Analyzing gluten in food

(1)

MSc Chemistry

Analytical sciences

Literature Thesis

Analyzing gluten in food

by

10291067 Sifra Vos

March 2020

12 ects

February 2020 - September 2020

Supervisor/Examiner:

2

nd

_reviewer:

Supervisor: Rob Haselberg

Examiner: Alina Astefanei Assessor: Andrea Gargano

Vrije Universiteit - Division of

BioAnalytical Chemistry

(2)

2

Abstract

This study examines if it is possible to accurately quantify gluten in food and whether food labels considering gluten content are reliable. Different analysis methods are available, each with their own advantages or disadvantages. The current approved method, ELISA, has the huge disadvantage that it is not able to measure certain modified gluten. This makes food labels for products with a large amount of modified gluten, like beer, unreliable. There is however a method that can be used to accurately quantify gluten in food. This method is RPLC-MS method in MRM mode. With this method it is possible to quantify gluten when using a marker for the specific type of gluten which depends on the cereal it originates from. This study has also reviewed at home devices for quantifying gluten and future developments regarding this field of research. There are a lot of possibilities for further development, but more research is needed before those are able to be used in industry.

(4)

4

1. Introduction

The last few years interest in health safety and a healthy lifestyle has increased.1,2_{People nowadays}

want to know exactly what they eat and scrap unhealthy foods off their diets. However, there are different opinions on what those unhealthy foods are. While everyone can agree that too much sugar and too much fat is unhealthy, a certain amount of either of the two is needed to fuel the body. This spiked interest in health has also lead to the misconception that gluten are bad and that a gluten-free diet is part of a healthy diet which leads to people cutting gluten out of their diets.2

While gluten are perfectly safe for most people to eat, there is a group of people that do get sick when eating gluten. People that do need to attain a gluten-free diet as to not get allergic or immune reactions have celiac disease (CD), wheat allergy (WA) or non celiac gluten sensitivity (NCGS). 2- 11_{But also people with gluten neuropathy (GN) or irritable bowel syndrome (IBS) may}

benefit from a gluten-free diet to stay symptom-free. 2, 8, 10-17 _{For both of these groups it is important}

to know which food is gluten-free and which foods are not.

To make it possible for the people at risk to follow their much-needed diet, it needs to be clear which food contains gluten and what food does not. In order to regulate gluten-free claims from food producers, in the EU commission Regulation No. 41/2009, the law distinguishes three types of food with reduced gluten (see table 1).3, 18 _{These regulations are not only in place to make life easier for}

consumers, but also to prevent fraud, misleading actions or adulteration of food products or procedures.1

For most food producers, the aim is to make as much money as possible and since gluten-free ingredients are usually more expensive than gluten-rich ingredients it is not attractive to make their food gluten-free. Also, to reduce cross-contamination, machines and spaces that are used to produce gluten-free food cannot be used to also make gluten-containing food. Considering this, food adulteration is usually economically motivated as to reduce production costs and increase market value. To prevent this adulteration European law requires that all food sold has a label that matches its contents.1_{Since analyzing for all different types of allergens in food does take a lot of time and}

money, an easy and accurate procedure to test for allergens in food would make it more appealing for food producers to not have fraudulent or adulteration actions.

Table 1. Law on food with reduced gluten

Claim Specifications

Gluten-free Contains 20 ppm or less gluten

Very low gluten Contains 21-100 ppm gluten

No gluten-containing ingredients Does not contain any ingredients that contain gluten either by nature or

cross-contamination

To give them good information the analysis of gluten in food needs to be reliable and preferably quick. This work gives an insight in the diseases that require a gluten-free diet and compares the different methods of analyzing gluten. It will compare different methods of characterization and quantification and whether it is possible to achieve the accuracy to meet the the requirements from table 1 for their respective claims.

(5)

5

2. Gluten and their function

2.1 What are gluten?

To be able to make a good estimation of the amount of gluten in food, it is necessary to have a clear definition of what gluten are. According to the Codex Standard 118-1979 this definition is: “ ...protein

fraction from wheat, rye, barley, oats, or their crossbred varieties and derivatives thereof, to which some persons are intolerant and that is insoluble in water and 0.5 M NaCl.” 3, 19 In practice, gluten

are a complex mixture of proteins from cereal seeds that mainly consists of glutenins and prolamins. The glutenin part of this mixture is soluble in diluted acid while the prolamins are soluble in water/alcohol mixtures. Usually the prolamins are given specific names for the types of cereal they are in (see table 2).3, 4, 19 _{While gluten is a complex mixture of proteins, it is established that the}

proteins in gluten are usually rich in glutamine and proline and are poor in basic amino acids.3 Table 2. The names of prolamins in different types of cereal

Type of cereal Prolamins name

Wheat Gliadins

Barley Hordeins

Rye Secalins

Oats Avenins

Maize Zeins

Cereals and grains can be divided into three groups. The first contains cereals and grains that contain gluten proteins such as wheat, barley, rye, spelt, einkorn and kamut. The second group contains gluten-like proteins and consists solely of oats. The last group contains no gluten proteins and consists of maize, rice, millet, buckwheat, linseed, sesame, tapioca, chestnut and tiger nut sedge. The first group causes reactions in people with CD or NCGS while the third group is safe for them to consume. The second group is cause for discussion since some people do react to oats while for others it is save to consume.4

When considering gliadins in wheat, they are primarily monomeric proteins. Gliadins can be divided into three groups (𝛼, 𝛾 and 𝜔) depending on their amino acid sequences and their mobility in gel electrophoresis.20_{Most 𝛼-gliadins contain six cysteine-residues that can form three intramolecular}

disulfide bonds and are potentially toxic for CD patients.3

In contrast to gliadins, glutenins are polymers. Glutenins can be divided in two categories based on their weight: low molecular weight glutenins (LMW-GS) and high molecular weight glutenins (HMW-GS)20_{. This distribution is also made by the analysis of glutenins in size exclusion}

chromatography (SEC) which will further be discussed in chapter 4.

The HMW-GS proteins consist of three parts: a repetitive central domain, a non-repetitive N-terminal domain and a non-repetitive C-N-terminal domain. The last two contain the most cysteine residues and are relatively small compared to the central domain which can differ in length considerably. This leads to a secondary structure of the central domain that is 𝛽-reverse dominant while the N- and C-terminal domains consist mostly of 𝛼-helices.21

2.2 Food and non-food products that contain gluten

Keeping strictly to a gluten-free diet can be quite challenging because a lot of food contains gluten. The most obvious types of food that contain gluten are bakery products that are usually made from wheat like pasta, pizza, bread, cookies and pastry. In these products gluten are mostly used to improve their structure and volume. 6, 9, 21, 22_{For these products gluten-free alternatives are usually available.}

Besides the most obvious products, gluten can also be used as a binding agent in food like candy, meat, soup or sauces. 6, 9 _{Since these products are less obvious to contain gluten, clear labels and}

(6)

6 experience with reading them is necessary. Besides these obstacles, food producers also have a tendency to improve their recipes. This may lead to a lot of frustrations and mistakes as products that have been gluten-free and used on a regular basis can suddenly contain gluten with their new recipe leading to yet another a long search of a gluten-free alternative.

Because of its function as a binding agent, gluten can also be found in non-food products like pet food, medicines, adhesives or cosmetics.8, 23 _{Especially gluten in medicine can cause serious}

problems. Pharmacists often do not think about checking for gluten in medicine before handing them out which can cause a reaction in people that do not handle gluten that well. Since diarrhea is one of the reactions such a person can get, it can cause the medicine to lose its effectiveness or even cause more harm than help. Gluten in pet food not only causes problems for people that are very sensitive to gluten handling the pet food, but also for pets themselves. Pets can, just like people, also be sensitive to gluten. For pets that are sensitive to gluten, eating it can cause similar symptoms as in humans.

