Clinical Nutrition 32 (2013) 1043e1049. Contents lists available at SciVerse ScienceDirect. Clinical Nutrition

Transcription

1 Clinical Nutrition 32 (2013) 1043e1049 Contents lists available at SciVerse ScienceDirect Clinical Nutrition journal homepage: Original article Evaluation of the safety of ancient strains of wheat in coeliac disease reveals heterogeneous small intestinal T cell responses suggestive of coeliac toxicity Tanja Suligoj a, *, Armando Gregorini b,d, Mariastella Colomba b,e, H. Julia Ellis a,c, Paul J. Ciclitira a,c a King s College London, Division of Diabetes and Nutritional Sciences, Gastroenterology, The Rayne Institute, St Thomas Hospital, Westminster Bridge Road, London SE1 7EH, UK b Università di Urbino Carlo Bo, via Maggetti 22 (loc. Sasso), Urbino (PU), Italy article info summary Article history: Received 10 July 2012 Accepted 8 February 2013 Keywords: Coeliac disease Ancient strains of wheat Small intestinal gluten-sensitive T-cell lines Proliferation assays Heterogenous T-cell responses Background & aims: Coeliac disease is a chronic small intestinal immune-mediated enteropathy triggered by dietary gluten in genetically predisposed individuals. Since it is unknown if all wheat varieties are equally toxic to coeliac patients seven Triticum accessions showing different origin (ancient/modern) and ploidy (di-, tetra- hexaploid) were studied. Materials and methods: Selected strains of wheat were ancient Triticum monococcum precoce (AA genome) and Triticum speltoides (BB genome), accessions of Triticum turgidum durum (AABB genome) including two ancient (Graziella Ra and Kamut) and two modern (Senatore Cappelli and Svevo) durum strains of wheat and Triticum aestivum compactum (AABBDD genome). Small intestinal gluten-specific T-cell lines generated from 13 coeliac patients were tested with wheat accessions by proliferation assays. Results: All strains of wheat independent of ploidy or ancient/modern origin triggered heterogeneous responses covering wide ranges of stimulation indices. Conclusion: Ancient strains of wheat, although previously suggested to be low or devoid of coeliac toxicity, should be tested for immunogenicity using gluten-specific T-cell lines from multiple coeliac patients rather than gluten-specific clones to assess their potential toxicity. Our findings provide further evidence for the need for a strict gluten-free diet in coeliac patients, including avoidance of ancient strains of wheat. Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. 1. Introduction Abbreviations: CD, coeliac disease; GFD, gluten-free diet; TCL, T-cell line; ttg, tissue transglutaminase; Tmp_AA, Triticum monococcum precoce (AA genome); Ts_BB, Triticum speltoides (BB genome); K_AABB, Triticum turgidum durum Kamut (AABB genome); GR_AABB, Triticum turgidum durum Graziella Ra (AABB genome); SC_AABB, Triticum turgidum durum Senatore Cappelli (AABB genome); S_AABB, Triticum turgidum durum Svevo (AABB genome); Tac_AABBDD, Triticum aestivum compactum (AABBDD genome); IEL, intraepithelial lymphocytes; PTG, peptictryptic digest of gluten; SI, stimulation index. * Corresponding author.. Tel.: þ44(0) ; fax: þ44(0) addresses: tanja.suligoj@gmail.com, tanja.suligoj@kcl.ac.uk (T. Suligoj), armando.gregorini@uniurb.it (A. Gregorini), mariastella.colomba@uniurb.it (M. Colomba), julia.ellis@kcl.ac.uk (H.J. Ellis), paul.ciclitira@kcl.ac.uk (P.J. Ciclitira). c Tel.: þ44 (0) d Tel.: þ39 (0) e Tel.: þ39 (0) Coeliac disease (CD) is a chronic small intestinal immunemediated enteropathy precipitated by exposure to dietary gluten in genetically predisposed individuals. 1 Treatment involves a longlife gluten-free diet (GFD) which excludes wheat, rye, barley and possibly oats. Gluten is however difficult to avoid because of its widespread use in food processing. It is used in the production of soups, sauces, meat products, potato chips, candies, ice cream and as excipients in medicines and vitamins supplements, etc. 2 Therefore the GFD is restricted, unpalatable and causes social disadvantage for the individuals resulting in poor compliance amongst CD patients. The symptoms include diarrhea, weight loss and fatigue, although a broad spectrum of clinical presentations occur. The condition is diagnosed by the presence of some degree of small intestinal villous atrophy with raised number of intra-epithelial /$ e see front matter Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.

