lable at ScienceDirect

ARTICLE IN PRESS

Clinical Nutrition xxx (2009) 1–6

Contents lists avai

Clinical Nutrition

journal homepage: ht tp: / / int l .e lsevierheal th.com/journals /c lnu

Original Article

T-cell response to different cultivars of farro wheat, Triticum turgidumssp. dicoccum, in celiac disease patients

Olimpia Vincentini a, Osvaldo Borrelli b, Marco Silano a, Laura Gazza c, Norberto Pogna c,Roberto Luchetti d, Massimo De Vincenzi a,*

a Division of Food Safety and Veterinary Public Health, National Institute of Health, Viale Regina Elena 299, 00161 Roma, Italyb Department of Paediatrics, University of Rome, La Sapienza, University Hospital Umberto I, Rome, Italyc CRA-QCE, Research Unit for Cereal Quality, Rome, Italyd Department of Gastroenterology and Internal Medicine, General Hospital S. Filippo Neri, Rome, Italy

a r t i c l e i n f o

Article history:Received 8 August 2008Accepted 25 March 2009

Keywords:Celiac diseaseFarro landracesGenetic variabilityProlaminsT-cell proliferation

Abbreviations: A-PAGE, acid-polyacrylamide geldisease; HMW-GS, high-molecular-weight glutenin sline; LMW-GS, low-molecular-weight glutenin subunmononuclear cells; SDS-PAGE, sodium dodecylelectrophoresis.

* Corresponding author. Tel.: þ39 0649906002; faxE-mail addresses: olimpia.vincentini@iss.it (O. Vinc

iss.it (M. De Vincenzi).

0261-5614/$ – see front matter � 2009 Elsevier Ltd adoi:10.1016/j.clnu.2009.03.013

Please cite this article in press as: Vincentinceliac disease..., Clinical Nutrition (2009), do

s u m m a r y

Background & aims: Celiac disease is a gluten sensitive disorder that occurs in genetically susceptibleindividuals. The present study deals with variation in the immune response of mucosal T-cells fromceliac children to prolamins extracted from nine landraces of farro wheat (Triticum turgidum ssp.dicoccum) with contrasting storage protein compositions.

Methods: The prolamin fraction from nine ‘dicoccum’ wheat landraces was subjected to peptic–trypticdigestion and supplied to T-cells from mucosal explants of four celiac patients. Immune reactions interms of cell proliferation and INF-g secretion by intestinal T lymphocytes were then determined.

Results: T-cell lines exposed to digested prolamins from landraces L5563, L5558 and L5540 showednegligible proliferative responses and released INF-g amounts similar to that of untreated control cells.By contrast, landraces Ersa 6, Ersa 8, Leonessa 4 and Leonessa 5 proved to be very active in triggering theimmune responses, whereas landraces Filosini and Prometeo exhibited an intermediate behavior. One-dimensional fractionations by A-PAGE or SDS-PAGE revealed distinctive prolamin patterns amongst thelandraces analysed.

Conclusions: ‘Dicoccum’ wheat represents a heterogeneous species showing a wide variation in bothprolamin composition and T-cell immunological activation, some ‘dicoccum’ landraces being poor in ordevoid of noxious gluten proteins.

� 2009 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.

1. Introduction

Exposure to dietary gluten causes celiac disease (CD), an intes-tinal inflammatory disorder that occurs in approximately 0.5–1% ofindividuals.1,2 This disease is characterized by a rapid cytotoxiceffect on the intestinal epithelium, followed by a T-cell-mediatedimmune response.3,4 The external factors that trigger CD areprolamins, a heterogeneous group of storage proteins accumulatedin the endosperm of cereal seeds.5 Harmful prolamins include

electrophoresis; CD, celiacubunit; iTCL, intestinal T-cellit; PBMCs, peripheral blood

sulphate-polyacrilamide gel

: þ39 0649387149.entini), massimo.devincenzi@

nd European Society for Clinical N

i O et al., T-cell response to di:10.1016/j.clnu.2009.03.013

gliadins and glutenin subunits, which are responsible for theunique viscoelastic properties of wheat gluten,6 as well as hordeinsand secalins, the alcohol-soluble proteins of barley and rye,respectively.7