Gluten are not only found in food by using ingredients that contain gluten, but also because of cross-contamination. Cross-contamination can occur in a factory that works with gluten-rich and gluten-free cereals causing gluten-free cereals like oats or buckwheat to contain gluten after processing.4 _{But cross-contamination can also occur when gluten-rich as well as gluten-free cereals}

are used in a kitchen at home or in a restaurant. When for example, a person makes their gluten-rich sandwich and does not thoroughly clean the counter, plates, knifes, towels, etc. and another person makes their gluten-free sandwich after that, it can be contaminated. Or when a person first heats their gluten-rich pizza in an oven and after that a person heats their gluten-free pizza in the same oven without thoroughly cleaning it, it will become contaminated. Because of this, patients that need to be on a gluten-free diet often have their own oven, cutlery, bread spreads and working space on the kitchen counter.3 _{These regulations are also in place for restaurants that offer gluten-free food causing}

higher prices for gluten-free food. This makes a gluten-free diet not only hard to maintain, but also expensive.

2.3 The function of gluten in food

The difference between gluten-rich and gluten-free food is usually not hard to tell. The results from an experiment I did showed that 80% of the participants was able to differentiate between gluten-free and gluten-rich cake and that 81% of them did so because they noticed the difference in structure (see appendix). A more dense and dry structure was mostly the reason to assign the cake gluten-free. This is because gluten proteins can form a gluten-network which gives structure and volume in most bakery products.6, 19_{This is why most gluten-free bread is very compact compared to gluten-rich}

bread.

As mentioned in chapter 2, a gliadin protein has six cysteine-residues. These cysteine-residues were thought to be the key in forming the gluten-network which gives the food its airy structure. It has been postulated in the past that the gluten-network is made through the disulfide-sulfhydryl exchange in the cysteine-components, but a later study showed that the tyrosine bonds that are formed during the process of mixing and baking wheat dough give the structure to the gluten-network.21

Because gluten-free cereals consist of proteins which are poor in certain amino acids that gluten-rich cereals have, they are less able to form gluten-networks causing the doughs to have less structure and volume.3 _{In gliadin, tyrosine residues occur 11-22 times throughout the length of the spiral backbone.}

It was estimated that the lack of ability to form tyrosine cross-links after the mixing peak of dough causes the dough to quickly lose its elasticity.21_{As a consequence of the different protein structure,}

gluten-free doughs tend to be less elastic and usually need more water as they tend to form hard-gel structures.24 _{These hard-gel structures cause bread to be compact and also have a hard structure in}

(7)

7 To pin down which part of the gluten free dough actually causes this hard-gel structure, a study has been done in which different types of gluten-free cereal doughs have been tested on their rheology properties.24_{This study shows that amaranth flour contributes to the hard-gel structure of}

gluten-free doughs. To increase the properties of the gluten-free dough, other rheology properties have also been studied. It has been found that cellulose increases the elasticity of the dough and pea protein isolate gives the dough a stiffer structure. However, the perfect mix of gluten-free dough to replace gluten-rich dough has not been found yet.24_{Current gluten-free products tend to be dry and}

hard of structure. They also have less nutritional values and are less tasty than their gluten-rich variants.3 _{This is also shown by the cake-tasting experiment I did in which 19% of the participants}

who picked the right cake to be gluten-free thought this was the case because they found the other one was tastier (see appendix).

The quality of dough-formation and bread making is thus mainly dominated by the glutenin subsets.21 _{After washing the wheat dough to remove starch and other soluble components, a rubbery,}

cohesive aggregate is left which is the gluten-part of the dough.19 _{To give dough even more foaming}

and emulsion abilities, gluten proteins can be deamidated. The deamidation of gluten proteins will be discussed in more detail in chapter 5.

(8)

8

3. Gluten intolerance and sensitivity

Besides a group of people that choose to eat gluten-free, there is also a group of people that have a disease that can only be treated by containing a gluten-free diet. In these cases, gluten cause a series of symptoms and only a gluten-free diet can lessen these symptoms. The diseases that show a correlation to gluten intake are celiac disease, wheat allergy, non-celiac gluten sensitivity, irritable bowel syndrome and gluten neuropathy.

3.1 Celiac disease

The most well-known disease that has a connection to gluten is celiac disease (CD). CD is clinically known since 1888, but the cereals that cause symptoms were not identified until the 1950’s.11_CD

affects about 0.6-1% of the world population with wide regional differences in worldwide. The only exception is the Suharawui population that lives in refugee-camps in North-Africa and have a prevalence 5%.3, 9-11 _{The symptoms in people with CD can be found in table 3.}9, 10, 12, 25, 26

In people with CD, an auto-immune disease, gluten bind to the HLA-DQ2 and/or HLA-DQ8 leukocyte antigens after which the complex interacts with T-cells causing an adaptive immune response. This reaction causes an increase in elevated immunoglobulin E and A (IgE or IgA) causing villous atropathy.2, 3, 6, 8, 10-12 _{The network of processes in the immune response in CD can be found}

in figure 1.27_{Since CD is characterized by the binding of gluten to the HLA-DQ2/DQ8 antigens, CD}

can only be developed by people that have the HLA-DQ2 or HLA-DQ8 antigens and is thus a good way to rule CD out in people that do not have either of the two antigens. 2, 3, 8, 9, 11, 12_{Not all people}

that have the antigens develop CD and not all people that have CD also have elevated IgE or IgA. For this reason, a diagnosis can only be made by confirming villous atropathy. This is done by performing a biopsy of the small intestine. However, before a biopsy will be performed a doctor needs to have a plausible cause. A lot of patients have already made the connection between eating gluten-free and a lessening of symptoms causing them to adjust to a gluten-free diet before going to their doctor in the first place. Since villous atropathy can heal during a 6-24 month period while being on a gluten-free diet, the patient needs to consume a daily amount of gluten for a period of 2 months before a biopsy is done. Also, the levels of IgE and IgA will no longer be elevated after adjusting to a gluten-free diet.8, 9, 11, 12, 25_{As a biopsy is usually done under sedation and it can be very stressful without, for a}

diagnosis in children the biopsy can be omitted if tT-IgA and EMA IgA levels are over a 10-fold higher than normal and HLA-DQ2 or HLA-DQ8 findings are positive.12

In contrast to an allergy - where people get severe reactions by even the tiniest amount of allergen - people with CD can consume 10-50 mg of gluten on a daily basis without it causing symptoms. To give an idea of how much that is, a slice of bread approximately contains 1.6 g of gluten. So, for a lot of people with CD even a crumb of gluten-rich bread can already cause symptoms.5, 9, 12 _{CD can cause a lot of symptoms making it hard to diagnose since patients usually}

(9)

9

Figure 1. Processes that gluten cause in patients with CD. Different processes include the binding of gluten peptides to intestinal epithelial cell (IEC) receptors, the production of interleukin-15 (IL-15), activation of intraepithelial lymphocytes (IEL), aptosis of IECs, production of interferon-gamma (IFN-𝛾), interleukin-21 (IL-21), binding to antigen

presenting cells (APCs) and reactions including transglutaminase-2 (TG-2)27 3.2 Wheat allergy

People that are affected by wheat allergy (WA) are allergic to wheat gluten. WA affects about 1% of the children, but almost half of the patients outgrow the allergy by the age of 8.4_{WA is triggered by}

the cross-linking of immunoglobulin (IgE) with the repeat sequence Ser-Gln-Gln-Gln-[Gln-]Pro-Pro-Phe in gluten peptides while non-gluten proteins induce the release of immune mediators like histamine.8, 11, 12 _{As WA is an allergy, it can lead to anaphylaxis and in extreme cases to death.}4