2 1044 T. Suligoj et al. / Clinical Nutrition 32 (2013) 1043e1049 lymphocytes which normalise when gluten is removed from the diet. In the submucosa of the small intestine the enzyme tissue transglutaminase (ttg), generally involved in tissue repair, deamidates gluten peptides, that allows high-affinity binding to the human leucocyte antigen class molecules HLA-DQ2 and HLA-DQ8, subsequently triggering an inflammatory reaction in patients with CD. 3 The activation of gluten-sensitive CD4þ T-cells is central to the inflammatory reaction, although innate immune mechanisms are also thought to be involved. 4 Gluten is a term that describes storage proteins in wheat endosperm and related cereals. Wheat storage proteins are divided into two major groups of proteins: gliadins and glutenins. Gliadins are further divided into a-, g- and u-gliadins and glutenins into high- and low-molecular weight glutenins. The a-gliadins are proteins encoded by a gene family at the Gli-2 loci, which may contain from 25 to 35 to even 150 a-gliadin genes per haploid genome, 5 although most of these (>70%) are presumably pseudogenes. The immunogenicity of many gluten peptides was assessed by activation of gluten-specific T-cells isolated from duodenal biopsies of CD sufferers. 6 Epitopes from gliadins, 7,8 in particular those that cluster within the N-terminal, including a stable 33mer fragment formed by physiologic digestion of a-gliadins, 9 are considered to have by far the highest clinical relevance with regard to the development of CD. It is unknown if all wheat varieties are equally toxic to individuals with CD. Since large variations exist in the amount of T-cell stimulatory peptides present in different wheat cultivars, 10 numerous accessions have been studied to identify those with a lower coeliac toxic profile. 2,10e12 It has been suggested that a diet based on wheat varieties reduced in T-cell stimulatory epitopes may help in prevention of CD since the amount and duration of gluten consumption are associated with the initiation of CD. The argument then goes that this would especially benefit children in which the onset of CD may be delayed or even prevented and in undiagnosed coeliac patients (which are the vast majority of all coeliac sufferers) might strongly reduce their symptoms. 2 Nevertheless this assumption is controversial and still subject to much debate and investigation. In an attempt to identify grains which were less toxic to patients suffering from coeliac disease, several scientists strongly focused on the analysis of grains considered forerunners of modern grains including Triticum monococcum (bearing the AA genome) and Triticum speltoides (considered the progenitor or otherwise very close to the ancestor of the BB genome). In particular, some ancient, that is ancestral, strains of wheat have been suggested to be less toxic for coeliac patients or even lack CD toxicity. For example gliadins of T. monococcum were reported to lack coeliac toxicity in an in vitro organ culture system suggesting new dietary opportunities for CD patients. 13 Some ancient wheat varieties including T. monococcum and T. speltoides were shown to be low in a-gliadin T-cell epitopes 5,10,11 or other gliadin and glutenin epitopes involved in the pathogenesis of CD. 10,11 More recently it has been suggested that some monococcum lines (Monlis and ID331) are to be considered toxic for coeliac patients. 14 Misinterpretations of the existing literature and contradictory evidence on the safety of T. monococcum accessions could potentially influence CD sufferers to consume ancient cereals that are unsafe for them. In the last decades, a huge number of durum wheat cultivars have been obtained by artificial selection, generally based on high yield, disease resistance and technological qualities. On the other hand, to preserve genetic variability and reduce genetic erosion it is extremely important to develop and maintain local crops, including old cultivars and landraces, which were not subjected to massive selective breeding or genetic modifications. Examples of this are Graziella Ra, an accession (not a cultivar) of durum wheat that recently appeared on the market as Graziella Ra Ò, and Kamut Ò. The latter is considered ancient relative of modern durum wheat. In their paper testing the hypothesis of a potentially reduced or absent coeliac toxicity of these strains of wheat, Gregorini et al. 15 reported that both Graziella Ra and Kamut, are not only putatively as CD toxic as the modern durum accessions analysed, but also contain greater amounts of a-gliadin. In the present study we sought to test whether the low immunogenicity hypothesis of ancient vs modern wheat varieties is confirmed or not, since any cereal from the tribe Triticeae has to be considered toxic unless thorough in vitro and in vivo evidence to the contrary is produced. The analysis was carried out by using 13 polyclonal gluten-sensitive T-cell lines (TCL) to assess their overall reactivity in relation to their (i) origin (ancient/modern) and (ii) ploidy. Wheat accessions were as follows: the diploid T. monococcum precoce and T. speltoides, representing potential ancient progenitors that might have hybridised into tetraploid strains of wheat 5,16 ; four accessions of the tetraploid (AABB genome) Triticum turgidum durum including two ancient (Graziella Ra and Kamut) and two modern (Senatore Cappelli and Svevo) durum wheat varieties 17,18 ; and the hexaploid (AABBDD genome) Triticum aestivum compactum. 2. Materials and methods 2.1. Subjects Small intestinal biopsies were obtained at endoscopy performed for diagnostic or clinical management purposes in subjects with suspected or known CD. Thirteen volunteer subjects diagnosed according to British Society of Gastroenterology guidelines 19 included twelve females and one male, median age 35 years, range 23e72 (Table 1). Biopsy specimens were taken from the second part of the duodenum. All participating subjects provided written informed consent. The study was approved by the St Thomas Hospital Research Ethics Committee (reference number 05/Q0207/167) Cereal sources of gliadins Ancient and modern wheat accessions with their correspondent ploidy, genomes and abbreviated names used in this manuscript Table 1 Details of CD volunteers included in the study: sex, age, time on a gluten-free diet, DQ status, and histology results according to Marsh-Oberhuber classification of duodenal histological lesions in CD. 20 Subject Sex Age at time of biopsy GFD (in years) DQ status Histology 1 F 35 0 DQ2 1 2 F 28 0 DQ2 3b 3 F 72 7 DQ2 3a 4 F DQ2 3a 5 F 23 0 DQ2 3a 6 F DQ2 1 7 F 32 1 DQ2 0 8 F 23 2 DQ2 0 9 F DQ8 3a 10 F 53 9 DQ2 3a 11 F 49 0 DQ2 3b 12 M 60 0 DQ2 3a 13 F DQ2 3a GFD, gluten-free diet; histology 0, normal mucosal architecture; 1, normal mucosal architecture with an increased number of intraepithelial lymphocytes (IEL); 3a, increased IEL with hyperplastic crypts and mild villous atrophy; 3b, increased IEL with hyperplastic crypts and moderate villous atrophy. Subject number 1 had positive CD serology 3 months prior to endoscopy and family history of CD.

