Natural co-occurrence of aflatoxins and ochratoxin A in liquorice roots

ABSTRACT The natural co-occurrence of aflatoxins (AFs) and ochratoxin A (OTA) in liquorice roots obtained from different herbal outlets of Karachi, Pakistan, was evaluated. A total of 60 samples were obtained during 2019‒2021 and analysed using HPLC with fluorescence detection. The method was validated according to the European Union (EU) regulation. The incidence of AFs was 52 in all tested samples (87%), with a contamination range of 1.2‒6.4 µg kg−1 and an average of 2.1 ± 0.3 µg kg−1. In all samples, the AFs contamination was below the maximum limit (ML) of 10 µg kg−1 for total AFs as set by the EU. The OTA contamination in all samples ranged 1.5‒60.3 µg kg−1, with a mean of 18.9 ± 0.8 µg kg−1. In 36 samples (60%) the OTA contamination was below the ML of 20 µg kg−1 as set by the EU. These results led to the recommendation to test liquorice root on mycotoxins, as it comes to food quality standards.


Introduction
In the pharmaceutical industry, the requirement of medicinal plants has massively increased during recent years.The root of leguminous Glycyrrhiza glabra Liquorice is a perennial herbaceous plant and native to Western Asia, North Africa, and Southern Europe.Liquorice is widely consumed around the world due to a range of pharmacological properties, such as antibacterial, antiviral, antispasmodic, anti-inflammatory, antiallergic, antitussive, expectorant, and inhibit or reduce the production of bacterial toxins (Wali et al. 2022;Yin et al. 2022).Liquorice and their derivatives are applied as dietary supplements and are used as a sweetening and flavouring agent in confectionery and various food products like beverages, herbal tea, chewing gums, candies, and sweets as a fresh, dry root or as a liquorice extract (Husain et al. 2021).
Climatic conditions such as temperature and relative humidity (RH) support microbial growth on liquorice roots (Khalesi et al. 2013).Traditional harvesting, handling, and storing conditions accumulate the exposure to fungal contamination (Asghar et al. 2022).In addition, fresh liquorice contains around 70% moisture content and generally is stored in open shade.Liquorice roots are dehydrated naturally over a period subjected to climate conditions such as temperature and RH, which play a dynamic role in fungal accumulation.Therefore, the drying period is of vital importance for mould activities and subsequently mycotoxins formation in the stored roots.
Mycotoxins are a large group of toxic compounds produced by certain fungal species as secondary metabolites that easily colonise and infect both field and harvested production (Patriarca and Pinto 2017).Amongst all the identified mycotoxins, aflatoxins (AFs) and ochratoxin A (OTA) are considered the most challenging and frequently found in several agricultural foodstuffs (Asghar et al., 2016;Ałtyn and Twarużek 2020;Yu et al. 2022).AFs are mainly formed by Aspergillus flavus and A. parasiticus.Presently, more than 20 AFs have been identified (Asghar et al. 2020), although the four most important are aflatoxin B 1 (AFB 1 ), B 2 (AFB 2 ), G 1 (AFG 1 ) and G 2 (AFG 2 ).However, AFB 1 is recognised as the most toxic with hepatotoxic and carcinogenic nature and the International Agency for Research on Cancer (IARC 1993) has categorised it as a group 1 carcinogen.The suitable conditions for the production of AFB 1 were reported as pH 3-7, humidity 85%, moisture content 15-25%, water activity >0.92 and temperature 25-35•C (Asghar et al. 2022).
Whereas OTA is another mycotoxin, that is mostly produced by two genera of fungi: Aspergillus (A. carbonarius and A. ochraceus) and Penicillium (P.verrucosum and P. nordicum) and first isolated from A. ochraceus Wilh.IARC has classified OTA as a possible human carcinogen (group 2B) based on adequate evidence for OTA-carcinogenicity attained from several animal investigations (Ostry et al. 2017).AFs and OTA have carcinogenic, teratogenic, genotoxic, immunotoxic, and neurotoxic effects on human and animal health as reported earlier (Cimbalo et al. 2020;Nourbakhsh and Tajbakhsh 2021).The ecological circumstances for optimum production of OTA have been revealed to differ slightly with those for aflatoxins, with the optima being pH 5.0-6.5, moisture content ≥27%, water activity >0.93 and temperature 15-25°C (Wang et al. 2016).
Due to the frequent occurrence and toxicity of mycotoxins, many countries have imposed strict guidelines regarding mycotoxins contamination in food commodities.For instance, the European Commission (2010) has set MLs for OTA of 20 and 80 µg kg −1 for liquorice root (ingredient for herbal infusion) and liquorice extracts (used in liquorice confectionery), respectively, in Regulation No. 105/2010.Since the EU has categorised liquorice as a spice (European Commission 2018), it has to apply to the ML of 10 µg kg −1 for total aflatoxins in spices (European Commission 2010).The USDA (Food and Drugs Authority of the USA) established an ML for AFs of 20 µg kg −1 in all food items (Liang et al. 2021).
Liquorice roots are more susceptible to OTA contamination in comparison to aflatoxins, as surveys reported from various regions of the world (Pietri et al. 2010;Wang et al. 2013).A previous study showed that only 18% of the samples were contaminated with AFB 1 , whereas 100% of the samples contained OTA, with a maximum level of 990 µg kg −1 .As a result, most of the researchers are focusing on the detection of OTA contamination in liquorice roots.These previous studies described higher contamination of OTA in liquorice roots and their products (Ariño et al. 2007;Khalesi 2015;Yu et al. 2022).In addition, to date only one report, with a very small number of samples (n = 4), is available on the simultaneous detection of AFs and OTA in Pakistani liquorice roots (Ahmad et al. 2014).Since liquorice roots might be contaminated concurrently with various fungal species, which produce AFs and OTA as well, multimycotoxin detection is necessary in liquorice roots.Therefore, the purpose of the current study was to simultaneously determine the AFs and OTA contamination in liquorice roots using high-performance liquid chromatography with a fluorescence detector and immuno-affinity column (IAC) extraction.Moreover, the levels of both mycotoxins were compared with previously described levels worldwide.The outcome of the study will help in the mycotoxins management programme, which will enhance the food safety in human consumable products.

