Outlook for Future Studies and Testing

While, for the alkenylbenzene methyleugenol, the bioactivation via SULTs could be described very well, corresponding experiments with SULT proficient test systems for other relevant alkenylbenzenes, such as estragole, safrole, or elemicin, are still widely lacking. Such experiments should be made up in order to better understand the influence of SULTs, thereby enabling a more realistic extrapolation to the human situation.

#### 3.3.3. Toxicity of Mixtures Is Still a Controversially Debated Issue

The Scientific Panel on food additives, flavorings, processing aids, and materials in contact with food (AFC) of EFSA commented regarding the genotoxic potential of estragole and tarragon that the modification of inherent toxicity of a naturally occurring substance by the matrix in which it is present (e.g., the herb) can be considered plausible [13]. However, the panel further stated that, besides a reduction in toxicity, effects related to additional compounds present in the corresponding matrix could also lead to unchanged as well as to increased toxicity, depending on the mode of action. Moreover, it was mentioned that research on individual substance/matrix interactions cannot be used to draw general conclusions about herbs and spices under all conditions of use, ingestion, and metabolism [13].

Discrepancies between experimental settings and real-life human consumption scenarios raise another important issue in this context. In animal studies, alkenylbenzenes, such as methyleugenol, were administered to test subjects as pure substances. However, this does not reflect the eating habits of the consumer, who mainly consumes methyleugenol via herbs and spices [182,183].

The group of Rietjens could demonstrate that flavonoids such as nevadensin—which, in addition to methyleugenol, is also present in certain matrices, such as herbs and spices exhibit SULT-inhibiting effects. In animal studies, it was shown that methyleugenol-derived

DNA adduct levels were lowered by simultaneous administration of methyleugenol and nevadensin in the liver of rats [137]. These experimental findings may indicate that such matrix-derived effects should be considered in the evaluation of genotoxicity studies to reliably assess the risk of adduct formation in humans. For methyleugenol, however, adducts could also be detected in human lung and liver samples [84,85,184] raising the question whether possible matrix effects would be sufficient to protect humans against methyleugenol derived mutagenicity. This should be taken seriously, especially since a copy number variation in humans exists for SULT1A1—the SULT form with the highest activity towards 1- -hydroxymethyleugenol [77]. An association of methyleugenol-mediated DNA adducts in human livers with this copy number variation and their expression levels has already been shown [85].

Together, these data and arguments show that the toxicity of mixtures is still a controversially debated issue. Therefore, further studies are needed to shed more light on this controversially debated issue.

#### 3.3.4. Transferability of Findings in Animal Studies to Human

Experimental animal models (e.g., mice or rats) are typically used to study toxicological effects of substances occurring in food, such as alkenylbenzenes, to assess their potential impact on human health through the oral intake of food [96,147,185]. This is in line with current recommendations of international scientific bodies, such as EFSA or OECD, published in corresponding testing guidelines [186,187]. However, the utilization of experimental animals and the transfer or extrapolation of the obtained results to a human setting poses different problems, leading to uncertainties regarding data interpretation and assignability.

