Gluten Detection by Real-Time PCR: an Alternative for Tracking Processed Foods

Abstract

This work reports a real-time PCR assay to specifically detect the presence of gluten in complex food matrices and to carry out an in-silico prospection of primers used in scientific research. The primers used were “tritprglut” and “Planta 18S” (reference gene), which had mean quantification cycle values (Cq) of 34.30 and 16.98, respectively. The real-time PCR protocol was validated in different meats (beef, chicken, pork, horse and lamb) with an average Cq of 25.69. Tests to verify fraud in industrialized foods were carried out with the following products: cereal bars, chocolate, crackers and two types of snacks. All foods complied with the information contained on the label, except for the cereal bar that was identified as “may contain gluten” and had a “high content” concentration (1,925 mg/kg). The LD value was 36 cycles and the LQ was 60 mg/kg, being within the “low content” classification range. The in-silico tests were performed using two software, MFE and NETprimer, and the content parameters GC, Tm (°C), ∆G (kcal/mol), dimer formation and hairpins. The “Wheat-w-Gliadin” primer showed the best average parameters: size= 24 bp; GC= 44%; Tm = 62.5 °C; ∆G= -32.25 kcal/mol; no dimer or hairpin formation; and a maximum primer rating (100). There were differences in results between the software used. The results highlight the potential of the real-time PCR technique in detecting gluten and/or allergens in foods with a complex matrix, such as chocolate and cereal bars tested in this study, proving to be sensitive and robust to detect the presence of potentially high gluten concentrations. harmful for celiac consumers.

Keywords:
gluten screening; allergen; molecular biology; vegetable protein; in-silico

HIGHLIGHTS

Gluten detection by real-time PCR.

Real-time PCR as an alternative in fraud detection.

Food rigged with gluten.

In silico prospecting of primers for gluten detection.

INTRODUCTION

Food allergy is an immune-mediated response that occurs reproducibly after exposure to a particular food, component or ingredient, with more than 200 proven allergenic foods [¹], the most common being cereals containing gluten, shellfish, egg, fish, peanuts, soy, milk and nuts [²], which can trigger gastrointestinal symptoms, respiratory or systemic problems, with a risk to life [³].

In recent years, cereals have gained interest as niche market crops in response to the demand for healthy foods [⁴] and, despite having functional properties, proteins present in grains, such as gliadin (wheat), hordein (barley) and secalin (rye), have been associated with human pathologies such as celiac disease, wheat allergy and non-celiac gluten sensitivity [⁵]. Thus, governments and regulatory agencies have recognized the need to focus allergen labeling regulations on a limited set of “priority allergens” [¹].

In Brazil, the Resolution of the Collegiate Board (RDC) No. 26, of July 2, 2015 [⁶], provides for the requirements for mandatory labeling of the main foods that cause food allergies. In addition, Normative Instruction (IN) No. 75 establishes the technical requirements for declaring nutritional labeling on packaged foods [⁷], requiring the mandatory labeling of ingredients that cause allergies or intolerances, including cereals containing gluten [⁶]. The most recent one, RDC No. 429 of October 8, 2020, also establishes some criteria for labeling foods with allergens [⁸].

As established by the Codex Alimentarius, products labeled “gluten-free” must not contain levels of this protein above 20 mg/kg for safe consumption by patients with celiac disease [²]. This means that analytical methods to detect the allergen need to be sufficiently sensitive, specific, suitable for routine analysis and validated by collaborative studies. Therefore, the authenticity analysis of cereal-based products is necessary to comply with labeling rules, avoid unfair economic competition and protect consumers [⁴,⁹].

To detect and/or quantify gluten, enzyme-linked immunosorbent assay (ELISA) is normally used, however, due to the limitations of the method, such as the possibility of false positive and negative results due to the complexity of the gluten protein and other ingredients present in foods [¹], other non-immunological techniques have also been researched, such as mass spectrometry (MALDI-TOF), sensors (electrochemical, fluorescence, optical, and biosensors), near-infrared spectroscopy coupled with chemometrics and molecular ones, based on PCR (Polymerase Chain Reaction) [⁵,¹⁰-¹⁵], ddPCR (droplet digital polymerase chain reaction) and LAMP (loop-mediated isothermal amplification) [¹]. We must not fail to consider that the world lacks rapid, sensitive, more specific and easy-to-use methods to detect gluten in food samples [¹²].

Identifying the presence of such cereal species in a product is of paramount importance, not only to prevent food safety risks in sensitive or allergic individuals but also to prevent economic fraud [¹⁶]. Two studies carried out in Canada identified the presence of gluten in food products that should not contain this allergen. In the first study, carried out in 2011 using ELISA, high levels of gluten were reported in 82% (n=131) of the oat samples tested [¹⁷]. The subsequent study, carried out two years later, detected the presence of gluten in 32% (n=640) of gluten-free flours sold in retail stores [¹⁸]. In Brazil, a study carried out in 2019 identified gluten contamination, through enzyme immunoassays, in 2.8% (n=180) of traditional Brazilian foods sold in 60 food services, of which 6.7% tested positive for the presence of gluten in meals [¹⁹]. Kryuchenko and coauthors (2021) [²⁰] detected fraud (gluten concentration greater than 20 mg/kg), using PCR and ELISA, in four of the nine samples of gluten-free sausages and flours.

Based on this, the objective was to standardize and validate the real-time PCR (qPCR) technique for tracking gluten in processed foods, as well as to prospect and analyze, in silico, primers with greater detection efficiency and accuracy of results.

MATERIAL AND METHODS

Evaluation and validation of primers

Sample preparation with gluten

To create the gluten model matrix, wheat flour containing gluten was added to different types of meat at a concentration of 50-100 mg/kg (5-10%). Certified reference meats (CRM) from LGC (UK) were purchased for each gluten-free animal species, 100% (w/w) of horse meat (LGC7220), bovine (LGC7221), swine (LGC7222), sheep (LGC7223) and chicken (LGC7224). As a positive control, commercial wheat flour was used.

Gluten detection

The primer sets (Invitrogen - Thermofisher Scientific) used in this study were described by Zeltner et al. (2009) [²¹] and the positive control of plant amplification according to sequences published by Allmann et al. (1993) [²²] (Table 1).

Table 1
Identification of primers, reference genes, forward and reverse sequence and positive control that were used in the experiment for gluten detection.

