Determination of patulin in products containing dried fruits by Enzyme‐Linked Immunosorbent Assay technique Patulin in dried fruits

Abstract The era of globalization causes that the export and import of food from different continents of the world are becoming more and more common, which may directly contribute to the increase in pollution in them. The presence of mycotoxin in food is an ubiquitous problem. There is very limited information on the possible influence of the composition of herbal mixtures on the presence of mycotoxins in them, which is an area where research can be expanded. The aim of this study was to determine patulin (PAT) in commercial products containing dried elderberry, rose, blueberry, rowan, hawthorn, and chokeberry fruits by enzyme‐linked immunosorbent assay technique. Research using this technique allowed for considering the possible influence of the composition of herbal mixtures on the concentration of patulin in them. Patulin was detected in all analyzed samples with wide range of <LOD ÷ 4,102.0 µg/kg. In 91% of the single‐ingredient products, the mean patulin concentration below 50 µg/kg was found. We observed that patulin content in products containing only rose, elderberry, blueberry, rowan, hawthorn, or chokeberry fruit was lower than in herbal blends. Research showed that adding dried rowan fruits to herbal blends may contribute to a decrease in PAT levels (r = 0.8581; p = .0031). Looking for such technological methods creating the most unfavorable conditions for the biosynthesis of patulin in medicinal raw materials is extremely important from the point of view of medical and pharmaceutical care.

different parts of the world is not a problem. Almost 90% of apples and raw ingredients for apple products are imported in Taiwan (Lien et al., 2020). In turn, in Afghanistan, raisins are the main export of all agricultural products. In this country, the value of the export market is estimated at 17% (McCoy et al., 2015). The export and import of food from different continents of the world is becoming more and more common, which may directly contribute to the increase in pollution in them (Fritsche, 2018;Kendall et al., 2018). A literature review shows that the presence of mycotoxin in food is an absolutely ubiquitously problem (Moretti et al., 2017). The presence of mycotoxins in food is associated with a twofold problem. The first is economic losses, as it turns out that up to 25% of food is exposed to mycotoxins, which directly translates into losses in agriculture (Moretti et al., 2017). The second, extremely important problem, is the immunotoxic, neurotoxic, and dermotoxic effect of patulin (PAT) on animals and human (Pal et al., 2017;Przybylska et al., 2019). In experimental animals, patulin causes hemorrhages, formation of edema, and dilation of the intestinal tract. In addition, in some cases it causes convulsions, dyspnea, pulmonary congestion, edema, ulceration, hyperemia, and distension of the gastrointestinal tract (Pal et al., 2017). Patulin also poses a significant threat to humans, causing allergic reactions (Sohrabi et al., 2021). Thus, the search for natural, eco-friendly ways to reduce the amount of patulin in plant raw materials seems to be extremely important from the economic and health point of view.
Due to its properties of patulin, the World Health Organization (WHO) recommends limiting the maximum PAT content in apple to 50 µg/kg, in apple puree 25 µg/kg, and in baby food to 10 µg/kg (Hussain et al., 2020). There are no directives relating to the presence of patulin in medicinal plant. The Joint Expert Committee on food Additives of the World Health Organization (JECFA) recommended that daily human exposure to PAT should be reduced to 0.4 µg/kg body weight per day (Hussain et al., 2020;Przybylska et al., 2019).
The major source of PAT in food are apples, pears, mango fruits, orange fruits, cherries, grapes, and fruit juices (Hussain et al., 2020;Pal et al., 2017). It turns out that PAT is also present in the products containing hawthorn berries (Ji et al., 2017;Li et al., 2007;Przybylska et al., 2019;Xiang et al., 2012;Zhou et al., 2012). The literature review shows that there are still no studies on the assessment of patulin content in commercial plant raw materials and herbal blends containing dried elderberry, rowan, chokeberry, blueberry, or rose fruits.
Taking into account the above considerations, the aim of this study was to determine PAT in dried herbal products containing only one dried component and herbal blends with dried different parts of medicinal plant. In addition, the studies attempted for the first time to determine the effect of the presence of individual components of the herbal blend on the final concentration of PAT in them.