Because of the danger, patients usually get antihistamine medication. To diagnose WA, a gluten-rich diet needs to be followed since the serum biomarkers are only reliable when eating gluten.9_Another

way to diagnose WA is by a skin prick test after discontinuing antihistamine medication and the final diagnosis is made by an oral challenge.2, 11, 12_{Since wheat is one of the most common used}

gluten-rich cereals, a gluten-free diet may be advised for people that suffer from WA.5, 9 _{Symptoms of WA}

can be found in table 3.4

3.3 Non-celiac gluten sensitivity

A newer diagnosis that has a direct connection to gluten is non celiac gluten sensitivity (NCGS). Where CD is an auto-immune disease characterized by an adaptive immune response, NCGS is caused by an innate immune response with increased levels of claudin-4, Toll-like receptors TLR2 and TLR4 and a and b intraepithelial lymphocytes and decreased levels of T-regulatory cells. 2, 8, 11-13_{This reaction is triggered because the non-adaptive peptide 31-43 facilitates the recognition of}

adaptive gliadin epitopes by the expression of interleukin-15 (IL15), CD83, cyclo-oxygenase-2 and CD-25 and promotes apoptosis of enterocytes. A-Gliadin peptides interact with the CXCR-3 receptor in the intestinal mucosa which releases zonulin that regulates gut-permeability and facilitates the moving of antigens starting an innate immune response.8 _{The processes of the immune response in}

NCGS can be found in figure 2. It is not clear yet what exactly causes NCGS butbecause amylase trypsin inhibitors (ATIs) in wheat induce the activation of the innate immune response TLR4 complex, ATIs could be the inducers of NCGS.2, 8, 10, 11, 13

(10)

10

Figure 2. Innate immune response in NCGS. The figure shows the processes that the different parts in the gluten peptide start.8

CD and NCGS have a lot of similarities in symptoms. Symptoms in people with NCGS can be found in table 3.8 _{NCGS is characterized by CD-like symptoms after eating gluten and can only be}

diagnosed when CD and WA are ruled out.7, 8, 10_{The diagnosis of NCGS can be only done after a}

supervised gluten challenge.2, 10-12 _{A difference between CD and NCGS is that patients with NCGS}

do not have villous atrophy and therefore do not develop other diseases like diabetes or colonial cancer. For this reason, the yearly follow-ups that are necessary for CD patients are not necessary for NCGS patients.

Because of the similarities between CD and NCGS, the two can be easily misdiagnosed by doctors that are not specialized in these diseases. It is not uncommon for patients with NCGS to be diagnosed with CD when the biopsy is omitted because of a lack of knowledge by the doctor. In these cases, the patient has high IgE or IgA levels which are seen by a blood test but no villous atrophy which is missed because of the omittance of the biopsy. The other way around is also possible. People with CD can have low or normal levels of IgE or IgA but have villous atrophy causing them to be diagnosed as NCGS when the villous atrophy is missed by omitting the biopsy. Misdiagnosis can also be caused because doctors do not know that the tests can only give positive results for CD after being on a gluten-rich diet and do the tests after the patient has been on a gluten-free diet for a while causing the IgE and IgA to decrease and the villous atrophy to heal. Figure 3 shows a pathway to getting a reliable diagnosis.

Since NCGS is a relative new disease, a lot of research is done on the origin of it. Besides the similarities with CD, there are also similarities between NCGS and irritable bowel syndrome (IBS) and some studies indicate that FODMAPs are the cause of trouble in NCGS and claim that NCGS is therefore a class of IBS.2, 7, 8, 11, 13 _{FODMAPs are a group of product containing Fermentable fructans,}

Oligosaccharides present in wheat, Disaccharide lactose in milk, Monosaccharide fructose in honey and fruits And Polyols like sorbitol in chewing gum. _{But other studies show that NCGS and IBS are}

caused by other proteins.8 _{Because there are still a lot of questions on NCGS, more research is needed}

to know more about this disease and how to treat it. For now, the only remedy for NCGS is a gluten-free diet. Because eating gluten with NCGS does not trigger other disease like colonial cancer, patients can eat gluten from time to time without it causing serious harm.

(11)

11

Figure 3. Pathway to getting a reliable diagnosis for CD, WA or NCGS. All tests are done during a gluten challenge. As can be seen, the only difference between the diagnosis of CD and NCGS is the presences of villous atrophy during the biopsy. 9

3.4 Irritable bowel syndrome

As mentioned before, NCGS and IBS have some similarities that makes the diagnosis of either of them a gray area. IBS is a chronic functional gastrointestinal disorder that is characterized by abdominal pain associated with abnormal bowel habits in the absence of any structural, physiological or biochemical abnormalities of the gastronomical tract. Just like CD and NCGS, IBS is more common in women than in men.14_{The diagnosis for IBS is however different than CD and NCGS}

since IBS does not involve an immune response. IBS can be diagnosed if the following is true: abdominal pain related to defecation or associated with altered bowel frequency or stool form accompanied by at least two of the following: altered stool passage, bloating, abdominal distension, abdominal tension, abdominal hardness, symptoms worsening by eating or passage by mucus. Symptoms should appear at least 4 days per month.14-16 _{The symptoms for these three diseases are}

T h e ne w e ngl a nd jou r na l o f m e dicine

n engl j med 367;25 nejm.org december 20, 2012

2424

supervision. A reduction in symptoms after the implementation of a gluten-free diet is not patho-gnomonic of celiac disease, since a placebo effect and other forms of gluten reaction have been de-scribed.44_{Patients with a wheat allergy may also}

benefit from a gluten-free diet. The distinction among celiac disease, gluten sensitivity, and wheat

allergy can be difficult to establish and should be based on several criteria (Table 2). Since serum biomarkers that are currently available for wheat allergy, like those for celiac disease, are reliable only when patients are exposed to gluten, a diag-nostic algorithm (Fig. 1) should be followed while the patient is still consuming gluten.45

Suggestive history, physical examination, and initial evaluation; consider

differential diagnosis

Wheat allergy _{gluten sensitivity}Celiac disease or

Specific skin-prick tests Wheat-specific serum

IgE test Gluten challenge

Confirm wheat allergy

tTG IgA test (with or without EMA) plus total IgA Deamidated AGA IgA

and IgG tests

Tests and challenge positive Rule out wheat allergy No Yes Perform EGD

with biopsy Perform EGDwith biopsy Positive tTG, deamidated AGA, or both Strong clinical suspicion No No Consider other diagnoses Consider gluten sensitivity Consider celiac disease (follow-up) No No No Yes

Confirm gluten sensitivity Perform double-blind gluten

challenge

Rule out gluten sensitivity; consider other diagnoses No Yes Biopsy positive Yes Biopsy positive Yes Yes

Confirm celiac disease

Yes Yes

Figure 1. Proposed Algorithm for the Differential Diagnosis of Gluten-Related Disorders.

AGA denotes antigliadin peptide antibodies, EGD esophagogastroduodenoscopy, EMA antiendomysial antibodies, and tTG tissue trans-glutaminase. Adapted from Sapone et al.45

The New England Journal of Medicine

Downloaded from nejm.org at UNIVERSITEIT VAN AMSTERDAM on September 14, 2019. For personal use only. No other uses without permission. Copyright © 2012 Massachusetts Medical Society. All rights reserved.

(12)

12 quite similar, causing confusion by doctors. The symptoms of IBS besides the symptoms needed for diagnosis can be found in table 3.15, 16_{Patients with IBS are at risk of also having CD. Differentiation}

between IBS and CD can be hard in adulthood.16

The big difference between IBS and the other two diseases is the role of FODMAPS. The symptoms in IBS can be worsened by the consumption of FODMAPs.14, 15 _{Which of the FODMAPS}

increase the symptoms does differ from person to person. Some people can handle fructans quite well while lactose causes an increase in symptoms while it can be the other way around for others. Diagnosis of the FODMAPs that cause symptoms is done by a low FODMAP-diet for a few weeks after which each of the FODMAPS will in turn be reintroduced in the diet.14

It is suggested in studies that gluten-rich food can trigger IBS-symptoms.16 _{This can be}

explained because the oligosaccharides are also present in gluten-rich cereals like wheat and thus can be easily confused. Also since NCGS is fairly new, many patients have been wrongly diagnosed with IBS while actually suffering from NCGS, especially in the earlier days.16_{Although both groups}

improve on a gluten-free diet, patients with NCGS suffer more frequently fromfatigue, headaches, anxiety, numbness, myalgia, weight loss, and even autism or hallucinations.16_{Treatment for IBS can}

be done by a gluten-free diet, medication like antibiotics or loperamide, antidepressants for patients that suffer from psychiatric disorders, peppermint oil, acupuncture or psychotherapy/hypnosis.16, 17

As can be seen, while for CD and NCGS the only treatment is a gluten-free diet, for patients with IBS other treatments might also work giving them the opportunity to eat gluten in higher quantities than those with CD or NCGS.

