Cytochrome P450-dependent biotransformation in (eco)toxicology

“Xenobiotic biotransformation” refers to the process by which a compound foreign to an organism is converted into a usually more polar, i.e. more hydrophilic, and thus more readily excretable metabolite (Parkinson et al. 2013). Biotransformation is conventionally divided into two distinct phases: Phase I reactions are functionalization reactions, which serve to incorporate functional groups into the molecular structure of a xenobiotic compound or expose (demask) already existing polar groups (Parkinson and Ogilvie 2008; Penner et al. 2012). This is achieved via oxidation, reduction or hydrolysis reactions catalyzed, e.g., by alcohol dehydrogenases, epoxide hydrolases, flavin-containing monoaminooxidases and by cytochrome P450-dependent monooxygenases (CYPs; Chen 2020; Penner et al. 2012; Testa 2008). Phase I metabolites may be the final products ready for excretion, but usually undergo phase II biotransformation in which already existing, incorporated or exposed functional groups serve as active sites for conjugation with endogenous polar molecules (Parkinson et al. 2013).

CYP enzymes constitute a functionally diverse superfamily of cysteine thiolate-ligated heme enzymes. Present in most tissues and organs, CYPs are by far the predominant phase I biotransformation enzymes. They catalyze aromatic and aliphatic hydroxylation, azo reduction, desulfuration, epoxidation, N-hydroxylation, O- and N-dealkylation, nitro reduction, oxidative dehalogenation and sulfoxidation on a broad range of compounds, ranging from small non-polar molecules to complex polypeptides (Guengerich 2001; Isin and Guengerich 2007; Qiang and Lu 2014).

Based on amino sequence similarities, CYPs are clustered into CYP families and subfamilies. CYPs sharing a minimum of 40% amino sequence similarity are grouped within the same CYP family (e.g., CYP1, CYP2, CYP3), and those sharing at least 55% similarity are grouped within a CYP subfamily (e.g., CYP1A, CYP2B, CYP3C; Nelson 2006; Nelson et al. 1993). Whereas most isoforms of the CYP families 1, 2, 3 and 4 primarily act on xenobiotic compounds (Nebert and Russell 2002; Rendic and Guengerich 2015; Taavitsainen 2002), most isoforms of the CYP families 5–51 predominantly act on endogenous substrates, many of which have critical roles in normal development, maturation and physiological homeostasis (Guengerich 2017; Nebert et al. 2013).

CYPs do not only modify physicochemical characteristics of their substrates, but may also impact their (eco)toxicological properties. Given the broad spectrum of substrates accepted by at least part of the CYP isoforms, the competition of different CYP isoforms for substrates and the complexity of CYP-mediated transformations, CYPs are involved in both detoxification and toxification processes: CYPs usually increase the polarity of xenobiotic compounds and enable phase II biotransformation and, thereby, contribute to an efficient excretion and detoxification (Guengerich 2003; Suter 2008). However, there are numerous examples for CYP-catalyzed reactions which result in the formation of metabolites or intermediates that are more reactive, teratogenic or even carcinogenic and/or toxic than the parent compounds (i.e. pro-carcinogens, protoxicants and proteratogens), a process referred to as xenobiotic bioactivation (Smith and Brian 1991; Stiborova et al. 1992; Weigt et al. 2011).

Over the past two decades, the zebrafish (zf, Danio rerio) and − specifically − zebrafish embryos (≤ 120 h post-fertilization; hpf) have emerged as pre-eminent model organisms with numerous applications not only in ecotoxicology, but also in toxicology and pharmacology (Bambino and Chu 2017; Barros et al. 2008; Hill et al. 2005; Kithcart and MacRae 2017; McGrath and Li 2008). At least in Europe, early developmental stages of zebrafish have received particular attention, since these are not regarded protected according to current EU animal welfare legislation (EU 2010; Strähle et al. 2012). Current examples of the use of zebrafish embryos within validated test protocols include the fish egg test (DIN 38,415–6; ISO 2016), which is a mandatory stand-alone component in routine whole effluent toxicity testing in Germany (Bundesgesetzblatt 2005; Norberg-King et al. 2018), and the fish embryo toxicity test (FET, OECD TG 236), which was originally designed for determination of the acute toxicity of chemicals on embryonic stages of fish (Busquet et al. 2013), be it as a full replacement for the acute fish toxicity test (AFT; OECD; TG 203) conducted with juvenile and adult (zebra)fish to provide data for regulatory decision making (Braunbeck et al. 2015; Scholz et al. 2013) or as a central component of weight-of-evidence approaches (ECHA 2017; Lillicrap et al. 2020; Moe et al. 2020; Paparella et al. 2021). In fact, the identification of a number toxicological outliers (i.e. ~ 30 compounds with more than ten times lower toxicity in the FET than in the AFT, e.g., allyl alcohol, cyclohexane, nonylphenol, dieldrin and permethrin (Klüver et al. 2014, 2015) has drawn attention to potential limitations of the FET and has led to criticism concerning its regulatory use as surrogate for the AFT (Sobanska et al. 2018).

Concerns have been raised that, due to potential limitations in phase I biotransformation and bioactivation capacities of zebrafish embryos, there might be a risk of underestimation the toxicity that pro-toxicants might pose to juvenile or adult fish, but not to embryos (Busquet et al. 2008; Saad et al. 2016b; Verbueken et al. 2017). Since then, there is an ongoing debate whether or not zebrafish embryos have sufficient capacities to biotransform and bioactivate xenobiotics. The clarification of this aspect is, however, of particular relevance for our understanding of the capabilities and limitations that eventually define the applicability domain of the FET (Sobanska et al. 2018).

Over the past two decades, the CYP system, and in particular the 56 isoforms identified for the zebrafish CYP families 1, 2, 3 and 4 (Table 1), have been subject to many studies that have led to a more detailed and sophisticated understanding of the development of the CYP-dependent biotransformation and bioactivation capacities in zebrafish. These studies provided insights into developmental CYP mRNA expression patterns, tissue- and organ distribution of CYP transcripts, metabolic activities towards fluorogenic and luminogenic mammalian CYP probe substrates, bioactivation activities towards pro-toxicants and pro-teratogens and xenobiotic metabolite formation. This review provides an in-depth coverage of the current state-of-knowledge on the CYP1-4 families and phase I biotransformation and bioactivation capacities of zebrafish in general, and zebrafish embryos in specific. Data will be critically discussed and, whenever possible, compared with findings for other fish species. Thereby, this review will identify gaps in our knowledge and discuss future directions for research.