Name	Reference gene	Sequence	Reference
tritprglut 3 F	HMW Glutenin B1-1	5´-CCTCTTTGCGGCAGTAGTYGT-3´	Zeltner et al. (2009)[21]
tritprglut 3 R		5´-CTCGCRCTCACABTGTAGTTG-3´
Plant amplification control	18S rARN	5´-TCTGCCCTATCAACTTTCGATGGTA-3´	Allmann et al. (1993)[22]
		5´-AATTTGCGCGCCTGCTGCCTTCCTT-3´

DNA extraction

Total DNA was extracted from 0.2 g of meat (intentionally contaminated) using the DNeasy® mericon® Food Kit [²³], according to the manufacturer's instructions. The quantification and purity of the extracted product were evaluated by a UV spectrophotometer (Nanodrop, Thermo Fisher Scientific) and kept under refrigeration (≤8 °C). The extracted DNA was standardized to a final concentration of 10 ng. All samples were extracted in duplicate, including the positive control (pure wheat flour).

Quantitative real-time PCR (qPCR)

Real-time qPCR analysis was performed using the QuantiNova™ SYBR® Green PCR kit and/or SsoFast EvaGreen Supermixes (Bio-Rad). The EvaGreen kit was used to perform the standard curve with flour and in the analysis with processed foods. The reaction mixtures followed the quantities proposed by the manufacturer's labels, with modification in the final volume to 10 µL and DNA at a concentration of 10 ng/PCR. For the reaction, 8 µL of the reaction mixture and 2 µL of the extracted DNA were used. The details of the reaction mixture were followed as described by the manufacturer [²⁴].

As it is a quick start protocol, the QuantiNova™ SYBR® Green PCR kit establishes an initial activation step at 95 °C for 2 minutes, 40 cycles of 95 °C for 5 seconds and a combination of annealing/extension steps at 60 °C for 10 seconds. The experiments were performed with the QuantStudio™ 5 Real-Time PCR System (Thermofisher Scientific) and the qPCR conditions followed those recommended by the kit manufacturer. The standard curve was constructed using DNA extracted from wheat flour at concentrations between 6.1 and 100 000 mg/kg ranging from trace to high gluten content. All analyzes were performed in duplicates.

The results obtained by the QuantStudio™ 5 Real-Time PCR System were evaluated by the Design and Analysis Software 2.6.0 (both Thermo Fisher Scientific). The evaluation of the Cq value of the analysis was carried out, both for samples of meat intentionally defrauded with gluten, and for processed and industrialized foods. The GraphPad Prism program (version 8.2) was used to perform basic and descriptive statistical analysis. Furthermore, a comparative analysis was carried out between the values ​​determined and those established by current legislation.

Evaluation of processed and industrialized foods

A total of five commercial food products were purchased from local businesses in the city of Lajeado (RS) with different characteristics in terms of target cereals, namely: cereal bar (n=1); crackers with salt and water (n=1 positive control); chocolate (n=1); and snacks (n=2). Table 2 includes a brief description of the foods.

Table 2
Description of processed and industrialized products obtained commercially.

Food	Code	Classification *	Ingredients
Cereal bar	BC	May contain gluten	Corn glucose, dehydrated pineapple, rolled oats, almonds, raisins, dehydrated coconut, cashew nuts, Brazil nuts, rice flakes, antioxidant and soy lecithin.
Water and salt biscuit **	BAS	Contains wheat and soy derivatives	Wheat flour enriched with iron and folic acid, vegetable oil, invert sugar, salt, sugar, potassium chloride, starch, baking powders, ammonium and sodium bicarbonate and soy lecithin emulsifier and esters of mono and diglycerides of fatty acids with diacetyl acid tartaric.
Chocolate	CH	Free gluten	Chocolate 41% cocoa: Maltitol natural sweetener, cocoa butter, polydextrose, cocoa mass, soy extract, soy lecithin and INS 476 emulsifiers, flavorings, salt and sucralose artificial sweetener).
Food snacks	SF	Free gluten	Chickpea flour, rice flour, potato starch, corn flour, sunflower oil, salt and natural flavoring.
Cheese snacks	SQ	Free gluten	Non-transgenic corn hominy, vegetable oils (corn, soy and cotton), low sodium salt, yeast extract, natural annatto coloring and aromas.

• ** Food used as a positive control.

In-silico tests

The in-silico analysis was used with the objective of prospecting and evaluating primers with greater potential for use in qPCR in the detection/quantification of gluten. For this, two screening algorithms called MFEprimer-3.1 [²⁵] (https://mfeprimer3.igenetech.com/spec) and NETprimer [²⁶] (http://www.premierbiosoft.com/NetPrimer/) were used. Initially, a search was conducted in electronic scientific databases (Science Direct, PubMed and Google Scholar) for articles and patents in English, Spanish or Portuguese, regardless of the year of publication. Works such as editorials, letters, commentaries, dissertations, theses and book chapters were excluded. The search terms used were “PCR”, “qPCR” and “gluten” and were combined with each other for greater coverage and relevance of the results.

The articles identified through the search strategy were evaluated independently. The initial phase of article selection consisted of analyzing the titles, followed by the abstracts and, finally, the methodologies for surveying the primers used in the studies. Articles that did not describe the primers used in the methodology were excluded from the analysis. Eleven studies published between 2003 and 2019 were selected and 23 sets of primers were analyzed. The basic specificities raised were: GC content (GC%), Melting temperature (Tm °C), initiator free energy (∆G Kcam/mol), staples, dimers and individual classification (rating) based on the calculation: Rating = 100 + (∆G (Dimer)*1.8+∆G(Hairpin)*1.4) [²⁶]. For the presentation of individual classification results, the average between the forward and reverse primers was performed.

RESULTS AND DISCUSSION

Evaluation and validation of primers for detecting gluten in artificially rigged meat

To validate the protocol and analysis conditions in the food matrix, five certified reference meats from different animal species were artificially contaminated with 5% gluten. In all samples the result was positive for the presence of gluten. It was observed that the Cq detection value remained within the same range, with an average value of 25.69 (Table 3) and the positive control (plant system) detected an average Cq value of 11.70 (Figure 1). The proximity of Cq values between the different meat types shows that the food matrix did not interfere in the process of extraction and amplification of gluten DNA, since the physical-chemical composition and the fatty acid profile showed differences between different species [²⁷].

Table 3
Qualitative detection of the glutenin target using different meats (CRM) artificially contaminated with 5% gluten.

