Assessment of urinary biomarkers of mycotoxin exposure in adults from Cameroon

In Cameroon dietary staples are contaminated with diverse toxic fungal metabolites, known as mycotoxins. Aatoxins and fumonisins are of particular public health concern, particularly in relation to cancer and/or early life stunting. Mixtures of these toxins are predicted from food measures, and in this work, the levels and frequencies of urinary mycotoxin biomarkers are reported in Cameroonian adults. were tested for eight distinct mycotoxins using measures of both parent compounds and/or their metabolites by Liquid Chromatography tandem mass spectrometry (LC-MS/MS). the relationships between dietary and study to determine the levels of urinary biomarkers of exposures in male and female adults in the city This study supports recent observations of frequent mycotoxin co-exposures in African populations based on urinary measures. This study observed 11 mycotoxin analytes, in 89 urine samples, while earlier studies in Nigeria, Cameroon, South Africa and Cameroon reported eight, eleven, four and seven mycotoxin analytes, respectively [38, 30, 39, 28], in roughly similar sized studies. The mean (maximum) concentration of AFM 1 [0.03 (0.21) µg L -1 ; 42%] in urine analysed in the present study was similar, albeit lower, compared with the mean (maximum) levels of AFM 1 previously reported in adult urine from Cameroon [0.05 (1.38) µg L -1 ; 10%] [30] and urine from households in Nigeria [0.3 (1.5) µg L -1 ; 14.2%] [38]; however, the AFM 1 incidence was higher in our present study than in the two previous reports. The FB 1 concentrations were also similar in the present study (mean 0.43 (max 0.83) µg L -1 , 10%) compared to (mean 0.33 (max 9.54) µg L -1 , 3%) previously reported in Cameroon [30], though the maximum level was somewhat higher. The detected mean (maximum) amounts of FB 1 in our study were, however, lower than the mean (maximum) levels of FB 1 [4.6 (12.8) µg L -1 ; 13.3%] reported in a Nigerian population [38]. These differences should not be over-interpreted given the relatively small numbers of samples involved.


Background
In Cameroon dietary staples are contaminated with diverse toxic fungal metabolites, known as mycotoxins. A atoxins and fumonisins are of particular public health concern, particularly in relation to cancer and/or early life stunting. Mixtures of these toxins are predicted from food measures, and in this work, the levels and frequencies of urinary mycotoxin biomarkers are reported in Cameroonian adults.

Methods
A single rst void urine sample was collected from 89 adults (aged range: 28-85, male 39, female 50) from the city of Yaoundé, Centre Region, Cameroon. Urines were tested for eight distinct mycotoxins using measures of both parent compounds and/or their metabolites by Liquid Chromatography tandem mass spectrometry (LC-MS/MS).

Results
Altogether seven distinct mycotoxins, a atoxin, fumonisin, deoxynivalenol, zearalenone, nivalenol, ochratoxin A and citrinin, (or their metabolites) were observed in urine samples. At least one mycotoxin was detected in all of the urine samples, 87 (97.8%) of which were above the method's quanti cation limit. A atoxin M 1 was detected in 42% (n.d. − 0.21 µg L − 1 ) of samples of which about a quarter additionally contained fumonisin B 1 . Of the remaining toxins deoxynivalenol, zearalenone, ochratoxin A, nivalenol and citrinin were present in 78%, 99% 95%, 53%, and 87% of the samples respectively. Alternariol was not detected in any sample. Mixtures of mycotoxins in the samples were frequently observed with 64 samples (72%) containing more than ve mycotoxins. Estimates of intake exceeded the TDIs for fumonisin (n = 4), deoxynivalenol (n = 1) and zearalenone (n = 2), no TDI is set for a atoxin.

Conclusions
This study reveals frequent co-exposure of Cameroonian individuals to a complex mixture of toxic and carcinogenic mycotoxins, with mixtures of a atoxin and fumonisin a particular priority from a public health standpoint.