3 T. Suligoj et al. / Clinical Nutrition 32 (2013) 1043e Table 2 Ancient and modern strains of wheat included in the study: Wheat accession, ploidy, genome and the name used in the manuscript are shown. Ancient wheat Modern wheat Wheat species (accession) Ploidy and genome Name in the manuscript T. monococcum precoce diploid, AA Tmp_AA T. speltoides diploid, BB Ts_BB T. turgidum durum (Kamut) tetraploid, AABB K_AABB T. turgidum durum (Graziella Ra) tetraploid, AABB GR_AABB T. turgidum durum tetraploid, AABB SC_AABB (Senatore Cappelli) T. turgidum durum (Svevo) tetraploid, AABB S_AABB T. aestivum compactum hexaploid, AABBDD Tac_AABBDD are presented in Table 2. Ancient wheat varieties included T. monococcum precoce (ITA ), T. speltoides (CGN 10684), T. turgidum durum: Kamut and Graziella Ra. Modern wheat varieties included T. turgidum durum: Senatore Cappelli and Svevo, and T. aestivum compactum (CGN 04210) Preparation of antigens for T-cell assays Kernels of wheat accessions were milled with an analytical grinder and sieved through a 0.5 mm mesh. The resultant flour was washed twice with phosphate buffer (0.067 M NaKH phosphate buffer in 0.4 M NaCl, ph ¼ 7.6) 21 to remove concomitant albumins and globulins. Gliadins were extracted twice with 60% EtOH, and the extracts combined. They were then dried and sequentially digested with agarose-bound pepsin (P0609, SigmaeAldrich Ltd, Poole Dorset, UK) and trypsin (T1763, Sigma-Adrich) according to Bolte et al. 22 Peptic-tryptic (PT) digests were freeze-dried. Prior to ttg deamidation protein concentrations were determined with BCA Protein assay kit (Novagen, EMD Chemicals, San Diego, CA, USA) according to the manufacturer s instructions. Deamidation mixes contained 100 mg/ml guinea pig liver ttg (T5398, Sigmae Aldrich) in PBS containing 1 mm CaCl 2 and 370 mg/ml PT digests of prolamins. Incubation was carried out for 4 h at 37 C Establishment of small intestinal T-cell lines The methods were as previously described. 23 Plasmocin (antmpt, Invivogen Ltd, Cayla, Toulouse-Cedex, France) was used at all stages. Small intestinal biopsies obtained from coeliac patients were incubated overnight in the presence of 5 mg/ml peptic-tryptic digest of industrial gluten (PTG) bought from Roquette Ltd, Corby Northants, UK (418, batch NW652). This was in order to stimulate activation by the whole range of gluten proteins that are likely to be encountered by individuals consuming a modern gluten containing diet. Biopsies were then mechanically disrupted to release lymphocytes and passed through a 70 mm cell filter (BD Falcon). The collected cells were cultured in RPMI medium with 10% heat inactivated autologous plasma plus /ml irradiated (22 Gy) autologous peripheral blood mononuclear cells. Human recombinant interleukin 2 (I7903, Sigma) was added on day 5 and subsequently twice a week at 10 U/ml. Cells were restimulated every seven days with PTG, pre-treated with ttg, at a final concentration of 200 mg/ml. Autologous irradiated (22 Gy) peripheral blood mononuclear cells acted as antigen presenting cells T-cell proliferation assays T-cell proliferation assays were performed a minimum of seven days after antigenic stimulation. The antigens tested were ttg deamidated PTG and ttg deamidated gliadins from the seven strains of wheat. Antigens were incubated overnight with antigen presenting cells ( )at100mg/ml prior to addition of T-cells ( ). Following incubation for 48 h, tritiated thymidine (NET355, Perkin Elmer, Boston, MA, USA) was added for 18 h prior to harvesting and measurement of thymidine incorporation. The stimulation index (SI) was calculated by dividing the mean counts per minute (cpm) in the presence of antigen by the mean cpm in the absence of antigen. An SI of 2 or more was considered positive. 23 The use of internal negative controls in proliferation assays is not standard practice. 3,11 Further, using another dietary coeliac non-toxic protein that has been treated in the same way as gluten proteins results in SI < Hence, we employed medium only as negative control Data analysis and statistics Data are presented as arithmetic mean of cpm standard deviation of at least triplicates with corresponding SI. Statistical analysis was performed with GraphPad Prism version 5.03, using the Friedman test. Spearman non-parametric correlation analysis choosing a twotailed P value was performed to investigate the correlation between SI values and the duration of the GFD in the subjects from which the TCLs were obtained. 