Sample ID	Meat matrix	Value Cq
8a	Bovine (LGC7221)	25.71
8b		25.06
9a	Chicken (LGC7224)	25.72
9b		25.83
10a	Swine (LGC7222)	25.56
10b		26.17
11a	Horse (LGC7220)	25.57
11b		25.06
12a	Sheep (LGC7223)	25.92
12b		26.26

By establishing a detection threshold (threshold line) of 0.030, the amplification curve of the systems glutenin (red) and plant positive control (18S) (green) showed similar Cq values, starting practically at the same point, as shown in Figure 1. This is important in the evaluation and interpretation of real-time PCR data, as DNA extracted from meat must be diluted at a low concentration, as high concentrations can impair the functioning of the DNA polymerase enzyme present in the PCR reaction. In this study, the Cq value obtained refers to a total concentration of 20 ng/PCR reaction.

Figure 1
Qualitative detection of DNA extracted from five meats of different animal species, artificially contaminated with 5% gluten, through the mastermix containing SYBR Green.

Food DNA analysis is a tool that can identify the ingredients of the product, reveal the composition of the product, in addition to helping to track possible fraud and adulterations [²⁸]. Therefore, there is a need for a rapid diagnosis for food inspection and health surveillance services.

Analyzes of processed foods by qPCR

The amplification efficiency of the qPCR assay using the “tritprglut” primers, to amplify the glutenin gene, was determined by analyzing the DNA of 100 ng/µL, serially diluted (1:4) in water to a concentration of 0.0015 ng/µL totalizing nine points, in triplicate. One point was removed for the best fit of the curve. From the Cq value obtained, a graph was plotted against the logarithm of the DNA concentration (Figure 2A). The curve showed a linear correlation coefficient (R2 = 0.989), with a slop (slope) of -3,283, corresponding to an amplification efficiency of 101.65%. The standard curve presented considerable parameters to be used in the process of quantifying gluten in processed and industrialized foods by qPCR.

Figure 2
(A) Efficiency of the gluten standard curve obtained from the amplification of the glutenin gene through the mastermix containing SYBR Green (100 ng/uL - 0.00015 ng/uL). (B) Amplification curve of the glutenin target gene from processed and industrialized foods through the mastermix containing SYBR Green. SF= snacks food; SQ= cheese snacks; BC= cereal bar; CH=chocolate; BAS= water and salt biscuit; FA= wheat flour (positive control).

In order to verify and identify fraud in industrialized commercial foods, different processed and industrialized products classified as “gluten-free”, “may contain gluten” and “contains wheat and soy derivatives” were purchased in local stores (Table 2). Figure 2B shows the amplification curves of DNA samples from the different foods tested.

With the qPCR technique it was possible to detect and quantify the concentration of gluten in the analyzed foods. Of the total of 5 food matrices investigated, only one (20%) did not comply with the classification indicated on the label. It can be observed, in Table 4, that the cereal bar had a concentration of 1,925 mg/kg of gluten, thus being classified as “high content”. However, on the label (Table 2) the manufacturer only informs that the product “may contain gluten”. This information can lead the consumer to confusion at the time of purchase, since it can be understood that if gluten is present it would be at a “low content” level (20 to 100 mg/kg), and could even be characterized as fraud. According to the IFS Food Standard (International Food Standard), version 7, October 2020, food fraud is “the intentional substitution, misleading labeling, adulteration or falsification of food raw materials or packaging materials made available on the market for economic gains” [²⁹]. Since the food trade was introduced, food fraud has existed and most have the purpose of obtaining financial benefits through deception and damage to consumers and/or non-compliance or violations of the food law by official control authorities [⁹]. Studies such as that of Tramuta and coauthors [³⁰] have been carried out using real-time PCR to detect the presence of allergens in food and to ensure compliance with allergen labeling regulations by food manufacturers.

A possible justification for the high concentration of gluten in the cereal bar is the presence of laminated oats in its composition. The positivity for gluten can be attributed to cross-contamination during the cultivation, harvesting and processing of cereal grains [⁵]. For this reason, oats were included among the mandatory allergens in European regulations, in addition to the high tolerance of celiac patients, around 80% of this population, to oat prolamins [³¹,³²]. García-García and coauthors (2019) [⁵] reported positive amplification of 19 (79%) oat-containing food samples in a real-time PCR assay and concluded that there is a high proportion of oat-based products showing wheat, barley and/or wheat contamination. or rye. In addition to these, 11 other foods were also detected with the presence of the offending target cereals. Furthermore, Zhao and coauthors [³³] were also able to validate a molecular detection method, real-time fluorescence quantitative PCR, for gluten allergens in infant formulas. The detection showed good specificity, sensitivity and accuracy.

Table 4
Quantitative detection (qPCR) of the glutenin target gene in processed and industrialized foods using a mastermix containing SYBR Green.

Food	Tritprglut
	%	mg/kg	Average mg/kg	Classification *	Labeling
Cereal bar	0,009	90	1.925	high content	nonconforming
	0,376	3.760
Water and salt biscuit	9,057	90.570	105.835	high content	conforming
	12,101	121.100
Chocolate	0,001	10	5	gluten free	conforming
	-	-
Food snacks	-	-	-	gluten free	conforming
	-	-
Cheese snacks	0,003	30	15	gluten free	conforming
	-	-
Flour **	32,514	325.140	287.370	high content	conforming
	24,960	249.600

• *Reference value <20 mg/kg - gluten-free; 20 to 100 mg/kg - low content; >100 mg/kg - high content. **Used for making the curve and as a positive control.

The other foods were in accordance with the classifications described on the product label. All foods with precautionary labeling indicating gluten-free analyzed in this study tested negative. These results, together with those found by García-García and coauthors (2018; 2019) [⁵,³⁴] and Pegels and coauthors (2015) [⁴], reveal that a low percentage of foods identified as allergens actually contain the harmful ingredient, causing unnecessary restrictions in food choice. of sensitive individuals.

