Wheat Gluten and Celiac Disease in Human

BY

AHMED ELFATIH ELDOLIEFY

Topic Advisor ASS. PROF. SENAY SIMSEK

PhD Supervisor PROF. MOHAMED MERGOUM

Outlines

Introduction

Wheat importance

Gluten

Celiac DiseaseCauses

Therapy

Research

Breeding efforts

Abstarct: Gluten is a vital component in bread wheat flour and wheat based products, yet in addition to its nutritional value, gluten showed some inheritable allergic disorder for some patients called celiac disease. Celiac Disease is a chronic enteropathy of the upper portion of the small intestine, which affects between 0.5 and 2 % of humans. The main concern regarding CD is the fast diagnosis, because if late, or undiagnosed, the risk of death is increased dramatically as statistically reported during the last 50 years in the United States of America (USA). Some vital environmental factors behind the increased CD specially in Asia.

Molecular techniques

Bread wheat (Triticum aestivum) is a naturally hybridized allohexaploid species, where T. turgidum (AB genomes) combined with wild species Aegilops tauschii (D genome). D Genome has improved the bread-making quality

Wheat importance

The wheat grain is a single-seeded fruit, or caryopsis, and consists of three distinct parts: the embryo, which forms the next generation; the starchy endosperm, a storage tissue which supports germination and early seedling growth; and the bran, comprising the pericarp (fruit coat), testa (seed coat) and outer endosperm (aleurone layer). The starchy endosperm cells are packed with insoluble storage components, mainly starch and protein, which account for over 80 % of the grain mass. During milling the endosperm is usually separated from the germ and the bran in order to obtain the white flour, which is the most suitable raw material for modern bread making.

At grain maturity, only the embryo and the aleurone cells remain viable, the subaleurone and central starchy endosperm cells having undergone a process of programmed cell death (PCD), which includes the degeneration of nuclei and cytoplasm (Young and Gallie, 1999).

The starch is distributed fairly evenly throughout this tissue, particularly in the dorsal part of the grain, with the exception of the sub-aleurone cells where the content of starch is significantly lower (Ugalde and Jenner, 1990a).

They are classically divided into monomeric gliadins and polymeric glutenins, with the latter comprising subunits which are further divided on the basis of their molecular masses into high-molecular-weight subunits of glutenin (HMW-GS) and low-molecular-weight subunits of glutenin (LMW-GS). Clear gradients exist in protein concentration across the starchy endosperm, being high in the sub-aleurone cells and lower in the central starchy endosperm cells.

The second major cell-specific components of the starchy endosperm cells, and the most important in terms of flour functional properties, are the gluten proteins. These belong to the prolamin family of cereal grain storage proteins and are characterized by their high content of the amino acids proline and glutamine.

What is Gluten

The most important proteins in wheat are the gluten. Gluten is almost equally comprised of gliadins and glutenins that represent 80% of storage proteins in wheat kernel.

Other proteins have lower representation such as albumins (12%) and globulins (8%). At the time that glutenin can be sub-classified into low molecular weight glutenin subunit (LMW-GS) and High MW-GS, our interest will be focused on the gliadins, the major effecter in CD, which can be distinguished as α, β, γ and ω proteins (Shewry and Tatham 1999).

high proline and glutamine contents protect the gluten proteins from proteolytic digestion (Shan et al. 2004). Partial digestion of gluten proteins produces T-cell stimulatory peptide epitopes (koning 2008). In regard to both adaptive and innate immune responses, the epitopes of α-gliadins were highly correlated with clinical relevance to CD development (Camarca et al. 2009).

How celiac disease occur

Celiac disease

It is an inheritable chronic enteropathy of the upper portion of the small intestine, which affects between 0.5 and 2 % of humans.

Gluten consumption is triggering the allergy resulting in intestinal inflammation, villous atrophy, and crypt cell hyperplasia, which leads to a significant decrease in intestinal wall surface area. Consequently, many symptoms are taking place, such as diarrhea, weight loss, anemia, and bone disorders (Green and Cellier 2007). Three factors unite together to manifest disease, which are the genetic makeup, environmental conditions, and immune response.

