Gluten and Celiac Disease

Literature-Based Guided Assessment (LGA)

Gluten and Celiac Disease

Introduction

You can't enter a restaurant or food store and not see foods labeled "gluten-free".  Some people have autoimmune reactions to dietary gluten, a grain protein. They can suffer from celiac disease, an autoimmune disorder, that affects a little less than 1% of the US population.  A Google Trends analysis shows that searches for celiac disease have remained generally level over the last 20 years while searches for gluten have soared 10-fold as illustrated in Figure \(\PageIndex{1}\) below. The data is scaled (normalized) to 100 for comparison purposes.

Figure \(\PageIndex{1}\):

Has the incidence of diagnosable celiac disease increased over time?  Or are there other factors that contribute to an increased interest in gluten?  Consumption of whole grains is associated with positive health effects, yet some experience negative ones.  These include Non-celiac gluten sensitivity, also referred to as gluten-sensitive enteropathy (GSE) or gluten intolerance (which has no easy diagnostic test), and a general wheat allergy, characterized by an IgE response (much as with ragweed). Many food additives like fillers and sauces also contain gluten.

This LGA will explore the biochemistry of the proteins that comprise wheat gluten.  We will see how some gluten protein fragments can trigger a T cell immune response in individuals with Celiac disease.

Gluten, a complex protein structure is a major component of grains.  We will focus our discussion on wheat grain, whose structure is shown in Figure \(\PageIndex{2}\):

Figure \(\PageIndex{2}\): Wheat grain structure.  Onipe, O.O., Jideani, A.I.O. and Beswa, D. (2015), Composition and functionality of wheat bran and its application in some cereal food products. Int J Food Sci Technol, 50: 2509-2518. https://doi.org/10.1111/ijfs.12935.  Creative Commons Attribution-NonCommercial License

Gluten is needed for bread, noodles, pasta, and baked goods.  It causes dough to be strong and elastic.  It is what is left when dough is washed with lots of water which removes starch and soluble components.  Doughs made from some grains don't have this proteinaceous substance. These include oats, quinoa, brown rice, corn, and buckwheat. 

Almost 100 years ago, Osborne studied proteins in seeds and came up with four main types of proteins based on the solvent required to fractionally extract them from the seeds:

• albumins (soluble in water)
• globulins (soluble in salt water)
• prolamins (soluble in 70% ethanol)
• glutelins (insoluble except in alkalai)

The names albumins and globulins are still used (consider blood proteins).  The name prolamins was derived from the fact that these proteins had lots of proline and amide groups, which we now know derive from a high abundance of glutamine. Gluten consists mostly of a class of prolamin proteins called gliadin (horde in barely and zein in corn) and another class of proteins, glutelins. Gluten is a source of amino acids for seed germination and development. They are both high in proline and glutamine.

As fairly insoluble proteins, the synthesis of the completed gluten in seeds must be complicated.  Individual proteins aggregate through noncovalent and covalent (disulfide bonds) to form the gluten matrix which encapsulates starch granules within the endosperm.  This matrix is visible in a microscope when the starch is removed by enzymatic degradation. These structures are illustrated in Figure \(\PageIndex{3}\) below.

Figure \(\PageIndex{3}\):  The origin of wheat gluten.  Shewry Peter, Frontiers in Nutrition, 6, 2019.  https://www.frontiersin.org/articles...nut.2019.00101.  Creative Commons Attribution License (CC BY)
      
Panel (A): Transmission electron microscopy of starchy endosperm cells at a late stage of grain development (46 days after anthesis)
shows that the individual protein bodies have fused to form a continuous proteinaceous matrix. Taken from Shewry et al. (9) with permission, provided by Dr. M.
Parker (IFR, Norwich, UK).

Panel (B): Digestion of a flour particle to remove starch reveals a continuous proteinaceous network. Taken from Amend and Beauvais (10) with
permission.

Panel (C): Transverse section of the lobe region of a developing wheat grain stained with Toluidine Blue to show the tissue structure and deposited protein (in
blue). The gluten is abundant in the subaleurone layer.  

When dough is kneaded, separate gluten matrixes contact others to form an extensive gluten mesh.

Figure \(\PageIndex{4}\)  below shows an optical reconstruction of gluten in whole noodles.

Figure \(\PageIndex{4}\):  3D reconstruction of gluten in whole noodle.  Ogawa, T., Matsumura, Y. Revealing 3D structure of gluten in wheat dough by optical clearing imaging. Nat Commun 12, 1708 (2021). https://doi.org/10.1038/s41467-021-22019-0.  Creative Commons Attribution 4.0 International License.  http://creativecommons.org/licenses/by/4.0/.

