Food Chemistry 448 (2024) 139148
2
are three to nine amino acids long and rich in proline and glutamine
(Shewry, 2019). The high repetition of glutamine and proline content in
gluten proteins makes them partially resistant to proteolytic cleavage by
human digestive enzymes. The resulting undigested gluten epitopes can
cross the small intestinal barrier and trigger immune reactions. The
deamidation of these peptides by transglutaminase 2 (TG2) in the gut
can enhance the binding afnity of these peptides to the antigen-
presenting heterodimers HLA-DQ2 or HLA-DQ8 and activate the
CD4+T cells (Brouns et al., 2019). This activation leads to the pro-
duction of anti-gluten and anti-TG2 antibodies and cytokines such as IL-
15 to damage the intestinal epithelial cells, leading to celiac enteropa-
thy. In addition, plant defence-related proteins such as alpha-amylase
trypsin inhibitors (ATIs), wheat germ-agglutinins, and serpins present
in wheat can also trigger immune responses in some people (Aziz,
Hadjivassiliou, & Sanders, 2015; Zevallos et al., 2017). Non-celiac wheat
sensitivity (NCWS) is increasingly reported by patients with intestinal
and extra-intestinal responses. Patients diagnosed with NCWS test
negative for CD and WA, but the key cause for this disease symptoms is
currently unknown (Brouns et al., 2019), and the role of gluten proteins
in NCWS pathology remains elusive. Although wheat consumption can
trigger adverse immune reactions in a small part of the population, most
of the human population still depend on wheat as their primary diet. As
such, it is important to understand the immunostimulatory peptide
contents precisely so that the majority of people can meet their regular
dietary requirements.
Studies have reported quantitative and qualitative differences in
epitope content and composition between the genotypes of different
Triticum species (Ribeiro et al., 2016) and T. aestivum varieties (Denery-
Papini et al., 2007; Prandi, Tedeschi, Folloni, Galaverna, & Sforza, 2017;
Ronga et al., 2020; Schalk, Lang, Wieser, Koehler, & Scherf, 2017). The
environmental variability of epitope-containing proteins has been
investigated (Juh´
asz et al., 2018; Juh´
asz, Haraszi, & B´
ek´
es, 2020;
Landol et al., 2021; Ronga et al., 2020) and more broadly, the inuence
of environment on protein composition and content (and thus impacting
epitopes) has been extensively studied (Johansson et al., 2020). In
addition, several studies have shown the immunostimulatory potential
of wheat products due to the presence of IgE-binding WA epitopes and
immunogenic celiac epitopes (Denery-Papini et al., 2007; Schalk et al.,
2017). In addition to the availability of >16 high-resolution bread wheat
genome sequence resources, advancements in mass-spectrometry (MS)-
based proteomics and bioinformatics workows have allowed us to
understand better the diversity of immunostimulatory peptides involved
in wheat-related disorders. The MS-based proteomics approach can
precisely identify and quantify immunogenic epitopes from complex
our samples and processed food products (Ribeiro et al., 2021). The
identication and quantitation of epitopes in the experimental samples
depend on the precise characterization of sub-classes of proteins upon
digestion (Landol et al., 2021; Ogilvie et al., 2020). For instance,
α
-gliadins generate the most diverse CD-related epitopes (Landol et al.,
2021), and these peptides were reported among the most immunodo-
minant (Juh´
asz et al., 2018; Tye-Din et al., 2010). However, the γ-gli-
adins have been shown to generate the largest number of CD epitopes
during digestion (Ogilvie et al., 2020). Two studies monitored the
release of immunogenic peptides during bread (Ogilvie et al., 2020) or
pasta (Mamone et al., 2015) digestion and showed that a few immu-
nogenic peptides survived the in vitro digestion, i.e., they resisted to
proteolysis.