3. Results All wheat varieties tested, irrespective of their ploidy or ancient/ modern origin triggered heterogeneous T-cell responses as assessed by a wide range of stimulation indices in 13 proliferation assays. Two gluten-sensitive TCLs responded to gliadins of all wheat samples with markedly high SI (48.9 and above to the extreme value of 411.7). All other lines responded with SI in the range from 0.4 to 41.1 (Table 3). As shown in Fig. 1, all accessions showed two TCLs as outliers. Amongst the four durum wheat accessions studied (ancient K_AABB and GR_AABB; modern SC_AABB and S_AABB) all evoked similar T-cell responses to gliadins as shown by similar sizes of boxes covering ranges of 1.7e22.5 for SI and comparable medians (5.8, 5.7, 5.5 and 4.3). In addition, all four boxplots for durum wheat accessions had similar SI values for the two outliers. Similar results apply to ancient Ts_BB, however with slightly lower median value (3.6). T-cell responses to ancient Tmp_AA were less dispersed as shown with a smaller box (SI in the range 2.4e14.5, with 7.6 median). Both boxplots of the results for ancient diploid accessions Tmp_AA and Ts_BB had outliers of similar magnitude (approximately 50 and 80). On the contrary, T-cell responses to modern hexaploid accession Tac_AABBDD were notably higher as shown with a larger box (SI in the range 3.9e 40.7). The latter notably exceeds the range of stimulation indices of all other wheat varieties (diploid and tetraploid). The median value for the Tac_AABBDD accession was 8.5, the distinctive feature for this variety is one of the two outliers presenting an extreme SI of Figure 2 shows a line-graph for individual TCL tested in proliferation assay with different wheat accessions. Stimulation indices for TCL4 and TCL6 to all wheat accessions are markedly higher (SI 48.9) than for all other T-cell lines. TCL2 and TCL13 respond in the middle range (SI between 10.5 and 40.2) whereas other TCLs responded to wheat strains with lower SI, with the exception of TCL8 for Tac_AABBDD (SI ¼ 14.3). As shown in Fig. 3 TCL10 tested negative (as defined by SI < 2) to gliadins of all ancient and modern varieties despite being reactive

4 1046 T. Suligoj et al. / Clinical Nutrition 32 (2013) 1043e1049 Table 3 Results of proliferation assays showing individual TCL tested with gluten (positive control) and gliadins from four ancient (Tmp_AA, Ts_BB, GR_AABB, K_AABB) and three modern wheat accessions (SC_AABB, S_AABB, Tac_AABBDD). Medium only Gluten Tmp_AA Ts_BB K_AABB GR_AABB SC_AABB S_AABB Tac_AABBDD TCL (1) 24, (5.2) 12, (2.7) 10, (2.2) 14, (3.0) 12, (2.7) 11, (2.3) 10, (2.2) 18, (3.9) TCL (1) (59.8) (10.5) (27.3) (21.7) (16.7) (15.0) (17.3) (40.2) TCL (1) 33, (33.2) (1.0) (1.2) (3.7) (1.1) (3.2) (0.4) 41, (41.1) TCL (1) 29, (448.7) (48.9) (81.8) (101.5) (111.4) (149.6) (69.7) 27, (411.7) TCL (1) (9.1) 10, (10.4) (8.8) (5.7) (6.5) (5.5) (4.3) (8.5) TCL (1) 28, (194.6) 12, (81.3) (55.4) 14, (95.6) 15, (104.6) 13, (91.8) 14, (96.9) 18, (126.3) TCL (1) (8) (2.8) (3.6) (4.4) (4.3) (4.7) (4.9) (6.1) TCL (1) (22.4) (8.4) (5.8) (6.6) (5.8) (5.7) (2.4) (14.3) TCL (1) (5.2) (7.6) (3.0) (7.8) (9.3) (7.1) (7.5) (6.3) TCL (1) (4.2) (0.9) (1.1) (1.0) (1.4) (1.2) (0.9) (1.4) TCL (1) (3.2) (2.1) (3.2) (3.9) (4.7) (3.2) (3.7) (3.8) TCL (1) (8.6) (4.0) (0.9) (1.7) (1.0) (0.7) (1.1) (1.0) TCL (1) 57, (57.9) 18, (18.5) 16, (16.3) 23, (23.3) 27, (27.5) 28, (28.6) 21, (21.7) 31, (31.3) Results (expressed as cpm) are presented as arithmetic mean standard deviation and in bracket corresponding stimulation index. TCL T-cell line; Tmp_AA Triticum monococcum precoce; Ts_BB Triticum speltoides; K_AABB Triticum turgidum durum Kamut; GR_AABB T. turgidum durum Graziella Ra; SC_AABB T. turgidum durum Senatore Cappelli; S_AABB T. turgidum durum Svevo; Tac_AABBDD Triticum aestivum compactum. Fig. 1. Stimulation indices of 13 gluten-sensitive T-cell lines tested in proliferation assays with four ancient (Tmp_AA, Ts_BB, GR_AABB, K_AABB) and three modern wheat accessions (SC_AABB, S_AABB, Tac_AABBDD). The line in the boxplot presents the median value; the box represents the interquartile range covering the middle 50% of stimulation indices per wheat tested. The two whiskers together with outliers represent the range of the other 50% of results obtained for stimulation indices per wheat accession. Tmp_AA, Triticum monococcum precoce; Ts_BB, Triticum speltoides; K_AABB, Triticum turgidum durum Kamut; GR_AABB, T. turgidum durum Graziella Ra; SC_AABB, T. turgidum durum Senatore Cappelli; S_AABB, T. turgidum durum Svevo; Tac_AABBDD, Triticum aestivum compactum. to industrial gluten (SI ¼ 4.2, Table 3). Similarly, TCL12 tested negative to gliadins of nearly all wheat accessions except one ancient accession, Tmp_AA (SI ¼ 4.0) and industrial gluten (SI ¼ 8.6, Table 3). Perhaps the most striking results of the proliferation assays are those