Panel a: Experimental setup for 2PEM imaging.

Panel b: Comparison of deep imaging of noodles treated with water (left) and SoROCS (right).

Panel c: 3D reconstruction of gluten for noodle treated with water (left) and SoROCS (right). Note that when rendering with FluoRender, gamma change (3.2) and alpha blending (500) were applied and a threshold was set to 146 and 90 for water and SoROCS, respectively (b, c).

Panel d-h: The fluorescent image quality of gluten depending on the depth from the surface of the noodles. Data are presented as mean values ± SD (N = 3) independently prepared noodle samples). One of the section images obtained from these three different samples is shown in (e–h), respectively. Note that the 2D imaging conditions with 2PEM were adjusted to be optimal for each depth (d–h). Optical sections at several depths: z = 550 μm (e), z = 1050 μm (f), z = 1550 μm (g) and z = 2050 μm (h). Scale bars indicate 100 μm (e–h). Source data underlying (d) are provided as a Source Data file.

 

Gluten Molecular Structure

Gluten contains two different types of proteins, glutenins and gliadins. 

• The glutenins are generally higher in molecular weight than gliadins, and are responsible for the elastic properties of wheat dough.  The protein has two repetitive motifs:  GQQPGQ and GQQPGQGQQGYYPTS.
• The gliadins are smallerl in molecular weight.  They likewise have repetitive motifs.  An internal 33 mer (amino acid 57-89 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF in the mature protein but 76-106 in AlphaFold Structure) is key in promoting inflammation in Celiac Sprue disease. 

The AlphaFold predicted structures for two examples of glutenins and one of gliadin are shown in Figure \(\PageIndex{5}\) below.

Glutenin - P10388, high molecular weight subunit DX5, 90.2 Kd	Glutenins - P10386, low molecular weight subunit 1D1, 34.9 Kd	Gliadin - P18573, Alpha/beta-gliadin MM1.  35.4 Kd

Figure \(\PageIndex{5}\): AlphaFold predicted structures of wheat glutenin and gliadin monomers, two components of gluten

 

Exercise 1

What is a predominant structural characteristic of these proteins?  How might that influence the formation of gluten? 

Answer

The proteins, especially high molecular weight glutenin, are very disordered and have extensive flexibility.  The proteins could easily flex and form complex intra- and intermolecular interactions on the formation of the gluten matrix.

 

Note that the low molecular weight glutenin is more structural homologous to gliadins than the high MW glutenin.

Figure \(\PageIndex{6}\) shows an interactive iCn3D AlphaFold model of the monomeric wheat gliadin (P18573).

Figure \(\PageIndex{6}\): AlphaFold model of the monomeric wheat gliadin (P18573). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...5eCAiuikUZok38

The 33 mer peptide (LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF) that is a main contributor to immune system activation in celiac patients (76-106 in the unprocessed AlphaFold Structure) is shown in spacefill.  The 105 glutamine side chains (of the 307 amino acids in the processed protein) are shown in purple stick or purple spacefill (in the 33 mer).  The main central predicted helix (QQQQQQQQQQQQKQQQQQQQQILQQILQQQ) is composed mostly of glutamines (25/30 amino acids)!

 

Exercise 2

Time for a review! What is so special about glutamine?  Name some of its biochemical features that you have studied.  The overall structure of gluten must depend to a significant degree on this amino acid given its abundance.

Answer

For one, it's the most abundant free (not part of a protein) amino acid in humans. 

From a reaction point of view:

• Glutamine can give up its amide nitrogen to form NH₄⁺ on conversion to glutamic acid, a reaction catalyzed by glutaminase.  The resulting glutamic acid can undergo oxidative deamination to form α-ketoglutarate (a TCA intermediate) and free NH₄⁺, a reaction catalyzed by glutamate dehydrogenase.  However, excess NH₃/NH₄⁺ is toxic to cells.  Also, glutamine is a prime energy source for intestinal cells.

From a structural perspective:

• It can form hydrogen bonds through its amide side chain which has hydrogen bond donors and acceptors
• It is found in both alpha helices and beta structure  The Chou-Fasman propensities for Gln for alpha helix and beta structure are 1.17 and 1.23, respectively.

 

 

Table \(\PageIndex{1}\) below shows some key properties of glutenins and gliadins that compose gluten.