3.5 Gluten neuropathy

The last disease that can be treated by a gluten-free diet is gluten neuropathy (GN). GN is defined as “an otherwise idiopathic neuropathy, in the absence of an alternative etiology despite extensive investigation and in the presence of serology evidence of gluten sensitivity.” GN does not lead to a severe disability, but can lead to pure sensory large-fiber axonal loss in patients with sensory ganglionopathy. Since this form of neuropathy is caused by the intake of gluten, a gluten-free diet has shown to improve patients with GN.28

(13)

13

Table 3. Symptoms in different gluten-related diseases 4, 8, 9, 10, 12, 15, 16, 25, 26, 28

Symptom CD WA NCGS IBS GN Chronic diarrhea x x Weight loss x x Abdominal distension x x x Bloating x x x Poor appetite x Nausea x Hard stools/constipation x x x Loose stools x Joint/muscle pain/disorders x x x Skin/oral lesions x x Neurological symptoms x x Depression x x x Vitamin deficiency x x

Recurrent abdominal pain x x x

Short stature x

High aminotransferase levels x x x

Chronic fatigue x x x

Reduced bone mineral density x

Dermatitis herpetitiformis x

Blistering rash x x

Gluten ataxia x x

Celiac crisis x

Metabolic/electrolyte disbalances x

Reproductive system disorders x x x

Hashimoto’s thyroiditis x

Other auto-immune diseases x

Down’s syndrome x Turner’s syndrome x IgA deficiency x Anaphylaxis x Anemia x ADHD x Autism x Schizophrenia x Peripheral neuropathy x x

(14)

14

4. Gluten characterization in flour

When dealing with gluten analysis two approaches exist; 1) determining the type of gluten present in the matrix and 2) establishing the gluten amount in foodstuffs. This chapter will focus on the characterization of gluten in flour while the next chapter will focus on the quantification of gluten in food.

Characterization of gluten is usually done with a flour sample. For the characterization of gluten various methods are available. ELISA, isoelectric focusing (IEF), gas chromatography (GC), gel electrophoresis approaches (GE), reversed phase liquid chromatography (RPLC), size exclusion chromatography (SEC) and capillary electrophoresis (CE) have all been employed to study purified gluten.1, 3 _{Of these methods, with the current experimental data available, GE, RPLC and SEC seem}

the most reliable and promising for the future. For that reason, these methods will be discussed in more detail.19

4.1 Sample preparation

To determine the type of gluten present in flour, the prolamins are characterized. This is done because they are different for each type of cereal. The easiest way to separate prolamins from glutenins is by using different solvents. Prolamins dissolve very well in alcohol while glutenins do not. Using alcohol in an extraction method is thus a good way to extract the prolamins from the flour after which these alcohol fractions can be further used for characterization.

If the aim is to characterize glutenins, the extraction of prolamins has to be followed by an extraction of the glutenins with 50% propanol with 1% DTT. This has to be done because otherwise the glutenins co-elute with the other prolamins. 3, 19_{These fractions can subsequently be analyzed to}

get information about the intact proteins. Knowing the exact amount of glutenins present in flour is also useful as glutenins are usually the glutenpart that is toxic.

In some cases, proteins need to be enzymatically digested to be analyzed. This is usually the case if detection is done by MS as analysis of intact gluten proteins seemed to be insufficient for identification by peptide-mapping. If enzymatic digestion is necessary, before analysis the protein extracts should be heated at 80 o_{C for 2 minutes to have them stabilized and the stabilized extracts}

need to be enzymatically digested with chymotrypsin.3, 18

4.2 Gel electrophoresis

Agarose Polyacrylamide Gel Electrophoresis (A-PAGE) and Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE) are two of the methods used for separation of gluten which are quite alike. But where A-PAGE separation is based on differences in protein charge density, SDS-PAGE is based on separation by size.3_{Since proteins can have a lot of different charges,}

A-PAGE should be a good method to analyze them. However, gluten have an unusual scarcity of negatively charged residue, making A-PAGE a less preferable method,29_{whereas SDS-PAGE is}

widely used.

SDS-PAGE is a separation method that separates proteins based on their size.30 _{The function}

of the SDS is to denaturate the proteins and contain the proteins in the wells of the gel. To separate the proteins, first the gel needs to be prepared, preferable a day before the measurements. Then the buffer with SDS is loaded in the wells of a polyacrylamide gel. At the day of the measurements, the samples also need to be prepared. Sample preparation for SDS-PAGE is different from the general sample preparation as discussed above as it separates intact proteins instead of digested proteins. This sample preparation consists of three steps: cell disruption, inactivation or removal of interfering substances, and solubilization of the proteins. Cell disruption is mostly done using reducing agents that block the forming of disulfide bonds between cysteines. The removal of interfering substances is

(15)

15 usually done by extraction and solubility enhancers are used to solubilize the proteins. After the samples are prepared they are loaded in the wells with the buffers. Together with the samples, a protein standard is also loaded to be able to estimate the sizes of the proteins. When the samples are loaded an electrical charge is put on the gel causing the proteins to move through the gel. Because the SDS is bound to the protein in a fixed ratio based on the size of the protein, the negative charge that SDS gives to the protein is proportional to its size. The electrical charge that is put on the gel causes the proteins with more negative charges and thus a bigger size to move further through the gel than proteins with less negative charges which are smaller. This separates the proteins based on size. After separation, proteins are stained with a dye to make them visible in the gel.30

The results of the separation of different types of wheat gluten by SDS-PAGE can be found in figure 4a. As can be seen in the figure, all three types of tested wheat gluten (Bagou, Elpa and Genoveva) show clear bands for the HMW-GS with a different combination of bands for each individual type making it possible to distinguish the types by looking at the bands with their respective molecular weights and intensities in the HMW-GS region (>75 kDa). What can also be noticed from the figure is that the LMW-GS tend to co-migrate in the gel as the bands below 75 kDa are close together. 31

Similarly, for other types of flour, different prolamins were identified using SDS-PAGE (figure 4b).32, 33 _{As can be seen, each type of prolamin has its own distinctive pattern of bands. The}

bands of hordein (lane a), secalin (lane b), oat globulins (lane c) and wheat gliadins (lane h) are mostly present in the HMW-region (>29,000 kDa) while both zeins (lane d and e), millet prolamins (lane f) and sorghum prolamins (lane g) have bands in the LMW region.

Concluded from these results, SDS-PAGE is a good method to separate HMW-GS, but a poor method to separate LMW-GS and that it can be used to distinguish between different types of prolamins.

Figure 4. a. Results of gluten separation by SDS-PAGE. The SDS-PAGE pattern shows clear lines for the Bagou wheat for 108 kDa (2), 99 kDa (6.5), 88 kDa (7.5) and 80 kDa (12). There are also clear lines shown for the Elpa wheat for 113 kDa (1), 105 kDa (5), 100 kDa (6), 86 kDa (8) and 82 kDa (10) as well as for the Genoveva wheat for 105 kDa (5), 98 kDa (7), 83 kDa (9) and 82 kDa (10).31_{b. SDS-PAGE of different cereals. The proteins studied are indicated by arrows} and given in parentheses in order of increasing mobility. a, Hordein (D hordein, C2 hordein, C1 hordein, B3 hordein, B1 hordein); b, secalin (HMW secalin, 75 k 𝛾-secalin 1, 75 k 𝛾-secalin 2, 𝜔-secalin, 40 k 𝛾-secalin); c, oat globulins (large

subunit 0f 12S globulin); d, zein-2 (Z15, Z10); e, Zein I (Z22, Z19); f, millet prolamins (band 1, band 2); g; sorhum prolamins (band 1, band 2, band 3); h, wheat prolamins (HMW subunit 5, HMW subunit 10; 𝜔-gliadin 1, 𝜔-gliadin 2,

S-rich gliadin 1, S-S-rich gliadin 2)33

Potravinarstvo Slovak Journal of Food Sciences

Volume 13 479 No. 1/2019

Figure 1 Comparison of the protein patterns of chosen varieties from chip electrophoresis (a) and SDS-PAGE (b).