At the genetic factor level, most patients are sharing a heterodimeric classII (HLA-DQ2 or HLA-DQ8) as a common genetic background (HLA-DQ region of 6p21.3) (Lundin et al. 1990). These classII molecules can confer susceptibility and be expressed on antigen-presenting cells (mainly macrophages, dendritic cells, and B cells). CD-related peptides are presented to T cells in the lamina propria of the small intestine, by the human leukocyte classII motifs. This triggers the gluten specific CD4+ T-helper 1 (Th1) cells (central effector cells for intestinal inflammation resulting in crypt hyperplasia and villous atrophy). Although 30-35% of predominant populations are expressing classII, only 2–5 % of gene carriers develop CD, which implicates the existence of other genes. Dubois et al. (2010) investigated 13 loci to be involved in CD, while Trynka et al. (2011) used genome-wide association analysis, to introduce 139,000 SNPs across ~12,000 cases and equal number of controls and discovered 36 loci including 13 new loci conferring susceptibility to CD. This indicates multiple genetic factors that are yet to be elucidated and may play an important genetic role in CD.

Fig1. Celiac disease pathogenesis. Gluten peptides that are highly resistant to intestinal proteases reach the lamina propria by loosening of tight junctions because of zonulin response or by epithelial transcytosis. Reaching lamina propria the epitopes get crosslinked and deamidated by tissue transglutaminase 2 (tTG2) and presented via HLA-DQ2 (-DQ8) on the surface of antigenpresenting cells (APC). Subsequently, CD4+ T cells are activated; as a result, the secretion of Th1 cytokines will trigger the release and activation of metal proteinases by myofibroblasts, finally resulting in mucosal remodeling and villous atrophy (Osori et al. 2012).

The third factor is the immune response; which is mainly depending on the protein composition. Gluten peptides are mainly without negatively charged amino acids, which is weakening its binding ability to HLA DQ2 or DQ8. This means that modification of the peptides is required for high affinity binding which is done by tTG2 (ubiquitous enzyme expressed in tissues associated with the extracellular matrix). The tTG2 targets certain glutamine residues in extra- or intracellular proteins by binding to the lysine residues to establish a crosslink between the two proteins and deaminates these glutamines to negatively charged glutamic acids (Sollid and Jabri, 2011). The spacing between the targeted glutamine and a C-terminal proline residue is crucial for the specificity of tTG2 binding. Therefore, the sequence QXP is often found in the peptide region where the T cell epitopes cluster. A proline at positions 6 and 8 in the peptide epitope facilitates the enzymatic conversion of glutamine into glutamic acid. This characteristic led to the conclusion that HLA-DQ2 is capable of binding multiple negatively charged residues, at position 1, 3, 5, 7, and 9 (Stepniak et al. 2008). Many deamidated gluten peptides have been identified that share a highly rich proline and glutamine epitope region, which make them exceptionally resistant to gastric, pancreatic, and intestinal digestive proteases.

The environmental factor is simplified in the immunogenic peptides of wheat, barley, and rye prolamins, which trigger the immune response in genetically predisposed individuals. Environmental factors were investigated such as, decrease in exclusive breastfeeding time to infants and early exposure to dietary gluten together with a change of the bacterial flora, as well as infection with enteropathic viruses, all can favor the development of CD at early ages (Collado et al. 2009).

The first important inquiry that may be proposed is: how the gluten components seep out of the gut to interact freely with genetically sensitized elements of the immune system? Studies on the mechanisms for transfer of immunogenic gluten peptides into the lamina propria revealed two pathways. First, gliadin fragments increase permeability of the intestinal epithelium by modifying the structure of the intercellular tight junctions that leads to an increased space between the epithelial cells allowing paracellular antigen traffic into the lamina propria (Tripathi et al. 2009). Second, intracellular transport of gliadins by transcytosis through the epithelial cells in apical-to-basal direction, i.e., the gliadin peptides are taken up into the epithelial cells and secreted towards the lamina propria (Heyman and Menard 2009).

Can wheat breeding practices truly affect CD incidence?