From the Cq values of the standard curve, which ranged from 100 ng/µL - 0.00015 ng/µL, the limit of detection (LD) and the limit of quantification (LQ) were determined. LD resulted in a fluorescence signal within a Cq of 36 cycles. The concentration of 0.0015 ng/µL was not considered in the calculation because it had an undetermined Cq. For Druml and coauthors (2015) [³⁵], the limit of 37 cycles is defined in the analytical routine in order to reduce the probability of obtaining false-positive signals due to cross-reaction. Furthermore, Mazzara and coauthors (2008) [³⁶] established in their study a LD95% which is sufficient to detect the presence of the analyte at least 95% of the time, with a probability of false-negative results of only 5%. In the test carried out by García-García and collaborators (2019) [⁵], the detection capacity of up to 10 mg/kg (0.001%) of the target cereals was determined, obtaining Cq values of approximately 32 and 35 cycles.

The LQ determined in the present study was 60 mg/kg, which falls within the low content classification limit. Values below this concentration are not accurate and reliable. Pöpping and Holzhauser (2004) [³⁷] share the information that, for the German Society of Allergology, an adequate analytical method for detecting gluten must have a sensitivity of 10 mg/kg, a value six times lower than that found in this study.

In the search for more accurate results that are closer to the reality of the food, the same test was carried out using the dye Eva Green to verify which dye would be more suitable for detecting gluten. Amplification using this dye as an alternative was not efficient and it was not possible to determine the Cq value. It was verified slope values (slope) of -2.009, correlation coefficient (R2) of 0.604 and amplification efficiency (Eff.) of 214.604%. All parameters were completely outside the ideal values for quantification, leading to the conclusion that the Eva Green dye is not a good option for this type of analysis. In view of these results, it is suggested that the fluorogenic SYBR® Green is the most suitable, compared to Eva Green, for analysis of gluten detection in a complex food matrix. In a study conducted by Andersen and researchers (2006) [³⁸], showed no significant difference in amplification results between four different chemicals, SYBR® Green, Taqman, MGB probe and molecular beacon, used for qPCR reaction.

Prospection and in silico analysis of primers for gluten detection

The detection limit of the “tritprglut” primer was up to “low gluten content” between 20 - 100 mg/kg (unpublished data). For this reason, an in-silico analysis was performed to prospect new sets of oligonucleotides in order to obtain greater sensitivity in detection. For this, 11 publications were surveyed during the prospect of primers for detecting gluten and/or allergens and 23 sets of primer sequences were selected for in silico analysis (Table 5). Even if the primer sequences are obtained from international publications, a prior evaluation is necessary to ensure greater efficiency and precision in DNA-based assays [³⁹], since the primer DNA fragment is unstable and can form intramolecular (hairpins) and intermolecular (self dimer and cross dimer) structures [²⁸].

The primers had an average size (forward/reverse) of 22 base pairs (bp), with three sets of primers with an average size between 18-19 bp (a2g; type 2; ITS2) that differed from WBR with the highest average of 26 sc. The length of DNA oligomers, which must be between 18 and 30 nucleotides to have good annealing efficiency to the target DNA, can influence the formation of hairpins that affect the PCR reaction [²⁸]. Analyzing the primers individually, all were within the length indicated by Fröder [²⁸], ranging from 18 to 30 bp between them.

Furthermore, the individual primers were classified according to the data provided by the NETprimer algorithm (Table 5) which selects primers to generate a single PCR product [⁴⁰]. The predicted efficiency of a primer sequence can be ranked quickly and makes it possible to screen the closest DNA oligomers for practical application. The higher the classification of a primer, the greater its amplification efficiency [²⁶]. Of the 23 sets of oligonucleotides analyzed, “Wheat-w-Gliadin” [⁴¹] presented the maximum classification of 100, followed by two others with average values of 96, namely “α2-gliadin” [⁴²] and “Wheat, barley and rye” [⁴³] and the lowest classification (77) for the primer “Agglutinin isolectin” [⁴⁴].

ISO/TS 20224 (2006) [⁴⁵] advises that the primer design should take into account the GC:AT ratio of 50:50, or the closest to this ratio, and the absence of concentration of Gs and Cs bases in short segments of primers due to high internal stability. The average GC content was 49%, varying between 36 (Barley system 1 - HV) and 59% (Agglutinin isolectin - Tri18 and Barley system 2 - hor) considering the set of primers (forward/reverse). In the individual analysis of the GC content, 13 of them were outside the ideal range, which would be between 40 and 60%. In addition, the ideal Tm should be between 60 and 65 °C [⁴⁶], so that the high negative value of Gibbs free energy (ΔG), required to break the secondary structure, indicates stable hairpins and undesirable. Tm values were close between the two softwares, in all sets of primers, presenting an average value of 60.7 and 61.0 °C for MFE and NETprimer, respectively, and not statistically different. The opposite was observed for the ∆G values that were discrepant between the algorithms, being -23.4 and -37.3 kcal/mol, respectively (p<0.05).

NETprimer presented a limitation in the analysis of degenerate primers because it was unable to evaluate them, which did not occur with MFE. A difference was also observed, between the two algorithms, in the detection of dimer formation, being identified only by the MFE in the primers “Tri25-F” and “Pr.GluteninR” (Figure 3). Both formed self-dimers (Table 5). These secondary competitive interactions with other pairs of primers or template DNA decrease the efficiency of PCR [⁴⁷]. Furthermore, the primer fragments that signaled the possibility of forming dimers can be an option to be used with probes, since they only produce a fluorescent signal if they have specific hybridization with the target DNA sequence. In this way, primer dimers or other nonspecific amplification products do not produce a signal [⁴⁸]. The presence of hairpin formation was not observed in either software.

Table 5
In silico analysis of primers used in gluten studies.