of TCL3 that tested negative with 3 out of 4 ancient wheat varieties (Tmp_AA, Ts_BB and GR_AABB) and 1 out of 3 modern wheat accessions (S_AABB). All other TCLs responded with a wide range of positive SI to all modern and ancient accessions tested with TCL4 and TCL6 being markedly higher than the rest (Fig. 3). Considering all wheat accessions tested, the Friedman test showed the SI mean values to be significantly different (P < 0.001). Fig. 2. Stimulation indices of 13 gluten-sensitive T-cell lines tested in proliferation assays with four ancient (Tmp_AA, Ts_BB, GR_AABB, K_AABB) and three modern wheat accessions (SC_AABB, S_AABB, Tac_AABBDD). The line in the graph presents stimulation indices obtained per cell line when tested with wheat accessions. A line at SI ¼ 2 distinguishes positive from negative results of proliferation assays. Tmp_AA, Triticum monococcum precoce; Ts_BB, Triticum speltoides; K_AABB, Triticum turgidum durum Kamut; GR_AABB, T. turgidum durum Graziella Ra; SC_AABB, T. turgidum durum Senatore Cappelli; S_AABB, T. turgidum durum Svevo; Tac_AABBDD, Triticum aestivum compactum.

5 T. Suligoj et al. / Clinical Nutrition 32 (2013) 1043e Fig. 3. Stimulation indices of 13 gluten-sensitive T-cell lines tested in proliferation assays with four ancient (Tmp_AA, Ts_BB, GR_AABB, K_AABB) and three modern wheat accessions (SC_AABB, S_AABB, Tac_AABBDD). The line in the graph presents stimulation indices obtained per wheat accession tested with 13 TCLs. A line at SI ¼ 2 distinguishes positive from negative results of proliferation assays. Tmp_AA, Triticum monococcum precoce; Ts_BB, Triticum speltoides; K_AABB, Triticum turgidum durum Kamut; GR_AABB, T. turgidum durum Graziella Ra; SC_AABB, T. turgidum durum Senatore Cappelli; S_AABB, T. turgidum durum Svevo; Tac_AABBDD, Triticum aestivum compactum. The Dunn s multiple pair-wise comparison post test determined that, with reference to SI, only Tmp_AA, Ts_BB and S_AABB were significantly different from Tac_AABBDD (P < 0.05, P < 0.01, P < 0.01, respectively) (data analysis not shown but available on request). Therefore, small intestinal T-cell responses to hexaploid wheat Tac_AABBDD were significantly higher compared to both diploid strains of wheat (Tmp_AA and Ts_BB) and one tetraploid wheat (S_AABB). The correlation between SI values and the duration of GFD in the subjects from which the TCLs were obtained was investigated. Nonparametric analysis showed no significant correlation (for P < 0.05) between SI of any of the antigens and duration of GFD (data not shown but available on request). 4. Discussion The only effective treatment available for CD patients is a strict exclusion of gluten from the diet. Nevertheless, the existence of thousands Triticum accessions and hundreds of alleles for gliadins and glutenins has raised the question of whether they are equally toxic to CD patients. Moreover, it has been speculated that ancient strains of wheat that have not been subjected to major genetic improvements may show potentially reduced or absent toxicity and therefore potentially better suited to be introduced into the diet of people suffering from wheat intolerances or allergies, including coeliac disease. To test this hypothesis, in the present study, we analysed how TCLs from 13 subjects with CD responded to seven wheat accessions selected according to their origin (ancient/modern) and ploidy (di-, tetra-, hexaploid). Our findings revealed that all the wheat varieties tested, irrespective of their ancient/modern origin or ploidy, evoked heterogeneous small intestinal T-cell responses. For example hexaploid wheat Tac_AABBDD evoked a wide range of intestinal T-cell responses which include negative (in 2 out of 13 gluten-sensitive TCLs), very high stimulation indices in 2 TCLs and a moderate SI in all other lines. A roughly similar trend was observed with all other modern and ancient wheat accessions. Our results are not surprising from a clinical point of view, since majority of wheat strains are likely to be toxic to coeliac patients. However, our findings highlight the occurrence of a heterogeneous response by different coeliac patients. Our results are in agreement with previous studies where Camarca et al. 8 showed that intestinal T-cell responses to gluten peptides are largely heterogeneous and Spaenij-Dekking et al. 10 who showed that wheat varieties either induced low, medium or high T-cell responses independent of the ploidy or genome background of the accession. Our results are in further agreement with Molberg et al. 11 who showed that polyclonal T-cell line reactivity patterns to ancient wheat varieties are heterogeneous despite the ancient strains of wheat being