Dibenzyl trisulfide binds to and competitively inhibits the cytochrome P450 1A1 active site without impacting the expression of aryl hydrocarbon receptor

Abstract

The toxicological manifestation of many pollutants relies upon their binding to the aryl hydrocarbon receptor (AHR), and it follows a cascade of reactions culminating in an elevated expression of cytochrome P450 (CYP) 1 enzymes. CYP1A1 and CYP1B1 are associated with enhanced carcinogenesis when chronically exposed to certain polyaromatic hydrocarbons, and their inhibition may lead to chemoprevention. We evaluated dibenzyl trisulfide (DTS), expressed in the ethnomedical plant, Petiveria alliacea, for such potential chemoprevention. Using recombinant human CYP1A1 and CYP1B1 bactosomes on a fluorogenic assay, we first demonstrated that DTS moderately inhibited both enzymes with half maximal inhibitory concentration (IC₅₀) values of 1.3 ± 0.3 and 1.7 ± 0.3 μM, respectively. Against CYP1A1, DTS was a reversible, competitive inhibitor with an apparent inhibitory constant (K_i) of 4.55 ± 0.37 μM. In silico molecular modeling showed that DTS binds with an affinity of −39.8 kJ·mol⁻¹, situated inside the binding pocket, approximately 4.3 Å away from the heme group, exhibiting interactions with phenylalanine residue 123 (Phe-123), Phe-224, and Phe-258. Lastly, zebrafish (Danio rerio) embryos were exposed to 0.08–0.8 μM DTS from 24–96 hours post fertilization (hpf) with the in vivo ethoxyresorufin-O-deethylase (EROD) assay, and, at 96 hpf, DTS significantly suppressed EROD CYP1A activity in a dose-dependent manner, with up to 60% suppression in the highest 0.8 μM exposure group. DTS had no impact on gene transcription levels for cyp1a and aryl hydrocarbon receptor 2 (ahr2). In co-exposure experiments, DTS suppressed CYP1A activity induced by both B[a]P and PCB-126, although these reductions were not significant. Taken together, these results demonstrate that DTS is a direct, reversible, competitive inhibitor of the carcinogen-activating CYP1A enzyme, binding in the active site pocket close to the heme site, and shows potential in chemoprevention.

1. Introduction

Members of the cytochrome P450 1 (CYP1) family of enzymes, in particular CYP1A1 and CYP1B1, have been associated with carcinogenesis through the activation of environmental anthropogenic chemicals such as benzo[a]pyrene (B[a]P), into their oxygenated metabolites that form DNA-binding adducts (Shimada et al., 1996; Shimada and Fujii-Kuriyama, 2004). The hepatic CYP1A2 and the extrahepatic CYPs 1A1 and 1B1 enzymes are inducible by the same classes of coplanar aromatic hydrocarbons via the cytosolic aryl hydrocarbon receptor (AHR). The AHR is a basic helix–loop–helix Per-Arnt-Sim (bHLH/PAS) transcription factor; upon ligand binding, the AHR translocates from the cytosol to the nucleus and forms a heterodimer with the aryl hydrocarbon nuclear translocator (ARNT) protein. This complex binds to xenobiotic response elements (XREs) in promoter regions, which subsequently leads to the transcriptional regulation of numerous cyp1 genes. Numerous compounds, including natural products, have been shown to be ligands of the AHR and/or substrates of CYP1, or antagonists of AHR and inhibitors of CYP1, or various combinations thereof (Billiard et al., 2006; Kikuchi and Hossain, 1999; Perepechaeva et al., 2017; Timme-Laragy et al., 2007). There has been emerging interest in the search for inhibitors of CYP1 enzymes given their potential use as chemopreventors, particularly via retardation at the carcinogen-activating stage for those with chronic exposure to pro-carcinogens. Furthermore, it has also been found that CYP1 enzymes are overexpressed in certain cancer cells, such as breast cancer, in comparison to normal surrounding tissue (McKay et al., 1995), and have a putative role in drug resistance. Thus, CYP enzyme inhibitors present an attractive target for cancer therapeutic development.

The use of herbal medicines is popular across the world, including in island states such as Jamaica, where 73% of its population reported self-medicating using ethnomedicine (Picking et al., 2011). 80% of Jamaicans use herbal medicine to treat their cancers, and 75% engage in concomitant use with allopathic drugs (Foster et al., 2017). This practice of using herbal medicine to treat cancer is also seen in other countries such as China (Chen et al., 2015) and Ethiopia (Erku, 2016). One popularly used plant is Petiveria alliacea (Guinea Hen Weed), used by 30% of Jamaican cancer patients (Foster et al., 2017). Several studies in vivo and in vitro have shown that extracts of P. alliacea and the key active ingredient, dibenzyl trisulfide (DTS), potently inhibited the growth of several tumors (Hernández et al., 2014; Urueña et al., 2008; Williams et al., 2007). Although DTS has been shown to be a moderate inhibitor of the activities of major drug-metabolizing enzymes CYPs 1A2, 3A4, and 2C19 in vitro (Murray et al., 2016), its impact on the activities of carcinogen-activating CYPs 1A1 or 1B1 remain unknown. We thus sought to evaluate such inhibitory activity using in vitro, in silico, and in vivo models.

Given the similarity of its CYP induction systems to humans, we used the well-characterized zebrafish (Danio rerio) model (Timme-Laragy et al., 2007) to better understand the in vivo impact of DTS on the AHR pathway using the EROD assay. Our objectives were, firstly, to determine whether DTS functioned as an AHR antagonist in combination with known disparate inducers(B[a]P) and 3,3’,4,4’,5-pentachlorobiphenyl (PCB-126), which are good and poor substrates of CYP1A, respectively (Billiard et al., 2006), and, secondly, to evaluate DTS’s impact on the mRNA expression of ahr2 (the functionally human-relevant AHR isoform in zebrafish) and the AHR-regulated cyp1a gene, and how such expressions compared to the protein’s enzymatic activity. Together with in silico and in vitro evaluations on its human counterparts, we hoped to gain an understanding of the nature of DTS interactions with CYP1A1 and its inducing partner AHR. Such insights garnered can provide guidance for future searches of molecular leads with improved potency and reduced toxicity.

2.0. Materials and Methods

2.1. Enzymes and reagents

Bactosomes overexpressing CYPs 1A1, 1A2, and 1B1, 3-cyano-7-ethoxycoumarin (CEC), furafylline, and ketoconazole were purchased from CYPEX Limited (Dundee, UK). 3,3’,4,4’,5-Pentachlorobiphenyl (PCB-126) was purchased from Ultra Scientific (North Kingstown, RI), benzo[a]pyrene (B[a]P) was purchased from Sigma-Aldrich (St. Louis, MO), and 7-ethoxyresorufin-O-deethylase (EROD) was purchased from MP Biomedicals (Solon, OH). DTS was purchased from International Laboratory (San Francisco, CA, USA). All chemicals were dissolved in 100% dimethyl sulfoxide (DMSO) from Fisher Scientific (Fair Lawn, NJ), aliquoted, stored as stock solutions in glass amber vials at −20 °C, and fully thawed and vortexed before use.

2.2. In vitro inhibition of CYP activity

Enzymatic assays were carried out as previously described (Murray et al., 2016) using substrates well-characterized for each enzyme. Briefly, 200 μL of enzyme mix containing heterologously expressed CYP1A1 (1.02 pmol/200 μL), CYP1A2 (2 pmol/200 μL), or CYP1B1 (2.57 pmol/200 μL) and 0.5 μM CEC, 5 μM CEC, or 0.4 μM 7-ethoxyresorufin (7-ER), respectively, were incubated at 37 °C with varying concentrations of DTS (0–12.12 μM). The reaction took place in the presence of 34 μM NADPH in 0.1 M potassium phosphate buffer (pH 7.4), with the exception of the CYP1B1 reaction, which had a buffer pH of 7.6. After 15 minutes, 75 μL of an 80% acetonitrile (ACN)/20% Trizma base solution (pH 10.3) was added to stop the reaction. A Varian Cary Eclipse Fluorescence Spectrophotometer (Victoria, Australia) was used to read the fluorescence of the metabolites produced in each well at the recommended wavelength. Each incubation had triplicates, and each experiment was conducted at least twice, along with controls for background fluorescence, stop solution effect and quenching of metabolite fluorescence.

2.3. Mechanistic assessments of CYP1A1 activity inhibition

2.3.1. Single-point NADPH-dependent inhibition

To characterize the mechanism of inhibition, the single-point NADPH dependence assay was utilized. To determine whether DTS is a mechanism-based inhibitor of CYP1A1, the enzyme was preincubated with or without DTS (1.7 μM) in the presence or absence of NADPH (340 μM) for 30 minutes. Following preincubation, a 10-fold dilution was performed, and incubations were carried out with a saturating concentration of CEC (1.5 μM) as described above.

2.3.2. Time-dependent inhibition and IC₅₀ shift assay

To assess whether the half maximal inhibitory concentration (IC₅₀) of DTS inhibition decreased over time, a modified method previously described (Naritomi et al., 2004) was employed, where CYP1A1 (1.02 pmol/200 μL) was incubated with varying concentrations of DTS (0–12.12 μM) under similar conditions to those described in section 2.2, except that the reaction was allowed to continue without the addition of a stop solution. The reaction was conducted in the fluorometer, fluorescence readings were taken at 5-minute intervals for 30 minutes, and IC₅₀ values were determined at each time point. CYP1A2 was tested with furafylline, which is a known time-dependent inhibitor of CYP1A2, to validate the use of this method.

To further assess the time-dependence properties of DTS with CYP1A1, an IC₅₀ shift assay was conducted. CYP1A1 (1.02 pmol/200 μL) was preincubated for 30 minutes with varying concentrations of DTS (0–12.12 μM) in the presence or absence of NADPH (34 μM). The metabolism of 0.5 μM CEC was measured as described in section 2.2. To validate the method, furafylline was tested with CYP1A2, and an IC₅₀ shift of 32.1 was obtained.

2.3.3. K_i determination

Following confirmation of reversible status, CYP1A1 (1.02 pmol/200 μL) was incubated with varying concentration of DTS under similar conditions to those previously described. Varying concentrations of CEC (1–10 μM) surrounding its Michaelis constant, K_m (0.5 μM), was used. The apparent K_i values were determined on the basis of visual inspection of Eadie–Hoftsee plots and various statistics to evaluate goodness of fit, such as the size of the sum of squares of residuals, Akaike’s information criterion, and standard error (Enzyme Kinetics Module, version 1.3). The data listed represent the average values from three different determinations.

2.4. In silico modeling of DTS binding to the CYP1A1 active site

To establish the preferred binding mode of DTS in CYP1A1, docking studies were performed using AutoDock Vina (Trott and Olson, 2010). Firstly, the three-dimensional protein structure of CYP1A1 was downloaded from the Protein Data Bank (PDB) with identifier 4I8V (Walsh et al., 2013). Polar hydrogens were added using AutoDock Tools (Morris et al., 2009), and the grid box was centered on the heme group with dimensions of 20 Å in the x-, y-, and z-planes.

The molecular structure of DTS was generated using Avogadro v 1.90.0 (Hanwell et al., 2012) and optimized, before being prepared with AutoDock Tools (Morris et al., 2009). Once both ligand and receptor files were ready, automated flexible docking was performed using AutoDock Vina (Trott and Olson, 2010), with no added restrictions. Only the best docking pose was selected for CYP1A1; the structure was visualized with VMD (Humphrey et al., 1996), through which an image of the binding mode was generated. Lastly, LigPlot+ v 2.2 (Laskowski et al., 2011) was applied to visualize the CYP1A1–DTS interactions, using default settings.

2.5. In vivo evaluation of DTS interactions with CYP1A using zebrafish

2.5.1. Animal husbandry

Adult wildtype AB zebrafish (Danio rerio) were maintained on a 14 h/10 h light/dark cycle in a recirculating Aquaneering system (San Diego, CA, USA) at a temperature of 28.5 °C. Adult fish were fed twice daily with GEMMA Micro 300 (Skretting, Westbrook, ME, USA) and bred in a 2:1 female-to-male ratio. Embryos were collected at 1 hour post fertilization (hpf), washed, and stored in 0.3× Danieau’s medium (17 mM NaCl, 2 mM KCl, 0.12 mM MgSO₄, 1.8 mM Ca(NO₃)₂, and 1.5 mM HEPES, pH 7.6) in an incubator under similar conditions to those used for the adult fish. All animal care and experiments were conducted in accordance with protocols approved by the University of Massachusetts Amherst Institutional Animal Care and Use Committee (IACUC; Protocol Number 2019–0067). Animals were treated humanely with due consideration to the alleviation of stress and discomfort.

2.5.2. Embryo exposures

At 24 hpf, the embryos were screened to ensure they were at the correct developmental stage before use in experiments. The embryos were dechorionated manually using Watchmaker forceps. Five embryos were placed in 20-mL glass scintillation vials containing 5 mL of 0.3× Danieau’s water and 0.5 μg/L 7-ethoxyresorufin-O-deethylase (7-ER). Each treatment group had two replicates. Embryos were treated under static medium conditions for 72 h with either single doses of DTS (0.08, 0.2 or 0.8 μM) or these DTS concentrations in combination with either 100 μg/L benzo[a]pyrene (B[a]P) or 5 nM 3,3’,4,4’,5-pentachlorobiphenyl (PCB-126). The final concentration of DMSO in each vial was 0.05%. At 96 hpf, zebrafish were rinsed 3 times with Danieau’s medium before imaging. Zebrafish collected for RT-qPCR were exposed in the same way as the zebrafish collected for imaging, except that 7-ER was replaced with DMSO. Upon collection for RT-qPCR, 3–4 vials (15–20 fish) were pooled together per exposure group as 1 sample before storage in RNAlater (Thermo Fisher Scientific, Waltham, MA) in 1.5-mL Eppendorf tubes at −80 °C. Each experiment was independently repeated 3 times.

2.5.3. Imaging

At 96 hpf, larvae were rinsed 3× with Danieau’s medium, placed in dishes, sedated using MS-222 (prepared as 4 mg/mL tricaine powder in water, pH-buffered, and stored at −20 °C until use), and staged on drops of 3% methylcellulose gel in a left-lateral orientation. The larvae were imaged on a Zeiss Stereo AxioZoom.V16 (Carl Zeiss Inc.) with an HXP 200C light source. Images were taken with a 10× zoom at 2× magnification (20× total magnification) under brightfield settings to visualize the whole fish, at 10× magnification (100× total magnification) using an RFP filter to visualize EROD intensity in the gut area, and at 10× magnification under brightfield settings to create final overlay images for visualization. The length, EROD intensity, and pericardial area of the larvae were analyzed using the Zen Lite program (Carl Zeiss Inc.). We measured pericardial area on all 96-hpf zebrafish as a quantitative measurement to represent overall deformities since pericardial edema is a characteristic outcome of fish embryos exposed to PCB-126 (King-Heiden et al., 2012) and other compounds that suppress CYP1A activity (Wincent et al., 2016).

2.5.4. RT-qPCR

Zebrafish larvae were thawed, transferred to lysis buffer, and sonicated by pulsing 3–5 times with an Emerson Industrial Branson Sonifier® (Danbury, CT). RNA isolation was performed using 2-mercaptoethanol (MP Biomedicals) and a GeneJET RNA Purification Kit (Thermo Fisher Scientific) following the manufacturer’s instructions. RNA quantity and quality were assessed using a BioDrop μLITE spectrophotometer (Cambridge, United Kingdom). Sample cDNA was prepared using an iScript reaction mix kit (Bio-Rad, Hercules, CA), diluted 1:9 with nuclease-free water, and stored at −80 °C until processing. Each RT-qPCR sample was prepared using 10 μL of 2× iQ SYBR® Green Supermix (Bio-Rad), 5 pmol (250 nM) each of forward and reverse primers (1 μL total), 5 μL of nuclease-free water, and 4 μL (10 ng) of cDNA. Samples were run on 96-well plates in a CFX Connect Real-Time PCR Detection System (Bio-Rad), and samples were analyzed using the CFX Manager software (Bio-Rad). RT-qPCR was carried out in duplicate for the cyp1a and ahr2 genes. The β-actin (actb) gene was used as a housekeeping gene, and its transcription did not change significantly across exposure groups.

2.6. Statistical Analyses

In vitro data:

Data were collected in triplicate, and each experiment was repeated at least twice. The mean and standard error of the mean (SEM) were calculated. IC₅₀ curves were generated with Sigma Plot (Version 10.0), using the four-parameter logistic (4PL) nonlinear regression model as shown in the equation below.

$$\%EnzymeActivity=\left[Min+\left(Max-Min\right)/\left(1+{10}^{((LogI{C}_{50}}{}^{-x)\times p)}\right)\right],$$

where Min and Max are the minimal ad maximal observed effects, respectively, x is the concentration of test compounds, IC₅₀ is the half maximal inhibitory concentration, and p is the slope parameter.

In vivo data:

Individual fish were analyzed for outliers (IQR method) per exposure group for body length, EROD, and pericardial area endpoints across all experimental replicates, and individual samples were removed completely from all analyses if they were flagged as outliers for at least one endpoint. For each type of experiment, 1–5% of samples were considered outliers. All endpoints were then averaged per vial so that each vial was n = 1. A one-way analysis of variance (ANOVA) with a Tukey–Kramer post-hoc statistical test was performed for experiments with JMP® Pro software version 14.1.0 (Cary, NC). Statistical significance was considered using a 95% confidence interval (α = 0.05). For RT-qPCR experiments, gene transcription fold-changes were calculated using the ΔΔCT method (Livak and Schmittgen, 2001).

3.0. Results

3.1. In vitro inhibition of CYP activity

The ability of DTS (chemical structure shown in Figure 1A) to inhibit the activity of the CYP1 family of enzymes was explored using an in vitro fluorescence-based assay (as described in section 2.2). DTS displayed a dose-dependent increase in inhibition of CYP1-catalyzed CEC and ethoxyresorufin O-deethylation using heterologously expressed human CYP1 enzymes. Inhibition was characterized as moderate with IC₅₀ values of 1.3 ± 0.3 and 1.7 ± 0.3 μM, respectively, against the activities of CYP1A1 and CYP1B1 (Figure 1B).

Figure 1. DTS inhibition of human CYPs 1A1 and 1B1 isoforms.

A) The structure of dibenzyl trisulfide (DTS), and B) the inhibition of the activity of human CYPs 1A1 and 1B1 isoforms by DTS. The metabolism of 0.5 μM CEC and 0.4 μM 7-ethoxyresorufin (7-ER) by heterologously expressed CYP1A1 (1.02 pmol/200 μL) and CYP1B1 (2.57 pmol/200 μL), respectively, was measured in the presence of varying concentrations of DTS (0.006–12.12 μM) using a fluorometric assay as described in the methods. The control specific activity of the uninhibited enzyme was 2.45 × 10⁻¹ μM of CEC/min/pmol of CYP1A1 and 1.05 × 10⁻¹ μM of RES/min/pmol of CYP1B1. Data are reported as the mean of triplicates, and the assay was conducted twice.

Given that the highest potency was against the activity of CYP1A1, and given the critical role this enzyme plays in carcinogenic activation, we further characterized the mechanism underlying this interaction (as described in section 2.3). Inhibition of CYP1A1 activity by DTS appeared to be independent of NADPH over a 30-minute period as shown in Figure 2A, as there was no significant difference in inhibition with preincubation with the cofactor. Time dependence was then evaluated in two ways. The impact on inhibition as DTS concentration increased was independent of the time of reaction over the 30-minute period, as shown in Figure 2B, which was also confirmed by Figure 2C, where a plot of IC₅₀ displays fair constancy over the half-hour period. There was an insignificant shift of 0.77 in IC₅₀ values in comparison to known time-dependent inhibitors that have IC₅₀ shift values of >1.5 (Zhuang et al., 2013). Following the affirmation from these experiments that DTS is likely to be a reversible inhibitor, we carried out further kinetic characterization to unveil DTS as being a competitive inhibitor of CYP1A1 with a K_i value of 4.55 μM (Figure 2D).

Figure 2. Kinetics of the inactivation of CYP1A1 by DTS.

(A) The absence of dependence on NADPH using a single-point concentration 1.7 μM of DTS, (B) the absence of time dependence of DTS inactivation with varying concentrations of DTS, (C) \the absence of time dependence as shown by a lack of significant shift in IC₅₀ values, and (D) Eadie–Hofstee plot of the inhibition of CYP1A1 by DTS. Each data point represents the mean of triplicates of two independent analyses.

3.2. In silico assessments of DTS binding to CYP1A1

Following the recognition that DTS is a competitive inhibitor to CYP1A1, molecular binding studies were carried out to visualize its binding with DTS (as described in section 2.4). Figure 3A displays the best theoretical binding pose achieved for CYP1A1, which demonstrated an in silico binding affinity of −39.8 kJ·mol⁻¹. In the best theoretical binding pose achieved, DTS was situated inside the binding pocket, approximately 4.3 Å away from the heme group. The closer benzyl group exhibited an edge-to-face π–π stacking interaction with phenylalanine residue 123 (Phe-123), as well as offset π–π stacking interactions with Phe-224 and Phe-258, representing common features of many CYP1A inhibitors (Gonzalez et al., 2012). The full list of residues forming potential hydrophobic interactions with DTS is presented in Figure 3B.

Figure 3. Best theoretical binding pose achieved for CYP1A1 with DTS.

A) The protein is shown in surface representation (gray), with phenylalanine (Phe) residues 123, 224, and 258 shown in stick representation (blue), as well as the heme group (red). DTS is also shown in stick representation, colored according to atom type (C—cyan; S—yellow). DTS is shown to make an edge-to-face π–π stacking interaction with Phe-123, as well as offset π–π stacking interactions with Phe-224 and Phe-258. B) Potential hydrophobic reactions between CYP1A1 and DTS according to LigPlot+. DTS is colored as in Figure 3A, and all residues initiating potential hydrophobic reactions (red arcs) are schematically displayed.

3.3. In vivo assessment of impact of DTS on CYP1A activity and expression

Given the inhibition patterns described by in vitro studies, and given the direct active site binding with the heme moiety by in silico studies, we undertook an evaluation of the in vivo significance of the DTS interaction with the CYP1 enzyme, employing a well-characterized zebrafish model (as described in section 2.5). Zebrafish embryos were exposed at 24 hpf to DMSO, 0.08–0.8 μM DTS, or 100 μg/L B[a]P as a positive control for the EROD assay, and then collected at 96 hpf to observe EROD light intensity (representative of CYP1A enzyme activity). B[a]P induced a significant 137% increase in CYP1A activity, and DTS significantly suppressed CYP1A activity in a dose-dependent manner, with the highest 0.8 μM group suppressing CYP1A activity 60% (Figure 4A). This reduction in CYP1A activity for the 0.8 μM group was not associated with increased pericardial area; however, total body length was significantly reduced (data not shown). The representative images of whole zebrafish and CYP1A activity in their gut regions can be seen in Figure 4B. These findings suggest that DTS acts as an inhibitor of CYP1A activity in vivo.

Figure 4. DTS single exposures in zebrafish, and co-exposures with either B[a]P or PCB-126.

Zebrafish embryos at 24 hpf were treated with 0.08, 0.2, or 0.8 μM DTS either alone or in combination with 100 μg/L B[a]P or 5 nM PCB-126 until 96 hpf. A) EROD activity for single exposures of embryos to DTS. B) Representative images of whole zebrafish and Cyp1a activity in their gut regions for single exposures to DTS or 100 μg/L B[a]P. C) EROD activity for co-exposures of DTS and 100 μg/L B[a]P. D) EROD activity for co-exposures of DTS and PCB-126. Gene expression for single exposures of DTS for E) ahr2 and F) cyp1a1. For EROD experiments, a one-way ANOVA (p < 0.05) with Tukey’s post-hoc test was used, with at least n = 6 vials per exposure group over 3–4 experiments. For RT-qPCR experiments, a one-way ANOVA (p < 0.05) with Tukey’s post-hoc test was used, with n = 3–4 pooled vials of 15–20 embryos each per exposure group across 3 experiments. Bars that share the same letter indicate that no significant differences occurred between those groups.

Subsequent experiments were conducted to investigate whether the CYP1A activity suppression from DTS could be observed in co-exposures with the cyp1a inducers B[a]P and PCB-126. In co-exposures with B[a]P, we observed that the highest 0.8 μM DTS concentration mildly decreased the CYP1A activity induced by B[a]P, but this decrease was not significantly different from the B[a]P group (Figure 4C). Pericardial area for the 0.8 μM group was not significantly increased compared to the DMSO group; however, total body length was significantly decreased (data not shown). In co-exposure experiments with PCB-126, we observed that both 0.2 and 0.8 μM DTS decreased the CYP1A activity induced by PCB-126, but these decreases were not significantly different from the PCB-126 group (Figure 4D). The pericardial areas for the 0.2 and 0.8 μM groups were not significantly decreased compared to the pericardial area of the PCB-126 group; however, total body length for the 0.8 μM group was significantly decreased compared to both the DMSO and the PCB-126 groups (data not shown).

Lastly, the effect of DTS on the gene transcription of cyp1a and ahr2 was investigated. We observed that B[a]P induced significant 1.7-fold and 47-fold increases in ahr2 and cyp1a expression, respectively; DTS did not induce significant gene transcription changes for either ahr2 or cyp1a (Figure 4E,F). These in vivo findings suggest that DTS is a partial CYP1A enzyme inhibitor in zebrafish, and they are congruent with the in vitro and in silico findings pointing to direct inhibition of the human CYP1A1 enzyme.

4.0. Discussion

In this research, we investigated DTS, a naturally occurring organic trisulfide (Figure 1A) expressed in the popular ethnomedical plant, Petiveria alliacea (Guinea Hen Weed). Interactions of DTS with the CYP1A enzyme were explored using in vitro, in silico, and in vivo methods, and DTS was found to directly bind to the CYP1A1 enzyme (Figures 1B and 2), with proximal binding to the heme active site (Figure 3). In vivo CYP1A activity of zebrafish embryos was significantly inhibited in a dose-dependent manner in the presence of DTS in single-exposure experiments; however, this inhibition did not occur at the gene transcription level on cyp1a or ahr2 (Figure 4), confirming that this inhibition occurred at the level of enzyme activity. Together, these research findings provide some insight into DTS’s use in chemoprevention.

Human CYP1A1 has been associated with chemical carcinogenesis given its key role in the catalysis of polycyclic aromatic hydrocarbons into activated epoxides, quinones and semiquinones, which are capable of DNA binding and can cause mutations leading to tumors (Shimada and Fujii-Kuriyama, 2004). Thus, direct inhibitors of the activity of CYP1A1, as well as those indirectly impeding the transcription and translation of cyp1a1 genes or AHR nuclear translocations, are believed to be of value as chemopreventors. A recent review (Li et al., 2020) categorized some flavonoids, alkaloids, and synthetic aromatics as indirect inhibitors and placed coumarins, stilbenes, and sulfur-containing isothiocyanates as direct inhibitors. Our finding of the sulfur-containing DTS as a direct inhibitor aligns well with this structure–activity-based categorization. A plethora of natural products have been identified as CYP1A1 inhibitors, including quassinoids (Shields et al., 2009), which were shown to exhibit hydrophobic interactions with the binding pocket, notably with residue Phe-123 (as seen with DTS), using one of the first active site models of CYP1A1. DTS was found to establish additional aromatic stacking interactions with Phe-224 and Phe-258, which constitute an important feature of many CYP1A inhibitors (Gonzalez et al., 2012).

The structural insights gained can help in future searches for improved candidates that may increase potency and decrease toxicity. Target molecules with increased specificity to CYP1A1 and enhanced binding affinity to the active site are logical considerations. Elongation of chain length could be considered as a means for increasing specificity (Shridar et al, 2017). The introduction of hydrogen acceptor functional groups into the candidate molecule’s rings to line up with CYP1A1’s hydrogen donor residues (Ser122, Asp313, Asp 320) can aid in increasing affinity for binding (Dutkiewicz et al, 2018).

The downregulation of CYP1 has also been shown to occur via activating negative regulatory mechanisms, such as the AHR repressor protein (AHRR), which competes for nuclear translocator ARNT and forms the heterodimer AHRR–ARNT (Perepechaeva et al., 2017). Naturally occurring quercetin has also been shown to impact transcription factors Oct-1 and С/EBPβ, which can inhibit CYP1A1 expression by binding to the 5’ region of the CYP1A1 gene independently of AHR (Perepechaeva et al., 2017). Whether DTS follows similar repressor pathways is yet to be determined. What is clear, however, is that DTS is unlikely to impact the expression of other enzymes that AHR induces given the absence of AHR attenuation. While in vitro findings previously displayed a nominal impact on drug-metabolizing enzymes such as CYPs 2D6 and 3A4 (Murray et al., 2016), an evaluation of the in vivo impact of DTS administration on drug-metabolizing phase I and phase II enzymes is also warranted. Such evaluations will permit the prediction of relevant pharmacokinetic-based interactions with concomitantly administered pharmaceutical drugs, providing valuable data for individuals self-administering the high-DTS-content Petiveria alliacea plant.

Zebrafish were used as an in vivo model to assess DTS’s CYP1A inhibition effects that were observed in vitro. We first exposed zebrafish from 24–96 hpf to single exposures of DTS using the well-established EROD bioassay, and, at 96 hpf, the larvae were assessed for CYP1A activity and morphological outcomes. We observed that single exposures of 0.2 μM and 0.8 μM DTS significantly reduced CYP1A enzyme activity by 27% and 60%, respectively. In subsequent co-exposure experiments with zebrafish, while we did observe 0.8 μM DTS to mildly decrease the CYP1A activity induced by either B[a]P or PCB-126, these reductions were not significant. While the highest concentrations of 0.8 μM DTS tested in zebrafish had the greatest inhibitory effect on CYP1A in both single and co-exposure experiments, they also negatively impacted overall fish growth. Further experiments would need to be conducted to elucidate the mechanism for this outcome, along with comparisons between embryonic and adult stages. There are no known reports of toxicity in human adult usage of the plant extract, and thus until results of a human clinical trial (currently on-going: https://clinicaltrials.gov/ct2/show/NCT04113096) on adults are published, direct comparisons across species, as well as across different development stages, are not recommended. It is known that DTS has cytotoxic effects; thus, for potential future use in vivo, with an abundance of caution the balance of CYP1A inhibition while avoiding other toxicological outcomes, if any, will need to be further characterized.

Overall, the 0.2 μM and 0.8 μM DTS concentrations demonstrating CYP1A inhibition were close to the 1.3 μM IC₅₀ concentration observed for the in vitro experiments. For the co-exposure experiments in zebrafish, we used previously optimized concentrations of B[a]P and PCB-126; future studies expanding these concentrations may allow us to more clearly observe additional aspects of DTS inhibition. According to the single-exposure EROD and gene expression data, it appears that DTS is a direct CYP1A protein inhibitor and not an AHR antagonist preventing or competing for AHR binding upstream.

Our findings conclude that DTS is a direct enzyme inhibitor of CYP1A1, which is known to play a role in carcinogenesis, as well as tumor development, and which has been identified as a target molecule. While AHR pathway and CYP1 inhibitors may mostly benefit individuals chronically exposed to CYP1A1 substrates, there is hope that direct inhibition of the CYP1A1 enzyme, found overexpressed in certain tumor microenvironments, may lead to the development of chemopreventors and the eventual overcoming of chemo-resistance in tumors.

Highlights.

• Dibenzyl trisulfide,DTS was shown to be a competitive inhibitor of CYP1A1 in-vitro

• DTS bound within the active site, 4.3 Å away from the heme

• In Zebra fish, DTS significantly inhibited the CYP1A activity in-vivo

• Gene expression of cyp1a and ahr2 were not modulated

• DTS from the medicinal plant Petiveria alliacea binds directly to CYP1A1

Abbreviations:

AHR

aryl hydrocarbon receptor

ANOVA

analysis of variance

B[a]P

benzo[a]pyrene

CEC

3-cyano-7-ethoxycoumarin

CYP1

cytochrome P450 1 enzyme family

DTS

dibenzyl trisulfide

7-ER

ethoxyresorufin

EROD

ethoxyresorufin-O-deethylase

hpf

hours post fertilization

IC₅₀

half maximal inhibitory concentration

PCB-126

3,3’4,4’,5-pentachlorobiphenyl

RT-qPCR

reverse-transcriptase quantitative polymerase chain reaction