Background
Foodstuffs around the world are frequently contaminated during crop growth and/or storage by several toxigenic fungi that produce poisonous secondary metabolites, known as mycotoxins. Poor agricultural techniques, under favourable conditions including temperature and relative humidity often exacerbate toxin production in many world regions [1,2]. In sub-Saharan Africa, poor and prolonged storage additionally contribute to the burden of crop contamination for some mycotoxins, such as a atoxins.
Many mycotoxins are heat resistant, such that traditional cooking practices with grains or nuts have little impact on plate-ready concentrations [3,4]. Consequently, mycotoxin exposure seems inevitable in many parts of rural Africa.
Acute and chronic ingestion of mycotoxins can be harmful to health, and on occasions be fatal [5].
Mycotoxins such as a atoxin B 1 (AFB 1 ) are proven human carcinogens [6]. Additionally, AFB 1 and other mycotoxins are implicated in a range of other conditions including kwashiorkor in children [7], infant stunting (AFB 1 and fumonisin B 1 , FB 1 ) [8,9,10], immunosuppression (AFB 1 and deoxynivalenol, DON) [11,12,13,14,15], and neural tube defects (FB 1 ) [16,17]. The diversity of these health effects creates signi cant concerns regarding the need to monitor and assess the potential risks posed to consumers of mycotoxin-tainted foods in regions where mycotoxin-prone crops are frequently consumed [8]. Maximum tolerable limits (MTLs) in food for some mycotoxins such as a atoxins (AFB 1 , and total AFs), fumonisins (FB 1 , and total FBs), ochratoxin A (OTA), deoxynivalenol (DON), and zearalenone (ZEN) have been established [18,19,20,21]. Tolerable daily intake (TDI) levels for some of the frequently occurring toxins including FB 1 has also been de ned by the Scienti c Committee on Food [22] and DON by the European Food Safety Agency [23], likewise ZEN [24]; but as a proven carcinogen, no recommended TDI is possible for AFB 1 .
In Cameroon, reports on mycotoxin contamination of raw and cooked foods are increasing in the past decade [7, 25, 26, 27 28, 29]. AFB 1 and FB 1 have been observed in food, and in some cases the concentrations exceeded MTLs stipulated by the Codex Alimentarius [18] and the European Union [20].
Limited data are available that measure multiple urinary markers of mycotoxin in exposure in Cameroon, including 175 adults [30] and 220 young children [28]. Here we report data from 89 Cameroon adults by a highly sensitive LC-MS/MS method to add to the growing data sets on individual mycotoxin measurements in Sub-Saharan Africa.

Study populations, recruitment of participants and sample collection
This study was carried out in 2013 in the city of Yaoundé, Centre Region, Cameroon. Targeted subpopulations were informed about the nature of the study. Signed informed consent forms were obtained from 89 adult male (n = 39) and female (n = 50) volunteers (age range: 28-85 years, body weight (range: 55-129 kg)) recruited in this study. Ethical approval was received from the Cameroon National Ethics The urine samples were immediately frozen and transported on dry ice to BOKU/IFA-Tulln, Austria, where eluent B was raised to 95% until min 18 followed by a hold-time of 4 min and subsequent 3 min column re-equilibration at 90% A. The ow rate was set to 100 µL min − 1 . After injection of 10 µL the needle was washed for 20 sec to minimize carry-over. The column e uent was transferred either to the mass spectrometer (minutes 5 to 22.5) or to the waste via a six-port valve. The analytes were separated on column at 35 °C.
ESI-MS/MS was performed in scheduled multiple reaction monitoring (sMRM) mode, with a 180 sec detection windows. At least two individual transitions were monitored for each analyte. One chromatographic run consisted of two MS/MS experiments where both ionization modes run simultaneously using fast polarity switching. All measurements were conducted using: source temperature 550 °C, curtain gas 30 psi (69 kPa of max. 99.5% nitrogen), ion source gas 1 (sheath gas) 80 psi (345 kPa of nitrogen), ion source gas 2 (drying gas) 80 psi (345 kPa of nitrogen), collision gas (nitrogen) high. Ion spray voltage was − 4500 V in negative mode while it was set to 4500 V in positive mode.