Sampling
Sixty samples of dried liquorice root were obtained randomly from various herbal outlets of Karachi, Pakistan in a time span of 3 years (2019-2021).Due to heterogeneous distribution of mycotoxins, the sampling process was based on the procedure as reported in AOAC official method no.977.16 (Horwitz and Latimer 2023).Briefly, 1 kg of each collected sample was chopped using mortar and pestle and then grinded with a stainless grinding mill (Cytcloec, model 1093, Tector;Höganäs, Sweden).The grinded samples were sieved using sieve no.20 and stored in a polyethylene bag at −20°C.A portion of 20 g was analysed.The homogeneity of the sample was confirmed by analysis of each sample in triplicate, since the obtained standard deviation fulfiled the statistical criteria.
The solution was evaporated at 45°C under a flow of N 2 gas and the residue re-constituted in 3 mL of MeOH:H 2 O (50:50 v/v).The final concentration of the intermediate solution was 100 ng mL −1 and stored at −20°C.Eight working standard solutions of each AFsin the range of 0.05 to 20 ng mL −1 were made in MeOH:H 2 O (50:50 v/ v) using the intermediate solution and kept at 4-8°C in the dark.For OTA, 49.0 µL of standard solution (10.2 µg mL −1 ) was dispensed in an amber vial and vaporised under a flow of N 2 gas.The residue was re-constituted in 5 mL of MeOH:H 2 O:CH 3 COOH (50:49:1 v/v/v) to obtain 100 ng mL −1 intermediate solution.Eight working standard solutions between 0.05 and 40 ng mL −1 were made in H 2 O:ACN:CH 3 COOH (47:51:2 v/v/v) solution using the intermediate standard solution and kept at 4-8°C in darkness.