The recent literature strongly supports the hypothesis that CD isa multiform disease derived from interaction of two polymorphousfactors i.e. the immunological pattern of the CD patients exposed togluten and the primary structure of prolamins present in the grainof the different cereal species or cultivars.8,9 A single cultivar ofcommon (bread) wheat Triticum aestivum, the hexaploid species ofthe Triticum genus, possesses hundreds of prolamin genes, eachexpressing multiple alleles responsible of the extremely high levelof polymorphism in the prolamin patterns of the thousands ofwheat cultivars grown all over the world.5,6 Polymorphism inprolamin composition has also been observed in other cultivatedwheat species such as Triticum turgidum ssp. durum (durum wheat),Triticum monococcum (‘monococcum’ wheat) and T. turgidum ssp.dicoccum (‘dicoccum’ or farro wheat).10–12 This polymorphism was

utrition and Metabolism. All rights reserved.

ifferent cultivars of farro wheat, Triticum turgidum ssp. dicoccum, in

O. Vincentini et al. / Clinical Nutrition xxx (2009) 1–62

ARTICLE IN PRESS

found to be paralleled by variation in either the present/absence orthe amount of T-cell-stimulatory prolamin sequences, indicatingthat some wheat varieties are likely to be less toxic for celiacs.13

Moreover, reduced lesions, if any, were observed in culturedintestinal mucosa from CD patients exposed to prolamin peptidesfrom ‘monococcum’ wheat and some durum wheat cultivars, aswell as from common wheat cultivars lacking some gliadin poly-peptides.14–17

Large differences in the harmfulness profile of prolamins havebeen found in the Triticum genus, which contains diploid, tetraploidand hexaploid species.9,13,18 Prolamins from T. turgidum ssp. dicoc-cum were shown to differ from those of spelta wheat (T. aestivumssp. spelta) and bread wheat cv. S. Pastore in lacking cytotoxiceffects against human colon cancer Caco-2/TC7 cells grown invitro.19

The tetraploid AABB-genome wheat T. turgidum ssp. dicoccumwas the main cereal crop during the roman period under the Latinname of ‘‘farrum’’, and is still cultivated in hilly areas of Italy,Greece, Turkey and Balkan countries. The farro wheat is charac-terized by hulled (non-free-threshing) kernels and low glutencontent compared to T. turgidum ssp. durum, its naked-kernelcounterpart grown in the Mediterranean countries for pasta andcous–cous.12

The development of wheat genotypes lacking gluten proteinswith harmful sequences has been proposed.9,13,16 These genotypescould be valuable to select and breed wheat varieties of goodbreadmaking or pasta-making quality for consumption by CDpatients.

In the present work, gliadin-responsive T-cell lines from celiacpatients were investigated for their reactions against peptic–trypticdigested prolamins extracted from nine ‘dicoccum’ landraces withcontrasting storage protein compositions. This investigation wasaimed at obtaining information about immune reactivity of anancient wheat crop that has been neglected by the breeding activityin the last century. The hope is that selected cultivars of ‘dicoccum’wheat could represent an alternative food for CD patients, andreduce the exposition of the general human population to harmfulgluten proteins.

2. Materials and methods

2.1. Plant materials

Seeds of farro wheat (T. turgidum ssp dicoccum, genome AABB,2n¼ 28) landraces Filosini, Prometeo, Leonessa 4, Leonessa 5, Ersa6, Ersa 8, L5540, L5558 and L5563 were used in the present study.Common wheat (T. aestivum, genome AABBDD, 2n¼ 42) cv. S.Pastore and durum wheat (T. turgidum ssp. durum, genome AABB,2n¼ 28) cv. Simeto were used as well. All the plant material wasfrom the collection maintained by the CRA-QCE, Research Unit forCereal Quality, Rome, Italy.

2.2. Extraction of prolamin, gliadin and total protein

After removing the hull, seeds (100 g) were ground on a Cyclotec1093 mill (0.5 mm screen) and added to 400 ml (w/v) of 0.5 M NaCl.Extraction of prolamin was performed as described previously.20

Gliadin polypeptides were extracted from single crushed seeds(25 mg) with 75 ml of 70% (v/v) ethanol for 1 h at room temperatureunder constant agitation. After centrifugation at 15,000g for10 min, the protein suspension (25 ml) was mixed with 25 ml of anaqueous solution containing 50% glycerol and 0.1% (w/v) pyronineY. An aliquot (25 ml) of the suspension was fractionated by A-PAGE.Total proteins from individual crushed seeds (25 mg) wereextracted with 0.5 ml of a solution containing 0.25 M Tris–HCl

Please cite this article in press as: Vincentini O et al., T-cell response to dceliac disease..., Clinical Nutrition (2009), doi:10.1016/j.clnu.2009.03.013

buffer (pH 6.8), 0.12% (w/v) SDS, 10% (v/v) glycerol, 0.2% (w/v)pyronine Y and 5% 2-mercaptoethanol and shaken for 1 h at roomtemperature. After incubation at 80 �C for 20 min and centrifuga-tion at 15,000g for 10 min, an aliquot (20 ml) of the proteinsuspension was fractionated by SDS-PAGE.

