In silico and in vitro evaluation of potential agonistic and antagonistic estrogenic and androgenic activities of pure cyanotoxins, microcystin-LR and cylindrospermopsin

In silico and in vitro evaluation of potential agonistic and antagonistic estrogenic and androgenic activities of pure cyanotoxins, microcystin-LR and cylindrospermopsin Antonio Casas-Rodríguez, Antonio Cascajosa-Lira, María Puerto * , Ana María Came´ an, Angeles Jos Area of Toxicology, Faculty of Pharmacy, University of Sevilla, Profesor García Gonz´ alez nº 2, Sevilla 41012, Spain ARTICLE INFO Edited by Dr. Hao Zhu Keywords: MC-LR CYN Estrogens Androgens Docking ABSTRACT The potential endocrine disruption activity of cyanotoxins, particularly their effects on estrogen and androgen receptors (ER, AR), remains poorly understood. In the present study, the potential agonistic/antagonistic estrogenic and androgenic activities of MC-LR and CYN have been determined for the first time with validated OECD Test Guidelines No. 455 and 458, respectively. The data show that only MC-LR demonstrated weak estrogenic agonistic effects (LogPC 10 value of −9.85 M), while both toxins displayed antagonistic effects on the ER, with LogIC 30 values of −4.4 and −6.4 for MC-LR and CYN, respectively. In addition, neither MC-LR nor CYN exhibited agonistic/antagonistic activities in AR. Docking studies revealed potential interactions between both toxins and AR, with CYN showing a higher predicted affinity for this receptor. In vivo studies, particularly those investigating androgen disruption, are warranted to confirm the endocrine disrupting potential of MC-LR and CYN. 1. Introduction The growth of harmful algal blooms, particularly those caused by cyanobacteria, has been increasing, and this increase is linked to climate change and generates concern about water quality and public health (Huisman et al., 2018, Chen et al., 2023). Cyanobacteria are capable of producing a variety of toxins, known as cyanotoxins such as microcystins (MCs), cylindrospermopsin (CYN), guanitoxin, anatoxin, saxitoxin, and nodularins, among others. Toxins can be classified on the basis of their chemical structure and poisonous effects on mammals. For example, peptides such as MCs and nodularins act as hepatotoxins, while alkaloids such as CYN have both hepatotoxic and cytotoxic properties. Others, such as anatoxin-A, are well known for their neurotoxic effects (Chorus and Welker, 2021; Ricciardelli et al., 2023). Among cyanotoxins, MCs are the most widely detected and studied in freshwater and food. They are cyclic heptapeptides composed of five common amino acids, with a general structure of cyclo(-D-Ala1-L-X2-Derythro-β-methylAsp(iso-linkage)3-L-Z4-Adda5-D-Glu(iso-linkage)6-Nmethyldehydro-Ala7). Structural modifications can occur in all seven amino acids of MCs, producing more than 279 structural variants of MCs (Bouai¨cha et al., 2019). Microcystin-LR (MC-LR) is the variant most frequently assessed due to its high toxicity in various models. However, there is growing interest for other variants (Testai et al., 2016; WHO, 2020a). MCs are potent hepatotoxins and tumor promoters. Their toxicity is attributed to strong inhibition of protein phosphatases 1 and 2 A (PP1 and PP2A), induction of oxidative stress, and effects on cell signaling pathways (Bouai¨cha et al., 2019; Puerto et al., 2011). Although the liver is the primary target organ, MCs have been shown to impact other tissues, including the reproductive system (Zhang et al., 2021) and endocrine function (Chen et al., 2021; Casas-Rodríguez et al., 2022). CYN is a water-soluble alkaloid with a unique three-ring structure containing guanidine and hydroxymethyluracil moieties. Its zwitterionic nature contributes to its high solubility in water (Pichardo et al., 2017). CYN toxicity is attributed to several mechanisms, including inhibition of protein and glutathione synthesis, induction of oxidative stress, and genotoxicity (Runnegar et al., 1995; Froscio et al., 2003; Puerto et al., 2011, 2018). Additionally, cytochrome P450 appears to play a role in mediating its toxic effects (WHO, 2020b; Buratti et al., 2017). Moreover, CYN acts as a broad-spectrum cytotoxin, targeting * Corresponding author. E-mail address: [email protected] (M. Puerto). Contents lists available at ScienceDirect Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv https://doi.org/10.1016/j.ecoenv.2024.117456 Received 24 June 2024; Received in revised form 7 November 2024; Accepted 30 November 2024 Ecotoxicology and Environmental