Gluten groups (% in gluten overall), MW	Names	AA Composition	Consensus
High Molecular Weight Glutenins (HMW-GS) (9%), 65-90K	HMW subunit of glutenin elasticity of dough form intermolecular S-S polymers Examples  DX5, 90.3K;  DY10, 69.6K	30-35% Q 10-16% P 15-20% G 0.7-1.4% K 0.5-1.5% C	PGQGQQ GYYPTSPorLQQ GQQ
Sulfur-rich Glutenin (24%), 30-45K	Low Molecular Weight Glutenins (LMW-GS), B-Type dough strength similar to gliadins but.. form intermolecular S-S bonds and polymers Example:  1D1, 34.9K	30-40% Q 15-20% P 2-3% C <1% K)	B type:  PPFS/PQQ(QQ)
Sulfur-rich gliadins (56%) extensibility of dough, 30-45K	α-/-β type (α and β) (35%) γ-type (21%) monomeric since .. form intramolecular S-S bonds	30-40% Q 15-20% P 2-3% C <1% K)	α type:  PForPQQQQ(QQ), PFPQ(Q)PQ(Q) γ type: PFPQ(Q), PQQ(PQQ)
Sulfur-poor gliadins (11%) 30-50K	ω type	40-50% Q 20-30% P 8-9% F 0-0.5% K 0 Cys	PRRPFPQQ, QQQFP

Table 1\(\PageIndex{5}\):   adapted from Peter R. Shewry, Peter S. Belton, What do we really understand about wheat gluten structure and functionality?, Journal of Cereal Science, Volume 117, 2024, 103895, ISSN 0733-5210, https://doi.org/10.1016/j.jcs.2024.103895.  Creative Commons BY 4.0 DEED.  Attribution 4.0 International

A simple summary of the properties of gluten proteins is shown in Figure \(\PageIndex{7}\) below.

Figure \(\PageIndex{7}\):  Wieser, H., Koehler, P., & Scherf, K. A. (2023). Chemistry of wheat gluten proteins: Qualitative composition. Cereal Chemistry, 100, 23–35. CC BY DEED

Classification of wheat proteins into different fractions and types. The percentages are typical values for each gluten protein type relative to the total amount of gluten determined by modified Osborne fractionation and reversed-phase high-performance liquid chromatography as reported by Lexhaller et al. GS, glutenin subunits; HMW, high molecular weight; LMW, low molecular weight; m, monomeric; MMW, medium molecular weight; MW, molecular weight; p, polymeric.

We will explore the polymeric structure that arises from covalent interchain disulfide bonds between separate glutenin monomers and how noncovalent intermolecular interactions of glutenin chains in the polymer with monomeric gliadins result in the elasticity of dough.  In addition, we will explore gluten's immunogenic properties that cause celiac disease.

The gluten matrix, which consists of many copies of glutenins cross-linked by sulfide bonds along with interspersed gliadins could probably be considered the largest protein (complex) in nature with effective molecular masses in the 10's of millions.  The largest single-chain protein in humans is the muscle protein titin, with a molecular weight of 3.9 million and around 34,000 amino acids).  Both proteins confer elasticity (to muscle tissue and dough). Both proteins have large numbers of repetitive sequences, which confer elastic properties to the proteins.  This allows a stretching and return to an unextended state. In titin, the domain that confers elasticity contains repeating PEVK motifs(in both heart and skeletal muscle). The repetitive "spring-like" domain in gluten proteins contains large quantities of Q, P, and G residues. The main function of gluten is for amino acid storage for the sprouting seed, not the anthropomorphic function of allowing elastic bread doughs.  The elasticity of gluten is not required for the plant but may arise secondarily from the tight packing of the protein in a relatively small volume matrix (compared to the full seed).    

Exercise 3

Other than for structural consideration in creating a compact matrix, what might be the nutritional value for the germinating seed of having such a high proportion of glutamines in the gluten proteins?

Answer

The glutamines could be deaminated and offer a vital source of ammonia for the growing sprout.  The nitrogen at the end of the glutamine side chain is already "fixed" with no immediate need for nitrogen fixation through root microbes and the nitrogenase complex which fixes atmospheric N₂.

The unique repeats are highly enriched in glutamine (Q, 35-55%), proline (P, 10-25%), and in the HMW glutenin protein glycine (G, 11-12%). Cysteines are found in the nonrepetitive sequences. These amino acids confer on gluten its properties, including insolubility and elasticity. 

First, let's explore the role of disulfide bonds in gluten monomers and overall gluten structure.  Both glutenins and gliadin can form intramolecular disulfide bonds but intermolecular S-S bonds are generally found between glutenins monomers. 

Figure \(\PageIndex{8}\) shows an interactive iCn3D AlphaFold model of wheat low molecular weight glutenin subunit 1D1 (P10386).  This protein forms both intramolecular and intermolecular S-S bonds. 