2.3. Peptic–tryptic digestion

The alcohol-soluble protein fraction was extracted from wholeflour and subjected to sequential digestion as described previ-ously.21 Prolamin peptic–tryptic digests were freeze-dried, lyophi-lized and stored at �20 �C until used. The different farro wheatdigests were freshly prepared by suspending them in RPMI 1640and filtered at 0.45 mm. Gliadin preparations were assayed forendotoxin by using the QCL-100 reagent kit (Biowhittaker, Wal-kersville, MD, USA) and found to have an endotoxin level lowerthan 0.5 E.U./ml.

2.4. Fractionation by A-PAGE and SDS-PAGE

Protein fractionation by acid-polyacrylamide gel electrophoresis(A-PAGE) was performed at pH 3.1 in a Hoefer SE 600 apparatus(Amersham) at 450 V constant voltage until 1 h after the dyereached the bottom of the gel. The separating gel (16 cm� 18 cm),1.5 mm thick, was prepared by mixing 7.5% acrylamide (T¼ 7.5%and C¼ 0.37%) with 80 mg of ascorbic acid, 8 ml of sodium lactatebuffer (3.4 g/l of 97% NaOH adjusted to pH 3.1 with lactic acid) and120 ml of 1% ferrous sulphate, brought to 80 ml with distilled water.After cooling at 4 �C, the acrylamide solution was mixed with 24 mlof 2.5% hydrogen peroxide and immediately poured into the gelcassette. The electrophoretic buffer was 0.17 g/l of 97% NaOHadjusted to pH 3.1 with lactic acid.

SDS-PAGE running gels were prepared with 15% (w/v) acryl-amide (T¼ 15% and C¼ 0.5%), 0.375 M Tris–HCl (pH 8.4) and 0.1%(w/v) SDS. Stacking gels contained 4.5% (w/v) acrylamide (T¼ 4.5%and C¼ 0.06%), 0.08 M Tris–HCl (pH 6.8) and 0.1% (w/v) SDS. Theelectrophoresis buffer was 0.025 M Tris–glycine (pH 8.3) and 0.1%(w/v) SDS. The gels (160�180�1.5 mm) were run at 18 mA/gel.Electrophoresis was stopped 1 h after the tracking dye had reachedthe bottom of the gel. A 0.25% (w/v) solution of Coomassie BrilliantBlue R250 in 6% trichloroacetic acid was used to stain both A-PAGEand SDS-PAGE gels.

2.5. Patients

Four patients (3F and 1M, mean age at enrollment 4.5 years)were included in this study. Celiac disease was diagnosed accordingto the revised ESPGHAN criteria.22 The study was approved by theNational Institute of Health Committee (CE-ISS-05/111, January 9th,2006) and written informed consent was obtained from the parentsof the patients. Biopsy specimens from distal duodenum wereobtained during upper gastrointestinal endoscopies performed fordiagnostic purposes in subjects under study for suspected orknown CD at the diagnosis. All patients were consuminggluten containing food and exhibited a positive titre of anti-transglutaminase antibody.

2.6. tTG deamidation

Deamidation mixes were set up as follows: 100 mg/ml Guineapig liver tTG, 2 mM CaCl2, 1 mg/ml of peptic–tryptic digest of thedifferent farro wheats. Incubation was carried out for 4 h at 37 �C.All reagents were purchased from Sigma Chemical (St. Louis, MO,USA).