Safety 289 (2025) 117456 Available online 3 December 2024 0147-6513/© 2024 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/bync-nd/4.0/ ). various organs beyond the primary targets of the kidney and liver. These include the eyes, spleen, lungs, thymus, and heart (Pichardo et al., 2017; WHO 2020b). However, the effects of CYN in the endocrine system have been scarcely investigated (Casas-Rodríguez et al., 2022). Endocrine disruptors (EDs) can be defined as external chemicals or mixtures that can interfere with the body’s hormonal system. This disruption can lead to various health problems in organisms, their offspring, or entire populations. (WHO/IPCS World Health Organization/International Programme on Chemical Safety, 2002). This concern is driven by the growing recognition of numerous compounds with diverse structures and applications (pesticides, plastics, metals, and food contaminants) that exhibit endocrine disrupting activity (ECHA and EFSA, 2018). To address this growing concern, the Organization for Economic Cooperation and Development (OECD) developed a conceptual framework for the testing and evaluation of EDs, comprising five different levels (OECD 150, 2018). Cyanotoxins have been proposed as potential EDs due to their estrogenic activity and the ability to interfere with intracellular receptor signalling in vertebrates, affecting hormone regulation, reproduction, and development (Rogers et al., 2011). Recent reviews (Casas-Rodríguez et al., 2022) highlight diverse endocrine disruption (ED) effects of MC-LR, disrupting the synthesis of steroid hormones, resulting in dysfunction of the endocrine system and reproductive toxicity and including pathological damage in related organs and cells. In vivo, MC-LR affects serum levels of hormones and their expression by affecting the HPG (Wang et al., 2012, 2016). Through this axis, gonadotropin-releasing hormones (GnRH) released from the hypothalamus stimulate the secretion of gonadotropin hormones (GtH), including follicle-stimulating hormones (FSH) and luteinizing hormones (LH), by the pituitary. The GtH are then transported to the gonads to induce steroidogenesis and produce sex steroid hormones, such as 17β-estradiol (E2) and testosterone (T). MC-LR can alter the levels of these hormones, which modulate the reproductive process. Furthermore, several studies have shown that MC-LR can alter steroidogenesis by causing changes in the expression of genes involved in this process (Chen et al., 2021; Hou et al., 2018). In the case of CYN, studies focusing on the potential endocrine disrupting effects of the toxin are very scarce. Some estrogenic effects have been detected in cyanobacteria extracts containing CYN, but information on the estrogenic potency of CYN is very limited (Casas-Rodríguez et al., 2022). In this context, our team recently investigated the potential estrogenic effects and thyroid dysfunction of both cyanotoxins in vivo using rats following the OECD test guideline 440 (Casas-Rodriguez et al., 2023). This work demonstrated that MC-LR and CYN were not estrogenic in the doses and conditions tested, but the effects of thyroid disruption cannot be ruled out. There is little research on potential interactions of cyanotoxins with signalling pathways of various molecular receptors, including estrogen (ER), androgen (AR), glucocorticoid (GR), and retinoid (RAR) receptors (Jonas et al., 2015; Mallia et al., 2020). To our knowledge, no studies have investigated the estrogenic and/or androgenic activities of purified MC-LR. With respect to CYN, only one study investigated its potential estrogenic activity using the yeast estrogen screen assay (Liu et al., 2018). Therefore, more research is needed to elucidate their potential interactions with various hormone receptors (ER, AR). Moreover, in silico prediction of hormone receptor interaction allows evaluation of how a given chemical compound may interact with these receptors and whether it has the potential to act as an agonist (activator) or antagonist (blocker). This is especially relevant in the assessment of chemicals, as certain compounds can have adverse effects by altering normal hormones function. By using in silico methods, such as molecular modeling and molecular docking, interactions between a chemical compound and hormone receptors can be predicted and evaluated. These computational approaches allow the structure and affinity of a compound for a specific hormone receptor to be analyzed, as well as for predicting possible bioactive effects. In