In a research study conducted by Van den Broeck et al. (2010), they tried to find out how much genetic diversity is existed in gluten proteins to produce two CD epitopes (Glia-α9 and Glia-α20) in 36 modern European wheat varieties compared to 50 landraces that been grown for nearly a century. Glia-α9

was chosen because it is the immuno-dominant epitope that is highly recognized in CD patients, while Glia-α20 was included as minor technical reference. They used two monoclonal antibodies generated against both epitopes (Mitea et al. 2008a) to screen for the T-cell stimulatory epitopes in gluten protein extracts from the plant materials. Glia-α9 epitope sequence (αI) has a proteolytic-resistance 33-mer in α-gliadins that has high T-cell stimulatory effect (Shan et al. 2005).

This research provides evidence that the diversity in banding patterns was lower in modern varieties which was concluded as a reduction in genetic diversity among these varieties. An implication toward non-diagnosed CD patients nowadays may encounter less diverse gluten proteins than preceding decades. It is necessary to note that they faced some technical shortage to accurately visualize T-cell recognition using the antibodies that might over-estimate toxicity. Though they have used protein extracts from a limited selection of wheat varieties they successfully demonstrated a large variation in immune responses based on antibodies binding affinity studies (table 2; Salentijn et al. 2009).

Defensively, the authors clarified that using only European varieties in their study does not affect the comparison as the decrease in the diversity of α-gliadins in the varieties was the same when only Euorpean landraces were considered. Moreover, they concluded that Glia-α9 epitope was highly presented in modern varieties, whereas Glia-α20 epitope was lower presented as compared to landraces. Their findings strongly raised the probability of high properties that have driven by modern breeding practices on increased exposure to CD epitopes. Though some of the modern varieties and landraces have showed relatively low concentrations of both epitopes, they emphasized this observation as a new start to breeding for “low CD toxic “wheats” as a new breeding trait that by consumption of such varieties can hopefully help decrease the CD incidence. Generating a non-CD-stimulatory wheat variety can be used in a differential study to test T-cell from large number of patients to confirm the safety levels for other new-released wheat varieties. Recently, application of second generation and high throughput sequencing technologies of DNA, RNA and proteomics of gluten proteins can help quantitatively assess for the presence of all CD/T-cell stimulatory epitopes in wheat varieties (Van den Broeck et al. 2010).

Do the Plant genomic factors affect immune-reactivity of CD?

As proved by Tye-Din et al. (2010) that gluten specific polyclonal T-cells were responsive for the same set of gluten peptides after feeding patients the same cereal, where wheat T-cells were different from barley and rye T-cell. Hypothetically, T-cell response to large number of gluten peptides but when response is evolve, it focuses only on the most immune-dominant sequences, which indicates that the type of cereal and the CD status are defining the T-cell activity responsive toward different gluten peptides in CD patients.

Mainly polyclonal gliadins/T-cell lines enabled to state that γ-gliadins responsive T-cell is more heterogeneous (large and various derived peptides) than α-gliadins responsive T-cell, which is in part, indicate less focus behavior and more genetic diverse of γ-gliadins T-cells (Camarca et al. 2009). Recently in 2012, a research paper was investigating a very interesting correlation between genome origin of wheat, transcript frequency and flanking protein variation and its effects on the immune-reactivity of T-cell epitopes produced by γ-gliadins.

This research paper is achieved by Salentijn et al. (2012) and strongly evident that other epitopes produced by γ-gliadins, as part of the gluten proteins, are playing a vital role in CD development. They studied the natural variation of CD epitopes (γ-gliadins) of bread wheat using two data sets. The first data- set includes 69 genomic γ-gliadins clones from diploid wheat species (T. monococcum, Ae. speltoides and Ae. tauschii) carrying the genomes (Ab, S and D) that act as ancestral homologes of bread wheat genomes (A, B and D). The second data-set includes transcripts of bread wheat γ-gliadins derived from NCBI gene bank (Fig.2; Salentijn et al. 2012).