| MATERIAL S AND ME THODS
Thirty-one commercial products were analyzed in this study (See Tables 1 and 2). All of them were purchased from supermarkets, herbal stores, and pharmacies in Bydgoszcz (Poland) and by internet sale in January 2020. All the samples (n = 31) were grouped into products single-component commercial products (SC; n = 22) and multicomponent commercial herbal blends containing different dried parts of various medicinal plants (MC; n = 9). Single-

component (SC) and multicomponent (MC) commercial herbs
were packed in collective packages or in sachets. Bags were selected at random for each packaged product, and bulk samples were taken from three different parts of the pack. These products contained dried parts of various medicinal plants belong to eight different families (Rosaceae, Adoxaceae, Ericaceae, Grossulariaceae, Berberidaceae, Malvaceae, Polygonaceae, Scabiosa) where three of them originated from organic farms but five of them were dietary supplements. All analyzed samples were stored in their original packaging until analysis. During the research, the use-by date of the analyzed products was taken into account. All the samples were analyzed at least twice.
Most of the reagents used in this study were contained in the Patulin ELISA Test Kit, which included microtiter plate with patulin standards, patulin-HRP conjugate, extraction buffer, wash solution, stop buffer, and TMB substrate. Methanol for HPLC was obtained from POCH (Poland).
To analyze the PAT content in dried samples, commercial Patulin ELISA Test Kit (Reagen TM , USA) was used. All steps of the analysis were carried out in accordance with the manufacturer's assay procedure. Four grams of samples was added to 20.0 ml methanol for HPLC (POCH, Poland), and the solution was shaken for 400 rpm. Then, 8.0 ml of the homogenized sample was centrifuged for 5 min at 3,500 × g for 5 min at room temperature (22℃). After that, extract was filtered with use of a filter paper and a Chromafil PES-45/25 syringe filter (Mecherey-Nagel, Germany). The supernatant (0.5 ml) was transferred to a tube and added 0.5 ml Sample Extraction Buffer. The mixture was mixed before analysis. One hundred µL of the patulin standard (0.0 ng/ ml ÷ 1,000.0 ng/ml) and test samples (100.0 µl/well) was added to the wells of microtiter plate and added 50.0 µl of patulin-HRP Conjugate (patulin horseradish peroxidase) to each well and mix well by gently rocking the plate manually for 60 s. Then, the plate was incubated for 60 min at room temperature (22℃). After the washing step (3 × 250.0 µl), 100.0 µl of the enzyme conjugate was added and incubated for 20 min at room temperature (22℃) in the dark. Following the addition of 100.0 µl of the stop reagent to each well, the absorbance was measured at 450 nm in ELISA reader (Thermo Scientific, Finland).
The bulk density (BD) of the tested samples was determined according to the method of Ogrodowska et al. (2011).
According to the REAGEN Patulin ELISA Test Kit guidelines, the recovery rate of patulin is between 75% and 95% and the specificity (cross-reactivity) is 100%. The limits of detection (LOD) for patulin were 0.1 ng/ml. Nine standard solution of patulin (0,0 to 1,000,0 ng/ml) have been used for the calibration curve (R2 = 0.994) (See Figure 1). The equation of the trendline was shown below:

| Statistical analysis
The obtained results were analyzed statistically with Statistica v.12 (StatSoft, USA). The results (duplicate samples) are presented as the mean of PAT content in analyzed products. In turn, the results of PAT concentration in the analyzed products groups were shown as the median and quartile range. The Shapiro-Wilk test was used for each data sets. The p-value <.05 was considered significant. To evaluate the difference between samples, a nonparametric Mann-Whitney U test was used (p < .05). We also used the Spearman correlation to determine relationships between the relevant variables and the concentration of patulin.

