimetre plugs were cut from stained gels and proteins reduced with 1% (w ⁄ v) DTT, alkylated with acrylamide, cleaved with sequencing grade trypsin (Promega Corporation, Sydney, NSW, Australia) and analysed by HPLC MS ⁄ MS.11 Mass spectral data sets were used initially to search a sequence database for common contaminants (trypsin, keratins) and
(a) 100 kDa-
M
gradient (0 min: acetonitrile 3% in 1% (v ⁄ v) trifluoroacetic acid, 2 min: 30%, 8.5 min: 31.2%, 17 min: 36.6%, 18.5 min: 34.8%, 20.5 min: 36%, 22 min: 37.2%, 43.5 min: 38.4%, 25 min: 39.6%, and 27.5 min: 60%). The total eluant, or hordein fractions enriched for particular hordein families, from cv Himalaya, corresponding to A280 peaks, were collected and lyophylised (Figures 1 and 2a). Solvent controls were similarly pooled from runs without an injection, and the FPLC column cleaned to prevent contamination between samples. The composition of the hordein fractions was established by SDS-PAGE (Figure 2a), two-dimensional electrophoresis and HPLC. Fraction 1 contained mostly D-hordein at 100 kDa with traces of C-hordeins at 47 and 48 kDa (Figure 2a, #1). Fraction 2, (Figure 2a, #2) contained C-hordeins. Fraction 3 (Figure 2, #3) contained both C- and B- hordeins. Hordein fractions 4, 5, 6, (Figure 2a, #4, #5, #6 respectively) contained mostly B- hordeins with a small amount of gammahordeins.
Figure 2 | (a) Analysis of hordein fractions: 20 lg of fraction 1 (#1), 2 (#2), 3 (#3), 4 (#4), 5 (#5), or 6 (#6) was analysed by SDS-PAGE. (b) Total hordeins were purified from the indicated lines and 20 lg separated by SDS-PAGE. Low Gluten Barley lines were G1, J1, 4BH, 5RB, 9RE; the parents Risø 56 and Risø 1508, and wild-type barley Bomi, Carlsberg II (CII), and Sloop. Line K8 was a wild-type sibling from the same cross as the Low Gluten Barley lines. Protein spots were cut from the gel as indicated, and identified by mass spectral analysis of tryptic peptides as (3) D-hordein, (4) B3-hordein, (5) gamma-3-hordein, (6) gamma-1-precursor, (7) gamma-1-hordein precursor, (8) gamma-3hordein. The position of molecular weight standards (M; BenchMark, Invitrogen, Mulgrave, Victoria, Australia) is indicated.
0.2
#1
0.0 0
#2 20
#3
#4
40 Fraction (1 mL)
#5
#6 60
0
Figure 1 | Reverse-phase FPLC of hordeins: A representative chromatogram showing the A280 (solid line) of the effluent, and the solvent composition (broken line) during isolation of six hordein fractions from a barley extract. The indicated fractions were pooled as shown (bold line) from sequential injections. 1186
then the Hordeum subset of the NCBI nonredundant protein database using SPECTRUMMILL software (Agilent Technologies Australia Pty Ltd, Forest Hill, Victoria, Australia; Software Rev A.03.02.060) allowing the possibility of oxidized methionine and one missed cleavage. The default ‘autovalidation’ criteria were used to accept identifications with summed MS ⁄ MS search scores over 20 and 2–4 distinct peptides. The MS ⁄ MS data that were not yet validated were then used to re-search the database, allowing the possibility of ‘semi-tryptic’ peptides. Additional hordein peptides, resulting from a trypsin Aliment Pharmacol Ther 2010; 32: 1184–1191 ª 2010 Blackwell Publishing Ltd