The duration and intensity of protein hydrolysis within the gastro-
intestinal tract can increase the generation of shorter peptides, resulting
in a lower immunostimulatory effect in the gut. Protein digestibility can
be inuenced by several factors, including human gastro-intestinal
conditions (Torcello-G´
omez et al., 2020), the complexity of the food
matrix (Freitas, G´
omez-Mascaraque, & Brodkorb, 2022), composition
(Wu, Taylor, Nebl, Ng, & Bennett, 2017) and the food processing steps
such as baking. Heat treatments applied during baking or cooking can
alter the protein structure by unfolding the proteins through subsequent
re-arrangements of disulde bonds (Ogilvie et al., 2021; Pasini, Simo-
nato, Giannattasio, Peruffo, & Curioni, 2001). Contradictory results
have been published on the effect of baking on the protein digestibility
of wheat. According to Bredariol, Carvalho, and Vanin (2020), there are
optimum time and temperature baking parameters that would improve
proteolysis during digestion. In addition, the impact of protein di-
gestibility varied considerably for the bread crumb and crust; the crumb
portion showed similar digestibility to the uncooked dough, while the
bread crust had lower digestibility (Pasini et al., 2001). The thermal
treatment also inuences the recognition of WA epitopes by the IgE
receptors (Lupi et al., 2019). This study reported that the thermal
treatment initially reduces gliadin recognition by IgE due to the pro-
duction of large aggregates. However, upon hydrolysis under acidic
stomach conditions, the epitopes can be unmasked and recognized by
the T cells in the intestine. From our to processed and baked products,
epitopes can undergo modications that would alter their immunosti-
mulatory potential (Liu et al., 2023). Thus, to study the diversity of
peptides released from wheat-based food during human digestion, the
choices of the food matrix as well as the digestion model are essential. In
addition to wheat our’s baking or cooking process, the protein di-
gestibility varies across wheat genotypes. Notably, the protein compo-
sitional variability in the grain our and, thus, breads between
genotypes have shown a different spectrum of digestibility (Lavoignat
et al., 2022). Although many studies have reported the impact of the
baking process and genotype-dependent variations that inuence the
digestibility and release of immunogenic epitopes, the combination of
genotype and environmental variations and their food processing on the
human food digestibility assessment remains unknown.
The present study aims to investigate how wheat genotype-
dependent variability can impact the bread protein digestibility and
epitope release during the human digestion process. We hypothesized
that peptides with a specic signature could discriminate between high-
and low-digestible genotypes. First, we performed the in vitro digestion
assay on the bread samples and collected and measured the peptides
generated during the gastric and intestinal digestion process. The
measured peptide abundances were associated with the our composi-
tion and protein content to separate the genotype with high digestibility
(HD) and low digestibility (LD). Moreover, we detected and compre-
hensively mapped the epitopes identied during the digestion process to
the different sub-types of gliadins, glutenins and albumins-globulins
using an in-silico approach. Together, this study establishes the pepti-
domic map of partially digested bread samples and the ngerprint of
epitope diversity from 16 wheat genotypes grown at two different
environmental conditions.
2. Materials and methods
The plant material and wholemeal our phenotyping were previ-
ously described in (Lavoignat et al., 2023).
2.1. Meteorological data
In total, 16 bread wheat genotypes were grown in the eld in two
different locations: Clermont-Ferrand (CF) and Estr´
ees-Mons (EM) in
duplicates. To assess the weather pattern for CF and EM, the maximal
(Tmax), minimal (Tmin) and mean (Tmean) temperatures (◦C) and cu-
mulative rainfall (CumR, mm) of the 2016–2017 growing season and the
last 30 years (Tmax30, Tmin30, Tmean30, CumR30) were retrieved
(https://agroclim.inrae.fr/). The cumulative growing degree days
(CumDD_MJJ) and the cumulative rainfall (CumR_MJJ) over the grain
lling period, i.e., May to July, were calculated for two locations based
on the heading date of genotype.
2.2. R5 competitive ELISA assay on wholemeal our
For the 64 grains samples (16 genotypes ×2 locations ×2
M. Lavoignat et al.