Figure 2 Electrophoreogram of chosen wheat variety Genoveva with HMW-GS identification.

Table 1 Determined molecular weights for proteins extracted by single-grain extraction by means of SDS-PAGE and Lab-on-chip analysis.

HMW – GS SDS – PAGE (kDa) LOC (kDa)

1 113 206 – 212 2 108 214 – 216 2* 108 205 5 105 216 – 224 6 100 180 6.5 99 159 7 98 169 – 188 17 90 171 18 89 138 7.5 88 146 8 86 148 – 153 9 83 130 10 82 126 – 134 12 80 129 – 131

MOLECULAR WEIGHTS OF CEREAL PROLAMINS BY SDS-PAGE

-137 :> -:> : > -:> :> :> :> :> - 116,000 - 97.400 _ 66 ,0 00 -- 29 .000 :> :> :> - 24,000 :> > - 18.200 :> - 12 ,3 0 0 :> a b c d e f g h

FIGURE 2. Polyacr ylamide gel electrophoresis, in the presence of sodium dodecyl sulpha te, of prolamin fractions from different cereals using the Tria/borate gelsystem. The protein s studied are indicated by arrows and given in parentheses in order of increasing mobilit y. a, Horde in (0 horde in, C2 hordein, CI hordein, B3 hordein, BI hordein) ; b, secalin (H MW secalin, 75 k y-secalin 1, 75 k y-secalin 2, ro-secalin,40k y-secalin); c, oa t glob ulins (large subunit of 128 globulin); d, zein-2 (ZI5, Z lO); e, Zein I (Z22, Z 19);f, millet prolamins (band I, band 2); g, sorghum prolamins (band I, band 2, band 3); h, wheat pro lamins (HMW subunit5.HMW su bunit 10; co-gliadin 1, co-gliadin 2, S-rich gliadin I, S-rich gliadin 2). The positions of M, standards (see Fig. I for details ) are indica ted on the right. Similar separa tions were given by the

other gel systems .

include the HMW subunits of wheat glutenin, HMW secalin of rye and' D' hordein of barley. Although the standard errors of the M rSdetermined for these were high, there were clear differences between the results obtained using th e three gel systems. Thus the Laemmli system gave higherM rS,by 10-20 % , than the Tr is/ b o rate system while the addi tion of urea resulted in a further increase. The total increase va ried from about 25 % to over 40 % . Even the lo werM rSdetermined by the Tris/borate system are considerably gre ater than those from sedimenta tion equilib rium ultra centrifugation. This gave M rS of 54,700 and 67,600 F or ' D' hordein and an HMW secalin band-", whilst theM rSof the HMW su bunits 2 and 12 of whe at glutenin were 69,600 and 63,000, respectively-s-w . There were no signific ant differences between theM rSdetermined by the Laemmli a nd

a

(16)

16 SDS-PAGE is a good method to separate the different gliadins in wheat according to their sizes.31 _It

is also a useful method to identify individual HMW-GS and 𝜔-type gliadins, but 𝛼- and 𝛾-gliadins and LMW-GS have a tendency to co-migrate.19 _{This method is however very time-consuming,}

especially in sample preparation time.31_{It should also be noted that prolamins may migrate}

anomalously in this system, leading to overestimation for molecular masses of prolamins and underestimation for zeins.34_{This can be seen in figure 4b where lane e (zeins) only has one band with}

low intensity that is almost at the same height as the more intense bands for millet (lane f) and sorghum (lane g).

It has however to be mentioned that it was found that prolamins tend to be overestimated and zeins tend to be underestimated with this SDS-PAGE compared to other methods.33_{Because of the}

high amount of proline in the polypeptide, it might be unable to bind an equal amount of SDS per gram compared to other polypeptides. It might also not have the same conformation as other peptides complexed with SDS as proline can hinder structural conformations. Because of this, the intrinsic net charge of the polypeptide does not have a negligible effect on the net charge of the polypeptide-SDS complex. This can lead to over- or underestimations of certain prolamins as the isoelectric effects are not thoroughly excluded.

Interestingly, SDS-PAGE seems to form a good basis lab-on-a-chip approaches. Lab-on-a-chip approaches are less time-consuming and show cleaner results than SDS-PAGE.The lab-on-a-chip method takes 30 min to automatically perform all steps of gel-based electrophoresis (sample separation, staining, destaining, imaging, band detection, data analysis) for a set of 10 samples where SDS-PAGE takes up 30 min to 1 hour to only do the separation step. This can be seen in figure 5 where the lab-on-chip method as well as the regular SDS-PAGE method is used for the same analysis. These results look very promising, but the lab-on-a-chip method still needs some development to be able to replace the current SDS-PAGE method. Compared to the current SDS-PAGE method, the lab-on-a-chip method shows higher molecular weights in the results. This might be explained by a lower electrophoretic mobility of the HMW-GS in the lab-on-a-chip method because of the interaction of the proteins with the capillary walls or hindered mobility of protein aggregates in the miniaturized system. This might also cause a different order of migration in comparison to SDS-PAGE. 31

Not only can different gluten-based proteins be separated, they can also be identified by an immunoblotting procedure.35 _{Moreover, they can be quantified by measuring spot size and color}

intensity or by MS. To quantify the proteins by MS, spots of interest can be cut out of the and are then in-gel digested after which they can be quantified by, for example, MALDI-MS.3_{In a study that}

chose MALDI-TOF MS, measurements were done in linear positive mode after the equipment was calibrated using singly and doubly charged signals of BSA or using creatine phosphokinase chain B from rabbit brain.34, 36 _{It was, however, noted that the spectra showed overlap in MW ranges of}

α/β-gliadins and γ-α/β-gliadins in comparison with the classification in an A-PAGE system.36_MALDI-TOF

(17)

17

Figure 5. Comparison of the protein patterns of chosen varieties from chip electrophoresis (a) and SDS-PAGE (b)31 4.3 Size exclusion chromatography

A method that is reliable and used more and more these days is liquid chromatography (LC). LC methods have the advantage that multiple peptide markers can be combined which leads to an analysis of the source of the type of gluten (i.e. barley, wheat, oats, etc).18 _{LC methods can also be used to}

separate on different protein properties such as size and hydrophobicity. In LCxLC they could even be combined in the same run.3_{Yet another plus is that it is easy to combine MS-detection on-line}

with LC, leading to a higher precision because there are less steps involved that need to be done by humans who are known to be less precise than computers.

In liquid chromatography (LC) a mixture of analytes dissolved in solvent (mobile phase) is injected on a column with a stationary phase. Because of a difference in affinity for the mobile phase and stationary phase analytes get different retentions and it will take some of the analytes longer to pass through the column than other analytes resulting in a separation of analytes. The stationary phase of the column is made of silica, metal oxide particles or polymeric support materials. The types of columns that can be used are cylindrical tubes, flat channel, pillar array or monolith, but a packed bed of porous silica is the most common choice. In high performance liquid chromatography (HPLC) the mobile phase is usually pushed through the column under pressure.

Size exclusion chromatography (SEC) is a form of liquid chromatography where analytes are separated based on their size. In SEC the stationary phase of the column contains porous silica particles with pores of different sizes. Retention in SEC is achieved because small molecules can enter the pores and take a longer route through the column than big molecules that cannot enter the pores and take the shorter route between the particles. This mechanism causes the big molecules to leave the column before the small molecules. Exclusion can be modified by changing the pore sizes of the silica particles.