ifferent cultivars of farro wheat, Triticum turgidum ssp. dicoccum, in

O. Vincentini et al. / Clinical Nutrition xxx (2009) 1–6 3

ARTICLE IN PRESS

2.7. Organ cultures and establishment of small intestinal T-cell lines

Short term intestinal T-cell lines (iTCL-2, iTCL-5, iTCL-6, iTCL-11)were generated from the mucosal explants of four celiac patientsand three normal donors. Polyclonal T-cell lines were establishedfrom antigen-challenged intestinal biopsy specimens as describedpreviously.23 Small intestinal biopsies from celiac patients werefirst oriented on a metal grid in a tissue culture plate (Falcon,Franklin Lakes, NJ, USA) and incubated with 1 mg/ml of prolamindigest in RPMI 1640 (GIBCO, Invitrogen, Carlsbad, CA, USA) for 24 h.Biopsies were then digested with 1 mg/ml collagenase A (Sigma, St.Louis, MO, USA) in RPMI for 1 h at 37 �C with gentle shaking toproduce a single-cell suspension and passed through a 70 mm filterto eliminate epithelial cells. Cell obtained from biopsies werecultured in a 96-well-plate in the RPMI 1640 medium supple-mented with antibiotics, non-essential amino acids, sodium pyru-vate, glutamine and 10% inactivated human serum, stimulated with1�106/ml irradiated (3500 rad) autologous peripheral bloodmononuclear cells (PBMCs) together with 10 ng/ml IL-15 (Pepro-tech, London, UK). On day 7th, iTCLs were restimulated with 50 mg/ml of deamidated prolamin digest and a feeder layer of autologusirradiated PBMCs. Every 3 days, fresh medium and 10 ng/ml of IL-15were added to the cell culture. Plasmocin (2.5 mg/ml, Invitrogen,Carlsbad, CA, USA) was added to the cell culture to prevent myco-plasma infection. On day 15th, all lines established from eachbiopsy were tested in the resting phase.

2.8. Proliferation assay and IFN-g determination

Triplicate wells containing autologous irradiated PBMCs(1�105) acting as antigen-presenting cells were plated in 96-wellU-bottom plates and prepulsed overnight with prolamin digests ofthe different farro wheats in a 100 ml volume. Mucosal T-cells(5�104 cells per well) were added in a final volume of 200 ml. Finalconcentration of farro wheat digests was 100 mg/ml. Each cell line

Fig. 1. A-PAGE fractionation of gliadins from single seeds of farro wheat landraces. (1–3) Lgliadin biotypes are indicated by Latin letters.

Please cite this article in press as: Vincentini O et al., T-cell response to dceliac disease..., Clinical Nutrition (2009), doi:10.1016/j.clnu.2009.03.013

was tested separately, and the results were expressed as themean� SD. The culture medium was used as the negative control.

T-cell proliferation was performed after 72 h of incubation usingthe ELISA 5-bromo-2-deoxyuridine (BrdU) cell proliferation test(Chemicon International, Temecula, CA, USA), and read at 450 nmon a microplate reader. All assays were performed two or moretimes with similar results. Proliferative responses to antigen wereregarded as positive if the stimulation index (SI) was higher than 3.SI was calculated as the mean absorbance in the presence of theantigen divided by the mean absorbance in the absence of theantigen. Aliquots of supernatants from each iTCL were collectedafter 48 h and stored at �80 �C for IFN-g determination usingcommercial ELISA Kit in accordance with the manufacturer’sinstructions (Biosource, Camarillo, CA, USA). Standards were run oneach plate. Sensitivity of assay was 15 pg/ml.

2.9. Statistical analysis

Data are presented as arithmetic means� S.D. of at least threeindependent experiments. Statistical analysis was performed byStudent’s t-test. P-values lower than 0.05 were considered to besignificant.

3. Results

3.1. Different gliadin patterns in farro wheats

Upon A-PAGE, gliadins from the farro wheat landraces appearedas 18–25 bands having a wide variability in mobility and intensity.These bands are classified as a-, b-, g- and u-gliadins according totheir mobilities as shown in Fig. 1. Fractionation of 10 single seedsfrom each landrace revealed 13 patterns, six landraces exhibitingtwo to four different gliadin compositions (biotypes). Biotypea (Fig. 1, lane 14) was found in Filosini and Leonessa 4, whereasbiotype b (lanes 10–12) was observed in Ersa 6, Ersa 8, Leonessa 4and Leonessa 5 (Table 1).

5563, (4–6) L5558, (7–9) Prometeo, (10–12) Ersa 6 and (13,14) Filosini. The different

ifferent cultivars of farro wheat, Triticum turgidum ssp. dicoccum, in

Table 1Gliadin and HMW-GS compositions of farro wheat landraces.