Table 1 List of zebrafish CYP1, CYP2, CYP3 and CYP4 genes (GRCz11 assessed by Ensembl genome browser; release 100)

Literature sources used for this review

Relevant studies were identified through searching the following databases: Science Direct, PubMed, Scopus, Google Scholar and Web of Science using the keywords bioactivation, biotransformation, cytochrome P450, CYP, Danio rerio, expression, fish, metabolism, metabolite, phase I, pro-teratogen, pro-toxicant, pro-mutagen, xenobiotic and zebrafish as well as combinations thereof. In addition, reference lists of relevant articles were manually searched for further potentially relevant publications. No restrictions were made regarding the date and language of the publication. The searches were undertaken between November 2017 and March 2021 and yielded an array of publications on the cytochrome P450 system and phase I biotransformation, including peer-reviewed studies, reviews, original research articles and academic theses.

Spatial and temporal CYP gene expression patterns

Most of the current knowledge regarding CYP-dependent phase I biotransformation capacities in different developmental stages of zebrafish has been derived from studies profiling temporal (i.e. developmental) and spatial (i.e. tissue and organ distribution) mRNA expression patterns of the 56 genes identified for the zebrafish CYP families 1, 2, 3 and 4 (Saad et al. 2016a). These studies have built up a comprehensive data set, comprising information on developmental expression patterns of the full complement of zebrafish CYP1-4 genes as well as on organ- and tissue-specific expression patterns of 20 CYP1-4 genes. Figure 1 gives an overview of the number of CYPs whose temporal trends in constitutive expression levels have been studied in zebrafish across different stages of development (for details, Table 2). Table 3 presents a collection of organ- and tissue-specific CYP expression patterns reported for embryonic (≤ 120 hpf), juvenile (≥ 120 hpf, ≤ 3 mpf) and adult zebrafish (> 3 mpf), respectively.

Fig. 1
figure 1

Numbers of CYP1-4 isoforms whose constitutive expression patterns have been determined throughout embryonic, juvenile and adult development of the zebrafish (Danio rerio). Illustrations by Karlotta Boßung

Table 2 Information available on the mRNA expression of CYP1, CYP2, CYP3 and CYP4 genes in embryonic, juvenile and adult zebrafish (Danio rerio)
Table 3 Tissue and organ distribution of zebrafish CYP1, CYP2, CYP3 and CYP4 families. mRNA transcripts in embryonic, juvenile and adult stages of zebrafish (Danio rerio)

Besides the use as an indicator of metabolic competence, data on spatiotemporal CYP 1–4 expression patterns can also be informative on characterizing functional (i.e. physiological) roles of CYP isoforms in zebrafish. For instance, the relatively high CYP3C1 expression level found in the brains of 48 and 120 h old zebrafish embryos (Corley-Smith et al. 2006), the lack of CYP3C1 mRNA in the brains of 21 d old zebrafish (Taylor 2005), and its prominent expression in the main xenobiotic-metabolizing organs, i.e. intestine and liver, in adult zebrafish (Corley-Smith et al. 2006) make it is reasonable to assume that the function of the CYP3C1 might not only be related to xenobiotic biotransformation, but also to brain development in early life-stages of zebrafish.

Spatiotemporal expression patterns of the CYP1 family

The zebrafish CYP family 1 contains five CYP genes, i.e. CYP1A, CYP1B1, CYP1C1, CYP1C2, and CYP1D1 (GRCz11, see Table 1), which all differ in their developmental expression patterns (Goldstone et al. 2009, 2010; Jönsson et al. 2007a; Verbueken et al. 2018), tissue and organ distributions (Jönsson et al. 2007b), responses to xenobiotic inducers and inhibitors (Jönsson et al. 2007a) and catalytic activities towards xenobiotic and endogenous compounds (Scornaienchi et al. 2010a, b; Stegeman et al. 2015). This suggests that each may have distinct physiological functions and/or roles in xenobiotic biotransformation.

For all genes of the CYP1 family, transcripts have been detected whenever studied, i.e. in zebrafish from at least 3 hpf onwards (Table 2; Goldstone et al. 2010; Jönsson et al. 2007a; Verbueken et al. 2018). CYP1A is the only member of the CYP family 1 for which transcripts have also been detected in unfertilized zebrafish oocytes, indicating a maternal transfer of CYP1A mRNA to the embryo (Goldstone et al. 2010; Verbueken et al. 2018). The constitutive expression of CYP1A fluctuates during the first 48 hpf (Glisic et al. 2016; Goldstone et al. 2010) and considerably increases around hatching, reaching a first peak in late embryogenesis (96–120 hpf; Glisic et al. 2016; Jones et al. 2010; Verbueken et al. 2018). Controversial observations exist on CYP1A expression patterns in juvenile zebrafish, making it difficult to draw conclusions about the potential existence of differences in the extent of xenobiotic biotransformation capacities of zebrafish embryos and juveniles. While Verbueken et al. (2018) documented CYP1A to be constitutively expressed at relatively stable levels from late embryogenesis until 30 dpf, both Jönsson et al. (2007a) and Goldstone et al. (2009) documented CYP1A expression levels to fluctuate throughout embryonic and juvenile development with peaks reached at 21 dpf (Jönsson et al. 2007a) and 57 dpf (Goldstone et al. 2009). The reasons for these variable results are not obvious given that all studies used quantitative real-time PCR as method for profiling CYP1A expression in wild-type zebrafish. In zebrafish embryos at 30 hpf, CYP1A is constitutively expressed in many organs and tissues across the whole body, including eyes, heart, intestine, skin, fin bud, cloaca, intersegmental blood vessels and at lower levels also in otic vesicles and the brain (cf. Table 3; Kim et al. 2013). This is in contrast to juvenile and adult zebrafish as well as other fish species, e.g., scup (Stenotomus chrysops; (Stegeman et al. 1991) and turbot (Scophthalmus maximus; Reinecke and Segner 1998), where CYP1A is most abundantly expressed in the liver and intestine (Goldstone et al. 2009; Jönsson et al. 2007b; Taylor 2005), the major organs relevant to xenobiotic biotransformation. However, expression of CYP1A does not only vary with age, but also with sex. While CYP1A mRNA accounts for 14.5% of the total hepatic CYP mRNA contents in male zebrafish, it accounts for only 5% in female zebrafish (Kubota et al. 2019).