Reference	Nucleotide sequence *	Identification*	Size (bp)	Classification **	Specificity
					GC (%)		Tm (°C)		∆G (kcal/mol)
					MFE	NETpr	MFE	NETpr	MFE	NETpr
Martín-Fernández et al., 2015[42]	α2-gliadin	Gly-F	21	100	57,1	57,1	65,4	70,5	-25,3	-42,9
		Gly-R	25	92	48,0	48,0	65,0	68,8	-27,7	-43,65
	Agglutinin isolectin	Tri18-F	23	78	52,2	52,1	62,7	62,2	-24,7	-38,6
		Tri18-R	24	89	45,8	45,8	63,3	60,9	-24,1	-38,3
	Thioredoxin h	Tri25-F	24	84	45,8	45,8	63,3	64,6	-25,6	-39,2
		Tri25-R	21	81	57,1	57,1	63,0	63,4	-35,6	-38,2
Martín-Fernandez et al., 2016[44]	Agglutinin isolectin	Tri18-FS	23	93	65,0	65,0	64,8	65,6	-24,0	-39,5
		Tri18-RS	24	61	52,7	52,1	63,1	65,7	-25,3	-38,9
García-García et al., 2019[5]	ITS2	WBR-ITSfw	18	88	61,1	61,1	61,6	60,6	-21,3	-34,1
		WBR-ITSrev	20	-	50,0	-	59,4	-	-21,9	-
Garrido-Maestu et al., 2018[49]	-	a2g-F3	19	91	52,6	52,6	59,2	58,6	-20,9	-33,4
		a2g-B3	18	86	55,5	55,5	57,1	54,6	-19,8	-30,5
		a2g-F	22	81	63,6	63,6	65,6	65,7	-25,8	-39,2
		a2g-R	23	83	52,1	52,1	66,4	70,1	-26,7	-42,56
Henterich et al., 2003[50]	-	Type 1F	22	77	45,4	45,4	59,9	59,7	-22,8	-35,8
		Type 1R	24	81	37,5	37,5	59,4	61,2	-23,5	-38,5
		Type 2F	19	100	63,1	63,1	63,0	60,5	-22,8	-34,3
		Type 2R	18	-	-	-	-	-	-	-
Mujico et al., 2008[51]	-	WBRF8	30	85	20,0	20,0	56,2	62,2	-24,2	-46,0
		WBRR8	23	89	43,4	43,4	59,7	60,8	-23,2	-37,9
Mujico et al., 2011[48]	-	W.F1	19	86	57,8	57,8	60,5	59,7	-21,9	-33,5
		W.R1	20	87	45,0	45,0	59,6	59,6	-21,5	-35,9
Piknová et al., 2008[43]	Wheat, barley e rye	pinbF	20	100	55,0	55,0	61,3	59,5	-22,3	-34,6
		pinbR2	20	92	45,0	45,0	56,9	56,4	-20,5	-32,5
	Eukaryotic	TR03	25	87	44,0	44,0	62,1	64,3	-25,4	-41,8
		TR04	25	70	56,0	56,0	70,9	76,1	-30,9	-52,65
Pinto et al., 2014[52]	Biotin	F	21	92	57,1	57,1	61,4	58,9	-23,2	-35,2
		R	19	86	47,3	47,3	53,1	49,3	-18,3	-29,4
Sandberg et al., 2003[41]	Wheat - w-Gliadin	TAG2315F	25	100	44,0	44,0	61,1	63,9	-25,3	-41,5
		TAG2473R	23	100	47,8	47,8	61,1	62,4	-23,8	-38,5
	Rye - w-Secalin	SEC2599F	26	90	34,6	34,6	59,8	64,2	-24,9	-42,5
		SEC2756R	24	92	50,0	50,0	62,4	63,0	-25,1	-39,6
	Barley - Hordein	HVH1618F	24	85	33,3	33,3	56,8	57,5	-21,9	-37,6
		HVH1763R	19	90	47,3	47,3	55,8	54,7	-19,7	-31,0
	Oats - Avenin	ASA1097F	19	91	52,6	52,6	57,3	53,7	-20,2	-31,2
		ASA1177R	24	100	29,1	29,1	54,9	56,4	-21,1	-37,8
Zeltner et al., 2009[21]	Glutenin system	Pr.GluteninF	21	100	38,1	38,1	53,9	48,4	-19,3	-29,5
		Pr.GluteninR	19	71	63,1	63,1	60,9	60,3	-21,5	-35,9
	Glutenin system	tritprglut	21	-	52,3	-	61,4	-	-23,1	-
		tritprglut	21	-	57,1	-	62,0	-	-23,9	-
	Barley system 1	HVF	25	89	36,0	36,0	57,6	61,29	-23,0	-41,2
		HVR	25	-	36,0	-	57,6	-	-22,9	-
	Barley system 2	hor1	19	87	57,8	57,8	60,9	59,1	-22,0	-33,5
		hor2	20	100	60,0	60,0	65,6	60,1	-23,2	-34,0
	Oat system	OATF	20	91	55,0	55,0	60,6	59,2	-22,2	-34,8
		OATR	19	82	52,6	52,6	58,7	58,7	-20,9	-33,3
Medium values			22	88	49,4	49,4	60,7	61,0	-23,4	-37,3

• * Data taken from articles. **Data taken from NETprimer software - Input primer sequences error. nf: It didn't. Rating is based on the calculation Rating = 100 + (∆G (Dimer) * 1.8 + ∆G (Hairpin) * 1.4)(NETPRIMER, 2022)[²⁶]. MFE: MFEprimer - https://mfeprimer3.igenetech.com/spec. NETpr: NETprimer - http://www.premierbiosoft.com/NetPrimer/AnalyzePrimerServlet .

Figure 3
Schematic representation of the dimers formed by the forward Tri25-F and reverse PR.GluteninR primers.

Considering the characteristics of the 23 pairs of primers determined by the two software used in this study, it can be inferred that after screening, the set of primers that would have greater efficiency and accuracy in the amplification and detection of gluten and/or allergens is the “Wheat-w- Gliadin” - TAG2315F and TAG2473R, described in the work by Sandberg and coauthors (2003) [⁴¹]. It presented values of GC, Tm, ∆G and pb content within the optimal limits for the real-time PCR technique, as well as the highest classification (100) (Table 5). In addition, the possibility of formation of secondary structures such as dimers and staples, which made it have the maximum classification value, was not observed, since the calculation takes these two aspects into account.

Vallone and Butler (2004) [⁴⁷] state that selecting short sequences of DNA oligonucleotides that have the ability to bind only to their intended target is of great importance in the development of detection technologies based on nucleic acids. Another oligonucleotide pair that also showed potential for use in PCR is the “Agglutinin isolectin” - Tri18-F and Tri18-R [⁴²] which presented excellent parameters, however, obtained a lower classification (83,5) to the one mentioned above. The ability and possibility to quickly screen sets of potential oligomers for their cross-reactivity reduces overall assay development time [⁴⁷] and ensures greater reliability of results. In short, the use of suitable software is the ideal strategy to automate the steps in PCR and minimize errors [²⁸].

CONCLUSION

The results obtained in this study confirm the sensitivity and robustness of the real-time PCR technique to detect, in foods, the presence of potentially harmful concentrations of gluten-containing cereals, supporting the results of immunological assays currently used for this purpose. In addition, it is possible to highlight the possibility of implementing real-time PCR as a complementary method to the traditional ones (immunological), in the qualitative and quantitative detection of gluten in a complex food matrix. However, for this, it is necessary to use sensitive oligonucleotide primers with specific characteristics for interaction with the target gene.