Salentijn et al. (2012)

Comparative sequence analysis of these transcripts assigned Gli-A1, B1, and D1 loci on the respective homologous genomes 1A, 1B and 1D. About 717 bread wheat γ-gliadins transcripts encodes for 26 different γ-gliadins protein isoforms of which 10 full-length γ-gliadins. In this respect, they found that 50% of bread wheat γ-gliadins transcripts were grouped with Ae. tauschii sequences (D genome, 354 transcripts) at the Gli-D1 locus which indicate that the majority of the γ-gliadins are expressed from the D genome. The reminder of the transcripts were equally assigned to Gli-D1 (178 transcripts) and Gli-B1 (185 transcripts). Interestingly, α-gliadins are highly expressed similarly from D genome (Glia-D2) (Salentijn et al. 2009).

On the other hand, they showed that γ-gliadins of bread wheat have in average 4-10 potential CD epitopes in the first variable domain. For example, 9% of γ-gliadins of bread wheat are fitting to the 26-mer γ-gliadins peptide that harbor 4 distinct CD epitopes that only exist on D genome. Accumulatively, the number of immunogenic peptides form γ-gliadins across all genomes exceeds the number of those from α-gliadins. This later finding may support the findings of Camarca et al. (2009); that T-cells respond to limited number of α-gliadins, while widely respond to large number of different γ-gliadins which strongly suggest its collective effect on CD rather than numerically effect by α-gliadins.

The 9-mer sequence of DQ2- γ-I epitope (PQQSFPQQQ) is most recognized by CD patients. The current study observed high genetic variation (about 6 variants) at the C-terminal flanking region that determines the deamidation pattern of epitope core in diploid and hexaploid wheat. A deeper study for these 6 variants to test its capacity to trigger specific T-cell showed that transglutaminase 2 (tTG2) deamidation of glutamine (Q) at the position 9 is essential for T-cell stimulation most probably due to the essential negative charge for HLA-DQ2 binding. In contrast to the position 5 if substituted with serine (S) instead of phenylalanine (F), the T-cell lose stimulation similarly, if tryptophan (W) was positioned at residue 2 which inhibits the T-cell response.

Moreover the presence of the positively charged residue arginine (R) at position 1 is diminishing the T-cell proliferation. In conclusion, substitutions within and outside the epitope core can influence the T-cell stimulatory properties of DQ2-γ-I epitope. As proved, D genome is major contributor for the CD toxicity, however, gluten derived from tetraploids (lacking D genome) wheat is not tolerated by CD patients. This indicates that by eliminating the D genome is not going to be sufficient to generate safe wheat. Though some evidences were proved for the A genome to acquire other isotypes of IFN- γ-gliadin, its ability to cause CD is not yet proved. Such inferences may solidify the proposal of producing new reduced-CD wheat varieties, even if not for consumption by patients it may help to prevent the onset of CD in potential genetically susceptible individuals. Again, large screening for differences in CD toxicity and epitope regions at RNA level within a large number of wheat varieties (especially tetraploids) could be achieved by high throughput and second generation sequencing technologies that may enable for accurate and fast method rather than wild relative introgression from diploid to hexaploid wheat.

What are the effects if gliadins were suppressed?

RNA interfering based on hairpin RNA (hpRNA) vector can down-regulate the expression of multi gene families of metabolic pathways in genetically engineered plants. An article in 2010 used this technology to down-regulate the γ-gliadins in bread wheat. A study conducted by Gil-Humanes et al. (2010), who prepared a set of vectors containing 361 bp of highly conserved fragment of α, ω and γ-gliadins that can be transcribed to interfere with the RNAs of the various gliadins in the transgenic plant. The conserved IR-fragment was controlled by endosperm specific promoter from γ-gliadin or D hordein (maize). Transgenic plants with both α/ω hpRNA vectors reduced gliadins as 85%, while those transformed by γ-promoter had reduction as 82%, whereas lines transformed by hordein promoter reduced expression by 74%.

These findings indicate that both promoters suppression were highly efficient to down-regulate gliadin genes and inefficient to elicit T-cell response. Although the percentage of homology between ω/α IR-fragments and γ-gliadins group were 69%, the ω/α construct showed higher suppression levels than that of γ-gliadins fragment, besides, only the γ-gliadins was the one which is suppressed while the ω or α- gliadins was not. This implants additional factors rather than the homologies between trigger and target genes that are affecting the gene silencing. Since high transcribed genes are better silenced than lower ones, then the mount of target mRNA may be one of these factors.