fact, molecular dynamics simulation has been applied to investigate the conformation and binding of MCs to serin/threonine protein phosphatase 1 (PPP1) (Jaeger-Honz et al., 2022). In addition, computer simulations, particularly those employing ADMET models, have become valuable tools in understanding the toxicity and the ease with which different microcystin variants enter and leave the body. These simulations strengthen the findings of previous laboratory and animal studies (Gonçalves da Silva et al., 2021). In silico prediction of hormone receptor interaction not only saves time and resources but also provides valuable information for decision-making in the design of new compounds and in the safety assessment of existing ones. This can contribute to the early identification of toxic compounds with undesirable hormone potential and help to avoid health and environmental risks. These in silico predictions have been applied to several and different EDs (Kenda et al., 2020; Durcik et al., 2022), but as far as we know, no previous studies have been carried out on cyanotoxins. Taking into account all these facts, the objective of the present work was to investigate for the first time the in vitro estrogenic and androgenic activities (agonist and antagonist) of pure MC-LR and CYN by applying the OECD Test Guideline No. 455 to detect ER agonists and antagonists and the OECD Test Guideline No. 458 for the detection of androgenic agonist and antagonist chemicals. The in vitro data obtained were also compared with in silico predictions using an autodock vina docking program. Thus, these assays are framed in levels 1 (in silico) and 2 (in vitro) of the OECD 150 Conceptual Framework for the endocrine disruption. 2. Material and methods 2.1. Test compounds and reagents The toxins were obtained from a commercial supplier (Enzo Life Sciences, Lausanne, Switzerland). Both toxins had high purities, with CYN at 95 % and MC-LR at 99 %. All the necessary cell culture reagents were provided by Gibco (Biomol, Sevilla, Spain). Finally, the chemicals used for the cytotoxicity assays were purchased from Sigma-Aldrich (Madrid, Spain). 2.2. Model system 2.2.1. Transactivation assay for the detection of estrogenic agonist and antagonist activity A specific human cervical cancer cell line, hER α -HeLa-9903 (JCRB 1318), was used to identify compounds that interact with the estrogen receptor. This cell line has been modified to include a reporter gene for the human alpha estrogen receptor (ER α ) and a luciferase gene. Luciferase activity can be measured to assess ER α activation (e.g., by agonists) or inhibition (e.g., by antagonists). These cells were typically cultured in an incubator set at 37◦C with 5 % CO 2 and 95 % humidity. A specialized culture medium (Eagle Minimum Essential Medium without phenol red) supplemented with kanamycin (60 mg/L) and charcoal treated fetal bovine serum (DCC-FBS) was used. For consistency, experiments were repeated at least three times using cells within a specific passage range (passages 5–20). 2.2.2. Transactivation assay for the detection of androgenic agonist and antagonist activity A specific cell line, AR-EcoScreen (JCRB 1328), was chosen to assess the potential of the toxins to activate (agonist) or inhibit (antagonist) AR. This cell line is engineered to include a reporter gene for AR, a firefly luciferase gene, and an additional gene expressing renilla luciferase. Renilla luciferase allows for the detection of potential cytotoxicity within the system. These cells are typically cultured in an incubator set at 37◦C with 5 % CO 2 and 95 % humidity. A specialized culture medium (phenol red-free DMEM/F12) supplemented with fetal bovine serum (FBS), zeocin, hygromycin, penicillin, and streptomycin was used. For A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 2 consistency, the experiments were repeated at least three times using cells within a specific passage range (passages 5–20). 2.3. Evaluation of cytotoxicity To assess cytotoxicity, hER α -HeLa-9903 cells (10,000 cells/well) were exposed to varying concentrations of MC-LR (0, 20, 50, 100 and 200 µM) and CYN (0, 0.5, 0.625, 1.25, 2.5 and 3 µM) for 24 hours after a brief incubation period (3 hours). Similarly, AR-EcoScreen cells (9000 cells/well) were exposed to different concentrations of MC-LR (0, 1, 5, 10, 15, 25, 50, 100, and 150 µM) and CYN (0, 1.5, 3, 4.5, 6, 7.5, 9, and 12 µM) concentrations for 24 hours. The cytotoxicity in hER α -HeLa9903 cells and AR-EcoScreen cells was assessed using Cell Counting Kit8 and MTS assays, respectively. 