Figure \(\PageIndex{8}\): AlphaFold model of wheat low molecular weight glutenin subunit 1D1 (P10386). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...N7ocA7ksT5Rac9 

 

Note that two cysteines (C 25 and C 233 in the AlphaFold model, but C2 and C210 in the final mature protein) are free and not engaged in an intramolecular disulfide. These are free to form intermolecular S-S bonds between adjacent glutenins to form the intermolecular polymer.  The amino acids +/- 7 amino acids from each cysteine are shown in Table \(\PageIndex{2}\)

Cysteines	-7AACys#+7AA	-7AACys#+7AA
25 and 233 (no S-S)	IAQMETRC25IPGLERP	QPQQLGQC233VSQPQQQ
C150 S-S C185	ILQQLNPC150KVFLQQC	CHVMQQQC185CQQLPQI
C158 S-S C178	KVFLQQQC158SPVAMPQ	QMLQQSSC178HVMQQQC
C186 S-S C283	HVMQQQCC186QQLPGIP	LRILPTMC283SVNVPLY

Table \(\PageIndex{2}\): Disulfide bonds in wheat low molecular weight glutenin subunit 1D1 (P10386)

 

Exercise 4

Using the Kyte-Doolittle scales for estimates of hydrophobicity of side chains, let's assume that these amino acids are hydrophobic:  A, V, L, I, F, M, and C.  Compare the relative hydrophobicities of the amino acids +/- 7 amino acids on either side of the designated Cys.  From this, offer an explanation that could account for the observation that Cys 25 and 233 engage in intermolecular S-S bonds.

Answer

In general, C 25 and C 233 in the AlphaFold model (C2 and C210 in the mature protein) have fewer hydrophobic amino acids flanking them. The others have more which could allow a collapse with other hydrophobic regions, bringing the central cysteines in each segment into closer proximity needed for disulfide bond formation. During synthesis and aggregation of the chains, C25/C2) might freely form interchain disulfide links with others before the complete folding of the protein monomers. In the AlphaFold predicted model, both are in more disordered regions that would allow the free cysteines to search for other free cysteines on other monomers to form intermolecular S-S links.

See this reference for more details:  Markgren J, Hedenqvist M, Rasheed F, Skepö M, Johansson E. Glutenin and Gliadin, a Piece in the Puzzle of their Structural Properties in the Cell Described through Monte Carlo Simulations. Biomolecules. 2020 Jul 23;10(8):1095. doi: 10.3390/biom10081095. PMID: 32717949; PMCID: PMC7465137.  Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/)

 

Figure \(\PageIndex{9}\) shows a model of how low molecular weight glutenin and α-gliadin could associate to form gluten.

Figure \(\PageIndex{9}\): A proposed model for the synthesis, folding, disulfide bond formation, and storage of α-gliadin and low molecular weight glutenin subunits in the wheat grain.  Markgren J, Hedenqvist M, Rasheed F, Skepö M, Johansson E. Glutenin and Gliadin, a Piece in the Puzzle of their Structural Properties in the Cell Described through Monte Carlo Simulations. Biomolecules. 2020 Jul 23;10(8):1095. doi: 10.3390/biom10081095. PMID: 32717949; PMCID: PMC7465137. Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/)

 

Exercise 5

Describe in your own words what happens in each step in the process.  Here is a key to the diagram:

• ribosome (dark circle)
• cysteines residues (yellow surfaces)
• hydrophobic (purple surfaces)
• intramolecular disulfide bonds (yellow areas)
• free Cys encircled in red

Answer

Panel (a) start of the synthesis of the α-gliadin at the ribosome (dark circle);

Panel (b) as soon as cysteine residues (yellow surfaces) are synthesized, adjacent sections of amino acids influence the 3D-shape of the molecule, and those which are hydrophobic (purple surfaces) bring cysteine residues close;

Panel (c) closeness of cysteine residues contributes to the formation of intramolecular disulfide bonds (yellow areas);

Panel (d) a fully synthesized α-gliadin with disulfide-bonded cysteine residues (yellow areas) and a disordered fold;

Panel (e) a fully synthesized LMW-GS that has undergone the same formation of intramolecular disulfide bonds (yellow areas) but with two cysteine residues (with neighboring sections of amino acids that are less hydrophobic) that are not crosslinked (here encircled in red);

(f) intermolecular disulfide bonds are formed after protein synthesis for the LMW-GS;

(g) protein bodies are formed by glutenins and gliadins through intermolecular disulfide bond crosslinks (LMW-GS) and hydrophobic interactions.