The constitutive expression level of CYP1B1 increases immediately after activation of the embryonic genome, reaching a peak level in zebrafish embryos at 36 hpf (Verbueken et al. 2018). This level is not reached again in any of the later embryonic or juvenile stages studied so far (i.e. up to 30 dpf; Verbueken et al. 2018). A very similar trend, albeit with a peak in expression reached at 3 dpf and almost negligible expression levels found in juvenile zebrafish at 57 dpf, was reported by Jönsson et al. (2007a). In zebrafish embryos, beginning at 24 hpf, CYP1B1 transcripts have been localized in ocular cells (Yin et al. 2008), where maximum levels are reached between 30 and 48 hpf (Yin et al. 2008), which coincides with the start of cardiac looping (Bakkers 2011) and the onset of melanin synthesis in the retinal pigment epithelium (Glass and Dahm 2004). Moreover, CYP1B1 is expressed in the embryonic retina and midbrain–hindbrain boundary regions, but not in branchial arches, the kidney and fin buds (Yin et al. 2008). Adult zebrafish constitutively express CYP1B1 in a variety of organs with highest levels in brain, eyes and heart. Lowest levels were documented in gonads and intestine (Jönsson et al. 2007b). These spatiotemporal expression patterns led Jönsson et al. (2007b) to suggest that CYP1B1 may primarily have physiological functions in zebrafish (Jönsson et al. 2007b). However, heterologously expressed zebrafish CYP1B1 could be demonstrated to catalyze oxidative biotransformation of a number of xenobiotic compounds including O-alkyl derivates of resorufin and coumarin (Scornaienchi et al. 2010a, 2010b; Stegeman et al. 2015). CYP1B1 might thus also contribute to extrahepatic biotransformation processes in zebrafish.

Discrepancies exist in literature regarding the developmental expression trends of the two paralogous genes CYP1C1 and CYP1C2. Jönsson et al. (2007a) documented constitutive expression levels of both CYP1C genes to fluctuate during embryonic and juvenile development, with both genes reaching minimum levels in zebrafish at 6 dpf and CYP1C2 again at 57 dpf. In contrast, in a more recent study, Verbueken et al. (2018) found expression levels of CYP1C1 and CYP1C2 to steadily increase during the first 5 to 10 dpf, then leveling off until 30 dpf. This pattern led Verbueken et al. (2018) to suggest that the biotransformation capacity might be immature during early development of zebrafish. In adult zebrafish, both genes are predominantly expressed in the brain, eyes and heart. CYP1C1 is also expressed in gills and CYP1C2 in the kidney (Jönsson et al. 2007b). Both were thus suggested to primarily have physiological functions (Jönsson et al. 2007b). However, as CYP1C1 and CYP1C2 have also been demonstrated to metabolize benzo[a]pyrene (B[a]P) to its metabolite B[a]P-7,8-diol-9,10-oxide, which involves the intermediate formation of the ultimate carcinogenic form of B[a]P (i.e. B[a]P-7,8-diol-9,10-oxide; Stegeman et al. 2015), both isoforms might also be involved in extrahepatic bioactivation processes at least in adult zebrafish.

Among the genes of the CYP1 family, the constitutive expression of CYP1D1 peaks earliest (Goldstone et al. 2009, 2010). Its maximum expression level is reached in zebrafish embryos at 9 hpf and is two to three times higher than expression levels found in all later embryonic juvenile stages tested so far (1–7 dpf and 57 dpf; Goldstone et al. 2009). When describing the early expression peak, it has been hypothesized that CYP1D1 might have endogenous functions in early developmental processes (Goldstone et al. 2009). However, CYP1D1 has also been found to catalyze oxidative biotransformation and bioactivation of benzo[a]pyrene and a number of synthetic CYP probe substrates, although in most cases with a catalytic efficiency smaller than that of all other CYP1 isoforms (Scornaienchi et al. 2010b; Stegeman et al. 2015).

Spatiotemporal expression patterns of the CYP2 family

The CYP2 family, by far the largest and most diverse CYP family in zebrafish, contains 42 genes (GRCz11, Table 1), all of which are being expressed from very early stages of embryonic development, i.e. from as early as 3 hpf (Goldstone et al. 2010). In addition, transcripts of CYP2AA4 and CYP2P6 have been found in unfertilized oocytes (Goldstone et al. 2010).

Whenever expression levels of CYP2 genes have been studied in both pre-hatch (≤ 48 hpf) and post-hatch embryonic stages (> 48 hpf), for all genes examined (i.e. CYP2AD2, CYP2J26, CYP2K6, CYP2K18, CYP2N13, CYP2Y3), except CYP2R1 (Peng et al. 2017), higher expression levels have been documented in post-hatch stages (Jones et al. 2010; Nawaji et al. 2020; Poon et al. 2017b; Wang-Buhler et al. 2005). In the case of CYP2K6, transcripts were not detected in zebrafish embryos before 72 hpf, after which expression levels increased, first reaching significant levels in zebrafish embryos at 5 hpf (Wang-Buhler et al. 2005). However, up to date, no functional data are available for these isoforms, with the exception of CYP2K6, which was shown to catalyze the bioactivation of the mycotoxin aflatoxin B1 (Wang-Buhler et al. 2005). Thus, it remains to be clarified whether quantitative differences in CYP2 expression levels may have functional implications for the xenobiotic biotransformation and bioactivation capacities of zebrafish or might be a consequence of physiological processes.

So far, CYP2R1 is the only CYP2 gene, whose mRNA abundance has been studied across all developmental stages (i.e. in zebrafish embryos, juveniles and adults; cf. Table 2). The constitutive expression of CYP2R1 peaks twice, once at 9 hpf and again at 28 hpf and was, therefore, suggested to have different stage-specific functions (Peng et al. 2017). One of these could be linked to vitamin D3 biotransformation (Peng et al. 2017). In adult zebrafish, CYP2R1 is expressed at levels much lower than those in zebrafish embryos or juvenile zebrafish (Peng et al. 2017).

Although information is available on sex-differences in constitutive expression levels of a number of CYP2 genes (cf. Table 2), nothing is yet known about the impact these differences might have on the susceptibility of female and male zebrafish to xenobiotic exposure. In a study by Kubota et al. (2013), variability between female and male zebrafish, although not significant, was noted with respect to the transcript abundance of CYP2AA1 and CYP2AA1 in a number of organs (e.g., liver, gonads and kidney). CYP2J1, which has been suggested to play a role in gonadal development and ovarian follicular development (Wang et al. 2007) and CYP2K6 (Wang-Buhler et al. 2005) were both found to be constitutively expressed in gonads of adult zebrafish, however, at higher levels in female than in male zebrafish. Moreover, by transcriptional analysis of liver samples, Kubota et al., (2019) and Zheng et al. (2013) identified several CYP2 genes having sex-biased expression levels. These include CYP2N13, CYP2K6, CYP2AD2 and CYP2AA4.