Landrace Gliadinpattern

HMW-GS composition

No. Biotype No. Subunits

Leonessa 4 3 a, b, c 3 S1þ S6; S2þ S17; S1þ S6þ S8Leonessa 5 1 b 5 S17; S2þ S17; S1þ S6þ S8; S2þ S6þ S9; S1þ S6Ersa 6 1 b 2 S1þ S6þ S8; S1þ S17Ersa 8 2 b, o 1 S17Filosini 4 a, g, h, p 3 S2þ S17; S1þ S6þ S8; S2þ S6þ S8Prometeo 3 i, j, k 1 S1þ S17L5558 1 l 1 S6þ S8L5563 2 m, n 1 S1þ S6

O. Vincentini et al. / Clinical Nutrition xxx (2009) 1–64

ARTICLE IN PRESS

When total proteins extracted from single seeds of each land-race were reduced with 2-mercaptoethanol and fractionated bySDS-PAGE, eight different HMW-GS patterns were observed (Fig. 2).In particular, Ersa 6, Filosini, Leonessa 4 and Leonessa 5 were foundto contain two to five genotypes having different HMW-GScompositions, whereas each of the remaining landraces showed thesame HMW-GS pattern in the 10 seeds analysed electrophoretically(Table 1). Furthermore, the same HMW-GS pattern was observed indifferent landraces. For example, subunits S1þ S6þ S8 occurred infour landraces, whereas subunits S1þ S6 and S2þ S17 were foundin three landraces each. On the contrary, compositions S2þ S6þ S9and S6þ S8 were typical of landraces Leonessa 5 and L5558,respectively. The farro wheats analysed here showed genetic vari-ability for low-molecular-weight glutenin subunits (LMW-GS) aswell (Fig. 2). Gliadins and total proteins of farro wheat landraceL5540 were not analysed.

The banding patterns of gliadins, HMW-GS and LMW-GS of theendosperm of farro wheats were quite distinctive with respect tothose of durum wheat. Moreover, HMW glutenin subunit pairS7þ S8 and gliadin bands g-45 and u-35, which occur in durumwheat cultivars of superior pasta-making quality,20 were not foundin the ‘dicoccum’ landraces.

3.2. Proliferative response and INF-g release

The biopsies of the enrolled CD patients generated fourprolamin-reactive cell lines. As expected, upon challenge withprolamin digests from both bread wheat and farro wheat landraces,

Fig. 2. SDS-PAGE fractionation of total proteins from single kernels of ‘dicoccum’ wheat landmolecular-weight glutenin subunits; LMW-GS¼ low-molecular-weight glutenin subunits. S

Please cite this article in press as: Vincentini O et al., T-cell response to dceliac disease..., Clinical Nutrition (2009), doi:10.1016/j.clnu.2009.03.013

mucosal T-cells from three normal donors failed to proliferate.T-cell stimulation was determined by measurement of cell prolif-eration and INF-g production. In particular, cell proliferation, asdetermined by incorporation of 5-bromo-2-deoxyuridine (BrdU)into T-cell DNA, was expressed as the mean SI of the T-cell lineobtained from each of the four biopsies, and the responses of theT-cell lines to the prolamin digests from the farro landraces werecompared with that elicited by the prolamin digest of bread wheatcv. S. Pastore (Fig. 3). Mean SI values as low as 1.5� 0.3, 1.3� 0.2 and1.2� 0.3 were found for landraces L5540, L5558 and L5563,respectively. By contrast, ‘dicoccum’ wheats Leonessa 4, Leonessa 5,Ersa 6 and Ersa 8 induced strong proliferative responses(SI¼ 4.6� 0.4–6.0� 0.7), whereas Filosini (SI¼ 3.0� 0.8) andPrometeo (3.4� 0.70) exhibited an intermediate behavior.

INF-g is the main cytokine involved in the inflammatoryresponse of CD.24 The ability of deamidated digests from the farrowheats to induce INF-g production was tested in intestinal poly-clonal T-cell lines derived from four patients with untreated celiacdisease. The iTCL response to the prolamin digest was consideredpositive when INF-g production was at least 2-fold higher than thatcaused by the culture medium. The differences in proliferative T-cell response observed amongst the farro landraces (Fig. 3) wereparalleled by similar differences in INF-g production (Fig. 4). Inparticular, T-cell lines challenged with the prolamin digests oflandraces Ersa 6 and Ersa 8 showed mean INF-g releases as high as0.57� 0.04 ng/ml and 0.61�0.03 ng/ml, respectively, whereasthose exposed to the prolamin digests from L5540, L5558 andL5563 exhibited mean INF-g values as low as 0.10� 0.02 ng/ml,0.15� 0.05 ng/ml and 0.06� 0.03 ng/ml, respectively, suggestingthat genotypes belonging to the same ‘dicoccum’ wheat species candiffer from each other in the amount of T-cell-stimulatory epitopesin their gluten sequences.