 

Gluten and the Elasticity of Dough

A key theme of all biochemistry courses is that structure determines function.  From the earlier iCn3D models, the predicted structure of HMW glutenin is extremely disordered (almost amorphous) while the LMW glutenin and the gliadins are significantly disordered.  First, let's consider the backbone of the gluten proteins in the unstretched state and the stretched state in dough, without considering the side chains.  By analogy, think about the stretching of a rubber band!  Both dough and rubber bands are elastic since both can be stretched with an applied force, and a restoring force returns them to their original state. 

Natural rubber contains tangled strands of the polymer cis-1,4-polyisoprene (molecular weights 100K to 1M), so it has just a backbone without side chains like a protein. You may also remember that on stretching, a rubber band warms.  

Exercise 6

For the reaction:

Rubber band relaxed ↔ Rubber Band stretched

what are the signs (+/-) for ΔG, ΔH, and ΔS for stretching?

Answer

Since the reaction is not spontaneous, ΔG_stretching is +.  Since heat is released on stretching (the rubber band warms), ΔH_stretching is -, so ΔS_stretching must be - (unfavored) as well.  Remember that ΔG_stretching = ΔH_stretching - TΔS_stretching.  

 

Figure \(\PageIndex{10}\) is a cartoon version showing the effects of stretching on the polymer chains in a rubber band. Note that some of the chains are "pinned" together at one of their ends.

Figure \(\PageIndex{10}\): Theodore A. Brzinski III and Karen E. Daniels.  Stretching Rubber, Stretching Minds: a polymer physics lab for teaching entropy.  [1508.00538] Stretching Rubber, Stretching Minds: a polymer physics lab for teaching entropy (arxiv.org). With permission of authors. CC BY-SA 4.0 license (https://creativecommons.org/licenses/by-sa/4.0/)

Exercise 7

Based on the diagram above, explain the origin of the - ΔS on stretching the rubber band.

Answer

In the relaxed (equilibrium state), the chains are less ordered and are similar to random coils in proteins.   On extension with an applied forces, the chains become more ordered. 

Stretching is entropically disfavored since the number of microstates (possible structures) in the entangled relaxed states is much greater than in the much more ordered stretched state. The release of tension causes the rubber band to contract in a process characterized by a - ΔG in an endothermic process favored by an increase in chain conformational entropy.

 

Protein polymers are more complicated than simple rubber band polymers.

Exercise 8

What are some differences between the stretching of a rubber band and bread dough?

Answer

We used the simple example of a rubber band as a polymer with no side chains.  Proteins have side chains that can interact with each other.  In addition, there is a lot of water in the dough (obvious to those who have made bread!) which would contribute to the overall thermodynamics of dough stretching and contraction.  There would be more types of noncovalent interactions in dough than in simple rubbers, which interact through just nonpolar-nonpolar interactions.

 

 

Now let's think about the same process only with dough comprised of disulfide-linked glutenins with monomeric gliadins embedded in the polymer.  You might want to envision a polyacrylamide gel matrix with embedded proteins in the pores of the gel. Figure \(\PageIndex{11}\) below shows a PAGE matrix without proteins.

Figure \(\PageIndex{11}\):  Polyacrylamide gel microstructure..  Li, J.; Zhou, W.; Qi, Z.; Luo, T.; Yan, W.; Xu, H.; Cheng, K.; Li, H. Morphology and Rheological Properties of Polyacrylamide/Bentonite Organic Crosslinking Composite Gel. Energies 2019, 12, 3648. https://doi.org/10.3390/en12193648.  CC BY 4.0

We must now consider the amino acid sequences, as well as the domain structures of the proteins as illustrated in Figure \(\PageIndex{12}\) below.

Figure \(\PageIndex{12}\): Schematic representation of different domains (a–c) and segments (I–V) of gluten protein types adapted from Wieser et al. (2020). Wieser, H., Koehler, P., & Scherf, K. A. (2023), ibid.

The following amino acid sequences were used without signal peptide (UniProtKB identifier in parentheses): HMW-GS x (Q6R2V1), HMW-GS y (Q52JL3), ω5-gliadin (Q402I5), ω1,2-gliadin (Q6DLC7), α-gliadin (Q9M4M5), γ-gliadin (Q94G91), and LMW-GS (Q52NZ4). GS, glutenin subunit; HMW, high molecular weight; LMW, low molecular weight. The numbers at the C-terminal side of the proteins show the numbers of amino acids per protein.

Note the similarity between LMW-GS and alpha/gamma gliadin.  (Apparently under certain conditions, some gliadins can form large molecular weight, disulfide-linked aggregates but in this discussion we will assume that only the glutenin can form interchains S-S bonds.) The HMW glutenins (HMW-GS) have an internal segment (dark orange) that contains the repetitive repeats and is highly enriched in glutamine, proline and glycine.  The N and C-terminal domains have a normal distribution of amino acids and contains most of the cysteines for cross-linking chains as well as charged amino acids.