Quantification of AFs
The quantification steps such as extraction, clean-up, and determination of AFs were performed according to the method reported in our previous study (Asghar et al. 2016).Ten grams sample was homogenised with 50 mL of ACN:H 2 O (60:40 v/v) using an explosion-proof blender (Model No. 8018; Eberbach Corp., Ann Arbor, MI, USA) at 5000 rpm for 2 min.The mixture was filtered using Whatman no. 1 paper (Whatman Inc., Buckinghamshire, UK) and an aliquot of the 2 mL filtrate was diluted with PBS solution (48 mL).The solution was cleaned with IACs, rinsed 2 times using 20 mL of DI-H 2 O and dried under vacuum.The AFs were eluted with 1.5 mL of MeOH and diluted with 1.5 mL of DI-H 2 O into an amber vial, and the mixture was filtered using HV 0.45 µm (Millipore, Bedford, MA, USA) before HPLC analysis.

HPLC-FLD for AFs
AFs were measured with an HPLC Hitachi system with post-column derivatisation using a Kobra Cell TM and fluorescence detection (Asghar et al. 2016).The mobilephase flow at 1.0 mL/min was composed of H 2 O:MeOH: ACN (60:20:20 v/v/v) containing KBr (119 mg) and HNO 3 (167 µL of 65%) per litre.Aliquots of 99 µL of each standard and sample were introduced via an autosampler.The column temperature was maintained at 40°C.The Kobra cell is functioned at a fixed current of 100 µA.The wavelengths in the fluorescence detector were fixed at 365 nm and 440 nm for excitation and emission, respectively.The analysis was performed in an isocratic mode, and the retention times were 11.5, 9.7, 8.5, and 7.3 min for AFB 1 , AFB 2 , AFG 1 , and AFG 2 , respectively.

Quantification of OTA
OTA extraction and clean-up was executed with IACs, following the manufacturer's instructions.Ten grams sample was combined with 100 mL of ACN:H 2 O (50:50 v/v) using a blender at 6000 rpm speed for 2 min and filtered with Whatman no. 1 paper.Two millilitres of the extract were mixed with 48 mL PBS buffer.The filtrate was passed through IACs affixed onto a vacuum system at a 1 drop/second flow rate until air dryness.The IACs were rinsed with 20 mL of DI-H 2 O, dehydrated under vacuum and afterwards OTA was slowly eluted by delivering 3 mL of absolute methanol.The eluate was vaporised at 45°C under N 2 gas to dryness, and the residue was dispersed in 2 mL of H 2 O:ACN:CH 3 COOH (47:51:2; v/v/v).The solution was filtered using HV 0.45 μm, prior to HPLC analysis.

HPLC-FLD for OTA
OTA analysis was conducted by a reversed-phase HPLC column using a H 2 O:ACN:CH 3 COOH (47:51:2; v/v/v) mobile phase with 1 mL/min of flow rate.Of each standard and sample 99 µL was injected.The column temperature was retained at 40°C.The fluorescence wavelength settings were 334 nm for excitation and 464 nm for emission.The presence of OTA contamination in positive samples was performed by derivatisation into the methyl ester (Copetti et al. 2012).The analysis was conducted in an isocratic mode and retention time was 8.35 min for OTA.