| RE SULTS AND D ISCUSS I ON
Since 1970, the popularity of the use of immunoenzymatic methods has increased in the analysis of mycotoxins (Singh & Mehta, 2020 (Dungkokkruad et al., 2017;Orina et al., 2020;Thirumala-Devi et al., 2001;Tonti et al., 2017;L. Zhang et al., 2018). The applied ELISA method for the determination of PAT in plant dried materials was described for the first time. It should be noted, however, that although ELISA technique give quick and economical measurements, they lack precision at low concentration (Singh & Mehta, 2020).
In own research, the PAT content in the analyzed commercial single-component (SC) and multicomponent (MC) products is shown in  Table 4. After the analysis, it was found that the median of PAT content in rowan fruits (SC) is almost three times higher than in the blueberry and chokeberry fruit and almost 10 times higher than rose fruit. Five products (Tables 1,   2 (Table 5). During the analysis, it was found that in fruits belonging to the Ericaceae (blueberry fruits), the median of PAT was over two times lower than in fruits belonging to Rosaceae and Adoxaceae.
All the multicomponent (MC) products used in the research were a mixture of dried fruit, flowers and leaves (Table 2). All the analyzed MC samples were contaminated by PAT at the wide range In addition, analytical results suggested that only in 26% of the analyzed dried commercial products concentration of PAT was higher than the maximum tolerance limit of 50 µg/kg recommended by WHO (Drusch & Ragab, 2003;Przybylska et al., 2019). See Figure 2. Moreover, in the 11 samples (36%) PAT content was found exceeded the WHO acceptable upper limit of 25 µg/kg that is recommended for apple puree. Also, our results showed that in the 15 analyzed samples (48%), the concentration of PAT was higher than 10 µg/kg, which is the tolerance upper limit indicated by WHO for baby food. In 5 samples of 31 (16%), an average concentration of PAT above 100 µg/kg was recorded (Figure 2).
The current studies have shown that lower concentration of PAT has been reported for products containing only one dried