The gliadins and more closely homologous LMW-GS have 5 domains.  The second half of domain I has different repetitive sequences that include

• QPQPFP and PQQPYP in α-gliadin (this monomer also has segment II with up to 18 glutamines)
• QQPQQPFP in γ-gliadin
• QQQPPFS in LMW-GS

The other domains in gliadins and LMW-GS have a normal amino acid composition.

Given that the cysteines used for interchain disulfide in the glutenins are generally in the end domains, let's model the glutenin polymers as straight lines that can form significant interactions through their central domain. 

Exercise 9

What are the structural consequences of having such a high percentage of glutamine, proline, and glycine in the central domain of glutenins?

Answer

Prolines and glycines are often found in tight turns in proteins.  Proline is a helix breaker given its rigidity as a cyclic amino acid.  Glutamine can form both alpha and beta structure and form hydrogen bonds through its side chains.  You would expect all of these characteristic to emerge as key in the structures we will see below.

 

Figure \(\PageIndex{13}\) below shows a simplified train-loop model created by Belten for the main chains of glutenins in the relaxed (train or equilibrium state) and in the deformed stretched shaped with loops (side chains not shown).  Imagine disulfide linkages between cysteines that are generally localized in the N and C-terminal ends of the chains, much like the "pinning" dots at the end of the chains in the rubber band polymers (Figure 10).

Figure \(\PageIndex{13}\): after Belton, Journal of Cereal Science. Volume 29, Issue 2, March 1999, Pages 103-107

Exercise 10

What interactions likely stabilize the train (equilibrium) shape?  Offer a possible reason to explain the easier movement (deformation) of the chains after some initial deformation?

Answer

There must be significant stabilization of the interacting polymers through interstrand hydrogen bonding using glutamines as donors and acceptors.  The shearing forces introduced by kneading the dough must provide energy to break hydrogen bonds but also some disulfide bonds, allowing easier subsequent sliding of the chains with respect to each other.

At first, it is easier to stretch since the chains move into the "pores" (like in polyacrylamide gels) in the matrix and they don't have to disentangle.  But then more stretching requires sliding and disentangling of the chains, which requires more force.  As the volume of the pores decreases, there are more intrachain contacts.

The N and C terminal segments of HMW glutenin are presumably more alpha-helical

 

Exercise 11

Use the Expasy ProtScale to predict structural features of the HMW-glutenin (Uniprot ID:  P10388).  Follow the prompts in the web program and download an image file.

a.  Chou-Fasman alpha helix with window size 5

b.  Chou-Fasman beta-turn with window size 3 or 5

  

 

Answer

a.  Here is the output for the alpha helix

Clearly, the ends of the protein have a much higher propensity to form alpha-helices

b.  Here is the output for the beta-turn

The disordered internal domain highly enriched in Q, P, and G clearly can form beta-turns.  This should come as no surprise since P breaks helices and P and G are abundant in turns.

 

Exercise 12

What is the likely origin of strength (resistance to stretching) and elasticity (in the presence of an applied force)?

Answer

The central disordered domain is critical.  The repeat regions likely form hydrogen bonds between each other in adjacent chains.  Interchain disulfide from cysteine at the ends of the glutenin can explain the strength of the dough, while the more breakable and deformable hydrogen bonds between central domains account for the elasticity.  These can reform after breaking them, a requirement for elasticity.  In contrast, S-S won't' reform their original pairs after breakage as there is no easy way to do that.  

 

 

 

Exercise 13

On relaxing back to the unstretched state, glutamines will return in register to form hydrogen bonds to glutamines on neighboring chains.  This must be entropically disfavored.  What might compensate for this entropy loss?

Answer

Probably a gain in entropy from released hydrogen-bonded waters on reformation of Gln--Gln hydrogen bonds.

 

Models have been constructed (Figure \(\PageIndex{14}\)) to show how the HMW and LMW-glutenins can form a 3D network.

Figure \(\PageIndex{14}\):  Model of the three-dimensional glutenin network made up of x- and y-type HMW-GS and LMW-GS modified from Wieser et al. (2014). Disulfide-linked HMW-GS are displayed horizontally, and LMW-GS are attached to this backbone via intermolecular disulfide bonds. HMW-GS, high-molecular-weight glutenin subunit; LMW-GS, low-molecular-weight glutenin subunit.   Wieser, H., Koehler, P., & Scherf, K. A. (2023). Ibid

It is now thought that on heating (during baking), the intrachain S-S bonds in gliadins break and form interchain S-S bonds with the glutenin in a disulfide interchange reaction, similar to forming isopeptide linkages in proteins. This gives backed bread its relative rigidity.