Spatiotemporal expression patterns of the CYP3 family

The five genes of the CYP3 family, i.e. CYP3A65, CYP3C1, CYP3C2, CYP3C3 and CYP3C4, have been studied to varying extent (cf. Table 2). While development-related trends in constitutive expression levels of CYP3A65 and CYP3C1 were repeatedly assessed in zebrafish embryos of different ages (Chang et al. 2013; Corley-Smith et al. 2006; Glisic et al. 2016; Goldstone et al. 2010; Shaya et al. 2014; Tseng et al. 2005) and also in juvenile zebrafish up to 30 dpf (Verbueken et al. 2018), expression profiles of all other CYP3 genes have only been once determined in zebrafish older than 48 hpf (Nawaji et al. 2020).

Four different methods have been used to evaluate CYP3A65 expression in zebrafish embryos (Fig. 2). Of these, quantitative real-time PCR (qPCR) and microarray analysis proved to be the most sensitive, allowing for detection of CYP3A65 mRNA in whole-body homogenates of zebrafish from as early as 1.5 and 3 hpf, respectively (Goldstone et al. 2010; Verbueken et al. 2018). In contrast, by means of whole mount in situ hybridization and the use of a transgenic zebrafish line expressing eGFP:CYP3A65 constructs, transcripts have not been detected until 72 and 24 hpf, respectively (Chang et al. 2013; Tseng et al. 2005).

Fig. 2
figure 2

Comparison of the embryonic expression trends reported in literature for zebrafish CYP3A65. Data were generated by (a) microarray analysis (Goldstone et al. 2010), (b) qPCR (Glisic et al. 2016), (c) whole mount in situ hybridization (WISH; Tseng et al. 2005) and (d) a transgenic zebrafish line expressing CYP-eGFP constructs (TL; Chang et al. 2013)

The constitutive expression level of CYP3A65 remains relatively low until hatching, when expression of CYP3A65 starts to increase markedly, reaching a first peak in zebrafish embryos at 120 hpf (Chang et al. 2013; Glisic et al. 2016; Goldstone et al. 2010; Nawaji et al. 2020; Tseng et al. 2005; Verbueken et al. 2018). Therefore, it might be hypothesized that biotransformation processes depending on CYP3A65 are immature during early embryonic development and thus significantly different from that of juvenile and adult zebrafish. In juvenile zebrafish, i.e. between 120 hpf and 30 dpf, the constitutive expression of CYP3A65 was documented to remain at an almost stable level slightly below the peak level measured in 120-h-old zebrafish embryos (Verbueken et al. 2018). Through immunohistochemical analyses, CYP3A65 transcripts could be localized in the corpuscle of Stannius, eyes and ears of juvenile zebrafish and, at much lower levels, in the heart, intestine and kidney (Taylor 2005; cf. Table 3). This contrasts the spatial expression patterns found in zebrafish embryos and adults, where CYP3A65 transcripts were almost exclusively restricted to liver and intestine (Tseng et al. 2005). In adult female zebrafish, CYP3A65 is the most abundantly expressed hepatic CYP isoform, making up 13.4% of the total amount of CYP mRNA. It was, therefore, suggested to play a central role in liver physiology and/or xenobiotic biotransformation (Kubota et al. 2019). However, in male zebrafish CYP3A65 mRNA accounts for only 6.5% of the total hepatic CYP mRNA content (Kubota et al. 2019). This sex-dimorphic expression is contrary to that found in adult killifish (Fundulus heteroclitus), where male fish displayed up to 2.5 -fold higher hepatic CYP3A65 mRNA and protein than female killifish (Hegelund and Celander 2003). Whether the sex-related differences in the hepatic mRNA abundance of CYP3A65 has consequences for the metabolic competence and/or the susceptibility of male and female zebrafish to xenobiotic-exposure remains to be clarified. By characterizing the catalytic activities of heterologously expressed CYP enzymes from zebrafish, Scornaienchi et al. (2010a) could demonstrate that CYP3A65 has activities towards compounds of both endogenous (e.g., 17β-estradiol) and exogenous origin (e.g., 7-benzyloxy-4-trifluroromethyl-coumarin), albeit with an efficiency that was much lower than that of most CYP1 isoforms (Scornaienchi et al. 2010a).

Transcripts of CYP3C1 have been detected in zebrafish embryos from as early as the 4—8 cell stage (Shaya et al. 2014; Verbueken et al. 2018), which is prior to activation of the zygotic genome (Kane and Kimmel 1993). This indicates that CYP3C1 transcripts are maternally deposited into oocytes and might play a role in earliest developmental processes (Goldstone et al. 2010). The constitutive expression level of CYP3C1 fluctuates slightly during the first 48 h(Goldstone et al. 2010; Shaya et al. 2014) before starting to steadily increase until a peak in juvenile zebrafish at 10 dpf is reached (Verbueken et al. 2018). CYP3A65 is significantly higher expressed in male and female adult zebrafish than in zebrafish embryos and juvenile zebrafish (up to 144 hpf; Nawaji et al. 2020). The spatial distribution pattern of CYP3C1 is sex-biased and age-dependent (cf. Table 3). Transcripts of CYP3C1 are distributed throughout the whole body of 12-h-old zebrafish embryos, are concentrated in brains of 48-h-old zebrafish embryos and additionally appear in the intestine and pharynx at 120 hpf (Corley-Smith et al. 2006). In juvenile zebrafish, kidney, gills, eye, pseudobranch, and oropharynx are the major sites of CYP3C1 expression, but transcripts are also present, albeit at much lower levels, in neurons, skin, ear, taste bud, intestine and liver (Taylor 2005). In adult zebrafish, CYP3C1 is constitutively expressed in many tissues and organs, male-biased in liver, eyes and kidney, female-biased in gonads and intestine, and without significant sex-differences in brain, gills, heart, skin, spleen, and olfactory rosette (Corley-Smith et al. 2006; Shaya et al. 2014). Hence, depending on the function of CYP3C1, toxicological profiles of CYP3C1 substrates might not only vary by age, but also by sex.

Transcripts of CYP3C2 and CYP3C3 genes were detected prior to activation of the zygotic genome, i.e. from 1 and 3 hpf onwards (Goldstone et al. 2010; Shaya et al. 2014). While CYP3C2 has a bimodal expression pattern with peaks reached at 5 and 48 hpf, the constitutive expression level of CYP3C3 gradually decreases within the first hours after fertilization, eventually reaching a minimum in zebrafish embryos at 25 hpf (Shaya et al. 2014). After hatching, expression levels of CYP3C2/C3 start to markedly increase, reaching a first peak at the end of embryogenesis. Expression of CYP3C2/C3 is dependent on age and sex, with male zebrafish showing significantly higher expression levels than embryonic, juvenile (at 144 hpf) and female zebrafish (Nawaji et al. 2020). Both genes are widely expressed in several tissues and organs of adult zebrafish (e.g., brain, eyes, gills, gonads, heart, intestine, kidney, liver, olfactory rosette, and spleen; cf. Table 3), with brain, eyes and heart showing female-biased expression levels (Shaya et al. 2014).