2.4. ER α transactivation assay Following the OECD 455 test guideline, estrogenic activity was evaluated. Lab proficiency was first established using a variety of compounds with varying estrogenic activity, including strong (17β-estradiol, E2), weak (17 α -estradiol), very weak (17 α -methyltestosterone), and negative control (corticosterone). For antagonist assays, tamoxifen served as a positive control and flutamide as a negative control. After proficiency was confirmed, the toxins were tested. Cells were seeded (10,000 cells/mL) in luminometer plates and treated with MC-LR, CYN, reference compounds, vehicle control, and positive control (agonist: 1 nM E2, antagonist: 25 pM E2). The highest non-cytotoxic toxin concentration was then used for the transactivation assay. Luciferase activity was measured using a commercially available kit according to the manufacturer’s instructions (Promega E2510; Promega E2920), and luminescence was measured with a microplate reader. The calculated endpoints were the concentration of toxins at which the measured activity in an agonist assay is 10 % (LogPC 10 ) or 50 % (LogPC 50 ) of the maximum activity induced by 1 nM E2. And the concentration of toxins at which the measured activity in an antagonist assay inhibits by 50 % the maximum activity induced by 25 nM E2 (LogIC 50 ). 2.5. AR transactivation assay An AR activity assay was performed following the OECD 458 guidelines to assess the potential androgenic and antiandrogenic effects of toxins. Laboratoy proficiency was first established using reference standards with known AR agonist (5 α -Dihydrotestosterone, Mestanolone) and antagonist activity (Hydroxyflutamide, Bisphenol A). A negative control (Di (2-ethylhexyl) phthalate) that did not have AR activity was also included. The cells were seeded (9000 cells/well) in specific well plates and incubated for 24 hours. Then, they were exposed to non-cytotoxic concentrations of the toxins, reference compounds, vehicle control (0.1 % DMSO), positive controls (for the agonist assay: 10 nM DHT; for the antagonist assay: 1 μ M HF), and a cytotoxicity control (cycloheximide 10 µg/mL). Following a further incubation period (24 h), the activity of luciferase was measured using a commercially available kit according to the manufacturer’s instructions. Luminescence was measured with a microplate reader. Finally, similarly to the ER α assay the measured endpoints were the LogPC 10 and LogPC 50 , for the agonist assay and LogIC 30 and LogIC 50 for the antagonist assay. 2.6. Molecular docking The molecular docking of MC-LR and CYN was performed according to Cascajosa-Lira et al., (2023). The structures of MC-LR and CYN were obtained using ChemDraw (version 22.0) and energy-minimized by using the PyMOL minimization protocol. The resulting conformation with the lowest energy was chosen as the initial conformation for the docking analysis. The docking was performed using an automated method to determine the appropriate complex structure between MC-LR or CYN and the ER or AR. Four different 3D crystalized structures of ER were downloaded from the Protein Data Bank. Two of these structures were in agonist conformation (1ERE and 3ERD), while the remaining two were in antagonist conformation (1ERR and 3ERT). Moreover, it is important to point out that these four structures have slight variations in the distance between atoms. Missing hydrogen atoms were added to the receptor molecules using PyMOL (version 2.5) for docking calculations. The bound ligands were removed from the receptor structures. For docking computations, AutoDock 3.0 software was used, setting a grid box dimension of 80x80x80 Å to ensure ample space for ligandreceptor interactions. The ER docking process was conducted to predict whether MC-LR or CYN possessed potential agonist or antagonist activity. This evaluation involves a comparison between the computed energy values obtained with the toxins with those of established agonists (such as E2 and diethylstilbestrol (DES)) and antagonists (including 4hydroxytamoxifen and raloxifene) in the same experimental model. 2.7. Statistical analysis All cell-based assays were performed in triplicate. Data are reported as the mean ±standard deviation (SD) of at least two independent experiments. For agonist assays, data were normalized to a control group (cells treated with 0.1 % DMSO). For antagonist assays, normalization was performed to a spiked control (cells treated with 0.1 % DMSO and a known agonist). Statistical analysis involved a common technique, oneway analysis of variance followed by a post hoc test (Dunnett test) to compare the effects of each toxin concentration to the appropriate control group. A threshold p-value (less than 0.05) was used to determine statistical significance. Statistical software was used to perform the analyses (GraphPad Prism). 