The simple train-loop model is also limited as it doesn't contain water (an obvious component of dough to those that make bread) and the more globular gliadins, which would fit into pores of the glutenin matrix like proteins in a polyacrylamide gel.

Molecular dynamics simulations support the development of trains on elongations, as shown in Figure \(\PageIndex{15}\) below.

Figure \(\PageIndex{15}\). The time dependence of 6 different quantities that are measured for gluten (as defined in section 1 of the S1 Text) during the elongation in the last stage of the simulation.  Mioduszewski Ł, Cieplak M (2021) Viscoelastic properties of wheat gluten in a molecular dynamics study. PLoS Comput Biol 17(3): e1008840. https://doi.org/10.1371/journal.pcbi.1008840. Creative Commons Attribution License

 

Gluten and Celiac Sprue Disease

This disease is an autoimmune disease of the small intestines.  The repetitive sequences in gliadins and glutenins appear to trigger T cells in this autoimmune disease.  The 33 mer from monomeric wheat gliadin shown in Figure 6 above is released by intracellular proteolysis and deaminated to some extent by a transglutaminase. This protein usually functions to make interchain covalent bonds between a glutamine on one protein and a lysine side chain on another. The released 33 mer is resistant to further proteolysis and appears to be a prime, but not the only trigger of celiac disease. The deaminated peptide (called an epitope - a small sequence or motif recognized by the immune system) binds more tightly to immune antigen-presenting proteins than the original peptide. T cells then interact with the deaminated peptide (epitope) displayed (or presented) by antigen-presenting proteins (MHC/HLA) on the surface of the cell, where it is available for immune recognition by the T cell. The 33 mer peptide is missing in nontoxic food grains.  

It's not just the 33 mer epitope that activates the immune system in celiac disease.  Peptides originating from many of the other repetitive sequences also act as epitopes as shown in Figure \(\PageIndex{16}\) below.

Figure \(\PageIndex{16}\):  The distribution of T-cell epitopes (shown as red bars) in representative wheat gluten proteins (identified by GenBank accession codes). Shewry Peter, What Is Gluten—Why Is It Special? Frontiers in Nutrition, 6, 2019. https://www.frontiersin.org/articles...nut.2019.00101.   Creative Commons Attribution License (CC BY) .

The epitopes are based on Sollid et al. (36). α-gliadin P18573: DQ2.5-glia-α1a, DQ2.5-glia-α1b, DQ2.5-glia-α2, & DQ8-glia-α1. γ-gliadin AAK84774: DQ2.5-glia-ω1/hor-1/sec-1, DQ8-glia-γ1a, DQ8-glia-γ2, DQ8-glia-γ4c, & DQ8-glia-γ5. ω-gliadin (A/D) AAT74547: DQ2.5-glia-γ5, DQ8-glia-γ1a, DQ2.5-glia-ω1/hor-1/sec-1, DQ8-glia- γ1b, & DQ2.5-glia- γ3. ω-gliadin (B) AB181300 no coeliac toxic epitopes present. LMW subunit AAS66085:DQ2.5-glut-L1. HMW Subunit (1Bx17) BAE96560: DQ8.5-glut-H1. HMW Subunit (1Dy10) AAU04841: DQ8.5-glut-H1. 

Figure \(\PageIndex{17}\) illustrates the steps involved in celiac disease.  The enterocytes (epithelial) cells that line the small and large intestines are shown.  TGG is transglutaminase.  APC is antigen-presenting cell.

Figure \(\PageIndex{17}\): Mechanisms by which ingested gluten triggers celiac disease: Modification of figures by Serena, G.; Camhi, S.; Sturgeon, C.; Yan, S.; Fasano, A. The Role of Gluten in Celiac Disease and Type 1 Diabetes. Nutrients 2015, 7, 7143-7162. https://doi.org/10.3390/nu7095329.  Creative Commons Attribution license.  (http://creativecommons.org/licenses/by/4.0/)

 

Exercise 14

Explain what occurs in steps 2, 4, 5, and 6.  

Answer

Enterocytes (epithelial) cells that line the small and large intestines, transport the gluten peptide (made by limited proteolysis by gut proteases) across the cell (2) where they are deaminated by transglutaminases (4) and then bind to MHC proteins on antigen presenting cells (APC) like macrophages (5).  The deaminated peptide (epitope) presented by the MHC protein next interacts with and activates T cells (6).

 

Exercise 15

The GI system contains the highest levels of proteases, which makes sense given their role in food digestion.  Input the sequences of the two epitopes discussed here into the ExPasy peptide cutter to see if they are resistant to most proteases.  