Barley hordeins are immunotoxic in coeliac disease cleavage at lysine or arginine residues, followed by a proline, were identified, but no further proteins were identified. Proteins from Spots 3–8 (Figure 2b) were identified as D-hordein (Spot 3, score 133), B3-hordein (Spot 4, score 220), gamma-3- hordein (Spot 5, score 157), gamma-hordein-1 (Spot 7, score 227) and gamma-3hordein (Spot 8, score 261) with between 16% and 48% sequence coverage. Lower levels of D-hordein were also detected in Spot 4, B-hordein in Spot 5 and gammahordein-1 in Spots 6 and 8. Hordein fractions were also dissolved in IPA ⁄ DTT and examined by RP-HPLC12 calibrated with prolamins isolated from barley lines Risø 56 or Risø 1508. Hordein measurement Alcohol soluble protein (hordein) was measured in duplicate water-washed flour samples dissolved in IPA ⁄ DTT.13 The hordein content of wild-type barley cv Sloop was 35.7 0.4 and Bomi 24.6 1.9 mg ⁄ g flour, for Low Gluten Barley lines (J1, G1, 9RE, 4BH, and 5RB) between 7.3 0.5 and 12.0 1.5 mg ⁄ g, for C-null Risø 1508, 9.5 0.1 mg ⁄ g, and B-null Risø 56 22.9 2.9 mg ⁄ g in general agreement with published hordein compositional data.10 The hordein content was used to convert the hordein dose-response curves into fresh weight flour curves (Figure 4). Deamidation of prolamins and preparation antigens for T-cell assays Prolamins and controls (50 mg ⁄ mL in 2 M urea) were diluted with PBS containing 1 mM CaCl2, to give either 62.5, 250, 625, 2500 or 6250 lg prolamin ⁄ mL and deamidated by adding 25 lL of each solution to 100 lL of guinea pig liver transglutaminase (Sigma, St Lousi, MO, USA) (25 lg ⁄ mL tTG in PBS containing 1 mM CaCl2) and incubated for 6 h at 37 C and added to PMBC as described.9 Nondeamidated solutions were similarly prepared, but incubated in the absence of tTG. Solvent controls were added as for the highest prolamin concentrations. Final prolamin concentrations were 2.5– 250 lg ⁄ mL with final urea concentration 50 mM. The deamidated x-gliadin ⁄ C-hordein peptide (QPEQPFPQPEQPFPWQP, 50 lg ⁄ mL) immunodominant after oral barley challenge in HLA DQ2+ CD volunteers was used as a relevant positive control;4 the peptide was custom synthesized (Mimotopes, Melbourne, Victoria, Australia) to 91% purity, with identity confirmed by MS and HPLC. Tetanus toxoid (Commonwealth Serum Laboratories, Melbourne, Victoria, Australia) (10 light forming units ⁄ mL) was used as a positive control antigen. Aliment Pharmacol Ther 2010; 32: 1184–1191 ª 2010 Blackwell Publishing Ltd
Statistical analysis Analysis of variance (ANOVA; Supporting Information) or t-tests using GenStat (GenStat for Windows 12th Edition, VSNi Software) were used to determine the significance of the differences observed for the mean SFU produced by T cells isolated from CD patients either before or after oral barley challenge and incubated with hordeins, prolamins and controls. The response curves for T cells isolated from postchallenge patients varied by as much as 200-fold and a large proportion of the variability was due to these differences. To take account of the differing magnitude of patient IFN-c ELISpot responses, a Residual Maximum Likelihood (REML) model was fitted (Supporting Information). The statistically powerful REML analysis agreed with the more biologically relevant hyperbolic curve fitting described below. Individual dose-response curves of mean IFN-c ELISpot responses also fitted hyperbolic curves typical of Michaelis–Menten enzyme kinetics (r2 > 0.9; Prism5 for Windows, GRAPHPAD Software Inc., La Jolla, California, USA), allowing estimation of Kd (i.e. dose required for half maximal response; analogous to ED50) and maximum spot numbers (Max SFU) (Table 1, Figures 3 and 4). More immunotoxic preparations are expected to have a lower Kd and higher Max SFU, and the ratio Max SFU ⁄ Kd may be used to rank immunotoxicity (Table 1).14