Expression of CYP3C4 fluctuates during embryogenesis and reaches a peak level in juvenile zebrafish at 144 hpf. This peak level was documented to be approximately double that of adult female and four times that of male zebrafish (Nawaji et al. 2020). In adult zebrafish, CYP3C4 transcripts were documented in several organs and tissues, with highest levels found in brain, eyes and intestine of male zebrafish and gills, gonads and kidney of female zebrafish (Shaya et al. 2014).

Spatiotemporal expression patterns of the CYP4 family

The zebrafish CYP4 family contains four genes, i.e. CYP4F3, CYP4T8, CYP4V7 and CYP4V8 (GRCz11, see Tab 1). For none of these, information on potential roles in xenobiotic biotransformation is available. Moreover, information on developmental expression patterns of the CYP4 genes is currently limited to zebrafish embryos ≤ 48 hpf. Although transcripts of all CYP4 genes have been detected in zebrafish at 3 hpf (Goldstone et al. 2010), and for all except CYP4V8, expression peaks are reached within the first 6 hpf, no conclusions can be drawn with regard to the development of CYP4-dependent biotransformation capacities in zebrafish.

Activities towards synthetic CYP probe substrates

Metabolism of cytochrome P450 probe substrates

Given the complexity of the CYP system, a single assay or probe substrate appears to be hardly sufficient to assess the CYP-dependent biotransformation capacity of organisms. Instead, a comprehensive testing strategy covering a range of CYP activity assays is needed. In fact, during the past two decades, a variety of routinely used fluorescence- and luminescence-based assays for assessing CYP activities in mammals have been modified for in vitro and in vivo use in fish species, e.g., common carp (Cyprinus carpio; Funari et al. 1987), fathead minnow (Pimephales promela;, Boehler et al. 2018), guppy (Poecilia reticulata; Funari et al. 1987), medaka (Oryzias latipes, Funari et al. 1987), rainbow trout (Oncorhynchus mykiss; Andersson and Goksøyr 1994; Burkina et al. 2018; Smith 2009), killifish (Fundulus heteroclitus; Smith 2009) and zebrafish (Danio rerio; Chng 2013; Gonzalez-Doncel et al. 2011; Jönsson et al. 2009; Otte et al. 2017; Verbueken et al. 2017; see Tables. 4, 5). These assays rely on the use of synthetic pro-luciferin acetals and non- or low-fluorescent O-alkyl derivates of coumarin, fluorescein, quinoline and resorufin, which are metabolized by CYP enzymes into active luciferin (Cali et al. 2006) and highly fluorescent products, respectively. Probe substrates that have been used in zebrafish are listed in Table 4. Included among these are specific (e.g., luciferin isopropyl acetal; Doshi and Li (2011) and selective (e.g., 7-benzyloxyresorufin and 7-methoxy-4-trifluorormethylcoumarin; Pastrakuljic et al. 1997; Stresser et al. 2002) probe substrates of mammalian CYP1, CYP2 and CYP3 isoforms. Yet, there are no studies investigating CYP activities in (zebra)fish by using mammalian CYP4 probe substrates such as luciferin-4A or lauric acid (Yamaori et al. 2018).

Table 4 Spatiotemporal patterns of CYP-dependent activities in embryonic, juvenile and adult stages of zebrafish (Danio rerio)

Due to the limited knowledge regarding isoform specificity of mammalian CYP probe substrates in (zebra)fish (Scornaienchi et al. 2010b), the interpretation of fluorogenic and luminogenic CYP-activity assays remains challenging. An additional complicating factor is the characterization of several CYP isoforms by broad and overlapping substrate specificities (Scornaienchi et al. 2010b; Stegeman et al. 2015). Many CYPs are capable of metabolizing biotransformation reactions of different xenobiotic compounds that are not necessarily structurally related, and, vice versa, many xenobiotic compounds are metabolized in the same way by different CYP isoforms, albeit often with distinct efficiencies. For instance, heterologously expressed zebrafish CYP1B1, CYP1C1, CYP1C2 and CYP3A65 isoforms, and in particular the isoform CYP1A, have been found to be active in oxidative biotransformation (i.e. O-dealkylation) of 7-benzyloxy-4-(trifluoromethyl)-coumarin (Scornaienchi et al. 2010a, b), which is a selective, but not specific human CYP3A4 probe substrate (Stresser et al. 2002).

Not all studies provide precise information on the experimental conditions used (e.g., ‘room temperature’, no information on the probe substrate concentration; Tables 4, 5), thus rendering reproducibility and comparison between studies difficult. When experimental conditions were reported, these varied considerably: incubation time ranged from 10 min to 24 h, temperature from 20 to 37 °C and probe substrate concentration from 0.3 µM to 1 mM. Activities towards CYP probe substrates were either assessed dynamically, i.e. by measuring increases in fluorescence over time, or statically by measuring intensities of fluorescence or luminescence at a single time-point. In cases where embryonic or juvenile zebrafish were used, CYP activities were monitored both in vivo and in vitro using living organisms, subcellular fractions (e.g., S9 fractions and microsomes) prepared from whole-body homogenates or subcellular fractions prepared from different tissues and organs, such as liver, gills, muscles and brain. In cases where adult zebrafish were used, four out of five studies used microsomes prepared from liver samples, one used whole gill arches and microsomes prepared from liver samples, and one used microsomes prepared from whole-body homogenates. To date, most studies have focused on investigating the level of CYP activity in one developmental stage or even at a single point in development (Fig. 3). To the best of our knowledge, up to now, four studies are available comparing CYP activity levels across different developmental stages of zebrafish (i.e. embryo vs. juvenile and embryo vs. adult). Out of these, only one has systemically assessed and compared CYP activity levels across all developmental stages. In this study, published by Verbueken et al. (2018), benzyloxymethylresorufin-O-deethylase (BOMR) activities were assessed in microsomes prepared from whole-body homogenates of embryonic, juvenile and adult zebrafish. Furthermore, the study by Verbueken et al. (2018) demonstrated a substantial risk of underestimating CYP (i.e. BOMR) activities when directly comparing activity levels of zebrafish whole-body microsomes with those of zebrafish liver microsomes, as has been done in most previous studies comparing CYP activity levels between embryonic and adult zebrafish (e.g., Saad et al. 2016b; Verbueken et al. 2017). Hence, allowing for conclusions to be drawn concerning potential developmental-related differences in CYP activity levels, there is an urgent need for more systematic approaches as the one chosen by Verbueken et al. (2018).