• 33 mer: LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF
• 18 mer: QQYPSGQGSFQPSQQNPQ

Answer

The predictor suggests that chymotrypsin and pepsin could cleave them in the gut but they must be more resistant than other peptides formed by gut endoproteases. 

 

Figure \(\PageIndex{18}\) shows an interactive iCn3D model of the MHC class II molecule HLA-DQ8 bound with a deamidated gluten peptide from Alpha/beta-gliadin MM1 (2NNA) .  The bound peptide (derived from amino acids 242-260, QQYPSGQGSFQPSQQNPQ) is different peptide epitope from the same protein, Alpha/beta-gliadin MM1 (P18573) as the 33 mer (amino acid 76-108) shown in Figure 6.  The first 4 and last amino acids are not seen in the structure since they are too disordered.

 

Figure \(\PageIndex{18}\): MHC class II molecule HLA-DQ8 bound with a deamidated gluten peptide from Alpha/beta-gliadin MM1 (2NNA) . (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...cmbE1Wbmkv6zk6

The alpha-chain of the MHC protein is in cyan and the beta-chain in gray.  The surface of both is shown along with the underlying secondary structure. The gluten peptide (spacefill, CPK colors) is an 18-mer but only amino acids 5-17 are ordered enough to see in the structure.  The peptide fits like a lock and key into the grove.

Exercise 16

Is the peptide deaminated? Use iCn3D for the answer.

Answer

Yes. Instead of the sequence in the native protein, amino acid 242-260, QQYPSGQGSFQPSQQNPQ, the sequence of the bound peptide is shown below.  2 glutamines (Q) are replaced with glutamic acid (E).  The protein was crystallized with the deaminated peptide.  The small letters in the sequence below show the amino acids that are too flexible and disordered to be seen in the X-ray structure.

 

 

Now, let's model the noncovalent interactions between the peptide and the alpha and beta chains of the MHC protein dimer and the T-cell receptor that interacts with them both.

Figure \(\PageIndex{19}\) shows an interactive iCn3D model of the T cell receptor-HLA DQ8-gliadin peptide complex in Celiac Disease  (4GG6) 

 

Figure \(\PageIndex{19}\): T cell receptor-HLA DQ8-gliadin peptide complex in Celiac Disease  (4GG6). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...enMmmrmfbjYh67

 

Orient the cyan chains at the top and the gray chains at the bottom.  The T cell receptor alpha and beta chains (on the responding immune cell) are in shades of cyan.  The antigen (peptide)-presenting MHC alpha and beta chains are in shades of gray. The gliadin peptide (spacefill, cpk colors) is sandwiched between them.

In the final exercise, you will construct an iCn3D model showings the noncovalent interactions of the peptide with both the T cell receptor and the MHC proteins.

 

Exercise 17

Identify noncovalent interactions between the gliadin peptide and both the T cell receptor and HLA DQ8 peptide presenting protein using iCn3D.

1. Open iCn3D and enter the id 4GG6 and hit return
2. Analysis, Sequence and Annotations
3. Choose Details Tab and uncheck Conserved Domains
4. Click the blue hyperlinked Protein 4GG6_A (the HLA alpha chain) to select it; then Color, Unicolor, Gray, Light Gray
5. Click the blue hyperlinked Protein 4GG6_B (the HLA beta chain) to select it; then Color, Unicolor, Gray, pick a darker color gray
6. Repeat 4 and 5 for the T cell alpha and beta chains using two different shades of cyan
7. Click the 4GG6_J (the peptide) to select and highlight it.
8. Select, Save Selection, and name it Peptide
9. Style, Side Chains, Stick
10. Color, Atom
11. Analysis, Interactions
12. Choose these defaults: 
13. Click 4.  3D Display Interactions
14. Select, Save Selections, name it interactions
15. Click 2D Interaction Map
16. View, Selection
17. File, Save, iCn3D image (of model)
18. Follow prompts to save interaction map image

.

Answer

Here is link to the interactions between the peptide and the Tcell receptor and the HLA protein

https://structure.ncbi.nlm.nih.gov/icn3d/share.html?BJvZvNY23dpgZXGFA

 

 

Note that there are interactions between all the chains and the peptide.  These are shown in a 2D diagram below. (For some reason Pro 8 does not appear in the peptide.)

The top row is the peptide.  The bottom row shows 2 interacting amino acids from the T cell receptor (2 shades of cyan) and the most from the peptide presenting MHC HLA chains (2 shades of gray).  Green lines are hydrogen bonds and blue lines are salt bridges (ion-ion interactions).

.