4. Discussion

In previous works,13,15,16 the ancient ‘monococcum’ and ‘dicoc-cum’ wheat crops were claimed to be potentially less toxic for theCD patients than the modern durum or bread wheat cultivars.

The farro landraces analysed in this study are grown in hillyregions of Central and Southern Italy using very low amounts ofnitrogen fertilizer. The de-hulled kernels are used in the prepara-tion of soups, traditional cakes, biscuits and pasta. Unfortunately,

races. (1–7) Ersa 6, (8–11) Leonessa 4, (12) Filosini and (13) Prometeo. HMW-GS¼ high-1–S17 indicate different HMW-GS.

ifferent cultivars of farro wheat, Triticum turgidum ssp. dicoccum, in

Fig. 3. Proliferative responses of intestinal T-cell lines iTLC-2, iTLC-5, iTLC-6 and iTLC-11 to prolamins from nine farro wheat landraces. The stimulation index (SI) value wascalculated by dividing the mean absorbance at 450 nm after stimulation for 72 h with each tTG-treated prolamin digest by the mean absorbance of T-cells exposed to the culturemedium alone (negative control). The experiments were performed in triplicate and the mean SI value� SD is shown. The SI values of T-cells exposed to prolamin digests werestatistically significant (*P< 0.05) with respect to the negative control, with the only exception of landraces L5540, L5558 and L5563. GPT¼ peptic–tryptic digested prolamins frombread wheat cv. S. Pastore (positive control).

O. Vincentini et al. / Clinical Nutrition xxx (2009) 1–6 5

ARTICLE IN PRESS

these landraces are characterized by low yield, reduced proteincontent and poor gluten quality when grown under ‘‘low input’’conditions.12 However, the rising demand of healthy and tasty foodand the need of supporting the economy of the rural areas haverecently increased the interest for this crop.

Fig. 4. IFN-g production by intestinal T-cell lines iTLC-2, iTLC-5, iTLC-6 and iTLC-11 followinare expressed as mean� SD of triplicate cultures. INF-g release from all wheat genotypes econtrol cells (ctr) exposed to the culture medium alone. GPT¼ peptic–tryptic digested prol

Please cite this article in press as: Vincentini O et al., T-cell response to dceliac disease..., Clinical Nutrition (2009), doi:10.1016/j.clnu.2009.03.013

When fractionated by A-PAGE and SDS-PAGE, the ‘dicoccum’landraces exhibited eight different HMW-GS compositions and 13gliadin patterns quite distinctive with respect to those of bothdurum and bread wheat. This observation is in agreement withearlier reports12,25 suggesting that T. turgidum ssp. dicoccum can be

g 48 h stimulation with prolamin digests (100 mg/ml) from nine farro landraces. Resultsxcept L5540, L5558 and L5563 were statistically significant (*P< 0.05) with respect toamins from bread wheat cv. S. Pastore (positive control).

ifferent cultivars of farro wheat, Triticum turgidum ssp. dicoccum, in

O. Vincentini et al. / Clinical Nutrition xxx (2009) 1–66

ARTICLE IN PRESS

considered as a rich source of novel prolamin proteins with unusualprimary structures. All farro wheats except L5558 consisted of twoor more genotypes having contrasting gliadin and glutenincompositions (Table 1), a fact that should be taken into accountwhen landraces are compared for their gluten characteristics andtoxicity profiles. Moreover, the presence of gliadin biotype b inLeonessa 4, Leonessa 5, Ersa 6 and Ersa 8 indicates that theselandraces are genetically related to each other. The ‘dicoccum’ lineswere found to lack gliadins g-45/u-35, HMW glutenin subunits‘‘7þ 8’’ and ‘‘type-2’’ LMW-GS, which are typical of durum wheatcultivars of superior pasta-making quality,20 a fact that couldaccounts for the poor technological quality of farro wheats grownin Italy.12