RESULTS All hordein fractions are recognized by T cells elicited by oral barley challenge T cells isolated from CD donors (n = 21) after oral barley challenge responded to prolamin and hordein fractions in a dose-dependent manner. As expected, the fitted means, on a log scale, for the normalized SFU data for total prolamins were consistent with three significant toxicity groups (Supporting Information): total prolamins from barley (cv Himalaya) induced the highest response and were significantly more toxic than wheat, while total maize and oats prolamins were the least active group (P < 0.05) (Figure 3a,b,c,d respectively). Similarly, each of the six hordein fractions was active: fraction #1 (D-hordein and trace C-hordein), #2 (C-hordeins) and #3 (C- and B-hordeins) (Figure 4a,b,c respectively) were significantly more active (P < 0.05) than fractions #4, #5 and #6 (B- and gammahordeins) (4D, E, F respectively). Before oral barley challenge, peripheral blood T cells specific for any prolamin or hordein fraction were not 1187
G. J. Tanner et al. Table 1 | T cell response to cereal prolamins Kd (lg ⁄ mL)
Max SFU
Max SFU ⁄ Kd
Cereal
Fraction
Identity
Barley cv Himalaya
Total prolamin
B-, C-, D, c- hordein
Wheat
Total prolamin
Gliadin + glutenin
Oats
Total prolamin
Avenin
Maize
Total prolamin
Zein
Barley cv Himalaya
Fraction #1
D- (C-) hordeins
Barley cv Himalaya
Fraction #2
C hordeins
Barley cv Himalaya
Fraction #3
C- + B- hordeins
Barley cv Himalaya
Fraction #4
Barley cv Himalaya
Fraction #5
Barley cv Himalaya
Fraction #6
B- + c-hordeins
21 4.9
53 3.4
2.5
Barley WT (pooled cv Sloop & Bomi)
Total prolamins
B-, C-, D, c- hordein
1.39 0.59
110 13.7
79.1
Barley cv Risø 56
B-null
C-, D, c- hordein
6.05 3.9
301 83.8
49.8
Barley cv Risø 1508
C-null
B-, D, c- hordein
177 131
983 649
ND*
Low gluten barley (pooled 5RB, G1, J1, 4BH)
B,C-null
D, c- hordein
27.3 14.3
195 54.6
10 2.7
79 5.1
7.9
100 35
58 8.3
0.58
23 7.1
19 1.6
0.83
14 14
6.7 1.7
0.48
4.7 2.2
73 6.7
15.5
7.4 4.0
67 8.0
9.1
17 4.0
81 5.0
4.8
B- + c-hordeins
15 1.4
46 1.1
3.1
B- + c-hordeins
16 2.3
51 1.9
3.2
7.14
* The scatter of data for Risø 1508 prevented deduction of significant values. Kd in mg flour ⁄ mL.
80
SFU
(a)
(b)
40
0
SFU
(c)
(d)
40
0
0
100 200 Prolamin (µ g/mL)
0
100 200 Prolamin (µ g/mL)
detected in the blood of a subset of nine HLA DQ2+ CD donors (P = 0.77) despite the positive control, tetanus toxoid, eliciting a definite response (mean S.E. 22.3 4.64 SFU ⁄ million PBMC; medium 1.40 0.45, P < 0.001). As expected, tTG-mediated deamidation of prolamins substantially increased their immunogenicity with postchallenge T cells (P < 0.05; Figures 3 and 4). 1188
Figure 3 | IFN-c response to cereals: ELISpot responses (SFU ⁄ well), isolated from coeliac disease donors 6 days after commencing oral barley challenge, to total prolamin preparations from (a) barley, (b) wheat, (c) oats or (d) maize with (d, n = 21) or without tTG pre-treatment (s, n = 13). Data are mean standard error (S.E.); error bars are not shown when S.E. was smaller than symbols.