Analysis of creatinine in urine samples
The concentration of urinary creatinine was determined on the same instrument by the rapid method described in [32]. Urinary mycotoxin concentrations were adjusted for creatinine and expressed as microgram of analyte per gram (µg g − 1 ) creatinine.

Results
There were no differences by hypertensive status in the age, weight, height or BMI of the participants (see supplement Table S1). In Table 1 basic anthropometric parameters are provided for all study participants. Table 1 Anthropometric measures of the studied population, n = 89 Mycotoxin concentrations were measured in urine samples from 89 adult Cameroonians (males, 39 and females, 50). The LC-MS/MS method measured seven parent mycotoxins and ve mycotoxin metabolites in the samples, limit of quantitation (LOQ) and limit of detection (LOD) values vary for each toxin and metabolite. The parent mycotoxins were FB 1 , OTA, DON, ZEN, NIV, CIT, and AOH, whilst AFM 1 (from AFB 1 ), DOM-1 (from DON), α-ZEL and β-ZEL (from ZEN), and DHC (from CIT) constituted the metabolic products. One or more of the mycotoxins or metabolites was observed in all 89 samples, with 87/89 samples having at least one of the toxins at a concentration above the LOQ (Table 2). AOH was not detected in any of the urine samples. AFM 1 was observed above the LOQ in 42% of samples (LOQ 0.001 µg L -1 : overall range n.d. − 0.210 µg L -1 ), and FB 1 was observed above the LOQ in 10% of samples (LOQ 0.15 µg L -1 : overall range n.d. − 0.83 µg L -1 ). Of the other mycotoxins, total DON (72%), total CIT (80%), OTA (80%) and total ZEN (82%) were observed most frequently above the LOQs, see Table 2.  Table 3. Intakes of each mycotoxin can be roughly estimated using the individual mycotoxin concentration (including parent and metabolites), an estimated average urinary output of 1.5 litre per day, mean estimates of transfer of the mycotoxins from the diet to urine, and individual bodyweight. In Table 4 [35]. b : Average of two mean urinary excretory FB 1 values: 0.5% [36] and 0.075% [37]. c : [33]. d : [34]. however, the AFM 1 incidence was higher in our present study than in the two previous reports. The FB 1 concentrations were also similar in the present study (mean 0.43 (max 0.83) µg L -1 , 10%) compared to (mean 0.33 (max 9.54) µg L -1 , 3%) previously reported in Cameroon [30], though the maximum level was somewhat higher. The detected mean (maximum) amounts of FB 1 in our study were, however, lower than the mean (maximum) levels of FB 1 [4.6 (12.8) µg L -1 ; 13.3%] reported in a Nigerian population [38]. These differences should not be over-interpreted given the relatively small numbers of samples involved.
DON (and its derivative DOM-1), ZEN (and its metabolites: α-ZEL and β-ZEL) and OTA were detected in urine, typically at higher frequencies than AFM 1 and FB 1 . Total DON was detected about twice as frequently (76%) in this study compared to an earlier Cameroon study [30], and much more frequently than in Nigeria (5%), where children rather than adults dominated the exposure [38]. In South Africa, a similar high frequency (100%) of total DON was reported as observed in the current study [39]. In these earlier studies, the mean concentrations were typically around 5-15 µg L -1 , and this is in line with many studies in regions outside of Africa [40,41]. However, while the mean [17.8 µg L -1 ] is similar, one individual sample [760 µg L -1 ] was notably higher in the current study than most previously reported HBM studies. Notwithstanding, the major metabolite of DON in human urine, DON-15-glucuronide [42,34] was not measured directly in this study as enzymatic deconjugation was applied [32].
The mean (maximum) concentration of OTA and ZEN were relatively lower in this study compared to previously reported data from Cameroon [30] and Nigeria [38]. However, the extremely high detection rate of 82% for total ZEN is somehow worrisome given the high xenoestrogenic potential of ZEN and its phase I biotransformation products [43]. Recent studies further highlighted that ZEN is prone to synergistic mixture effects [44,45] and able to pass the placental barrier and thus exposure of mothers is likely to result in in utero exposure of the unborn child [46]. The impact of this chronic low-dose exposures on the endocrine system and related disease should be investigated in future studies.