Another factor could be the spreading from the target region. Traces of ω-gliadins in the suppressed γ-gliadins transgenic lines showed strong T-cell response may referred to the native sequence of DQ2 γ-VII epitope (9-mer core region: QQPQQPFPQ). Intentionally, the authors gave some focus on the baking quality that been remained almost the same as the wild lines, as shown by the SDS-sedimentation test (Table 3; Gil-Humanes et al. 2010). Because this test is highly correlated with the bread-making quality, transgenic lines containing ω/α constructs acquired values comparable with wild types. While 5 lines showed lower values than wild types, they showed comparable values with medium-quality bread wheat. Additionally, CD is also faintly triggered by the glutenin proteins, this makes its share in the transgenic lines is very low to stimulate T-cell.

Encouraging study proposed that daily gluten intake between 10-100 mg would be safe for most CD patients, suggesting that these transgenic lines could be used as a safe food tolerated by many CD patients. If not, it provides a tool to successfully targeting CD-related gene families and reduces expression of CD-related T-cell epitopes.

therapy and solutions

Among the therapies discussed in the three main research papers argued in this report, we can summarize the different approaches as: one approach tries to degrade the immunogenic peptides by managed enzymes in the stomach and the intestinal lumen of celiac patients. Another approach tries to prevent the trans-epithelial pathway by reducing its permeability. A third one tries to reduce the adaptive immune response. Additional studies to induce tolerance to gluten in patients that have shown CD in the early childhood. One option to eliminate these immunogenic proteins will be to silence/mutagenize wheat cysteine protease inhibitor (cathepsin B), which would allow advanced accumulation of cysteine proteases in developing wheat grains and thereby detoxify immunogenic prolamins. Another way of reducing the immunogenicity of gluten peptides is to use a mixture of selected sourdough lactobacilli and fungal proteases that are able to proteolize the proline-glutamine-rich peptides, decreasing the immunotoxicity of wheat flour (Gobbetti et al. 2007). The kinetics of the hydrolysis of the 33-mer by lactobacilli were highly efficient where the bread-making properties of the pretreated wheat flour can be improved by combining it with the proper amount of oat, buckwheat, and millet flours, which will bake into breads with normal texture.

Another strategy used to degrade the ingested peptides of the flour or wheat derived products that trigger the inflammatory response before they reach the lamina propria. One effective method is the use of prolyl endopeptidase (PEP), enzymes that are able to cleave the immunodominant proline-rich (Marti et al. 2005) regions and are expressed in various microorganisms such as F. meningosepticum (FM), Sphingomonas capsulate (SC), Myxococcus xanthus (MX), and Aspergillus niger (AN) (Mitea et al. 2008b). Therefore using enzymes with a broader activity spectrum or a combined enzyme therapy could be helpful. Cysteine endoprotease B2, a glutamine specific protease of germinating barley, has been tested together with different endopeptidases as combination enzyme therapy for celiac sprue. The enzyme was analyzed under the proteolytic conditions of the gastrointestinal tract, which are given by pepsin in the stomach, and trypsin and chymotrypsin in the small intestine. The recombinant EP-B2 remains active during the gastric phase of gluten digestion but rapidly degraded in the duodenum (Bethune et al. 2006).

In combination with PEP from S. capsulate, it efficiently breaks down whole wheat gluten in vitro as well as in a rat model in vivo, thus largely repealing its immunogenic potential, as assessed with several gluten-specific T cell lines. Both enzymes were active and stable at acid pH and can therefore be administered as lyophilized powders or simple capsules or tablets (Gass et al. 2007). Oral enzyme therapy will probably not be sufficiently degrade immunogenic epitopes of normal daily gluten ingestion (13 g), but rather will eliminate the detrimental effect of a few hundred milligrams to a few grams of gluten in patients with high gluten sensitivity or stubborn CD type 1. Nutraceutical approach to express barley EP-B2 and FM-PEP in large quantities in wheat endosperm to detoxify immunogenic gluten proteins seems promising approach that is based primarily on the molecular and cell biological experiences of synthesis, storage, and use of a heat-stable (1,3-1,4)-β- glucanase in transgenic barley grains (Von Wettstein, 2007).