The Experimentally, the primer set (Tritprglut) proved to be efficient in detecting gluten even in particularly complex food matrices, such as chocolate and cereal bars. The oligomer showed specificity for the target gene in different artificially contaminated meat samples, obtaining results close to the percentage of contaminants inserted in the sample. It presented a LD close to the optimum, which would be 37 cycles and a determined LQ of 60 mg/kg, being within the “Low content” classification range, being slightly above the recommended, which would be 10 mg/kg. Only the cereal bar showed non-compliance with the classification on the label.

After in silico screening, the “Wheat-w-Gliadin” primer showed greater potential to be used in the detection of gluten in foods. Even if the primer sequences are obtained from international publications, a prior evaluation, in silico, is necessary to ensure greater efficiency and precision in molecular analyses. This tracking in relation to specificity parameters, especially in relation to possible cross-reactivity, can reduce the overall analysis time and ensure greater accuracy of results. Among the two-software used in this study, MFE would be the most suitable for primer quality analysis. As future perspectives, it is intended to test the primers that had the best results in the in-silico analysis to verify the possibility of detection at even lower gluten levels than those obtained in this study. Furthermore, the experimental assay must be always added to the in-silico evaluation, as it is a prerequisite for validation studies.

Acknowledgments

The authors would like to thank the Pro-Rectory of Research at the Federal Institute of Education, Science and Technology Sul-rio-grandense (PROPESP, Pelotas, Brazil) for granting scholarships and financial support for carrying out this research and to Laboratório Unianálises for its support with the infrastructure.