shown to be low in certain T-cell epitopes. In relation to assessing the safety of ancient and modern wheat varieties for consumption by individuals with CD, the present study further supports the necessity of testing them with TCLs from several different individuals. Due to the wide heterogeneity of T-cell responses there is a realistic chance they would test negative if small numbers of TCLs were employed and results extrapolated to the safety of ancient wheat varieties for all coeliac subjects. This is particularly the case if clones as opposed to TCLs were used. Glutenspecific clones cover only a limited number of specificities and thus do not represent the whole repertoire of gluten-reactive T-cells within coeliac lesions. This is well demonstrated by the results obtained in the present paper. For example, the ancient diploid accession Ts_BB and modern hexaploid accession Tac_AABBDD have been tested with selected clones and shown to be low in some gluten epitopes, including the immunodominant peptide from a- gliadins and two other g-gliadin peptides. 10 In our experiments, testing gliadin extracts from these two strains of wheat with polyclonal T-cell lines showed that the majority of TCLs (10 out of 13 for Ts_BB and 11 out of 13 for Tac_AABBDD) responded with a positive SI, with no overall trend in reduction of polyclonal T-cell responses. In fact, responses to diploid Ts_BB were observed to be in a comparable range to those of all four durum strains of wheat (ancient K_AABB and GR_AABB; modern SC_AABB and S_AABB). T- cell responses to Tac_AABBDD, which was shown be low in the immunodominant epitope 10 were significantly higher than to both diploid strains of wheat (Tmp_AA and Ts_BB, P < 0.05 and P < 0.01, respectively) and one tetraploid wheat (S_AABB) (P < 0.01). This might indicate that other gliadin peptides from Ts_BB and Tac_AABBDD which were not tested in Spaenij-Dekking s study may also trigger T-cell responses, perhaps peptides from u-gliadins 7,8 or other a- and g-gliadins showing different amino-acid sequences. 6 Additionally, we cannot exclude that other unknown CD-triggering epitopes in gliadins might have triggered T-cell responses in our polyclonal TCLs. The literature suggests that, compared to diploid and tetraploid wheat, hexaploid wheat, might be increased in T-cell stimulatory epitopes that exacerbate CD because of the presence of the D genome. Indeed, with respect to T-cell toxicity as far as induced by a-gliadin epitopes, the D genome should be considered as the most relevant as it codes for several a-gliadin toxic epitopes. 5 This might explain higher stimulation indices obtained for our modern wheat accession Tac_AABBDD. On the other hand, this particular accession has been shown to be low in immunodominant and other CD toxic epitopes. 10 Such contrasting results obtained by the two groups working on the same wheat accession further supports the need to use polyclonal T-cell lines to assess the CD toxicity of wheat varieties as opposed to gluten-sensitive T-cell clones. This is because a lack of certain T-cell stimulatory sequences does not imply that their gluten proteins may not have any T-cell stimulatory properties. With reference to TCL responses to tetraploid accessions, the gliadins from the two ancient (K_AABB and GR_AABB) and two modern accessions (SC_AABB and S_AABB) triggered similar ranges of polyclonal T-cell responses. This is interesting particularly in relation to the amount of a-gliadins these four durum accessions

6 1048 T. Suligoj et al. / Clinical Nutrition 32 (2013) 1043e1049 contain and the relevance of a-gliadins in the pathogenesis of CD. As Gregorini et al. 15 showed e by using a monoclonal antibody specific for the immunodominant a-gliadin peptide, 24 a monoclonal antibody to another coeliac toxic a-gliadin peptide (A-gliadin 31e49) 25 and a commercial gliadin kit e gliadins from the two ancient accessions K_AABB and GR_AABB occur in greater amount than in modern durum varieties. We have shown in the present set of experiments that our T-cell lines reacted similarly to the gliadins of the four tetraploid strains of wheat which might indicate that (i) the response is not only dependent on the a-gliadin amount; (ii) other gliadin peptides contribute to the intestinal T-cell response 7 ; and (iii) different specificities of T-cells tested overall resulted in comparable stimulation indices. Most attention in relation to the safety of ancient diploid wheat varieties and potentially new dietary opportunities for coeliac patients has been focussed on T. monococcum. Molberg et al. 11 showed that fragments identical or equivalent to the immunodominant 33mer fragment are encoded by a-gliadin genes from wheat chromosome 6D, and are thus absent from gluten of diploid einkorn (including T. monococcum). Pizzuti et al. 13 showed lack of toxicity of T. monococcum gliadin in an in vitro coeliac small intestinal organ culture system. Vincentini et al. 26 and De Vincenzi et al. 27 showed that T. monococcum did not exhibit any negative effect on Caco-2/TC7 and K562(S) cells. T. monococcum was also suggested to be well tolerated for consumption by coeliac sufferers. 