Investigations of the cell biology of wheat storage protein biosynthesis and mobilization indicate that the concepts obtained in barley apply equally to wheat. Thus expectedly, small proportion of the transgenic grains containing large amount of “glutenases” will be good enough to detoxify the consumed gluten, and will work as a perfect therapeutic alternative for CD.

Intraluminal therapies were concluded to reduce gluten immunogenicity by preventing the uptake of immunogenic gluten fragments across the intestinal epithelium into the lamina propria by sequestering or degrading the gluten peptides. Early research was to develop wheat variants with low or null immunogenicity to achieve a decrease in the number of T cell epitopes, without having any influence on their baking properties. Complete deletion of α-gliadin locus on 6DS not only decrease the total immunogenic epitopes but also seriously affected the baking properties, whereas deletion of γ-gliadins, ω-gliadins, and low molecular weight glutenin subunits on 1DS lowered the immunogenecity without any adverse influence on baking properties (Van den Broeck et al. 2010). This means that deletion of gliadins is one of the strategies to develop safe wheat lines. However, it needs to be compensated with safe sources of gliadins genes like oats/maize through breeding or genetic modification in such a way that the ratio between gliadins and glutenin subunits remains the same to maintain baking quality (Osorio et al. 2012).

A challenging strategy was to identify a product’s suitability for a group of celiac patients where research was targeting different gliadins groups using RNA interference (RNAi). Initially, by focusing on targeting specific groups of gliadins using hairpin RNAs designed from the conserved regions identified for these gene families. The results were silencing of α-gliadins (up to 63 %) in the wheat cultivar Florida (Becker et al. 2006), silencing for γ-gliadins (up to 80 %) in wheat cultivar Bobwhite (Gil-Humanes et al. 2008), selective silencing of α-gliadins as DQ8-α-Glia-(206-217) epitope encoded by the loci on 6B and 6D in the wheat cultivar Cadenza leaving the other α-gliadins unaltered (Van Herpen et al. 2008). The drawback in these silencing strategies is that the partial down-regulation (from 33 to 80 %) may not be sufficient to claim these lines nontoxic for celiac patients.

Gil-Humanes et al. (2010) used chimeric hairpin RNA capable of silencing both α- and ω- gliadins in the wheat cultivar Bobwhite which enabled a chimeric hairpin construct capable of silencing α-, γ-, and ω- gliadins with low molecular weight glutenin subunits (LMWgs) to be controlled by an endosperm specific promoter with a hope to obtain lines showing significant amount of reductions in the content of all immunogenic prolamins. Recent studies were derived to epigenetically eliminate gliadins and LMWgs by silencing of wheat DEMETER homoeologues (encoding for 5-methyl cytosine DNA glycosylase/ lyase) responsible for transcriptional activation of gliadins and LMWgs in wheat grains by RNAi. Interestingly, promoters of HMWgs required for baking are protected from methylation by the presence of CpG islands (Rustgi et al. 2010), and their accumulation solely depends on tissue-specific transcription factors.

A different approach to reduce the immunogenicity of gluten peptides is to use the specificity of epitopes for tTG2. By this, gliadins were incubated with tTG2 and lysine methyl ester, which result in cross-linking of the tTG2 target sequence in gliadins with the terminal amino group of the lysine methyl ester (Gianfrani et al. 2007). So that, lysine-modified gliadins lose their affinity to bind with HLA-DQ2, which in turn stops IFN-γ production by intestinal T cell lines derived from HLA-DQ2-positive celiac patients. Similarly, when the whole wheat flour treated with low molecular weight transglutaminase derived from Streptomyces moboraensis.

The treatment does not have any adverse influence on the baking properties (Yokoyama et al. 2004). Sequestration of gluten residues can also be used as a therapy by using a resin that acts as a gluten-sequestering polymer. The use of a copolymer of polyhydroxy methacrylate and polystyrene sulfonate to bind gliadin in a specific manner was also successful and able to prevent in vitro gliadin-induced epithelial toxicity and intestinal barrier dysfunction in HCD4/DQ8 mice (Pinier et al. 2009).

How to solve the problem

Effect of gluten