**3. Wheat screening through** *in vitro* **digestion and LC‐MS analysis**

The *in vitro* digestion of the prolamin extract described in Section 2 was performed on a wide spectrum of wheat samples, to observe eventual differences in the production of immunogenic and toxic sequences after gastrointestinal digestion. Briefly, the prolamin fraction was extracted with 70% ethanol and submitted to simulated *in vitro* digestion using the three main proteases of the gastrointestinal tract: pepsin in the gastric phase (3 h) and chymotrypsin/trypsin in the intes‐ tinal phase (4 h). All the enzymes were used at their optimal temperature (37°C) and pH (respec‐ tively 2 and 7.2), in an enzyme:substrate ratio of 1:100. The quantification of the gluten‐derived peptides was achieved using reverse‐phase ultra‐high‐performance liquid chromatography (UPLC) coupled with single quadrupole mass spectrometer (SQD), using the isotopically labelled internal standard method (the peptide LQLQPF(*d*<sup>5</sup> )PQPQLPY, one of the most abundant, was used as standard). In this way, it has been possible to evaluate the influence of the cultivation area (i.e. the soil—climatic conditions) and of the genotype of the wheat (genetic influence).

#### **3.1. Influence of the cultivation region**

To investigate the role of soil and climatic conditions on the total amount of toxic and immu‐ nogenic sequences, durum wheat samples harvested in three different Italian regions were submitted to prolamin extraction and *in vitro* digestion. The three harvesting area were cho‐ sen to maximize the soil and climatic differences. Argelato is located in Northern Italy, in the Po plain with a temperate sub‐oceanic climate: mean annual temperature is 11.0–13.0°C (months with mean temperature below 0°C: January) and precipitations are 690–1200 mm (May and October are the rainy months, whereas July and August are the driest months). Falconara is located in the hills of the Central Italy, with Mediterranean sub‐oceanic climate: mean annual temperature is 12.5–16.0°C (no months with mean temperature below 0°C) and precipitations are 700–1000 mm (November is the rainy month, whereas July and August are the driest). Lucera is located in the Capitanata area, with Mediterranean subtropical climate: mean annual temperature is 12.0–17.0°C (no months with mean temperature below 0°C) and precipitations are 400–800 mm (October and November are the rainiest month, May to September are the driest).

The total amount off peptides containing immunogenic and toxic sequences is reported in **Figure 1**, mediated for each harvesting area. Statistically significant differences were deter‐ mined with analysis of variance (two ways ANOVA), with *p* < 0.05. Immunogenic and toxic peptides show the same trend: there were no statistically significant differences among the three different locations, with the exception of Argelato (BO), that show a slightly higher content of peptides containing sequences involved in celiac disease. Thus, the soil and climate conditions do not have a determining role on the amount of immunogenic and toxic gluten sequences. The high intra‐region variability, indeed, suggests that there are other factors that are playing an important role.

#### **3.2. Influence of the genotype**

The accurate molecular characterization of the *in vitro* digested prolamin mixtures is an interesting tool for the screening of different wheat lines aimed to identify those produc‐ ing a smaller amount of pathogenic peptides [31]. Genetic selection operated by breeders, to achieve the desired rheological properties, has led to a decrease in genetic biodiversity of wheat varieties nowadays present on the market. Thus, 25 accessions from a durum wheat panel were chosen to maximize genetic biodiversity of the samples: the prolamin fraction was extracted with 70% ethanol and submitted to *in vitro* digestion. From a molecular point of view, the results obtained confirm what has been previously assessed using genetic and immunologic approaches, that there is a strong influence of the genotype in the final amount of peptides containing sequences involved in celiac disease.

and toxic sequences after gastrointestinal digestion. Briefly, the prolamin fraction was extracted with 70% ethanol and submitted to simulated *in vitro* digestion using the three main proteases of the gastrointestinal tract: pepsin in the gastric phase (3 h) and chymotrypsin/trypsin in the intes‐ tinal phase (4 h). All the enzymes were used at their optimal temperature (37°C) and pH (respec‐ tively 2 and 7.2), in an enzyme:substrate ratio of 1:100. The quantification of the gluten‐derived peptides was achieved using reverse‐phase ultra‐high‐performance liquid chromatography (UPLC) coupled with single quadrupole mass spectrometer (SQD), using the isotopically labelled

used as standard). In this way, it has been possible to evaluate the influence of the cultivation area