Before analysis, the extractable and unextractable proteins first need to be separated based on their differences in solubility in alcohol. After separating the extractable from the unextractable proteins, exclusion of gluten from other analytes can be done by SEC.19, 23 _{A typical example of this}

is shown in figure 6. In this example a diode array detection (DAD) with UV absorbance spectroscopy at 210 nm was used as a detector. In this chromatogram LPP are large polymeric proteins, SPP are small polymeric proteins, LMP are large monomeric proteins and SMP are small monomeric proteins.23_{As can be noticed, the amounts of LPP, SPP and LMP are much higher in the extractable}

part than in the unextractable part. While it is hard to see a clear difference between the LPP and SPP in both extractable as well as unextractable parts, the differences between SPP and LMP and the difference between LMP and SMP are clear. This makes SEC a good method to separate gluten based on their size.

Potravinarstvo Slovak Journal of Food Sciences

Volume 13 479 No. 1/2019

Figure 1 Comparison of the protein patterns of chosen varieties from chip electrophoresis (a) and SDS-PAGE (b).

Figure 2 Electrophoreogram of chosen wheat variety Genoveva with HMW-GS identification.

Table 1 Determined molecular weights for proteins extracted by single-grain extraction by means of SDS-PAGE and Lab-on-chip analysis.

HMW – GS SDS – PAGE (kDa) LOC (kDa)

1 113 206 – 212 2 108 214 – 216 2* 108 205 5 105 216 – 224 6 100 180 6.5 99 159 7 98 169 – 188 17 90 171 18 89 138 7.5 88 146 8 86 148 – 153 9 83 130 10 82 126 – 134 12 80 129 – 131

(18)

18

Figure 6. Detection with a DAD at 210 nm of gluten proteins separated with SEC with line (a) represents the proteins after SDS-extraction (extractable part), line (b) represents the proteins after SDS-extraction (unextractable part) and sonication for 30 s and line (c) represents the proteins after SDS-extraction (unextractable part) and sonication for 90 s.23

Besides UV absorbance also fluorescence detection (FLD) can be used to detect gluten by making use of intrinsic protein fluorescence. As is known, the amino acids tyrosine and tryptophan have fluorescent properties and can therefore be used as a sensitive indicator to characterize proteins. The emission maximum for wheat gluten was found the be at 355 nm which is the characteristic fluorescence profile for tryptophan residues in a relatively hydrophobic environment as gluten peptides are known to be.37_{Using fluorescence detection the results in figure 7 were obtained with A}

being the results for gliadin, B for glutenin and C for gluten. In this experiment the gluten free wheat (GfW) was spiked with wheat flour for matrix calibration to obtain 10, 20, 50 and 100 mg gliadin/kg for A, 6, 11, 29, 57 mg glutenin/kg for B and 16, 31, 79, 157 mg gluten/kg for C. Based on the shapes of the peaks, characterization by this method is possible. However, SEC is also a good method for the quantification of gluten. In this experiment, peaks after 12.8 min were not considered to be relevant for gluten quantification. These results give useful data in how to quantify gluten. The obtained peak area can then be relatively compared to the peak area of this experiment and the gluten content of the sample with an unknown amount of gluten can be determined.38

Figure 7. Detection with FLD with chromatogram A being the results for gliadin with GfW spiked with wheat flour to obtain 10, 20, 50 and 100 mg gliadin/kg, B being the results for glutenin with GfW spiked with wheat flour to obtain 6, 11, 29, 57 mg glutenin/kg and C being the results for gluten with GfW spiked with wheat flour to obtain 16, 31, 79, 157 mg gluten/kg 38

P.Nordqvistetal./IndustrialCropsandProducts51 (2013) 51–61 53

0.00 0.40 0.80 AU 1.20 5.00 10.00 15.00 Minutes 20.00 25.00 30.00 (a) (c) (b) LPP SPP LMP SMP

Fig.1.ExampleofSE-HPLCchromatogramsof(a)SDS-extractable,(b)SDS-unextractable(sonication30s),and(c)SDS-unextractable(sonication30+60s)proteins.The chromatogramsaredividedintofoursectionsrepresentinglargepolymericproteins(LPP),smallerpolymericproteins(SPP),largemonomericproteins(LMP),andsmaller monomericproteins(SMP).AU,absorbanceunitsoftheUVdetector.

thepelletsfromthefirstextractionwerere-suspendedinthe SDS-phosphatebuffersolution(1.4ml),sonicatedfor30s(amplitude5) withanultrasonicdisintegrator(SanyoSoniprep150),and there-aftercentrifuged(30min,10,000rpm).Inthethirdextraction,the pelletsfromthesecondextractionweretreatedasthepelletsfrom thefirstextractionbutwiththeexceptionthatthesuspensions weresonicatedfor30+60s.

Thesupernatesfromeachextractionstepwereinjected(20!l) ontoaSE-HPLCPhenomenexcolumn(Biosep-SEC-S4000PEEK) usingaWatersHPLCsystembeingcomprisedofaWaters2690 SeparationsModuleandaWaters996PhotodiodeArrayDetector setat210nm.Separationwasachievedin30minwithaneluent of50%(v/v)acetonitrileandwatercontaining0.1%trifluoroacetic acidatanisocraticflowrateof0.2ml/min.

Thechromatogramsweredividedintofoursections(Fig.1): largepolymericproteins(LPP),smallerpolymericproteins(SPP), largemonomericproteins(LMP),andsmallermonomericproteins (SMP)(Johanssonetal.,2001;Kuktaiteetal.,2004).Larroqueetal. (1996)haveearlierpresentedthemolarmassrangesoftheSE-HPLC chromatograms.LPPandSPPconsistmainlyofglutenins,whileLMP andSMPconsistmainlyofgliadins,peptidesandaminoacids.The areasofthedifferentsectionswereintegrated.

2.5. Amountofextractedprotein

Todeterminethetotalpercentageofextractedproteinofeach sample,thenitrogencontentsoftheoriginalsampleandthepellet obtainedafterthethreeextractionstepsweredetermined accord-ingtotheDumasmethodusingaCarloErbaNA1500elemental analyzer(CarloErbaStrumentazione,Milan,Italy).Theamount ofnitrogenoftheextractedproteiniscalculated asthe differ-encebetweentheamountofnitrogenoftheoriginalsampleand theamountofnitrogenofthepellet.Thepercentageofextracted proteinofthesampleisthenobtainedfromtheratiobetween theamountofnitrogenoftheextractedproteinandtheamount ofnitrogenoftheoriginalsample.Thedeterminationswere con-ductedinduplicate.

2.6. 13_C_NMR_analysis

UntreatedWGandsomeof thetreatedsamples(WG-dh-0, WG-dh-0.8,WG-dh-5.5,WG-50-1h,WG-90-1h,WG-90-1h-s, WG-90-8h,WG-90-24h;detailedinTables1and2)wereanalyzedby13_C

NMR.Thesamplesweredispersedin0.1MNaOH(approximately 10–20%proteinpowder(w/w))andthendilutedwithD2O

(approx-imately6–12%proteinpowder,and37–50%D2Oineachsolution

(w/w)).Theproton-decoupled13_C_NMR_spectra_were_recorded_at

26◦_C_on_a_Bruker_Avance₄₀₀_MHz_instrument_equipped_with_a

10mmprobe.A45◦_pulse,_a_delay_of₅_s_between_pulses,_inversed

gateddecoupling,andeither6000(approximately11h)or10,000 (approximately18h)scans,wereused.Alinebroadeningof5Hz wasusedbeforetheFourierTransformofthefreeinductiondecay. Dioxane,67.4ppm,wasusedasanexternalreferenceforthe chem-icalshiftscale.

2.7. PreparationofdispersionsoftreatedWGandwood substrates,andevaluationthereof

TheenzymaticallyhydrolyzedorheattreatedWGsamples,and untreatedWG,weredispersedina sodiumhydroxidesolution (0.1M;pH13)toaconcentrationof23%(w/w)and46%(w/w)for thehydrolyzedsamples,20%(w/w)fortheheattreatedsamples, and20%(w/w)and23%(w/w)foruntreatedWG.Theviscosityof thedispersionswasmeasuredwithaBrookfieldViscometer(Model DV-II+Pro,VWRInternationalAB,Stockholm,Sweden)theday afterthedispersionswereprepared.Thepresettimewasusedto eliminatetimeeffects.