3. 3. Results 3.1. Cytotoxicity assay Preliminary tests were carried out to determine the appropriate range of concentrations (non-cytotoxic) of the two cyanotoxins to be tested, according to the OECD test guidelines 455 and 458. The results of the viability tests for the different cell lines are shown in Fig. 1. Respect to the hER α -HeLa-9903 cells, MC-LR (Fig. 1A) significantly decrease cell viability at the highest concentration tested (200 µM). For this toxin, 100 µM was chosen as the highest concentration for the ER α transactivation assay. On the other hand, for CYN (Fig. 1B), a significant decrease in viability was observed at 2.5 and 3 µM compared to the negative control. Finally, 0.5 µM was chosen as the highest noncytotoxic concentration for the ER α transactivation assay. In the case of AR-EcoScreen cells, a significant decrease in viability was found after exposure to 50 µM of MC-LR (Fig. 1C). However, EC 50 could not be calculated, as viability was always higher than 50 % compared to negative control. In this case, EC 20 was 54.65 ±1.21 µM and 54 µM was chosen as the highest concentration for the AR transactivation assay. Furthermore, CYN (Fig. 1D) caused a significant decrease in viability compared to control cells from 3 µM. EC 20 was 2.37 ±0.55 µM and 2 µM was chosen as the highest concentration for the AR transactivation assay. 3.2. Effects of MC-LR and CYN in the ER α using the hER α -HeLa-9903 cell line Before the toxins assays, a proficiency test was performed in triplicate using various reference standards (OECD 455, 2021). For ER α agonist assays, the proficiency test met the acceptable criteria with fold-induction values of 4.3, 4.6 and 4.7 at 1 nM E2 (Table 1), higher than the value of at least 4 times the mean of the vehicle control. The LogPC 50 , LogPC 10 , LogEC 50 and the Hill slope of reference chemicals fell A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 3 Fig. 1. Cytotoxicity assay on hER α -HeLa-9903 (A, B) and AR-EcoScreen (C, D) cells after 24 h of exposure to different concentrations of MC-LR (A, C) and CYN (B, D). Values are expressed as mean ±SD. All experiments were performed at least three times and at least in sextuplicate per concentration. Significant levels observed are *p <0.05, * **p <0.001 and * ** *p <0.0001 compared to the control group. C.solv: MeOH 0.1 %. Table 1 Proficiency test for hER α -Hela-9903 STTA assay. Acceptable criteria Results Agonist test LogPC 50a LogPC 10b LogEC 50 Hill Slope LogPC 50 LogPC 10 LogEC 50 Hill Slope E2 −11.4 to −10.1 <−11 −11.3 to −10.1 0.7–1.5 −11.2 −11.3 −11.4 <−11 <−11 <−11 −11.2 −10.7 −10.8 0.9 0.9 1.3 17α -estradiol −9.6 to −8.1 −10.7 to −9.3 −9.6 to −8.4 0.9–2−8.1 −8.4 −9.1 −10.6 −10.2 −10.7 −8.48 −8.43 −8.8 1.3 1.0 0.9 17α -methyltestosterone −6.0 to −5.1 −8.0 to −6.2 - - −5.9 −5.1 −6.0 −8.0 −7.9 −8.0 - - Corticosterone - - - - - - - - Fold Induction of 1 nM E2 ≥4 4.7 4.3 4.6 Antagonist test LogIC 50c LogIC 50 Tamoxifen −5.942 to −7.596 −6.9 −6.7 −6.8 Flutamide - - Fold Induction of 25 nM E2 ≥4 4.2 4.5 6.4 a The concentration of a test chemical at which the measured activity in an agonist assay is 50 % of the maximum activity induced by the 1 nM E2. b The concentration of a test chemical at which the measured activity in an agonist assay is 10 % of the maximum activity induced by the 1 nM E2. c The concentration of a test chemical at which the measured activity in an antagonist assay inhibits by 50 % the maximum activity induced by 25 nM E2. A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 4 within the acceptable criteria and these results presented a concentration-response curve that comprises the baseline and activity values. In that sense, for the positive control, the LogPC 10 were <-11, and the LogPC 50 values were −11.2, −11.3 and −11.4 (Table 1 and Figure S1A). Regarding the quality control of the assay, the fold-induction corresponding to the PC 10 value of the concurrent positive control (1 nM E2) was greater than 1 +2 SD of the concurrent vehicle control (DMSO). With respect to toxins, the criteria of OECD 455 considered a positive result when the maximum level of response induced by the toxin