T-cell responses to prolamins varied between individual donors by as much as 200-fold; this variability could not be attributed to plate-to-plate variability, as tetanus toxoid responses did not vary between plates (P = 0.193). Solvent contaminants in the prolamin preparations did not cause suppression or enhancement of antigen-induced T cell responses. In PBMC isolated from blood after oral Aliment Pharmacol Ther 2010; 32: 1184–1191 ª 2010 Blackwell Publishing Ltd
Barley hordeins are immunotoxic in coeliac disease
SFU
80 (a)
(d)
40
40
0
0
Generation and relative toxicity of ‘Low Gluten Barley’ lines To examine further the importance of B- and C-hordeins to the overall T-cell stimulatory activity of hordeins,
ª 2010 Blackwell Publishing Ltd
(f)
40
barley challenge, the positive control C-hordein peptide was clearly stimulatory (mean SFU S.E. 29.55 4.38) and unaffected by the addition of solvent control (P = 0.13). However, there was a small but significant reduction in SFU (P < 0.001) when solvent controls were added to medium (medium only: 2.75 0.67 SFU ⁄ well and medium + tTG: 1.49 0.24; medium + solvent 2.64 0.23 and medium + tTG + solvent 2.75 0.23 respectively). To rank the relative toxicity of prolamin preparations, dose-response curves of mean IFN-c ELISpot responses were fitted to hyperbolic curves and Kd and Max SFU deduced (Table 1). Of the cereals, total barley hordeins (cv Himalaya) were clearly the most immunotoxic preparation with the highest Max SFU and lowest Kd. Total prolamins from wheat, oats and maize had similar ratios of Max SFU ⁄ Kd (Table 1). Similarly, each of the six hordein fractions was immunotoxic with low values of Kd and similar values of MaxSFU.
Aliment Pharmacol Ther 2010; 32: 1184–1191
(e)
80 (c)
SFU
Figure 4 | Interferon-c response to hordein fractions: ELISpot responses (SFU ⁄ well), isolated from coeliac disease donors 6 days after commencing oral barley challenge, to hordein fraction #1 (a), #2 (b), #3 (c), #4 (d), #5 (e), and #6 (f), in the presence (d, n = 21), or absence (s, n = 13) of tTG pretreatment. Data are mean standard error (S.E.); error bars are not shown when S.E. was smaller than symbols.
SFU
80 (b)
100 200 Hordein (µ g/mL)
0
100 200 Hordein (µ g/mL)
double-null hybrid plants, largely devoid of both B- and C-hordeins, were produced by conventional crossing. Electrophoresis and mass spectral analysis showed that many of these plants unexpectedly accumulated at least one B-hordein protein in addition to the expected D- and gamma-hordeins (Figure 2b). However, the hordein concentration in these lines was reduced approximately 5-fold compared with wild-type cv Sloop and Bomi (see Methods). The T-cell stimulatory activity of Risø 56 (B-Hordein null) and of Low Gluten Barley lines (B-, and C-hordein null) was considerably reduced and required 5-fold and 20-fold respectively more flour than wild type cv Sloop, to produce a half maximal response (Table 1). A similar conclusion applies to the ratio Max SFU ⁄ Kd, which suggests that Riso 56 and Low Gluten Barley flour had 1.6-fold and 11-fold respectively lower immunotoxicity than Sloop flour. Hence, reduced expression of B- and C-hordeins achieved by crossing null strains substantially reduces the immunotoxicity of Low Gluten Barley lines.