Theonedayolddispersionswereusedtobondtogethertwo panelsofbeech.Ononesideofeachpanel,180g/m2_of

disper-sionwasapplied.Apresstemperatureof110◦_C,_a_press_time_of

15min,andapressureof0.7MPawereused.Thebondedpanels wereconditionedandevaluatedaccordingtoslightlymodified ver-sionsoftheEuropeanStandardsEN204andEN205(2001).They werebondedtogetherwithwhataccordingtothestandardsis clas-sifiedasathinbond-line(adhesivelayer0.1mmthick).Thewood substrateswerecutintotestpieces,whichweretreated accord-ingtotheconditioningsequencesshowninTable3.Asummaryof theminimumvaluesofbondstrengththatmustbereachedforthe classificationofthermoplasticadhesivesintothedurabilityclasses D1toD3isalsopresentedinTable3.

Thelengthofthetestpieceswas100mminsteadof150mm, whichisthestandard(EN205).Tentestpiecesweretestedfor

Table3

Conditioningsequencesandminimumvaluesofadhesivestrengthforthin bond-lines.

Conditioningsequences Durationandcondition

Adhesive strength(MPa) Durabilityclasses 7Daysa_in_standard atmosphereb ≥10 D1,D2,andD3 7Daysa_in_standard atmosphereb_,₃_h_in_water at(20±5)◦C,7daysa_in standardatmosphereb ≥8 D2 7Daysa_in_standard atmosphereb_,₄_daysa_in waterat(20±5)◦_C ≥2 D3 a₁_Day₌₂₄_h.

b₍₂₀_±₂₎◦Cand(65±5)%relativehumidity.

3.7% ωb-gliadins, 33.8% HMW-, and 62.5% LMW-GS. In com-parison, FLD277/345was particularly suitable for the detection of

HMW-GS, resulting in 2.2% ωb-gliadins, 55.2% HMW-, and 36.4% LMW-GS (Figure 2D). HMW-GS contain the highest tyrosine amounts (5.1−6.4 mol%)51_{of all gluten protein types} and were thus detected with high sensitivity. Corresponding observations were made using GP-HPLC. The shape of the gliadin peak was similar both by DAD210 and FLD277/345

(Figure 3), but HMW-GS (ﬁrst peak in glutenin extract, retention time 6.5−7.8 min) were detected with higher sensi-tivity than LMW-GS (second peak in glutenin extract, retention time 7.8−12.8 min) by FLD277/345compared to DAD210. The

DAD210 chromatogram yielded 17.8% HMW- and 82.2%

LMW-GS (Figure 3A), whereas the distribution changed to 36.0% HMW- and 64.0% LMW-GS by FLD277/345(Figure 3B).

Because interfering peaks from the starch matrix were detected in RP-HPLC-FLD in the retention time window of 18−23 min, thus coeluting with gluten proteins (not shown), GP-HPLC-FLD was chosen for all further analyses of WSt samples. To compare the sensitivity of FLD277/345 vs DAD210, linear

dilu-tions of PWG-gliadin and gluten soludilu-tions (0.0025−0.5 μg) were injected (n = 3), detected by both FLD and DAD in one run, and the peak areas compared using the most sensi-tive photomultiplier voltage setting (superhigh) of the FLD. While the LOD of the FLD was at 0.005 μg PWG-gliadin or 0.0025 μg gluten, the LOD of the DAD was much higher at 0.1 μg PWG-gliadin or gluten. The comparison of peak areas at 0.25 μg, where both detectors showed a clear signal, revealed a 36-fold (PWG-gliadin) and 113-fold (gluten) higher sensitivity of FLD compared to DAD. The diﬀerence in sensitivity between PWG-gliadin and gluten may also be explained by the highly sensitive detection of HMW-GS within gluten due to their high tyrosine content. Having ascertained the general suitability of the GP-HPLC-FLD method to detect gluten proteins with better sensitivity compared to UV detection, it was subsequently applied to the quantitation of gliadin, glu-tenin, and gluten contents of WSt.

Identiﬁcation of the Proteins within the Peak (Retention Time 6.5−12.8 min) of a Reduced Protein Extract of GfW4 by LC-MS. GfW4 was GF by sandwich R5 ELISA (gluten content 12.2 ± 1.1 mg/kg), but the GP-HPLC-FLD method yielded a higher gluten content of 44.9 ± 1.0 mg/kg (see below). To ascertain that the “gluten” peak (6.5−12.8 min) used for quantitation really contained gluten, not only inferred from similarity of retention times (Figures 3 and 4), the proteins contained within this peak were collected from several GP-HPLC runs and identiﬁed by LC-MS after chymotryptic digestion.40 All compounds with retention times beyond 12.8 min were not considered to be relevant for gluten quan-titation, because of low molecular weights (Mr< 30 000) and

nondose-dependent behavior (Figure 4). The majority of pep-tides in the gluten extract of GfW4 were derived from LMW-GS, but peptides from α-gliadins, HMW-GS and starch synthases were identified as well (Table 1). LMW-GS were assigned as fi_{rst protein hit by the database in most cases, but many of} these peptides may be found both in LMW-GS and γ-gliadins (e.g., SIILQEQQQGF also in γ-gliadin P04729.1), because these gluten protein types share similar sequence sections.2_The proteins corresponding to the identified peptides had Mrin the

range from ≈30 000 to ≈86 000 as expected on the basis of GP-HPLC retention times. Some peptide sequences with single amino acid exchanges (e.g., GQQPQQQKL, GKQPQQQQL, GQQPEQQQL, or LQPHKIAQL or VQQQLPVVQPSIL or

SQQQQPVIPQQPSF; exchanged position in bold) were detected as well. Most of the peptides seemed to be typical of HMW-GS, LMW-GS/γ-gliadins, and α-gliadins, because they matched 4 up to 572 other UniProtKB database entries and had high protein scores calculated by the Mascot software. Several protein sequences had two or more matching peptides (e.g., the peptides LQPHQIAQL, GQQPQQQQL, VLPQQQIPF, and SHHQQQQPIQQQPQPF were all found within LMW-GS ACA63873.1). Four peptides contained HLA-DQ2-mediated CD-immunogenic epitopes (underlined or in italics inTable 1), but since the LC-MS method employed only provided qualita-tive results, no estimation of peptide quantities was attempted. These ﬁndings are in accordance with earlier studies that reported the presence of a multitude of proteins in commercial Figure 4. GP-HPLC-FLD chromatograms of gliadin (A), glutenin (B), and gluten (C) extracts of gluten-free GfW4 (Gf Wgf) and the Gf Wgf

samples spiked with wheat ﬂour for matrix calibration to obtain 10, 20, 50, and 100 mg gliadin/kg (Gf Wgf + 10/20/50/100 mg/kg) (A),

corresponding to 6, 11, 29, and 57 mg glutenin/kg (Gf Wgf+ 6/11/

29/57 mg/kg) (B) and to 16, 31, 79, and 157 mg gluten/kg (Gf Wgf+

16/31/79/157 mg/kg) (C). The peak with retention times between 6.5 and 12.8 min was used for quantitation of gliadins, glutenins, and gluten. FLU, ﬂuorescence units.