tested (RPC max) is equal or exceeds 10 % of the response of the positive control (1 nM E2). In the present work, MC-LR showed positive ER α transactivation agonistic activity, and the luciferase signal increased with LogPC 10 value of −9.85 M, corresponding to 0.14 nM (Fig. 2A). On the contrary, CYN did not show ER α agonist effects and, consequently, LogPC 10 cannot be calculated (Fig. 2A). For the ER α antagonist assay, the fold-induction of 25 pM E2 was 4.2, 4.5 and 6.4, higher than 4 times, criteria of the OECD 455 (Table 1). Additionally, the LogIC 50 values of tamoxifen fell within the acceptable criteria and, in the case of flutamide, IC 30 could not be calculated (Figure S1B). To ensure the precision of this test, quality control measures were implemented according to the OECD 455 guidelines. This involved using reference compounds with known effects. Strong estrogen (1 nM E2) triggered significant relative transcriptional activation (RTA) that exceeded 100 %, indicating that the test system effectively detects estrogenic activity. On the contrary, the RTA remained below 40.6 % when exposed to a range of concentrations (0.1 μ M-1 μ M) of 4hydroxytamoxifen, a known anti-estrogen. This confirms the ability to identify anti-estrogenic compounds. Finally, a non-estrogenic compound (100 μ M Digitonin) produced an RTA of less than 0 %, demonstrating the precision of the assay in distinguishing between estrogenic, anti-estrogenic and neutral substances. These quality control steps ensure reliable results for further analysis. In the case of the toxins tested, the decision criteria for considering a positive response is the possibility of calculating the IC30 in at least 2/2 or 2/3 runs (OECD 455, 2021). Both toxins, MC-LR and CYN, exhibited antagonistic effects, showing LogIC 30 values of −4.4 and −6.4 (39.8 µM and 0.39 µM) for MC-LR and CYN, respectively (Fig. 2B). 3.3. Effects of MC-LR and CYN in AR using the AR-Ecoscreen cell line Before the application of the assay to cyanotoxins, a proficiency test was conducted in triplicate using various reference standards (OECD 458, 2021). In the agonist assay, the average luciferase activity triggered by a set concentration of DHT (10 nM) needed to be at least 6.4 times higher compared to the average activity observed with the control group (vehicle control). Similarly, for the antagonist assay, the average luciferase activity produced by a specific low concentration of DHT (500 pM) had to be at least 5.0 times greater than the control group. The results obtained in the present work are presented in Table 2 with values of 8.20, 8.80, and 9.97 for the agonist assay and values of 6.89, 7.32 and 8.12 for the antagonist assay. LogPC 10 , LogPC 50 , LogIC 30 and LogIC 50 of the reference chemicals were within acceptable criteria and these results presented concentration-response curves comprising baseline and activity values (Table 2 and Figure S2). In the agonist assay, OECD 458 criteria considered a positive result when the maximum level of response induced by the toxin tested (RPC max) is equal to or exceeds 10 % of the response of the positive control (10 nM DHT). In the present work and for the MC-LR concentrations tested, only values close to 1 % of the response of the positive controls were found (Fig. 3A). Similarly, CYN was considered negative in the AR transactivation agonist assay (Fig. 3A). Regarding the AR antagonistic Fig. 2. Estrogen receptor agonist (A) and antagonist (B) effect of MC-LR and CYN. Data are represented as mean ±SD of at least two independent repeats. The significant levels observed are *p <0.05 compared to the SC group. E2: 17β-estradiol; SC: spike in control; RTA: Relative Transcriptional Activity. Table 2 Proficiency test for AR-EcoScreen STTA assay. Acceptable criteria Results Agonist test LogPC 10a LogPC 50b LogPC 10 LogPC 50 5α -Dihydrotestosterone (DHT) −12.08 to −9.87 −11.03 to −9.00 −10.44 −10.51 −10.70 −9.53 −9.60 −9.72 Mestanolone −10.92 to −10.41 −10.15 to −9.26 −10.74 −10.82 −10.89 −9.64 −10.08 −10.07 Di(2-ethylhexyl)phthalate (DEHP) - - - - Fold Induction of 10 nM DHT ≥6.4 8.20 8.80 9.97 Antagonist test LogIC 30c LogIC 50d LogIC 30 Log IC 50 Hydroxyflutamide −8.37 to −6.41 −7.80 to −6.17 −7.78 −8.22 −7.73 −7.12 −7.32 −7.38 Bisphenol A −7.52 to −4.48 −7.05 to −4.29 −5.80 −5.85 −5.76 −5.50 −5.55 −5.47 Di(2-ethylhexyl)phthalate (DEHP) - - - - Fold Induction of 500 pM DHT ≥5 6.89 7.32 8.12 a The concentration of a test chemical at which the measured activity in an agonist assay is 10 % of