Method validation and quality assurance
The validation of the HPLC method was assessed according to Regulation No. 401/2006/EC (European Commission 2006), including linearity, selectivity, limit of detection (LOD), limit of quantification (LOQ), intraday and inter-day relative standard deviations (RSD), recovery and measurement uncertainty.The method's linearity was assessed in a designated concentration series with multiple calibration points and data fixed by the least squares method (total AFs = 0.05 − 80 ng/mL and OTA = 0.10 − 40 ng/mL).All measurements were done in triplicate.The calibration graphs were plotted between peak areas versus concentrations and the correlation coefficients (R 2 ) values were calculated.LOD and LOQ were measured from the calibration curves according to signal-to-noise ratios 3 and 10, respectively (Miller and Miller 1993).The precision of the method was estimated by the evaluation of intra-and inter-day RSD data analysing QC samples on the same and five consecutive days, respectively.Furthermore, the effectiveness of the method was investigated by the standard addition method.Briefly, 25 g of mycotoxin free (confirmed before via ELISA) were fortified with three different concentrations of the standard solution of AFs and OTA separately and subject to the analytical procedure as described above.
The measurement uncertainty was considered as a combined uncertainty (Uc) with covering factor k = 2 (so 2Uc) at a 95% confidence level.Laboratory performance was verified by participation in LGC proficiency test no.PT-AF-05 (LGC Limited 2021).In addition, for AFs and OTA analysis, QC samples no.04376QC and 17205QC from the Food Analysis Performance Assessment Scheme of 2020 (FAPAS®, The Food and Environment Research Agency, Sand Hutton, York, UK) were applied (Food Analysis Performance Assessment Scheme, 2020).Furthermore, the food and feed safety laboratory is ISO-17025:2017 accredited by the Pakistan National Accreditation Council (PNAC), which is a member of the International Laboratory Accreditation Cooperation (ILC).

Statistical data analysis
Statistical data analysis was assessed using the Statistical Analysis Tool of Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA, USA).Samples containing AFs and OTA were considered positive above LOD.

AFs contamination
During 2019-2021, a total of 60 samples of liquorice root samples were obtained and tested for AFs contamination, of which the frequency is given in  Table 4. Fifty-two samples (87%) contained the AFs levels above the LOD (0.11 µg kg −1 ).The maximum contamination was 6.4 µg kg −1 , with a mean level of 2.1 ± 0.3 µg kg −1 of AFs.In eight samples (13%), the AFs contamination was below the LOD.The AFs contamination in 49 samples (82%) was in the range of 1.2-3.7 µg kg −1 .Total AFs contamination was 4.7-6.4µg kg −1 in only three samples (5%), lower than the ML of 10 µg kg −1 in spices as set by the European Commission (2010).However, none of the samples exceeded the ML of 20 µg kg −1 for AFs as assigned by USDA (2000).Few studies also confirmed the present finding that AFs contamination was low in liquorice roots (Table 5).Ahmad et al. (2014) reported from Pakistan that 50% of the samples were contaminated, with a maximum of 9.3 µg kg −1 AFs.A survey from Italy reported that about 18% of the samples contained AFs, with a mean content of 0.4 µg kg −1 and a maximum level of 2.4 µg kg −1 (Pietri et al. 2010).In addition, Wang et al. (2013) reported from China 14% positive samples, with a maximum level of 18.7 µg kg −1 .