F I G U R E 2
Comparison of average of PAT in the analyzed samples where concentration of patulin was higher than the maximum tolerance limit of 10 µg/kg recommended by WHO for baby food (Hussain et al., 2020) four enniatins (2.44 ÷ 11.70 µg/kg), and beauvericin (4.5 µg/kg) (Reinholds et al., 2019). However, it should be noted that the studies did not include the assay in the PAT samples analyzed. Extensive research in Lithuania shows that 45% of "teas" samples of various species shows the presence of deoxynivalenol at levels between 129 and 5,463 µg/kg. Moreover, six "teas" containing rose fruits (dog fruits), midland hawthorn fruits, St. John's Wort, purple coneflower, and herbal blend (mixture of birch, bearberry, knotgrass, rest harrow, parsley, nettle, yarrow, elderberry) contained deoxynivalenol over 2000 µg/kg (Reinholds et al., 2020).
In our own research, in dried elderberry fruits (Vaccinium myrtillus) PAT was detected at a range of 2.1 ÷ 79.0 µg/kg (Table 5). For comparison, in the fresh highbush blueberry (Vaccinium corymbosum) that were grown in the Republic of Belarus (Zenkova & Pinchykova, 2019) and in raspberries, blueberries, blackberries, and such cherries from Czech Republic patulin was not found (Vaclavikova et al., 2015). Byssochlamys nivea-were not determined (Zhang et al., 2016).
Research shows that Aspergillus contamination of plant-based raw materials (garlic) comes from the field and persists through processing, including washing and drying (Amoah et al., 2020).
The presence of patulin-producing species of fungi is not the only condition that determines the increase in PAT biosynthesis. The biosynthesis of mycotoxins, including PAT, depends on many different physicochemical parameters. Some of them are water activity, and the temperature maintained during the storage of food products and their pH (Tannous et al., 2016). The intensity of PAT biosynthesis also depends on variety of fruits used in the production process, content of organic acids leached from the vacuole and to the mycelium of P. expansum, as well as abiotic factors such as humidity, temperature, and sunlight (Barad et al., 2016;Drusch & Ragab, 2003). The wide range of PAT concentration in the tested products may result from the use of different apple varieties, as the research of Barad et al. (2016) where suggested that cultivars of apple trees are an important factor influencing the inhibition of PAT biosynthesis. Nevertheless, noncontrolled use of apple leather to the herbal blends can be additional and significant source of PAT in products. The research of Montaseri et al. (2013) showed that apple leather can be infected with PAT in a wide range, up to <10 ÷ 2,559 µg/kg. By analyzing content of PAT in MC commercial herbal blends, we claimed that, according to the manufacturer's declaration, the sample P23 containing the most of hibiscus petals and dried rowan fruits characterized the lowest average concentration of PAT -5.80 µg/kg. Similar results were obtained by Przybylska et al. (2019). In this study, in the sample where the largest share, according to the producer's declaration, was hawthorn fruit, lemon balm leaf, and hibiscus flower, the concentration of PAT was significantly lower compared with products containing more dried fruits than leaves. It should be added that the presence of deoxynivalenol, aflatoxin B1, HT-2, T-2, zearalenone, sterigmatocystin, ochratoxin, enniatins, and beauvericin was not found in the bearberry leaves, eucalyptus leaves, linden flowers, calendula, yarrow, and yellow everlasting flower tea (Reinholds et al., 2019), which are a rich source of biologically active compounds. In recent years, attention has also been paid to the unusual properties of propolis. Studies have shown that 2 mg/ml propolis extract reduces the growth of PAT in apples juices (Silici & Karaman, 2014). Propolis is collected by honey bees, making it a rich source of aromatic acids, aromatic esters, volatile compounds, aromatic compounds, hydrocarbons, steroids, flavonoids, acids, micro-and macronutrients, vitamins, and essential oils (Hemmami et al., 2020;Pobiega et al., 2019). The abundance of bioactive compounds present in propolis or cinnamon oil which could reduce the expression of genes involved in PAT biosynthesis (Lai et al., 2021) suggests that the composition of herbal blends may also play a role in controlling optimal PAT growth conditions. The exogenous amino acid L-glutamate has strong properties that inhibit the development of blue rot caused by Penicillium expansum in the postharvest pear fruit (Jin et al., 2019). Literature data confirm that Camelia sinensis is source of γ-aminobutyric acid (Przybylska et al., 2021), but rowan fruits (Sorbus aucuparia L.) are richer source of amino acids than hawthorn fruits (Crataegus sanguinea Pall.) or cinnamon fruits (Rosa cinnamomea L.) (Sergunova et al., 2020). In turn, hibiscus (Hibiscus L.), rowan fruits, blackcurrant fruits (Ribes nigrum), or raspberry (Rubus idaeus) are richer source of amino acids compared with apples (Malus domestica) or redcurrant (Ribes rubrum) (Kunachowicz et al., 2018;Sergunova et al., 2020). In current study, the authors found the highest average PAT content in dried rowan fruit. See Table 4. Interestingly, at the same time, the lowest PAT concentration was recorded in multi-ingredient products with the addition of dried rowan. The authors speculate that it may be caused not only by the presence of amino acids in the dried fruit, but also by competition from microorganisms in the product. It should also be noted that in the P29 product, in which the presence of PAT above 4,000 µg/kg was found, there was no dried rowan fruit.
Interaction between mycotoxin and selected herbs and species components was described by Do et al. (2015). One of the most important mechanisms in biocontrol is competition for the living space and nutrients of numerous yeasts and molds, what may explain the significant decrease in PAT in herbal blends with dried rowan fruit (Spadaro et al., 2013). Numerous studies prove PAT degradation using a biocontrol mechanism. For example, strain of Metschnikowia fructicola AL27 is more effective than M. pulcherrima MACH1 and GS9 in the control of blue mold rot which results in a stronger reduction in PAT on four cultivar of apples (Spadaro et al., 2013). Thus, it can be speculated that in the samples of plant raw materials being a mixture of various plant parts, competition between species of mold and yeast for living space and nutrients may result in the presence of mold species responsible for the production or inhibition of PAT biosynthesis. However, further research is needed to clarify this hypothesis. In additional, PAT may act synergistically with other mycotoxins such as citrinin, causing more extensive effect on human tissues and organs (Qin et al., 2020).
The presence of mycotoxins other than PAT in food is also a serious health problem (Moretti et al., 2017).

| CON CLUS IONS
The

ACK N OWLED G M ENTS
The support from the Nicolaus Copernicus University, Collegium Medicum (Bydgoszcz, Poland), by the internal grant DS-UPB-449/2021 is kindly acknowledged.

CO N FLI C T O F I NTE R E S T
No potential conflict of interest was reported by the authors.

E TH I C A L A PPROVA L
This study does not involve any human or animal testing.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data obtained during the research appear in the submitted article.