28 T. monococcum, however, possess large genetic diversity in multiple traits including grain storage proteins, 29 therefore using different varieties to study coeliac toxicity might not result with the same answer, in particular when the methodology used to assess toxicity is not consistent. The present study using gluten-reactive polyclonal T-cell lines however suggests that T. monococcum is not safe for individuals with CD. Tmp_AA triggered positive T-cell responses of most TCLs tested (11 out of 13). Our results are in agreement with those of Vaccino et al. who used genomic approach to conclude that einkorn has the full potential to induce CD 30 and Gianfrani et al. who showed that the monococcum lines Monlis and ID331 activate the CD T cell responses. 14 We conclude that the T-cell responses to ancient and modern wheat varieties are indeed heterogeneous. Further, we have shown that any wheat variety that is suggested to be low in CD toxicity needs to be tested in multiple individuals with CD using T-cell lines. This is because detecting CD toxic epitopes with single gluten-specific monoclonal antibodies or T-cell clones to a selected epitope can give results which do not fully reflect the toxicity of wheat for the overall population of CD patients. Although heterogeneous intestinal T-cell responses to ancient and modern wheat accessions were observed, our findings provide additional evidence for the necessity of a strict lifelong gluten-free diet in CD patients, without exception. Our results underline strongly the need for all cereals from the tribe Triticeae to be considered coeliac toxic. Conflict of interest The authors declare that they have no conflict of interest. Author contribution The work presented here was carried out in collaboration amongst all authors who co-worked in experimental design and writing the paper. T S performed the experiments, data analysis and interpretation and drafted the paper. AG and MC performed strains of wheat selection, data analysis and interpretation, manuscript writing. HJE designed the experiments, performed data interpretation and manuscript writing. PJC undertook the clinical work, supervised the project and provided intellectual input. All authors read and approved the final manuscript. Funding The funders of this study had no involvement in the study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the manuscript for publication. Acknowledgements Authors wish to thank the Istituto di Genetica e Sperimentazione Agraria N. Strampelli (Vicenza, Italy) for providing T. monococcum precoce (ITA ), the Centre for Genetic Resources (Wageningen, The Netherlands) for providing T. speltoides (CGN 10684) and T. aestivum compactum (CGN 04210), the Alce Nero Cooperative (Isola del Piano, PU, Italy) for supplying Graziella Ra, Senatore Cappelli and Svevo, and Molini del Conero (Osimo, AN, Italy) for providing Kamut. In addition, authors are thankful to Dr Derek Cooper for advice with statistical analysis, Dr Suzanne Donnelly for helpful discussions. T S wishes to thank Rosetrees Trust and Clinical Research Trust for financial support. References 1. Ludvigsson JF, Leffler DA, Bai JC, Biagi F, Fasano A, Green PH, et al. The Oslo definitions for coeliac disease and related terms. Gut 2012 [Epub ahead of print]. 2. van den Broeck HC, de Jong HC, Salentijn EMJ, Dekking L, Bosch D, Hamer RJ, et al. Presence of celiac disease epitopes in modern and old hexaploid wheat varieties: wheat breeding may have contributed to increased prevalence of celiac disease. Theor Appl Genet 2010;121:1527e Molberg O, McAdam SN, Korner R, Quarsten H, Kristiansen C, Madsen L, et al. Tissue transglutaminase selectively modifies gliadin peptides that are recognized by gut-derived T cells in celiac disease. Nat Med 1998;4:713e7. 4. Jabri B, Kasarda DD, Green PHR. Innate and adaptive immunity: the Yin and Yang of celiac disease. Immunol Rev 2005;206:219e van Herpen T, Goryunova SV, van der Schoot J, Mitreva M, Salentijn E, Vorst O, et al. Alpha-gliadin genes from the A, B, and D genomes of wheat contain different sets of celiac disease epitopes. BMC Genomics 2006;7:1. 