OTA contamination
All 60 tested samples were contaminated with OTA, on average with 18.9 ± 0.8 µg kg −1 in an extensive concentration range from 1.5 to 60.3 µg kg −1 (Table 6).However, the average contamination was slightly lower than the ML of 20 µg kg −1 as set by the European Commission (2010).In 36 samples (60%) the OTA level was below this ML, ranging between 1.5 and 18.2 µg kg −1 .On the other hand, in 23 samples (38%) the OTA levels were between 20.3 and 48.6 µg kg −1 and only 1 sample (1.7%) showed the maximum OTA contamination of 60.3 µg kg −1 .The present findings show   that the OTA contamination in 60% of the samples were below the EU ML, so fit for human consumption.Moreover, there was a major difference during the yearly (2019-2021) OTA contamination.For instance, in 2019 the mean contamination of OTA was 23.0 µg kg −1 , with a maximum level of 60.3 µg kg −1 .The OTA contamination decreased in 2020 and the mean and maximum contamination of OTA was quantified as 20.0 µg kg −1 and 48.6 µg kg −1 , respectively.In 2021, the cultivation and storage conditions were improved on direction of the Department of Plant Protection of Pakistan.As a result, the contamination level of OTA was further decreased in 2021 in comparison to 2019 and 2020, since the mean OTA contamination was 13.8 µg kg −1 , with a maximum of 38.3 µg kg −1 in 2021.It was also observed that the highly contaminated samples with AFB 1 and OTA were found in July and August, when compared to other months.As mentioned earlier in our previous study, these months are counted to the monsoon period in Pakistan, where the usual temperature/RH was 33°C/50% and 31°C/58%, respectively.These climatic situations are thought of as promising conditions for the fungal growth and ultimately production of mycotoxins (Asghar et al. 2017).Furthermore, most liquorice root samples contained both AFs and OTA and co-occurrence of mycotoxins may exert synergistic lethal effects, although at levels far lower than the acceptable limits.As studied by Streit et al. (2012), the presence of mycotoxins in combination may exert synergetic, preservative or antagonistic effects.A study in feed (Boermans and Leung 2007) also reported that co-occurrence of AFs and OTA enhances their toxic properties synergistically.These impacts are more dangerous even at low levels of single mycotoxins, as their adverse effect on human and animal health can be increased synergistically.
Various studies also reported that liquorice root is more vulnerable to OTA contamination (Table 7).For instance, Ariño et al. (2007) detected OTA in 100% of liquorice root samples in Spain.The average contamination was 63.6 µg kg −1 , ranging from 1.4 to 253 µg kg −1 .Surveys from Italy and the Czech Republic also reported 100% contamination with OTA, with mean levels of 89.6 and 15.8 µg kg −1 and ranges from LOD-990 and 3.8-36.7 µg kg −1 , respectively (Pietri et al. 2010).Another survey from China reported only 5% of the samples were contaminated with OTA out of 21 tested samples, ranging from LOD to 26.1 µg kg −1 .Ahmad et al. (2014) tested 4 samples for OTA in Pakistan and 3 of them contained OTA, ranging from LOD-13.1 µg kg −1 .In addition, they referred to higher contamination in Spain (253 µg kg −1 ) and Italy (990 µg kg −1 ) when compared with the present study (60.3 µg kg −1 ).The fungal species A. carbonarius is able to produce OTA at temperatures between 15°C and 20°C (Passamani et al. 2014).These environmental situations are similar to European countries, whereas the climate of Pakistan is tropical, with temperatures around 25°C.Consequently, OTA contamination was lower when compared with other countries.Furthermore, the diverse level of OTA contamination in other countries depends on various factors, such as soil type, climatic conditions, microbial flora, pre-harvest practices, and post-harvest handling.Additionally, social and commercial features of human handling of food, such as unsuitable storage and span of storage in shops and homes might be prime factors in increased mycotoxin contamination.
The above mentioned previous literature and the current study report that liquorice roots are more susceptible to OTA contamination when compared with AFs.The presence of OTA contamination might be due to the tropical and sub-tropical environment of the regions in Pakistan, with an average temperature of 23°C, more than 60% humidity in the rainy season and a 425 mm rainfall pattern (Asghar et al. 2016).Consequently, strict procedures should be implemented during the various stages of harvesting, drying, handling, packing, storing, and transport of the product to minimise fungal attack and eventually mycotoxin production.In addition, awareness and management training programmes regarding the toxicity potential of mycotoxins should be conducted on a regular basis, because OTA in liquorice roots presents a potential threat to human's health.Finally, good producing and storage conditions must be implemented along with Hazard Analysis and Critical Control Points (HACCP) food safety plan.

Table 1 .
Method validation data for aflatoxins and ochratoxin A.

Table 2 .
Recoveries of aflatoxins and ochratoxin A from spiked liquorice root samples (n = 3).

Table 3 .
Performance data of the LGC proficiency testing programme and QC samples from FAPAS during 2020.

Table 5 .
Comparison of aflatoxin contamination in liquorice root samples in several countries.

Table 7 .
Comparison of ochratoxin A contamination in liquorice root samples in several countries.