The mean NIV level recently reported in a Nigerian study [32] was approximately 10 times greater than the level reported here for the Cameroonian population. Urinary CIT and its metabolite, DHC, were quanti ed in this study for the rst time in Cameroon. The detected mean (maximum) concentration of total CIT [2.3 (98) µg L -1 ; 80%] in this present study were lower than those in Nigeria [6.0 (241) µg L -1 ; 66%] [32], although our study had higher incidence. Comparison of urinary mycotoxin concentrations by either sex, or by hypertensive status did not reveal any signi cant differences (p < 0.05), noting limited study size would preclude meaningful comparisons.
One urine sample contained only one mycotoxin, while 20 combinations of two or up to seven mycotoxin urinary biomarkers were observed; more than 70% of the urines contained ve or more different mycotoxins. Complex mixture toxicology remains poorly examined though several groups have recently examined combined effects in vitro [47,48,49,50,51,52,53,54,55,56,57,44,45], with animal studies being more limited [58,59]. These studies remain hard to interpret for public health decisions, but some suggest more than additive effects, thus the mixtures reported here and elsewhere highlight signi cant knowledge gaps. It will be important to conduct longitudinal studies to better understand typical patterns and seasonal variation to better inform our understanding of mixture exposures. An interesting example for such longitudinal mycotoxin co-exposure assessment was recently published for an infant that was exclusively fed by breastmilk, which was tested for 29 mycotoxins [60]. However, it will be even more relevant to consider other food-and environment-related exposures beyond mycotoxins as proposed by the exposome concept [61,62,63].
From the mean (maximum) levels of some of the major urinary mycotoxins in this present study, an estimated average dietary exposure was calculated on the basis of each participant's estimated dietary exposure using each participant's urine mycotoxin exposure amount, individual weight, an assumed 1.5 L urinary output per day and estimated urine excretion rate for each mycotoxin. For data with urinary concentration below the LOQ, either half the LOQ or half the LOD was used. This is generally used in food safety risk assessment (e.g. by European Food Safety Agency, EFSA) as it provides conservative estimates for calculation of exposure assessment [64]. Any dietary AFM 1 is considered to be of concern, as no exposure level of AFM 1 is tolerated based on the conclusions of the Scienti c Committee on Food [65,22]

Conclusion
This study has further revealed that mycotoxin exposure is prevalent in city of Yaounde, Centre Region, Cameroon. This is evident by the detection of 11 mycotoxins (seven mycotoxins representative of AF, FB, DON, ZEN, NIV, OTA and CIT, and four of their derivatives: α/β-ZEL, DOM-1 and DHC) in 89 adult urines in this region. Most importantly is that every single urine sample contained at least one mycotoxin. For the rst time urinary CIT and its metabolite, DHC, were quanti ed in urine samples from Cameroon. The coexistence of as much as seven mycotoxins in up to 20 different patterns worsens the scenario and predicts potential health risk for the population. The presence of a atoxin biomarkers in 42% of samples and of those about a quarter additionally contained fumonisin B 1 is a concern. The potential risk derived from additional mixture effects remains poorly de ned, but as further studies add to these data sets their putative contributions may be understood, while a atoxins and fumonisins remain a priority in populations such as Cameroon with a high incidence of liver disease [67,68,69] and stunting [70,71].

Competing interests
The authors declare they have no competing nancial interests. Funding