Prolamins from nine farro landraces were extracted andscreened for the occurrence of T-cell-stimulatory epitopes. Inter-estingly, these genotypes revealed distinct differences in theirimmune activity against intestinal T-cells from CD patients. In fact,prolamins from landraces L5540, L5558 and L5563 induced negli-gible responses in terms of cell proliferation and INF-g productionby intestinal T-cell lines from four celiac children. In this context itis noteworthy that diploid and tetraploid hulled wheat species T.monococcum and T. turgidum ssp. dicoccum have been found to bepoor in or devoid of noxious prolamins.9 Moreover, prolaminpeptides from Filosini and Prometeo were weakly toxic, whereasthose from Leonessa 4, Leonessa 5, Ersa 6 and Ersa 8 proved to bevery active in triggering both T-cell proliferation and INF-g release.The immune reactions of these latter ‘dicoccum’ genotypes werecomparable to that of hexaploid common wheat, suggesting thatgenes coding for toxic prolamins are not confined to the D genome.The very reactive genotypes Ersa 6, Ersa 8, Leonessa 4 and Leonessa5 were found to share gliadin biotype b, which is absent in theinactive farro genotypes L5540, L5558 and L5563. This biotype andbiotypes a, c and o present in Ersa 8 and Leonessa 4 are currentlybeing compared for their proliferative responses against intestinalT-cells. The present results suggest that the contrasting immuneresponses exhibited by the ‘dicoccum’ wheat landraces are likelydue to variation in their storage protein compositions.

‘Dicoccum’ landraces have been grown for many years in Centraland Southern Italy, and likely represent the ‘farrum’ wheat pop-ulations cultivated in Italy during the ancient roman period. Takinginto account the unique prolamin patterns of ‘dicoccum’ wheat, itseems unlikely that they had played any significant role as parentalgenotypes of durum or bread wheat cultivars bred in Italy duringthe last century. The present findings suggest that (i) the humanpopulation has long been exposed to toxic wheat flour and (ii) farrowheats are a heterogeneous group of genotypes showing a widevariability in eliciting immune reactions by T-cell of CD individuals.These conclusions can raise the prospect of breeding wheat specieswith low levels of harmful gluten proteins with the attractive targetof developing non-toxic wheat cultivars with superior breadmak-ing or pasta-making quality.

Conflict of interest

The authors state that they have no conflict of interests.

Acknowledgments

All authors contributed to the preparation of the manuscript.O.V. participated in the conception design and writing, establishedthe T-cell line and performed the proliferation tests; O.B. and R.L:were responsible for patients recruitment and screened thepatients enrolled in the study; M.S. performed the proliferationtests, contributed to the interpretation of the data and drafted themanuscript; L.G. performed the peptic–tryptic digests and the

Please cite this article in press as: Vincentini O et al., T-cell response to dceliac disease..., Clinical Nutrition (2009), doi:10.1016/j.clnu.2009.03.013

fractionation of storage proteins by A-PAGE and SDS-PAGE, andparticipated in writing the manuscript; N.P. revised the draft of themanuscript, and added important points to the discussion; M.D.V.conceived the study, interpreted the data and critically revised themanuscript.

References

1. Marsh MN. Mucosal pathology in gluten sensitivity. In: Marsh MN, editor.Coeliac disease. Oxford: Blackwell Scientific Publications; 1992. p. 136–92.

2. Jabri B, Sollid LM. Mechanism of disease: immunopathogenesis of celiacdisease. Nat Clin Pract Gastroenterol Hepatol 2006;3:516–25.

3. Sollid LM. Molecular basis of celiac disease. Annu Rev Immunol 2000;18:53–81.

4. Maiuri L, Troncone R, Mayer M, Coletta S, Picarelli A, De Vincenzi M, et al. Invitro activities of A-gliadin-related synthetic peptides: damaging effect on theatrophic coeliac mucosa and activation of mucosal immune response in thetreated celiac mucosa. Scand J Gastroenterol 1996;31:247–53.

5. Shewry PR, Tatam AS. The prolamins storage proteins of cereal seeds: structureand evolution. Biochem J 1990;267:1–12.

6. Payne PI. Genetics of wheat storage proteins and the effect of allelic variationon breadmaking quality. Annu Rev Plant Physiol 1987;38:141–53.

7. Kilmartin C, Wieser H, Abuzakouk M, Kelly J, Jackson J, Feighery C. Intestinal Tcell response to cereal proteins in celiac disease. Dig Dis Sci 2006;51:202–9.

8. van Herpen TW, Goryunova SV, van der Schoot, Mitreva M, Salentijn E, Vorst O,et al. Alpha-gliadin genes from the A, B, and D genomes of wheat containdifferent sets of celiac disease epitopes. BMC Genomics 2006;7:1–13.