DISCUSSION The seminal oral feeding studies in patients with CD by Willem Dicke established that wheat, rye and barley must 1189
G. J. Tanner et al. be excluded from the diet for remission to occur.1–3 GFD is the only treatment for CD, but the diet has been associated with reduced fibre, increased fat, high cost and poor palatability.15 In the present study, we sought to understand the importance of specific hordein fractions that account for barley’s toxicity in CD. We selected a group of patients bearing the most common coeliac disease susceptibility haplotype HLA DQ2 to ensure consistency in the specificity of the hordein-specific T-cell response. Wheat and wheat-derived gliadins in gluten have been the focus of studies of gluten-peptide recognition by CD4+ T cells in CD, predominantly in HLA DQ2+ patients.16–20 In the present study, we extended our previous work assessing gluten peptides recognized by T cells to assessment of fractionated hordeins and prolamins from other cereals. By using sophisticated protein separation techniques and customizing FPLC gradients, protein fractions composed predominantly of a single hordein family were isolated, although some low level cross-contamination was encountered. All hordein fractions stimulated secretion of the pro-inflammatory T-cell specific cytokine, interferon-c, but fractions highly enriched for C- and Dhordeins stimulated interferon-c secretion from a greater number of T-cells from patients and were active at lower concentrations than fractions enriched for B- and chordeins. However, B-hordeins still had significant activity. The importance of B- and C-hordeins together was confirmed by showing that the alcohol-extracted prolamin fraction from barley lines with reduced accumulation of B- and C-hordeins were up to 20-fold less active per gram of flour than wild-type barley. Our findings confirm that all prolamin fractions from barley are toxic in CD. Despite using an assay relevant to the pathogenic T-cell response to gluten to monitor the toxicity of barley prolamins, it is apparent that monitoring total prolamin content may be as effective as the initial screen for potentially ‘nontoxic’ barley. Furthermore, it should be stressed that ex vivo T-cell stimulatory activity does not necessarily predict toxicity in vivo. Others have shown that avenin-specific T-cell lines can be raised from donors with CD who do not manifest intestinal damage while consuming oats.21 In this study, avenins were also found to be stimulatory for T cells isolated from blood after oral barley challenge. Maize prolamins were also weakly stimulatory producing 8% of the maximum spot forming units of barley (Table 1). An inherent weakness in all ex vivo, in vitro and in silico approaches to mapping T-cell stimulatory peptides and 1190
proteins is that they do not address proteolysis or modification of proteins and peptides in vivo after ingestion or during food preparation, and we suggest that chymotryptic ⁄ tryptic proteolysis during digestion normally destroys the maize epitopes, which provoked this weak response. Application of these findings to breed barley for inclusion in the GFD suitable for patients with CD is some way off. However, this is the first report of a cereal with reduced immunotoxicity being developed by crossing strains deficient in distinct prolamin families.
ACKNOWLEDGEMENTS The authors thank Malcolm Blundell for technical assistance. Funding from the Australian Grains and Research Development Corporation, the NHMRC (grant #361646), and a Victorian State Government Operational Infrastructure Support Grant is gratefully acknowledged. The co-operation of the Coeliac Society of Victoria and assistance of volunteers is also gratefully acknowledged. JT-D held a National Health and Medical Research Council Postgraduate Medical Scholarship and RPA holds a Lions Cancer Council Fellowship. Declaration of personal interests: GT and CH are employees of CSIRO and co-inventors of a patent pertaining to the use low gluten barley in the food and beverage industry. JT-D and RPA are employees of Melbourne Health and co-inventors of patents pertaining to the use gluten peptides in therapeutics, diagnostics and nontoxic gluten. RPA is a substantial shareholder, director and CEO of Nexpep Pty Ltd, a company developing peptide-based therapeutics and diagnostics for coeliac disease. JT-D is a shareholder and consultant to Nexpep Pty Ltd. RF is a CSIRO honorary fellow, and CH an employee of CSIRO. Declaration of funding interests: This study was funded in part by the Australian Grains and Research Development Corporation, grant no. CSP00088, and in part by the NHMRC grant #361646, and a Victorian State Government Operational Infrastructure Support Grant. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Data S1. Mean normalized SFU data. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by authors. Any queries (other than missing material) should be directed to the corresponding author for the article. Aliment Pharmacol Ther 2010; 32: 1184–1191 ª 2010 Blackwell Publishing Ltd
Barley hordeins are immunotoxic in coeliac disease
REFERENCES 1. Van De Kamer JH, Weijers HA, Dicke WK. Coeliac disease. IV. An investigation into the injurious constituents of wheat in connection with their action on patients with coeliac disease. Acta Paediatr 1953; 42: 223–31. 2. Dicke WK, Weijers HA, Van De Kamer JH. Coeliac disease. II. The presence in wheat of a factor having a deleterious effect in cases of coeliac disease. Acta Paediatr 1953; 42: 34–42. 3. Dicke WK. Coeliac Disease. Investigation of the Harmful Effects of Certain Types of Cereal on Patients With Coeliac Disease [MD]. The Netherlands: University of Utrecht, 1950. 4. Tye-Din JA, Stewart JA, Dromey JA, et al. Comprehensive, quantitative mapping of T cell epitopes in gluten in celiac disease. Sci Transl Med 2010; 2: 41ra51. 5. Shewry PR, Tatham AS, Halford NG. The prolamins of the Triticeae. In: Shewry PR, Casey R, eds. Seed Proteins. London: Klewer, 1999: 35–78. 6. Doll H. Inheritance of the high-lysine character of a barley mutant. Hereditas 1973; 74: 293–4. 7. Karell K, Louka AS, Moodie SJ, et al. Hla Types in Celiac Disease Patients Not Carrying the Dqa1*05-Dqb1*02 (Dq2) Heterodimer: results From the European Genetics Cluster on Celiac Disease. Hum Immunol 2003; 64: 469–77. 8. Walker-Smith J, Guandalini S, Schmitz J, Shmerling D, Visakorpi J. Revised criteria for diagnosis of coeliac disease.