REFERENCES

• ¹ Sena-Torralba A, Pallás-Tamarit Y, Morais S, Maquieira Á. Recent advances and challenges in food-borne allergen detection. TrAC Trends Anal Chem. 2020;132:116050. https://doi.org/10.1016/j.trac.2020.116050
  » https://doi.org/10.1016/j.trac.2020.116050
• ² World Health Organization (WHO). Risk Assessment of Food Allergens. Part 1: Review and validation of Codex Alimentarius priority allergen list through risk assessment: Meeting report. Food & Agriculture Org.; 2022.
• ³ Palhares M de PP, Santos PPB dos, Santos ALS dos, Goddard CL, Souza SVC de. [Allergenic foods from a regulatory perspective: a review]. Res Soc Dev. 2021;10:e7310111541. https://doi.org/10.33448/rsd-v10i1.11541
  » https://doi.org/10.33448/rsd-v10i1.11541
• ⁴ Pegels N, González I, García T, Martín R. Authenticity testing of wheat, barley, rye and oats in food and feed market samples by real-time PCR assays. Food Sci Technol. 2015;60:867-75. https://doi.org/10.1016/j.lwt.2014.10.049
  » https://doi.org/10.1016/j.lwt.2014.10.049
• ⁵ García-García A, Madrid R, Sohrabi H, de la Cruz S, García T, Martín R, et al. A sensitive and specific real-time PCR targeting DNA from wheat, barley and rye to track gluten contamination in marketed foods. Food Sci Technol. 2019;114:108378. https://doi.org/10.1016/j.lwt.2019.108378
  » https://doi.org/10.1016/j.lwt.2019.108378
• ⁶ BRASIL. Resolução RDC no 26, de 02 de julho de 2015. [Provides the requirements for mandatory labeling of the main foods that cause food allergies]. 2015.
• ⁷ Brazilian Ministry of Health, Agência Nacional de Vigilância Sanitária. Instrução normativa N° 75, de 8 de outubro de 2020. [Establishes the technical requirements for the declaration of nutritional labeling on packaged foods]. 2020.
• ⁸ Nacional I. Resolução de Diretoria Colegiada - RDC No 429, DE 8 DE OUTUBRO DE 2020 - DOU - Imprensa Nacional. n.d. Available from: https://antigo.anvisa.gov.br/documents/10181/3882585/RDC_429_2020_.pdf/9dc15f3a-db4c-4d3f-90d8-ef4b80537380
  » https://antigo.anvisa.gov.br/documents/10181/3882585/RDC_429_2020_.pdf/9dc15f3a-db4c-4d3f-90d8-ef4b80537380
• ⁹ Visciano P, Schirone M. Food frauds: Global incidents and misleading situations. Trends Food Sci Technol. 2021;114:424-42. https://doi.org/10.1016/j.tifs.2021.06.010
  » https://doi.org/10.1016/j.tifs.2021.06.010
• ¹⁰ Valdés I, García E, Llorente M, Méndez E. Innovative approach to low-level gluten determination in foods using a novel sandwich enzyme-linked immunosorbent assay protocol. Eur J Gastroenterol Hepatol. 2003;15:465. https://doi.org/10.1097/01.meg.0000059119.41030.df
  » https://doi.org/10.1097/01.meg.0000059119.41030.df
• ¹¹ Xhaferaj M, Alves TO, Ferreira MSL, Scherf KA. Recent progress in analytical method development to ensure the safety of gluten-free foods for celiac disease patients. J Cereal Sci. 2020;96:103114. https://doi.org/10.1016/j.jcs.2020.103114
  » https://doi.org/10.1016/j.jcs.2020.103114
• ¹² Jain S, Lamba BY, Dubey SK. Recent advancements in the sensors for food analysis to detect gluten: A mini-review [2019-2023]. Food Chem. 2024;449:139204. https://doi.org/10.1016/j.foodchem.2024.139204
  » https://doi.org/10.1016/j.foodchem.2024.139204
• ¹³ Mahgoub YA, Shawky E, Abdallah II. Authentication of oats and discrimination from their gluten-containing adulterants using NIR diffuse reflectance spectroscopy and multivariate analysis. Food Biosci. 2024;60:104271. https://doi.org/10.1016/j.fbio.2024.104271
  » https://doi.org/10.1016/j.fbio.2024.104271
• ¹⁴ Marra D, Acunzo A, Fulgione A, De Luca M, Villalonga R, Pisani F, et al. Advances in Gluten Detection: A Rapid Colorimetric Approach Using Core-Satellite Magnetic Particles. Biosens. Bioelectron. X. 2024;21:100545. https://doi.org/10.1016/j.biosx.2024.100545
  » https://doi.org/10.1016/j.biosx.2024.100545
• ¹⁵ Salman F, Zengin A, Çelik Kazici H. Simple detection of gluten in commercial gluten-containing samples with a novel nanoflower electrosensor made of molybdenum disulfide with comparison of the ELISA method. J Food Sci. 2024;89:2747-60. https://doi.org/10.1111/1750-3841.17043
  » https://doi.org/10.1111/1750-3841.17043
• ¹⁶ Amigo CD, Popping B. Labeling regulations, detection methods, and assay validation. J AOAC Int. 2012;95:337-48.
• ¹⁷ Koerner TB, Cléroux C, Poirier C, Cantin I, Alimkulov A, Elamparo H. Gluten contamination in the Canadian commercial oat supply. Food Addit Contam Part A. 2011;28:705-10. https://doi.org/10.1080/19440049.2011.579626
  » https://doi.org/10.1080/19440049.2011.579626
• ¹⁸ Koerner TB, Cleroux C, Poirier C, Cantin I, La Vieille S, Hayward S, et al. Gluten contamination of naturally gluten-free flours and starches used by Canadians with celiac disease. Food Addit Contam Part A. 2013;30:2017-21. https://doi.org/10.1080/19440049.2013.840744
  » https://doi.org/10.1080/19440049.2013.840744
• ¹⁹ Farage P, Zandonadi RP, Gandolfi L, Pratesi R, Falcomer AL, Araújo LS, et al. Accidental Gluten Contamination in Traditional Lunch Meals from Food Services in Brasilia, Brazil. Nutrients. 2019;11:1924. https://doi.org/10.3390/nu11081924
  » https://doi.org/10.3390/nu11081924
• ²⁰ Kryuchenko EV, Chernukha IM, Kuzlyakina YA, Zamula VS. Detection of unlabelled gluten in meat products and gluten-free flour by PCR and ELISA methods. IOP Conf Ser Earth Environ Sci. 2021;854:012047. https://doi.org/10.1088/1755-1315/854/1/012047
  » https://doi.org/10.1088/1755-1315/854/1/012047
• ²¹ Zeltner D, Glomb MA, Maede D. Real-time PCR systems for the detection of the gluten-containing cereals wheat, spelt, kamut, rye, barley and oat. Eur Food Res Technol. 2009;228:321-30. https://doi.org/10.1007/s00217-008-0937-4
  » https://doi.org/10.1007/s00217-008-0937-4
• ²² Allmann M, Candrian U, Höfelein C, Lüthy J. Polymerase chain reaction (PCR): a possible alternative to immunochemical methods assuring safety and quality of food. Detection of wheat contamination in non-wheat food products. Z Lebensm Unters Forsch. 1993;196:248-51. https://doi.org/10.1007/bf01202741
  » https://doi.org/10.1007/bf01202741
• ²³ DNeasy mericon Food Handbook - QIAGEN n.d. Available from: https://www.qiagen.com/us/resources/resourcedetail?id=bd9cc2a8-aa71-4cb5-b6f7-97b3d7fc306d⟨=en (accessed August 16, 2023).
  » https://www.qiagen.com/us/resources/resourcedetail?id=bd9cc2a8-aa71-4cb5-b6f7-97b3d7fc306d⟨=en
• ²⁴ QuantiNova SYBR Green PCR Manual - QIAGEN n.d. Available from: https://www.qiagen.com/us/resources/resourcedetail?id=61d2de10-2b7c-472c-96d9-a818e200fbda⟨=en (accessed August 16, 2023).
  » https://www.qiagen.com/us/resources/resourcedetail?id=61d2de10-2b7c-472c-96d9-a818e200fbda⟨=en
• ²⁵ Wang K, Li H, Xu Y, Shao Q, Yi J, Wang R, et al. MFEprimer-3.0: quality control for PCR primers. Nucleic Acids Res. 2019;47:W610-3. https://doi.org/10.1093/nar/gkz351
  » https://doi.org/10.1093/nar/gkz351
• ²⁶ NetPrimer Manual. PREMIER Biosoft International Corina Way, Palo Alto, CA. https://www.premierbiosoft.com/netprimer/ (accessed August 16, 2023).