To investigate the role of soil and climatic conditions on the total amount of toxic and immu‐ nogenic sequences, durum wheat samples harvested in three different Italian regions were submitted to prolamin extraction and *in vitro* digestion. The three harvesting area were cho‐ sen to maximize the soil and climatic differences. Argelato is located in Northern Italy, in the Po plain with a temperate sub‐oceanic climate: mean annual temperature is 11.0–13.0°C (months with mean temperature below 0°C: January) and precipitations are 690–1200 mm (May and October are the rainy months, whereas July and August are the driest months). Falconara is located in the hills of the Central Italy, with Mediterranean sub‐oceanic climate: mean annual temperature is 12.5–16.0°C (no months with mean temperature below 0°C) and precipitations are 700–1000 mm (November is the rainy month, whereas July and August are the driest). Lucera is located in the Capitanata area, with Mediterranean subtropical climate: mean annual temperature is 12.0–17.0°C (no months with mean temperature below 0°C) and precipitations are 400–800 mm (October and November are the rainiest month, May to

The total amount off peptides containing immunogenic and toxic sequences is reported in **Figure 1**, mediated for each harvesting area. Statistically significant differences were deter‐ mined with analysis of variance (two ways ANOVA), with *p* < 0.05. Immunogenic and toxic peptides show the same trend: there were no statistically significant differences among the three different locations, with the exception of Argelato (BO), that show a slightly higher content of peptides containing sequences involved in celiac disease. Thus, the soil and climate conditions do not have a determining role on the amount of immunogenic and toxic gluten sequences. The high intra‐region variability, indeed, suggests that there are other factors that

The accurate molecular characterization of the *in vitro* digested prolamin mixtures is an interesting tool for the screening of different wheat lines aimed to identify those produc‐ ing a smaller amount of pathogenic peptides [31]. Genetic selection operated by breeders, to achieve the desired rheological properties, has led to a decrease in genetic biodiversity of wheat varieties nowadays present on the market. Thus, 25 accessions from a durum wheat

(i.e. the soil—climatic conditions) and of the genotype of the wheat (genetic influence).

)PQPQLPY, one of the most abundant, was

internal standard method (the peptide LQLQPF(*d*<sup>5</sup>

**3.1. Influence of the cultivation region**

316 Wheat Improvement, Management and Utilization

September are the driest).

are playing an important role.

**3.2. Influence of the genotype**

**Figure 1.** Total average content of immunogenic (left panel) and toxic peptides (right panel) of wheat samples harvested in three different regions (Argelato, Falconara and Lucera). Total amount of peptides is expressed in μg of peptide for gram of sample. Bars with different letters mean statistically different samples. Adapted with permission from Ref. [31].

As shown in **Figures 2** and **3**, there are great differences among the different samples. More specifically, the peptides that are more affected by genetic features are those eliciting the adaptive immune system (immunogenic peptides). This relies on the fact that toxic peptides derive from the N‐term region of gliadins, which is much more conserved than the region that originates immunogenic peptides. In the latter case, the difference is surprisingly high: there is a 6‐fold difference between the highest and the lowest scoring sample (600 μg/g vs 100 μg/g). These data confirm the huge variability in gluten‐coding genes, since also among accessions of the same genetic group, there are noticeable differences, for example, in the first group. Recent studies demonstrated that number of subjects that lost the immunological tolerance to gluten in their adulthood is increasing and among the possible causes there is also the amount and the quality of ingested gluten [15]. This means that the use of less immunogenic wheat varieties (especially in the preparation of baby foods) can reduce the exposure to gluten, pos‐ sibly decreasing the incidence of the disease. And, moreover, it would be possible to operate a varietal selection aimed to have the same gluten content (thus comparable rheological proper‐ ties), but expressing different gliadin isoforms, with a reduced content of immunogenic and toxic peptides, to reduce the exposure of genetically predisposed subjects, and possibly to reduce the risk of celiac disease development. These data take in consideration the molecular point of view, so it would be really interesting to cross the data with immunological tests (such as T cell proliferation assays or K562 cells agglutination) on the samples to verify the quality of the correlation between pathogenic peptides content and immune response.

**Figure 2.** Total content of toxic peptides (expressed in μg of peptide for gram of sample) in 25 samples from a Durum Panel collection (upper panel). Samples were grouped on the bases of phylogenetic affinity on dendrogram (lower panel). Adapted with permission from Ref. [31].

**Figure 3.** Total content of toxic peptides (expressed in μg of peptide for gram of sample) in 25 samples from a Durum Panel collection (upper panel). Samples were grouped on the bases of phylogenetic affinity on dendrogram (lower panel). \*\* Statistically different group. Adapted with permission from Ref. [31].