9. Molberg O, Uhlen AK, Jensen T, Flacte NS, Fleckenstein B, Arentz-Hanson H,et al. Mapping of gluten T-cell epitopes in bread wheat ancestors: implicationsfor celiac disease. Gastroenterology 2005;128:393–401.

10. Saponaro C, Pogna NE, Castagna R, Pasquini M, Cacciatori L, Raedelli R. Allelicvariation at the Gli-A1m, Gli-A2m and Glu-A1m loci and breadmaking quality indiploid wheat Triticum monococcum. Gen Res Camb 1995;66:127–37.

11. Kudryavtsev AM. Genetics of gliadin of spring durum wheat (Triticum durumDesf.). Russ J Gen Camb 1994;30:69–75.

12. Galterio G, Cappelloni M, Desiderio E, Pogna NE. Genetic, technological andnutritional characteristics of three Italian populations of ‘‘farrum’’ (Triticumturgidum ssp. dicoccum). J Genet Breed 1994;48:391–8.

13. Spaenij-Dekking L, Kooy-Winkelaar Y, Van Veelen P, Drijffou JW, Jorker H, vanSoest L, et al. Natural variation in toxicity of wheat: potential for selection ofnon-toxic varieties for celiac disease patients. Gastroenterology 2005;129:797–806.

14. Auricchio S, De Ritis G, De Vincenzi M, Occorsio P, Silano V. Effects of gliadin-derived peptides from bread and durum wheat on small intestine cultures fromrat fetus and coeliac children. Pediatr Res 1982;16:1004–10.

15. Pizzuti D, Buda A, D’Odorico A, D’Inca R, Chiarelli S, Curioni A. Lack of intestinalmucosal toxicity of Triticum monococcum in celiac disease patients. Scand JGastroenterol 2006;41:1305–11.

16. De Vincenzi M, Luchetti R, Giovannini C, Pogna NE, Saponaro C, Galterio G. Invitro toxicity testing of alcohol-soluble proteins from diploid wheat Triticummonococcum in coeliac disease. J Biochem Toxicol 1996;11:313–8.

17. De Vincenzi M, Dessı M, Luchetti R, Pogna NE, Radaelli R, Galterio G. Toxicity ofbread wheat lines lacking prolamins encoded by the Gli-B1/Gli-B5/Glu-B3 andGli-D1/Glu-D3 loci in coeliac disease as determined by their agglutinatingactivity. ATLA 1996;24:39–48.

18. Anderson RP, van Heel DA, Tye-Din JA, Jewell DP, Hill AV. Antagonists and nontoxic variants of the dominant wheat gliadin T cell epitope in celiac disease. Gut2006;55:485–91.

19. Vincentini O, Maialetti F, Gazza L, Silano M, Dessi M, De Vincenzi M. Environ-mental factors of celiac disease: cytotoxicity of hulled wheat species Triticummonococcum, T. turgidum ssp. dicoccum and T. aestivum ssp. spelta. J Gastro-enterol Hepatol 2007;22(11):1816–22.

20. Pogna NE, Autran JC, Mellini F, La Flandra D, Fillet L. Cromosome 1B encodedgliadins and glutenins subunits in durum wheat: genetics and relationship togluten strength. J Cereal Sci 1990;11:15–34.

21. De Ritis G, Occorsio P, Auricchio S, Gramenzi F, Morisi G, Silano V. Toxicity ofwheat flour and protein-derived peptides for in vitro developing intestine fromrat fetus. Pediatr Res 1979;13:1255–61.

22. Walker-Smith J, Guandalini S, Schmizt J, Shmerling D, Visakorpi J. Report ofWorking Group of European Society for Pediatric Gastroenterology andNutrition: revised criteria for diagnosis of celiac disease. Arch Dis Child1990;65:909–11.

23. Molberg O, McAdam SN, Lundin KEA, Sollid LM. Studies on gliadin specific Tcells in celiac disease. In: Marsh MN, editor. Methods in molecular medicine.Celiac disease: methods and protocols, vol. 41. Totowa NJ: Humana Press; 2000.p. 105–24.

24. Lahat N, Shapiro S, Karban A, Gerstein R, Kinarty A, Lerner A. Cytokine profile inceliac disease. Scand J Immunol 1999;49:441–6.

25. Vallega V, Waines PI. High molecular weight glutenin subunit variation inTriticum turgidum var. dicoccum. Theor Appl Genet 1987;74:706–10.

ifferent cultivars of farro wheat, Triticum turgidum ssp. dicoccum, in