Table 5 Spatiotemporal patterns of CYP-dependent activities in ecotoxicologically relevant model fish species
Fig. 3
figure 3

Number of studies characterizing CYP-dependent activities by fluorescent or luminescent-based assays in embryonic, juvenile and adult life-stages of zebrafish (Danio rerio)

Mammalian CYP1-like activities

The 7-ethoxyresorufin-O-deethylase (EROD) activity assay is the most commonly used method for analyzing the presence of CYP1 activities in vertebrates (De Almeida et al. 2011; Parente et al. 2008; Whyte et al. 2000). In zebrafish, all five enzymes of the CYP1 family (i.e. CYP1A, CYP1B1, CYP1C1, CYP1C2 and CYP1D1) are principally capable of O-deethylating 7-ethoxyresorufin (ER). Of these, CYP1A has the highest catalytic efficiency, which is 1–4 orders of magnitude higher than that of all other zebrafish CYP1s (Scornaienchi et al. 2010a, 2010b). In vitro and in vivo studies have documented the presence of EROD activity from as early as the blastula (i.e. 2.5 and 5 hpf; Otte et al. 2017; Saad et al. 2016a, b) and the gastrula stage (i.e. 7 and 8 hpf; Otte et al. 2010; Verbueken et al. 2018). To our knowledge, the documentation of EROD activity happened much earlier in zebrafish than in other fish species (Table 4 vs. Table 5). So far, the earliest report of EROD activity in fathead minnow and medaka is at 144 hpf (Boehler et al. 2018) and 2 dpf (Gonzalez-Doncel et al. 2011), respectively. During zebrafish embryogenesis, the whole-body EROD activity peaks within the first 10 h of embryonic development, then decreases to a minimum reached in the pharyngula period and subsequently increases again around the time of hatching (Otte et al. 2010; Saad et al. 2016b; Verbueken et al. 2018). The early peak in EROD activity has been suggested to be a consequence of maternal CYP mRNA transfer (Saad et al. 2016a, b). At the end of embryogenesis, EROD activity tends to decrease again as indicated by EROD activities measured in early juvenile, which were lower rather than in embryonic stages of zebrafish (Otte et al. 2010; Saad et al. 2016a, b; Verbueken et al. 2018). The extent to which juvenile zebrafish possess EROD activities is still not clear, as the two studies currently available in literature show inconsistent results: While Pauka et al. (2011) documented constitutive and significantly inducible EROD activities in subcellular fractions prepared from whole-body homogenates of 2-weeks-old zebrafish, a more recent study published by Verbueken et al. (2018) could not detect EROD activity by epifluorescence microscopy in juvenile zebrafish at 14 dpf (Verbueken et al. 2018). The ability of adult zebrafish to O-deethylate ER and, thus, the constitutive presence of CYP1-dependent activity has been confirmed in vitro using liver microsomes and whole gill arches (Jönsson et al. 2009; Saad et al. 2016b). However, owing to differences in sample preparation and data normalization (i.e. resorufin gill arch−1 min−1 vs. resorufin × mg protein−1 × min−1), the actual values of hepatic and gill EROD activities cannot be compared to each other directly. Although in other fish species, such as killifish (Smith and Wilson 2010), the level of hepatic EROD activity was documented to depend on sex, this could not be confirmed in zebrafish (Saad et al. 2016a, b).

Mammalian CYP2-like activities

The fluorescent probe substrates 7-benzyloxyresorufin (BR), 7-ethoxycoumarin (EC), 7-methoxycoumarin (MC) and 7-pentoxyresorufin (PR) have all been used to monitor mammalian CYP2-like activities in zebrafish (cf. Table 4). However, since both resorufin derivates have been demonstrated to undergo O-dealkylation catalyzed by heterologously expressed zebrafish CYP1A, CYP1C1, CYP1C2 and CYP1C2 enzymes, and in addition BR also by CYP1B1 (Scornaienchi et al. 2010b), at least BR and PR might not be specific for monitoring activities of CYP2 isoforms in zebrafish. Up to now, there are no data available as to which CYP isoforms are active in catalyzing O-dealkylation of EC and MC in zebrafish.

By monitoring O-dealkylation of 7-methoxycoumarin (i.e. formation of 7-hydroxycoumarin), Loerracher et al. (2020) could document that zebrafish embryos exhibit mammalian CYP2-like activities from as early as 5.5 hpf. This was much earlier than expected from all previous studies. In the same study, the use of a confocal laser scanning microscope allowed for detailed visualizing the developmental pattern of 7-methoxycoumarin-O-demethylase (MCOD) activity throughout zebrafish embryogenesis (Fig. 4).

Fig. 4
figure 4

Developmental pattern of 7-methoxycoumarin-O-demethylase (MCOD) activity in zebrafish (Danio rerio). Lateral views of zebrafish embryos exposed to 1 mM 7-methoxycoumarin for 3 h

Otte et al. (2017) examined 7-benzyloxyresorufin-O-debenzylase (BROD) and 7-pentoxyresorufin-O-depentylase (PROD) activities in whole-body microsomes of different embryonic stages starting at 2.5 hpf. However, for both substrates, the rates of resorufin formation remained below the limit of detection (i.e. 0.7–1.3 fmol resorufin/min/embryo) until 120 hpf, when BROD activity became detectable, but could still not be quantified (< 1.4–2.6 fmol resorufin/min/embryo; Otte et al. 2017). Jones et al. (2010) documented 96 h old zebrafish embryos to metabolize 7-ethoxycoumarin and to excrete the metabolite, 7-hydroxycoumarin, into the surrounding medium. Earlier embryonic stages were not investigated in their study.

Up to now, 7-ethoxycoumarin is the only CYP2 probe substrate that has been used to monitor mammalian CYP2-like activities in juvenile or adult zebrafish. In juvenile zebrafish, 7-ethoxycoumarin-O-deethylase (ECOD) activity was investigated in microsomal fractions prepared from different tissues and organs. Of all tissues tested (i.e. liver, muscle and brain), the liver showed the highest constitutive level of ECOD activity (Wu et al. 2014). Funari et al. (1987) assessed ECOD activities in liver microsomes of adult zebrafish and found the levels to be comparable to those of rainbow trout (Oncorhynchus mykiss), common carp (Cyprinus carpio) and bluegill sunfish (Lepomis macrochirus).