  » https://www.premierbiosoft.com/netprimer
• ²⁷ Almeida JVP, Cellupi Neto R, Freitas RJS de, Waszczynskyj N, Stertz SC, Rosa MIS. [Physicochemical composition and fatty acid profile in differentiated meats, beef and pork]. Hig Aliment. 2009:159-64.
• ²⁸ Fröder H, Righi E. Quantification of animal species using quantitative real-time PCR (qPCR) to verify fraud in meat products. Res., Soc. Dev. 2023;12(1):e1512138972. http://dx.doi.org/10.33448/rsd-v12i1.38972
  » http://dx.doi.org/10.33448/rsd-v12i1.38972
• ²⁹ IFS Food. Standard for assessing product and process compliance in relation to food safety and quality. Food 2020. Available from: https://www.ifs-certification.com/en/food-standard (accessed August 16, 2023).
  » https://www.ifs-certification.com/en/food-standard
• ³⁰ Tramuta C, Decastelli L, Ingravalle F, Barcucci E, Fragassi S, Bianchi DM. Performance Evaluation of a Commercial Real-Time PCR Method for the Detection of Lupin Traces in Food. Foods. 2024;13:609. https://doi.org/10.3390/foods13040609
  » https://doi.org/10.3390/foods13040609
• ³¹ Comino I, Moreno M de L, Sousa C. Role of oats in celiac disease. World J Gastroenterol. 2015;21:11825-31. https://doi.org/10.3748/wjg.v21.i41.11825
  » https://doi.org/10.3748/wjg.v21.i41.11825
• ³² Fric P, Gabrovska D, Nevoral J. Celiac disease, gluten-free diet, and oats. Nutr Rev. 2011;69:107-15. https://doi.org/10.1111/j.1753-4887.2010.00368.x
  » https://doi.org/10.1111/j.1753-4887.2010.00368.x
• ³³ Zhao S, Wang Y, Li C. Validation of Molecular Detection Methods for Gluten Allergens in Infant Formula. Asian Agric Res. 2024. https://doi.org/10.22004/ag.econ.344157
  » https://doi.org/10.22004/ag.econ.344157
• ³⁴ García-García A, Madrid R, García T, Martín R, González I. Use of multiplex ligation-dependent probe amplification (MLPA) for screening of wheat, barley, rye and oats in foods. Food Control. 2018;84:268-77. https://doi.org/10.1016/j.foodcont.2017.07.037
  » https://doi.org/10.1016/j.foodcont.2017.07.037
• ³⁵ Druml B, Mayer W, Cichna-Markl M, Hochegger R. Development and validation of a TaqMan real-time PCR assay for the identification and quantification of roe deer (Capreolus capreolus) in food to detect food adulteration. Food Chem. 2015;178:319-26. https://doi.org/10.1016/j.foodchem.2015.01.003
  » https://doi.org/10.1016/j.foodchem.2015.01.003
• ³⁶ Mazzara M, Savini C, Delobel C, Broll H, Damant A, Paoletti C, et al. Definition of minimum performance requirements for analytical methods of GMO testing. Ispra Jt Res Cent. 2008:8. https://doi.org/10.2788/65827
  » https://doi.org/10.2788/65827
• ³⁷ Popping B, Holzhauser T. [Detection methods for food allergens]. Dtsch Lebensm-Rundsch. 2004;100:285-93.
• ³⁸ Andersen CB, Holst-Jensen A, Berdal KG, Thorstensen T, Tengs T. Equal performance of TaqMan, MGB, molecular beacon, and SYBR green-based detection assays in detection and quantification of roundup ready soybean. J Agric Food Chem. 2006;54:9658-63.
• ³⁹ Wobi A, Sebah D, Kraemer I, Losher C, Fischer G, Busch U, et al. A multiplex real-time PCR method for the quantification of beef and pork fractions in minced meat. Food Chem. 2015;169:305-13. https://doi.org/10.1016/j.foodchem.2014.07.139
  » https://doi.org/10.1016/j.foodchem.2014.07.139
• ⁴⁰ Rozen S, Skaletsky H. Primer3 on the WWW for General Users and for Biologist Programmers. In: Misener S, Krawetz SA, editors. Bioinforma. Methods Protoc., Totowa, NJ: Humana Press; 1999, 365-86. https://doi.org/10.1385/1-59259-192-2:365
  » https://doi.org/10.1385/1-59259-192-2:365
• ⁴¹ Sandberg M, Lundberg L, Ferm M, Malmheden Yman I. Real Time PCR for the detection and discrimination of cereal contamination in gluten free foods. Eur Food Res Technol. 2003;217:344-9. https://doi.org/10.1007/s00217-003-0758-4
  » https://doi.org/10.1007/s00217-003-0758-4
• ⁴² Martín-Fernández B, Costa J, Oliveira MBPP, López-Ruiz B, Mafra I. Screening new gene markers for gluten detection in foods. Food Control. 2015;56:57-63. https://doi.org/10.1016/j.foodcont.2015.02.047
  » https://doi.org/10.1016/j.foodcont.2015.02.047
• ⁴³ Piknová Ľ, Brežná B, Kuchta T. Detection of gluten-containing cereals in food by 5′-nuclease real-time polymerase chain reaction. J Food Nutr Res. 2008;47(3):114-9.
• ⁴⁴ Martín-Fernández B, Costa J, de-los-Santos-Álvarez N, López-Ruiz B, Oliveira MBPP, Mafra I. High resolution melting analysis as a new approach to discriminate gluten-containing cereals. Food Chem. 2016;211:383-91. https://doi.org/10.1016/j.foodchem.2016.05.067
  » https://doi.org/10.1016/j.foodchem.2016.05.067
• ⁴⁵ 14:00-17:00. ISO/TS 20224-4:2020. ISO n.d. Available from: https://www.iso.org/standard/75477.html (accessed August 16, 2023).
  » https://www.iso.org/standard/75477.html
• ⁴⁶ Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. Clin Chem. 2009;55:611-22. https://doi.org/10.1373/clinchem.2008.112797
  » https://doi.org/10.1373/clinchem.2008.112797
• ⁴⁷ Vallone PM, Butler JM. AutoDimer: a screening tool for primer-dimer and hairpin structures. BioTechniques. 2004;37:226-31. https://doi.org/10.2144/04372ST03
  » https://doi.org/10.2144/04372ST03
• ⁴⁸ Mujico JR, Lombardía M, Mena MC, Méndez E, Albar JP. A highly sensitive real-time PCR system for quantification of wheat contamination in gluten-free food for celiac patients. Food Chem. 2011;128:795-801. https://doi.org/10.1016/j.foodchem.2011.03.061
  » https://doi.org/10.1016/j.foodchem.2011.03.061
• ⁴⁹ Garrido-Maestu A, Azinheiro S, Fuciños P, Carvalho J, Prado M. Highly sensitive detection of gluten-containing cereals in food samples by real-time Loop-mediated isothermal AMPlification (qLAMP) and real-time polymerase chain reaction (qPCR). Food Chem. 2018;246:156-63. https://doi.org/10.1016/j.foodchem.2017.11.005
  » https://doi.org/10.1016/j.foodchem.2017.11.005
• ⁵⁰ Henterich N, Osman AA, Méndez E, Mothes T. Assay of gliadin by real-time immunopolymerase chain reaction. Food Nahr. 2003;47:345-8. https://doi.org/10.1002/food.200390079
  » https://doi.org/10.1002/food.200390079
• ⁵¹ Mujico JR, Méndez Cormán E. Method for the simultaneous detection and quantification of DNA from wheat, barley, rye and crosses and mixtures thereof in food, elements required to perform same and use thereof. Organización Mundial de la Propriedad Intelectual. WO 2008110655 (A1). 2008.
• ⁵² Pinto A, Polo PN, Henry O, Redondo MCB, Svobodova M, O’Sullivan CK. Label-free detection of gliadin food allergen mediated by real-time apta-PCR. Anal Bioanal Chem. 2014;406:515-24. https://doi.org/10.1007/s00216-013-7475-z
  » https://doi.org/10.1007/s00216-013-7475-z

• Funding:
  This research was funded by Pro-Rectory of Research at the Federal Institute of Education, Science and Technology Sul-rio-grandense (PROPESP, Pelotas, Brazil), grant number PE09220722/030.

Edited by

• Editor-in-Chief:
  Bill Jorge Costa

• Associate Editor:
  Bill Jorge Costa

Publication Dates

• Publication in this collection
  03 Feb 2025
• Date of issue
  2025

History

• Received
  17 Aug 2023
• Accepted
  26 Sept 2024