Journal of Agricultural and Food Chemistry Article

DOI:10.1021/acs.jafc.6b02512

J. Agric. Food Chem. 2016, 64, 7622−7631

7626

3.7% ωb-gliadins, 33.8% HMW-, and 62.5% LMW-GS. In com-parison, FLD277/345was particularly suitable for the detection of HMW-GS, resulting in 2.2% ωb-gliadins, 55.2% HMW-, and 36.4% LMW-GS (Figure 2D). HMW-GS contain the highest tyrosine amounts (5.1−6.4 mol%)51_{of all gluten protein types} and were thus detected with high sensitivity. Corresponding observations were made using GP-HPLC. The shape of the gliadin peak was similar both by DAD210 and FLD277/345 (Figure 3), but HMW-GS (ﬁrst peak in glutenin extract, retention time 6.5−7.8 min) were detected with higher sensi-tivity than LMW-GS (second peak in glutenin extract, retention time 7.8−12.8 min) by FLD277/345compared to DAD210. The DAD210 chromatogram yielded 17.8% HMW- and 82.2% LMW-GS (Figure 3A), whereas the distribution changed to 36.0% HMW- and 64.0% LMW-GS by FLD277/345(Figure 3B). Because interfering peaks from the starch matrix were detected in RP-HPLC-FLD in the retention time window of 18−23 min, thus coeluting with gluten proteins (not shown), GP-HPLC-FLD was chosen for all further analyses of WSt samples. To compare the sensitivity of FLD277/345 vs DAD210, linear dilu-tions of PWG-gliadin and gluten soludilu-tions (0.0025−0.5 μg) were injected (n = 3), detected by both FLD and DAD in one run, and the peak areas compared using the most sensi-tive photomultiplier voltage setting (superhigh) of the FLD. While the LOD of the FLD was at 0.005 μg PWG-gliadin or 0.0025 μg gluten, the LOD of the DAD was much higher at 0.1 μg PWG-gliadin or gluten. The comparison of peak areas at 0.25 μg, where both detectors showed a clear signal, revealed a 36-fold (PWG-gliadin) and 113-fold (gluten) higher sensitivity of FLD compared to DAD. The diﬀerence in sensitivity between PWG-gliadin and gluten may also be explained by the highly sensitive detection of HMW-GS within gluten due to their high tyrosine content. Having ascertained the general suitability of the GP-HPLC-FLD method to detect gluten proteins with better sensitivity compared to UV detection, it was subsequently applied to the quantitation of gliadin, glu-tenin, and gluten contents of WSt.

Identification of the Proteins within the Peak (Retention Time 6.5−12.8 min) of a Reduced Protein Extract of GfW4 by LC-MS. GfW4 was GF by sandwich R5 ELISA (gluten content 12.2 ± 1.1 mg/kg), but the GP-HPLC-FLD method yielded a higher gluten content of 44.9 ± 1.0 mg/kg (see below). To ascertain that the “gluten” peak (6.5−12.8 min) used for quantitation really contained gluten, not only inferred from similarity of retention times (Figures 3 and 4), the proteins contained within this peak were collected from several GP-HPLC runs and identified by LC-MS after chymotryptic digestion.40 All compounds with retention times beyond 12.8 min were not considered to be relevant for gluten quan-titation, because of low molecular weights (Mr< 30 000) and nondose-dependent behavior (Figure 4). The majority of pep-tides in the gluten extract of GfW4 were derived from LMW-GS, but peptides from α-gliadins, HMW-GS and starch synthases were identified as well (Table 1). LMW-GS were assigned as fi_{rst protein hit by the database in most cases, but many of} these peptides may be found both in LMW-GS and γ-gliadins (e.g., SIILQEQQQGF also in γ-gliadin P04729.1), because these gluten protein types share similar sequence sections.2_The proteins corresponding to the identified peptides had Mrin the range from ≈30 000 to ≈86 000 as expected on the basis of GP-HPLC retention times. Some peptide sequences with single amino acid exchanges (e.g., GQQPQQQKL, GKQPQQQQL, GQQPEQQQL, or LQPHKIAQL or VQQQLPVVQPSIL or

samples spiked with wheat ﬂour for matrix calibration to obtain 10, 20, 50, and 100 mg gliadin/kg (Gf Wgf + 10/20/50/100 mg/kg) (A),

DOI:10.1021/acs.jafc.6b02512

J. Agric. Food Chem. 2016, 64, 7622−7631

7626 3.7% ωb-gliadins, 33.8% HMW-, and 62.5% LMW-GS. In

com-parison, FLD277/345was particularly suitable for the detection of

HMW-GS, resulting in 2.2% ωb-gliadins, 55.2% HMW-, and 36.4% LMW-GS (Figure 2D). HMW-GS contain the highest tyrosine amounts (5.1−6.4 mol%)51_{of all gluten protein types} and were thus detected with high sensitivity. Corresponding observations were made using GP-HPLC. The shape of the gliadin peak was similar both by DAD210 and FLD277/345

(Figure 3), but HMW-GS (ﬁrst peak in glutenin extract,

retention time 6.5−7.8 min) were detected with higher sensi-tivity than LMW-GS (second peak in glutenin extract, retention time 7.8−12.8 min) by FLD277/345compared to DAD210. The

DAD210 chromatogram yielded 17.8% HMW- and 82.2%

LMW-GS (Figure 3A), whereas the distribution changed to 36.0% HMW- and 64.0% LMW-GS by FLD277/345(Figure 3B).

Because interfering peaks from the starch matrix were detected in RP-HPLC-FLD in the retention time window of 18−23 min, thus coeluting with gluten proteins (not shown), GP-HPLC-FLD was chosen for all further analyses of WSt samples. To compare the sensitivity of FLD277/345vs DAD210, linear

dilu-tions of PWG-gliadin and gluten soludilu-tions (0.0025−0.5 μg) were injected (n = 3), detected by both FLD and DAD in one run, and the peak areas compared using the most sensi-tive photomultiplier voltage setting (superhigh) of the FLD. While the LOD of the FLD was at 0.005 μg PWG-gliadin or 0.0025 μg gluten, the LOD of the DAD was much higher at 0.1 μg PWG-gliadin or gluten. The comparison of peak areas at 0.25 μg, where both detectors showed a clear signal, revealed a 36-fold (PWG-gliadin) and 113-fold (gluten) higher sensitivity of FLD compared to DAD. The diﬀerence in sensitivity between PWG-gliadin and gluten may also be explained by the highly sensitive detection of HMW-GS within gluten due to their high tyrosine content. Having ascertained the general suitability of the GP-HPLC-FLD method to detect gluten proteins with better sensitivity compared to UV detection, it was subsequently applied to the quantitation of gliadin, glu-tenin, and gluten contents of WSt.

Identiﬁcation of the Proteins within the Peak (Retention Time 6.5−12.8 min) of a Reduced Protein Extract of GfW4 by LC-MS. GfW4 was GF by sandwich R5 ELISA (gluten content 12.2 ± 1.1 mg/kg), but the GP-HPLC-FLD method yielded a higher gluten content of 44.9 ± 1.0 mg/kg (see below). To ascertain that the “gluten” peak (6.5−12.8 min) used for quantitation really contained gluten, not only inferred from similarity of retention times (Figures 3 and 4), the proteins contained within this peak were collected from several GP-HPLC runs and identiﬁed by LC-MS after chymotryptic digestion.40 _{All compounds with retention times beyond} 12.8 min were not considered to be relevant for gluten quan-titation, because of low molecular weights (Mr< 30 000) and

nondose-dependent behavior (Figure 4). The majority of pep-tides in the gluten extract of GfW4 were derived from LMW-GS, but peptides from α-gliadins, HMW-GS and starch synthases were identified as well (Table 1). LMW-GS were assigned as fi_{rst protein hit by the database in most cases, but many of} these peptides may be found both in LMW-GS and γ-gliadins (e.g., SIILQEQQQGF also in γ-gliadin P04729.1), because these gluten protein types share similar sequence sections.2_The proteins corresponding to the identified peptides had Mrin the

range from≈30 000 to ≈86 000 as expected on the basis of GP-HPLC retention times. Some peptide sequences with single amino acid exchanges (e.g., GQQPQQQKL, GKQPQQQQL, GQQPEQQQL, or LQPHKIAQL or VQQQLPVVQPSIL or

samples spiked with wheat ﬂour for matrix calibration to obtain 10, 20, 50, and 100 mg gliadin/kg (Gf Wgf+ 10/20/50/100 mg/kg) (A),

DOI:10.1021/acs.jafc.6b02512 J. Agric. Food Chem. 2016, 64, 7622−7631

Analyzing gluten in food

MSc Chemistry

Analytical sciences

Analyzing gluten in food

by

March 2020

12 ects

February 2020 - September 2020

Supervisor/Examiner:

2

reviewer:

Vrije Universiteit - Division of

BioAnalytical Chemistry

Table of contents

Abstract

1. Introduction

2. Gluten and their function

3. Gluten intolerance and sensitivity

4. Gluten characterization in flour

_reviewer: