BaP or AFB1 exposure increases DNA damage in HepaRP cells in a concentration-dependent manner

In order to evaluate the genotoxic effect of BaP or AFB1, the number of γH2AX foci per nucleus was evaluated in differentiated HepaRP cells exposed to increasing concentrations of BaP (Fig. 1A) and AFB1 (Fig. 1B), together with the relative cell count (RCC%). Both compounds induced a significant concentration-dependent increase in γH2AX foci formation, underlining their genotoxic properties. In BaP-treated cells, a significant increase in the mean number of γH2AX foci was observed starting at 1 µM, corresponding to an approximately 2.5-fold increase compared with control. At the highest tested concentration (15 µM), γH2AX foci levels reached a 6.7-fold increase relative to the DMSO control. No significant decrease in RCC% was observed across all tested BaP concentrations, indicating the absence of cytotoxicity under these conditions. In contrast, AFB1 exposure resulted in a marked induction of γH2AX foci at lower concentrations. A significant increase was detected at 0.1 µM, with an approximately 2-fold change compared with negative control, reaching up to a 13-fold increase at the highest concentration. This was accompanied by a significant reduction in RCC%, with cell viability decreasing to 93% at 0.5 µM and to approximately 79% at 2 µM, indicating concentration-dependent cytotoxicity. Concentration-response data were analyzed using the covariate approach implemented in PROAST (Supplementary Information 1), at benchmark response (BMR) levels of 300% and 400%, and benchmark concentrations (BMCs) were derived for both compounds. The black curve represents the experimental concentration-response data of BaP while the red curve represents the experimental concentration-response data of AFB1.

Fig. 1

γH2AX immunofluorescence analysis of HepaRP cells exposed for 24 h to benzo[a]pyrene (BaP) or aflatoxin B1 (AFB1). The number of γH2AX foci per nucleus observed after 24 h exposure to BaP (A) or AFB1 (B) are presented. Bars show the mean number of γH2AX foci per nucleus ± SEM (n = 3), normalized to DMSO control and presented as fold change. Statistical comparisons to DMSO control were performed using one-way ANOVA: p < 0.05 (*), p < 0.01 (**). Dots indicate the mean percentage of relative cell count (RCC%) ± SEM (n = 4), calculated from the number of counted nuclei compared to untreated controls. Statistical comparisons to DMSO control for cell counts were performed using one-way ANOVA: p < 0.05 (#), p < 0.01 (##). Representative images are shown in (C), with staining γH2AX foci in grey or the merge image with DAPI and γH2AX staining. Scale bar = 20 μm

Combined exposure to BaP and AFB1 leads to a synergistic increase of DNA damage

Based on the BMC values calculated for each compound, mixtures of BaP and AFB1 were subsequently prepared and tested at different ratios. For each BMR, BaP and AFB1 were evaluated individually (1:0 and 0:1 ratio, corresponding to their respective BMCs) as well as in mixtures with equipotent (1:1) and non-equipotent ratios (3:1 and 1:3), in order to assess the relative contribution of each compound to the observed effect (Table 1).

The mean number of γH2AX foci per nucleus and the distribution of cells according to γH2AX foci number (0–1, 2–5, or > 5 foci) were evaluated in BaP and AFB1 treated cells, either individually or in mixtures, at the different tested BMR levels.

At BMR 300%, as expected, cells treated with BaP or AFB1 alone exhibited same mean γH2AX foci numbers per nucleus (approximately 2.5) (Fig. 2A). On the opposite, mixture treatments significantly increased γH2AX foci formation, with mean values rising from approximately 3.7 foci per nucleus for the 3:1 ratio to 4.2 foci per nucleus for the 1:3 ratio. Consistent with this, the proportion of cells containing more than five γH2AX foci increased from near 12% in BaP or AFB1 treated cells to 19% and 22% in the 3:1 and 1:3 mixtures, respectively (Fig. 2B). These increased DNA damages were associated with a decrease in RCC in the three mixtures tested, while BaP and AFB1 alone did not significantly affect cell viability.

Fig. 2

Mean number of γH2AX foci per nucleus and cell classification based on foci number after BaP and AFB1 treatment at BMR300% or BMR400%. The mean number of γH2AX foci after 24 h treatment with BaP and AFB1, applied alone or in mixtures, and the mean percentage of relative cell count (RCC%) are presented at BMR300% and BMR400% (2-A and 2-C respectively). Bars represent mean ± SEM (n = 3), and dots represent RCC% ± SEM (n = 3). The percentage of cells presenting 0–1, 2–5 or > 5 foci of γH2AX for the different mixture ratios are presented for BMR300% and BMR400% (2-B and 2-D respectively). Statistical analyses were performed by one-way ANOVA (for γH2AX foci mean number) or two-way ANOVA (for % of cell classes) comparing BaP alone with mixtures and AFB1 alone with mixtures. Comparisons of BaP with mixtures are indicated by black symbols, whereas comparisons of AFB1 with mixtures are indicated by green symbols (* for mean γH2AX foci, # for RCC%). p < 0.05 (* or #) and p < 0.01 (** or ##)

At BMR 400%, a comparable pattern was observed. BaP and AFB1 alone induced mean γH2AX foci numbers of approximately 3 per nucleus, whereas mixture exposures resulted in significantly higher values, ranging from approximately 4.4 foci per nucleus at the 3:1 ratio to 4.9 foci per nucleus at the 1:3 ratio (Fig. 2C). The proportion of cells with more than five γH2AX foci increased from 15% for individual compounds to 22% in mixture-treated cells (Fig. 2D). Concomitantly, RCC values declined from approximately 103% and 84% for BaP and AFB1 alone to 78% and 65% in the 3:1 and 1:3 mixtures, respectively.

To evaluate the interaction between BaP and AFB1 through the principle of additivity, the experimental mixture data were analyzed using the PROAST dose-addition model in R. As shown in Fig. 3, the experimental mixture responses for γH2AX foci induction were plotted against the predicted concentration–response curve derived under the assumption of additivity. The experimental points obtained with the mixtures displayed clear up- and leftward deviations from the predicted additive curve, indicating synergistic interactions within the mixtures.

Fig. 3

Comparison of BaP and AFB1 mixture responses to additive model. BaP and AFB1 responses were measured by γH2AX immunofluorescence (mean number of γH2AX foci), modeled in R using dose-addition model 15. Black curve represents predicted concentration-response based on additivity. Mixtures responses are presented (dark blue downward triangles) at BMR300% and BMR400% (from left to right) with their corresponding 95% confidence intervals

In conclusion, our findings indicate that combined exposure to BaP and AFB1 produces a synergistic enhancement of genotoxicity compared to the single-compound exposure, supporting the presence of a biological interaction between these compounds at the cellular level.

Metabolic activation loop after BaP-AFB1 co-exposure induces synergistic genotoxic effects

As our results support the synergistic interaction between BaP and AFB1, we aimed to elucidate the biological mechanistic basis underlying this effect. We hypothesized that BaP-mediated AhR activation enhances CYP1A1 and CYP1A2 expression and thereby increases the metabolic activation of AFB1, ultimately contributing to the observed increased genotoxicity. To test this hypothesis and evaluate the involvement of the AhR pathway, mechanistic modulators, including a non-genotoxic AhR agonist (BNF) and a selective AhR antagonist (CH223191), were used.

The mean number of γH2AX foci per nucleus (Fig. 4A) and the distribution of cells according to γH2AX foci number (Fig. 4B) were analyzed for the different conditions. As expected, CH223191 and BNF alone did not significantly affect γH2AX foci formation compared with negative control cells. CH223191 pre-treatment significantly reduced BaP-induced γH2AX foci formation, decreasing from 2.6 to 1.5 the mean number of foci per cell. Consistently, after AhR antagonist treatment, the proportion of cells exhibiting 2–5 and > 5 γH2AX foci was significantly reduced, from 13% to 14% to 4% and 7% respectively. A similar effect was observed in cells exposed to the BaP-AFB1 (1:1) mixture, with CH223191 reducing the mean number of γH2AX foci from 3.1 to 1.8 and decreasing the percentage of cells with 2–5 and > 5 foci from 8% to 4% and 18% to 8% respectively. In contrast, no significant differences were observed between AFB1-treated cells and cells pre-treated with CH223191 prior to AFB1 exposure.

Fig. 4

Mechanistic analysis with AhR pharmacological modulators. Mean number of γH2AX foci per nucleus (4-A) and distribution of cell classes based on γH2AX foci number (4-B) following 24-hour treatment with BaP and AFB1, either alone or in a 1:1 ratio combination at a BMR300%, as well as AFB1 combined with β-naphthoflavone (BNF, 1 µM), with or without a 2-hour pre-treatment with CH223191 antagonist (10 µM). Bars represent mean ± SEM (n = 3). Statistical comparisons between conditions with and without CH223191 pre-treatment were performed using one-way ANOVA (for the mean number of γH2AX foci) and two-way ANOVA (for the percentage of cell classes): p < 0.05 (*), p < 0.01 (**)

To clarify the role of AhR in the mixture effect, we used BNF as a well-established AhR agonist to determine whether activation of this pathway could enhance AFB1-induced genotoxicity. Co-exposure resulted in a significant increase in γH2AX foci formation compared with AFB1 alone, with the mean number of foci per cell rising from 2.3 to 3.3 and the proportion of cells with > 5 foci increasing from 9% to 12%, indicating that AhR activation by BNF exacerbates AFB1-induced DNA damage. This potentiation was significantly attenuated by CH223191 AhR antagonist, which reduced the mean number of γH2AX foci to 2.7 and decreased the percentage of highly damaged cells (> 5 foci) to 7.5%. Together, these findings demonstrate that the observed exacerbation of AFB1-induced genotoxicity is mediated by AhR signaling.

Having established that BNF, a potent AhR agonist, exacerbates AFB1-induced genotoxicity, we next sought to determine whether AhR could be activated by other compounds. In the context of dietary co-exposure to PAH mixtures, we focused on non-genotoxic PAHs such as fluoranthene and phenanthrene which are known to activate AhR signaling. We therefore evaluated their effects, alone or in combination with AFB1, by quantifying a biomarker of DNA damage. The mean number of γH2AX foci per nucleus (Fig. 5) was analyzed under the different conditions. Although fluoranthene and phenanthrene alone did not significantly induce γH2AX foci formation compared with control cells, co-exposure with AFB1 led to a marked increase in γH2AX foci formation compared to AFB1 alone, with the mean number of foci per cell rising from 2.2 to 4.6 and 5.5 for fluoranthene and phenanthrene respectively. These results indicate that non-genotoxic AhR-activating PAHs can also exacerbate AFB1-induced DNA damage, supporting an AhR-dependent mechanism.

Fig. 5

γH2AX immunofluorescence analysis after AFB1 treatment in mixture with non genotoxic PAHs. Mean number of γH2AX foci per nucleus following 24-hour treatment with Fluoranthene (50µM) and Phenanthrene (50µM) alone or in combination with AFB1 (0.301µM). Bars represent mean ± SEM (n = 3). Statistical comparisons between conditions were performed using one-way ANOVA: p < 0.05 (*), p < 0.01 (**)

To further examine the involvement of AhR in BaP-AFB1 interactions, CYP1A2 and CYP1B1 protein levels were analyzed by western blot 8 h after exposure to BaP, AFB1, or their 1:1 mixture at BMC300. Cells were also treated with 1 µM BNF agonist alone or in combination with AFB1 as controls, and all treatment conditions were performed with or without a 2-h pre-incubation with CH223191 antagonist. Treatment with CH223191 alone did not significantly alter CYP1B1 or CYP1A2 protein expression (Fig. 6A and B). Exposure to BaP and the 1:1 BaP–AFB1 mixture (BMC300) markedly induced CYP1B1 levels (7.4 and 7-fold, respectively), and CYP1A2 levels (7.8 and 6.9-fold, respectively). Pre-incubation with CH223191 attenuated these inductions, reducing CYP1B1 expression to 4.4 and 2.3-fold, and CYP1A2 expression to 5.2 and 3.5-fold, respectively (Fig. 6A). Similarly, BNF and the BNF-AFB1 combination increased CYP1B1 protein levels (4.5 and 4.7-fold, respectively) and CYP1A2 levels (5.3 and 4.4-fold, respectively). These effects were also diminished in the CH223191 pre-treatment condition, with CYP1B1 induction reduced to 2.6 and 2.9-fold and CYP1A2 to 2.7 and 2.2-fold, respectively (Fig. 6B). In contrast, AFB1 alone did not significantly affect CYP1B1 or CYP1A2 protein expression, further underlining the role of BaP and AhR in these CYP450 modulations.

Fig. 6

CYP1B1 and CYP1A2 protein expression after 8 h exposure. Cells were treated with BaP and AFB1 individually or as a 1:1 mixture at a BMR300% (A), or with BNF (1µM) and AFB1 (0.301µM), alone or in combination, with or without 2-hour pre-treatment with CH223191 (10µM). Protein levels were normalized to GAPDH and expressed as CYP450/GAPDH ratios. Data represent mean ± SEM of three independent experiments

Finally, CYP1A1/1B1 enzymatic activities were measured to determine whether changes in protein level correlated with an altered function. BaP and the BaP–AFB1 mixture significantly increased CYP1A1/1B1 activities compared to control, with 4.7 and 5.9-fold inductions respectively (Fig. 7). CH223191 pre-incubation significantly attenuated CYP1A1/1B1 induction, greatly reducing activity by 49% and 66.4% in BaP and BaP–AFB1-treated cells, respectively. No significant changes in CYP1A1/B1 activities were observed in AFB1-treated cells, regardless of CH223191 pre-treatment. Finally, as the mRNA level of CYP1A1 in differentiated HepaRP cells is low compared to CYP1A2 mRNA level, and given that the protein level does not seem modulated after AhR antagonist or agonist treatment (Supplementary Information 2), this suggests that CYP1A1 do not participate to the observed effect.

Fig. 7

CYP1A1/1B1 activity after 8 h exposure. Cells were treated with BaP or AFB1 individually, or as a 1:1 mixture at a BMR of 300%, with or without 2-hour pre-treatment with the AhR antagonist CH223191 (10µM). Data are presented as mean ± SEM (n = 3). Statistical comparisons between conditions with and without CH223191 were performed using one-way ANOVA: p < 0.05 (*), p < 0.01 (**)

Collectively, these findings indicate that BaP increases CYP1B1 and CYP1A2 protein expression, as well as the CYP1A1/1B1 enzymatic activities in an AhR-dependent manner, with CYP1B1 playing a major role in addition to CYP1A2. Moreover, the results also support a key function for AhR signaling in mediating the synergistic genotoxic effects observed following co-exposure to BaP and AFB1.