Mammalian CYP3-like activities

The pro-fluorescent and pro-luminescent CYP probe substrates 7-benzyloxy-4-(trifluoromethyl) coumarin (BFC), N-ocytoxymethylresorufin (OOMR), luciferin-6’-benzylether (luciferin-BE), luciferin-6’-pentafluorobenzyl ether (luciferin-PFBE) and luciferin-isopropyl acetal (luciferin-IPA) are all selective, if not specific for human CYP3A isoforms (Cali et al. 2006, 2009; Renwick et al. 2000). Studies on these substrates allowed for conclusions with regard to the presence of mammalian CYP3A-like activities in zebrafish, which is of great significance considering the central role of the human orthologue CYP3A4 in biotransformation of pharmaceuticals (Wienkers & Heath, 2005). However, for most of these probe substrates, the isoform specificity has not yet been clarified in zebrafish. An exception is BFC, which was shown to undergo O-dealkylation by heterologously expressed CYP1A, CYP1B1, CYP1C1 and CYP1C2 enzymes (Scornaienchi et al. 2010b). One clear gap, as becomes evident from Table 4, is the lack of studies investigating the presence of mammalian CYP3-like activities in zebrafish embryos < 48 hpf. Moreover, there is only one study available that yields information about mammalian CYP3-like activity levels at different points in zebrafish development. In the study published by Li et al. (2011), the authors demonstrate 3-day-old zebrafish embryos to have higher, although not statistically significantly higher constitutive activities towards the luminogenic probe substrates luciferin-BE, when compared with two days old zebrafish embryos. Jones et al. (2010) used the probe substrate OOMR to study mammalian-like CYP3 activities in 96-h-old zebrafish embryos and demonstrated their ability to metabolize OOMR and to excrete its metabolite 7-hydroxyresorufin into the surrounding medium. Another in vivo study conducted by Chng (2013) provided evidence that 120-h-old zebrafish embryos possess both constitutive and inducible mammalian CYP3-like activities. To date, however, no study has been conducted to examine the level or presence of mammalian CYP3-like activities in juvenile zebrafish (cf. Table 4), and the only study that evaluated the presence of mammalian CYP3-like activities in adult zebrafish by monitoring the turn-over of luciferin-IPA to D-luciferin did not find any activity above the limit of detection (Verbueken et al. 2017).

Current state of knowledge on CYP gene expression patterns vs. CYP activities

Yet, our functional understanding of the zebrafish CYP system is far better than that at the gene expression level. Some reasons for this may include the current lack of knowledge about the CYP isoform specificity of mammalian CYP probe substrates in zebrafish (Scornaienchi et al. 2010b), the complexity of data interpretation and the missing standardization, which hinder comparison of data across different studies. Although zebrafish embryos of different ages have been documented to biotransform fluorogenic and luminogenic CYP1, CYP2 and CYP3 probe substrates in a way similar to mammals, the lack of data for juvenile and adult zebrafish hinders quantitative conclusions concerning potential age-related differences in functional biotransformation activities of zebrafish.

Bioactivation of pro-toxicants and pro-teratogens

Another common approach to assess the functionality of the CYP system in (zebra)fish has been monitoring effects of pro-teratogens and pro-toxicants in organisms. However, this indirect way of assessing biotransformation activities faces one big challenge: How to distinguish between direct effects of pro-teratogens and pro-toxicants, and those (i.e. indirect) mediated by their bioactivated metabolites? Admittedly, while for mammals, for instance, it is known that that several pro-teratogens and pro-toxicants (e.g., aflatoxin B1, carbamazepine, cyclophosphamide and phenytoin) need to undergo CYP-mediated bioactivation prior eliciting their ultimate toxic or teratogenic effects (Dohnal et al. 2014; Hill et al. 2010), corresponding information is lacking for zebrafish. Hence, today, we still rely on read-across, i.e. utilization of mechanistic information gained from studies in mammals, and extrapolation.

An example of such a read-across approach is provided by the study of Klüver et al. (2014), who investigated the acute toxicity of allyl alcohol in zebrafish at different ages. So far, allyl alcohol is the only pro-toxicant known to be less toxic to zebrafish embryos than to juvenile or adult zebrafish due to lack of bioactivation in embryos (Klüver et al. 2014). In mammals, allyl alcohol is biotransformed via oxidation into acrolein, a reactive toxic aldehyde metabolite (Auerbach et al. 2008; Ohno et al. 1985). Based on read-across, in zebrafish embryos the reduced toxicity of allyl alcohol could be documented to be caused by a lack of the alcohol dehydrogenase 8a enzyme (Klüver et al. 2014).

Concerning pro-teratogens known to undergo CYP-mediated bioactivation, at least in mammals there is ample evidence, but not yet conclusive proof that zebrafish embryos possess sufficient biotransformation capacities to bioactivate these compounds to a meaningful extent. Although Weigt et al. (2011) demonstrated that ten well-known mammalian pro-teratogens, with eight of them being pharmaceuticals (carbamazepine, phenytoin, trimethadione, cyclophosphamide, ifosfamide, tegafur, and thio-TEPA) were teratogenic to zebrafish embryos when exposed for 3 days (Weigt et al. 2011), it has not yet been demonstrated that bioactivated metabolites were actually formed. Furthermore, it cannot be ruled out that even the parent compounds themselves might be teratogenic to zebrafish embryos.

Xenobiotic metabolite profiles in zebrafish

Over the past ten years, there was a clear upward trend in the number of studies assessing biotransformation activities in zebrafish by monitoring formation of phase I and phase II metabolites. These studies provide conclusive evidence that zebrafish at different developmental stages do have metabolic competencies to perform several types of phase I and phase II biotransformation reactions (Table 6). These include inter-aromatic hydroxylation (Alderton et al. 2010; Brox et al. 2016b; Chng et al. 2012; Poon et al. 2017a, b; Saad et al. 2017; Zindler et al. 2020), N-dealkylation (Alderton et al. 2010; Saad et al. 2017), O-dealkylation (Alderton et al. 2010; Saad et al. 2017), glucuronide conjugation (Alderton et al. 2010; Kantae et al. 2016; Le Fol et al. 2017b), sulfation (Brox et al. 2016a, b; Kantae et al. 2016; Le Fol et al. 2017b; Li et al. 2015) and O-acylation (Zindler et al. 2020). Whole xenobiotic metabolite spectra, however, were only rarely determined (e.g., by Brox et al. 2016a; Zindler et al. 2020), probably because such studies are particularly demanding in terms of expertise, resources and funding. Moreover, as most studies today have focused on late embryonic stages of zebrafish (i.e, 72–120 hpf; e.g., Alderton et al. 2010; Jones et al. 2012; Kantae et al. 2016; Le Fol et al. 2017a; Zindler et al. 2020) or adult zebrafish (e.g., Chng et al. 2012; Li et al. 2015; Poon et al. 2017a, b; Saad et al. 2018; Wang et al. 2016), vital information concerning metabolic activities in early embryonic and juvenile stages is currently lacking. Therefore, it remains necessary to explore from which developmental stages certain biotransformation pathways are present and sufficiently developed to biotransform and bioactivate xenobiotic compounds to a biologically relevant extent.