Results of rodent studies regarding safrole showed that genotoxic effects were mediated via its active metabolites, such as the proximate carcinogen 1- -hydroxysafrole or, rather, 1- -sulfoxysafrole [80,98,102]. However, findings of a comparative study performed in rats and humans showed that the safrole metabolite 1- -hydroxysafrole was only found in rat but not in human urine [50]. These findings suggest species-specific differences regarding the metabolisms of rodents vs. humans, and—as discussed by Bode and Dong in 2015—this discrepancy raises the question of whether the genotoxic effects observed in experimental animals are also expectable in humans [101]. In line with this interpretation, absence of carcinogenic alkenylbenzene metabolites in human urine was used by Smith et al. as argument against a potential cancer risk to humans through the consumption of food containing methyleugenol and estragole [107]. However, such results might be influenced by the study design, e.g., by the dose administered. Indeed, the formation of 1- -hydroxy metabolites is also possible at relevant dose levels in humans, as the 1- -hydroxy metabolite of estragole has already been detected in the urine of human volunteers after drinking fennel tea [58]. Moreover, *N*-acetyl-*S*-[3- -(4-methoxyphenyl)allyl]-L-cysteine the mercapturic acid formed from 1- -sulfoxy estragole—was also found in the urine of human volunteers after drinking fennel tea [52], and DNA adducts of 1- -sulfoxymethyleugenol were detected in human liver samples [84,85]. This indicates that the intake of relatively low doses of alkenylbenzenes via food in humans may lead to the generation of instable cations. In addition, it should be noted in this context that a lack of certain metabolites in urine gives no information regarding the presence of metabolites, especially of phase I intermediates, in liver or other metabolizing tissues. In addition, metabolites with certain structural characteristics, such as some sulfates or even carbo cations, may have only a short half-life due to their reactivity and are therefore difficult to detect analytically.

Differences between rodent and human metabolism influencing potential genotoxic effects of alkenylbenzenes were also indicated by others. Sulfoconjugation-mediating SULTs are known to play an important role in the generation of ultimate carcinogenic metabolites of different compounds, including alkenylbenzenes (e.g., metabolites of safrole and methyleugenol) or heat-induced food contaminants, such as furfuryl alcohol [77,89,188]. Under the conditions of the test system used, human SULT1A1 and murine Sult1a1 activated the test compounds at lower concentrations than other members of the SULT family did [77,188]. In this context, the efficacy of human SULT1A1, regarding the activation of methyleugenol, was demonstrated in DNA adduct studies to be higher than that of its murine orthologue in vitro and in vivo [77,180]. Of note, quantitative comparisons should be handled with caution, as the level of adduct formation depends on the level of SULT enzymes. SULT expression varies in a tissue-specific manner and depends on the selected species (e.g., transgenic humanized versus wild-type mice). Nevertheless, these data indicate that SULTs may influence the genotoxic effects of alkenylbenzenes and other genotoxic compounds in a species-dependent manner. Further substantiating this, Al-Malahmeh and colleagues also described, in 2017, species-specific differences between rat and human regarding the metabolism of the alkenylbenzene myristicin and generation of its genotoxic 1- -sulfoxy metabolite. However, physiologically based kinetic modelling indicated that these differences were within a default factor of four [89].

Therefore, it has not yet been fully clarified to what extent carcinogenicity data from animal studies regarding alkenylbenzenes is transferable to humans.

#### **4. Conclusions**

In this review, we summarized several aspects regarding the occurrence, toxicokinetics, and toxicity of alkenylbenzenes.

The currently available information summarized in this article clearly show that a number of different alkenylbenzenes, such as safrole, methyleugenol, and estragole, have genotoxic and carcinogenic properties. Although the toxicological relevance for these well-investigated derivatives is still under discussion when these substances are taken up in low amounts via herbs and spices, it seems very clear, from a toxicological point of view, that high intake levels—as may result from specific plant food supplements, for example—should be avoided.

However, there are still several uncertainties impeding a reliable evaluation of the health risks that may result from the intake of different alkenylbenzenes via food. These uncertainties are based on the following data gaps, which need to be closed by appropriate research:


The aforementioned uncertainties and associated discussions underline that it is currently not possible to perform a conclusive evaluation of possible adverse effects to human health related to the consumption of alkenylbenzene-containing foods.

**Author Contributions:** Conceptualization, A.E. and B.S. (Bernd Schäfer); writing—original draft preparation, A.E., M.E.G., B.S. (Benjamin Sachse), B.H.M., K.H. and B.S. (Bernd Schäfer); writing review and editing, A.E., M.E.G., B.S. (Benjamin Sachse), B.H.M., K.H., and B.S. (Bernd Schäfer). All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