6. Sollid LM, Qiao S-W, Anderson RP, Gianfrani C, Koning F. Nomenclature and listing of celiac disease relevant gluten T-cell epitopes restricted by HLA-DQ molecules. Immunogenetics 2012;64:455e Tye-Din JA, Stewart JA, Dromey JA, Beissbarth T, van Heel DA, Tatham A, et al. Comprehensive, quantitative mapping of T Cell epitopes in gluten in celiac disease. Sci Transl Med 2010;2:41e Camarca A, Anderson RP, Mamone G, Fierro O, Facchiano A, Costantini S, et al. Intestinal T cell responses to gluten peptides are largely heterogeneous: implications for a peptide-based therapy in celiac disease. J Immunol 2009;182: 4158e Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Vader W, Kooy YM, 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;191:603e Spaenij-Dekking L, Kooy-Winkelaar Y, Van Veelen P, Drijfhout JW, Jonker H, Van Soest L, et al. Natural variation in toxicity of wheat: potential for selection of nontoxic varieties for celiac disease patients. Gastroenterology 2005;129:797e Molberg O, Uhlen AK, Jensen T, Flaete NS, Fleckenstein B, Arentz-Hansen H, et al. Mapping of gluten T-cell epitopes in the bread wheat ancestors: implications for Celiac disease. Gastroenterology 2005;128:393e van den Broeck H, Chen H, Lacaze X, Dusautoir J-C, Gilissen L, Smulders M, et al. In search of tetraploid wheat accessions reduced in celiac disease-related gluten epitopes. Mol Biosystems 2010;6:2206e Pizzuti D, Buda A, D Odorico A, D Inca R, Chiarelli S, Curioni A, et al. Lack of intestinal mucosal toxicity of Triticum monococcum in celiac disease patients. Scand J Gastroenterol 2006;41:1305e Gianfrani C, Maglio M, Aufiero VR, Camarca A, Vocca I, Iaquinto G, et al. Immunogenicity of monococcum wheat in celiac patients. Am J Clin Nutr 2012;96:1339e Gregorini A, Colomba M, Ellis HJ, Ciclitira PJ. Immunogenicity characterization of two ancient wheat alpha-gliadin peptides related to coeliac disease. Nutrients 2009;1:276e Ehtemam MH, Rahiminejad MR, Saeidi H, Tabatabaei BE, Krattinger SG, Keller B. Relationships among the A genomes of Triticum L. species as evidenced by S S R markers, in Iran. Int J Mol Sci 2010;11:4309e25.

7 T. Suligoj et al. / Clinical Nutrition 32 (2013) 1043e Colomba M, Vischi M, Gregorini A. Molecular characterization and comparative analysis of six durum wheat accessions including Graziella Ra. Plant Mol Biol Rep 2012;30:168e Colomba MS, Gregorini A. Are ancient durum wheats less toxic to celiac patients? A study of alpha-gliadin from Graziella Ra and Kamut. TheScientificWorldJournal 2012;2012: The Management of Adults with Coeliac. BSG guidance on coeliac disease; Available from: [accessed July 5, 2012]. 20. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol 1999;11:1185e Wieser H, Antes S, Seilmeier W. Quantitative determination of gluten protein types in wheat flour by reversed-phase high-performance liquid chromatography. Cereal Chem 1998;75:644e Bolte G, Osman A, Mothes T, Stern M. Peptic-tryptic digests of gliadin: contaminating trypsin but not pepsin interferes with gastrointestinal protein binding characteristics. Clin Chim Acta 1996;247:59e Ellis HJ, Pollock EL, Engel W, Fraser JS, Rosen-Bronson S, Wieser H, et al. Investigation of the putative immunodominant T cell epitopes in coeliac disease. Gut 2003;52:212e Nassef HM, Bermudo Redondo MC, Ciclitira PJ, Ellis HJ, Fragoso A, O Sullivan CK. Electrochemical immunosensor for detection of celiac disease toxic gliadin in foodstuff. Anal Chem 2008;80:9265e Ellis HJ, Rosen-Bronson S, O Reilly N, Ciclitira PJ. Measurement of gluten using a monoclonal antibody to a coeliac toxic peptide of A-gliadin. Gut 1998;43: 190e Vincentini O, Maialetti F, Gazza L, Silano M, Dessi M, De Vincenzi M, et al. Environmental factors of celiac disease: cytotoxicity of hulled wheat species Triticum monococcum, T-turgidum ssp dicoccum and T-aestivum ssp spelta. J Gastroenterol Hepatol 2007;22:1816e De Vincenzi M, Luchetti R, Giovannini C, Pogna NE, Saponaro C, Galterio G, et al. In vitro toxicity testing of alcohol-soluble proteins from diploid wheat Triticum monococcum in celiac disease. J Biochem Toxicol 1996;11:313e Zanini B, Petroboni B, Not T, Pogna N, Lanzini A. Is Triticum monococcum a coeliac-safe wheat? A phase II, single blind, cross-over study on the effect of acute administration on intestinal permeability. Gut 2011;60(Suppl. 1):86e Jing H-C, Kornyukhin D, Kanyuka K, Orford S, Zlatska A, Mitrofanova OP, et al. Identificationofvariation in adaptivelyimportanttraitsandgenome-wideanalysis of trait-marker associations in Triticum monococcum. JExpBot2007;58:3749e Vaccino P, Becker H-A, Brandolini A, Salamini F, Kilian B. A catalogue of Triticum monococcum genes encoding toxic and immunogenic peptides for celiac disease patients. Mol Genet Genomics 2009;281:289e300.