Aliment Pharmacol Ther 2010; 32: 1184–1191 ª 2010 Blackwell Publishing Ltd
9.
10.
11.
12.
13.
14.
Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch Dis Child 1990; 65: 909–11. Anderson RP, van Heel DA, Tye-Din JA, et al. T cells in peripheral blood after gluten challenge in coeliac disease. Gut 2005; 54: 1217–23. Doll H. Barley seed proteins and possibilities for their improvement. In: Gottschalk W, Muller HP, eds. Seed Proteins: Biochemistry, Genetics, Nutritive Value. The Hague: Martinus Nijhoff, 1983: 207–23. Campbell PM, Cao AT, Hines ER, East PD, Gordon KHJ. Proteomic analysis of the peritrophic matrix from the gut of the caterpillar, Helicoverpa armigera Insect Biochem. Molec Biol 2008; 38: 950–8. Larroque OR, Bekes F, Wrigley CW, Rathmell WG. Analysis of gluten proteins in grain and flour blends by RPHPLC. In: Shewry PR, Tatham AR, eds. Seventh International Workshop on Gluten Proteins, 2-6 April 2000, Bristol, UK. Bristol: Institute for Arable Crops Research, 2000: 136–9. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilising the principle of dye binding. Anal Chem 1976; 72: 248–54. Cornish-Bowden A. Fundamentals of Enzyme Kinetics. 3rd edn. London: Portland Press, 2004; 422 pp.
15. Kupper C. Dietary guidelines and implementation for celiac disease. Gastroenterology 2005; 128 (4 Suppl 1): S121–7. 16. Qiao SW, Bergseng E, Molberg O, Jung G, Fleckenstein B, Sollid LM. Refining the rules of gliadin T cell epitope binding to the disease-associated DQ2 molecule in celiac disease: importance of proline spacing and glutamine deamidation. J Immunol 2005; 175: 254–61. 17. Vader LW, Stepniak DT, Bunnik EM, et al. Characterization of cereal toxicity for celiac disease patients based on protein homology in grains. Gastroenterology 2003; 125: 1105–13. 18. Vader W, Kooy Y, Van Veelen P, et al. The gluten response in children with celiac disease is directed toward multiple gliadin and glutenin peptides. Gastroenterology 2002; 122: 1729–37. 19. Arentz-Hansen H, McAdam SN, Molberg O, et al. Celiac lesion T cells recognize epitopes that cluster in regions of gliadins rich in proline residues. Gastroenterology 2002; 123: 803–9. 20. Arentz-Hansen H, Korner R, Molberg O, et al. The intestinal T cell response to alpha-gliadin in adult celiac disease is focused on a single deamidated glutamine targeted by tissue transglutaminase. J Exp Med 2000; 191: 603–12. 21. Arentz-Hansen H, Fleckenstein B, Molberg O, et al. The molecular basis for oat intolerance in patients with celiac disease. PLoS Med 2004; 1: e1.
1191