Table 6 Metabolic profiles of xenobiotics in embryonic, juvenile and adult stages of zebrafish (Danio rerio)

Numerous studies have indicated that different developmental stages of zebrafish may vary in their biotransformation capacities, either in respect to rates of metabolite formation (i.e. quantitatively) or in the biotransformation pathways they use (i.e. qualitatively; e.g., Brox et al. 2016; Chng et al. 2012; Saad et al. 2017, 2018). In fact, there is growing evidence that the metabolic competence of zebrafish progressively develops with more phase I and phase II metabolites being detected, and less abundant metabolites being enriched as embryonic development progresses. For instance, Brox et al. (2016a) analyzed the metabolite profile of clofibric acid in extracts of zebrafish embryos. They could show that sulfate-containing (phase II) metabolites are formed from very early on (i.e. 7 hpf), while others such as glucuronide conjugates only reached detectable levels at 52 hpf (Brox et al. 2016a).

Whether juvenile zebrafish dispose a fully developed xenobiotic biotransformation capacity is generally not a subject of debate. To date, however, only little information is available on metabolite formation in juvenile zebrafish (Table 6). The only comprehensive study that focused on metabolite formation in juvenile zebrafish was performed by Alderton et al., (2010). In their study, juvenile zebrafish at 168 hpf could be documented to perform phase I (e.g., oxidation, N-dealkylation and O-dealkylation) and phase II reactions (e.g., glucuronidation and sulfation) similar to humans with a variety of pharmaceutical compounds. However, the metabolites recovered accounted for a very small fraction of the parent compounds administrated (Alderton et al. 2010). Furthermore, this study provided evidence that the metabolic competence of juvenile zebrafish might at least be quantitatively different from that of zebrafish embryos. While 168-h-old zebrafish were documented to metabolize cisapride to the phase II metabolite cisapride-sulfate, no such metabolite formation was observed in zebrafish embryos at 72 hpf (Alderton et al. 2010).

Comparative studies between zebrafish embryos and adult zebrafish, especially those based on microsomal preparations, are beset with problems regarding the comparability of the results. In particular, when studies use microsomes prepared from whole-body homogenates of zebrafish, as has been commonly done when assessing biotransformation activities in embryonic stages of zebrafish, there is a risk of underestimating the level of biotransformation activity obtained in liver microsomes or in vivo (Verbueken et al. 2018). The fact that zebrafish embryos and adult zebrafish may qualitatively differ in the metabolic pathways they use has been indicated by (Chng et al. 2012). In this in vitro study, differences were documented with respect to phase I biotransformation of testosterone. While in liver microsomes of adult zebrafish testosterone was metabolized to seven hydroxylated metabolites, only two hydroxylated metabolites were detected in whole-body microsomes of 5-d-old zebrafish embryos, whose main metabolite was unique (i.e. not found in adult zebrafish liver microsomes nor in human liver microsomes (Chng et al. 2012)). However, as indicated by studies of Saad et al. (2017, 2018), the metabolism of testosterone varies not only with age, but also with sex.

Quantitatively different, but qualitatively similar biotransformation activities in zebrafish embryos and adult zebrafish have been documented in an in vivo study by Le Fol et al. (2017). They documented that the biotransformation of benzophenone-2 was more extensive in adult zebrafish, if compared to embryos. However, expect for a single phase II metabolite, i.e. benzophenone-2-diglucuronide, which was only detected in adult zebrafish, five identical metabolites were produced in both developmental stages (Le Fol et al. 2017).

Conclusions and recommendations

Embryonic, juvenile and adult stages of zebrafish have been studied to a very different extent with respect to their biotransformation and bioactivation capacities. Especially juvenile zebrafish have been neglected so far. As a consequence, our knowledge about the development of the xenobiotic biotransformation capacity in zebrafish can be called—at best—fragmentary.

At the mRNA level, there is ample evidence that the vast majority of zebrafish CYP1, CYP2, CYP3 and CYP4 genes are constitutively expressed from earliest embryonic stages of development. At a first glance, this might be interpreted as an indication of a fundamental existence of an early competence for xenobiotic biotransformation. Since, however, extrapolation of gene expression levels to biochemical functionality is technically not possible to date, it is questionable whether this interpretation holds. One of the fundamental questions remaining to clarified is whether or not the developmental- and sex-related differences in CYP1 to 4 mRNA expression levels will lead to age- and sex-related differences in xenobiotic biotransformation capacities and, eventually, to differences in outcomes of toxicological studies.

Overall, studies on CYP gene expression patterns are not as conclusive as studies providing indirect or direct evidence of functional biotransformation activities. However, with respect to CYP-dependent activities, our knowledge is even far more fragmentary as it is for CYP expression patterns in zebrafish (Fig. 5). This is due to the fact that only very few studies allow quantitative comparisons with regard to CYP activities at different developmental stages of zebrafish. Moreover, the diversity of methodological approaches between studies does not allow an adequate comparison of results.

Fig. 5
figure 5

Availability of data on the cytochrome P450 system and phase I biotransformation in zebrafish (Danio rerio)

Evidence on the functionality of the CYP-system in zebrafish have come from studies monitoring formation of fluorescent and luminescent metabolites or assessing effects of mammalian pro-toxicants and pro-teratogens known to be activated by CYP enzymes in mammals. Again, there is evidence that zebrafish embryos from earliest stages of development do have functional biotransformation pathways. In many cases, the data available suggest that the biotransformation pathways of zebrafish embryos are at least qualitatively similar to those of juvenile and adult zebrafish as well as those of humans. Only for rare exceptions, such as in case of allyl alcohol, a lack of biotransformation could be found, which was due to a lack of the enzyme required for bioactivation. There is a need for systematic screening from which developmental stages CYP-dependent biotransformation capacities are sufficiently developed to biotransform and bioactivate xenobiotic compounds to a biologically meaningful extent. However, even when formation of a metabolite, bioactivated or not, is documented, the question remains whether the concentration of a metabolite—or the concentration of the parent compound—is sufficiently high to exert a toxicological effect in the respective developmental stage or sex of (adult) zebrafish.

There is a growing number of studies that use chemical analyses to obtain quantitative and/or qualitative information on biotransformation pathways and activities. Such studies, especially those examining full metabolite spectra, are a major challenge in terms of expertise, resources, and funding, but can assist in obtaining much better knowledge and more in-depth understanding of, e.g., the substrates accepted by zebrafish CYP isoforms and factors (e.g., age and sex) that may influence biotransformation activities and, eventually, the outcome of toxicological studies.