the maximum activity induced by 10 nM DHT. b The concentration of a test chemical at which the measured activity in an agonist assay is 50 % of the maximum activity induced by 10 nM DHT. c The concentration of a test chemical at which the measured activity in an antagonist assay inhibits by 30 % the maximum activity induced by 500 pM DHT. d The concentration of a test chemical at which the measured activity in an antagonist assay inhibits by 50 % the maximum activity induced by 500 pM DHT. A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 5 assay, the decision criteria to consider a positive response is the possibility of calculating IC 30 in at least 2/2 or 2/3 runs (OECD 458). In this case, both toxins slightly inhibited AR activation, but no significant differences were found in the experiments performed and LogIC 30 cannot be calculated (Fig. 3B). 3.4. Molecular docking: MC-LR and CYN with the ER and AR receptors The molecular docking process, which utilizes the crystal structures of MC-LR and CYN in complex with the four conformations of the estrogen receptor (1ER1, 3ERD, 1ERR and 3ERT), has been conducted and revealed potential binding conformations. All molecular models are represented in supplementary figures. The molecular docking analysis with MC-LR (Figure S3) suggests that this toxin interacts more favorably with the receptor in its agonist conformation. Specifically, when MC-LR was docked with the receptors 1ERE and 3ERD, the free binding energies were −7.1 kcal/mol and −8.2 kcal/mol, respectively. On the contrary, MC-LR exhibits weaker interactions with the receptor in its antagonistic conformation, with binding energies of −6.4 kcal/mol for 1ERR and −7.8 kcal/mol for 3ERT. MC-LR has three amino acids in its cyclic chemical structure that potentially interact with the amino acids of the estrogen receptor 1ERE. The nitrogen in the arginine chain of MC-LR can interact with ASP-321 and GLU-323 of 1ERE. Furthermore, the glutamic acid of MC-LR could form a binding interaction with GLY-442 within 1ERE. Additionally, the amino acid methyl aspartic acid of MC-LR can establish an interaction with TRP-393 in 1ERE. Similarly, these three amino acids interact with the 3ERD receptor in the same way. The interaction pattern between MC-LR and 1ERR resembled that of 1ERE and 3ERD. However, in this case, an additional interaction emerged. Specifically, the terminal nitrogen of leucine within MC-LR formed a binding interaction with ASN-435 of 1ERR. When MC-LR was subjected to docking with 3ERT, the resulting interactions were as follows: the nitrogen atom within the residual chain of arginine in MCLR established bonds with ASP-321 and MET-522, while the hydroxyl group of the methyl aspartic acid of MC-LR formed a binding interaction with VAL-533. CYN shows a propensity to form stable complexes with receptors in their agonist conformations (Figure S4). Specifically, the binding energies of CYN with receptors 1ERE and 3ERD are −6.7 kcal/mol and −7.0 kcal/mol, respectively. These relatively low binding energies suggest that CYN interacts strongly with these agonist conformations, indicating a preference for stabilizing the receptor in its active state. In comparison, CYN interactions with the antagonist conformations of the estrogen receptor (1ERR and 3ERT) are slightly weaker, with binding energies of −6.6 kcal/mol for 1ERR and −6.1 kcal/mol for 3ERT. While CYN still forms stable complexes with the antagonist conformations, slightly higher binding energies indicate a reduced affinity compared to the agonist conformations. When the 1ERE is docked with CYN two significant interactions are involved, the guanidine group of CYN could bind to the following amino acids: SER-305, ALA-307, GLY-366, and ASP-369, while the sulfate group could bind to ARG-363. Similarly, the interaction between CYN and 3ERD is mediated by binding between the guanidine and sulfate groups with ARG-363. Furthermore, this complex could bind to LYS 362. The docking results between CYN and the 1ERR receptor show possible interactions between the guanidine group with SER 512 and the sulfate group with HIS 526. On the contrary, when using the 3ERT receptor, the findings indicate interactions involving the guanidine group with HIS-547, the sulfate group with GLU-542, and the hydroxyl group with LYS-362. The molecular docking procedure, with crystal structures of MC-LR and CYN in complex with the androgen receptor AR, has been executed, revealing potential bindings. All molecular models are represented in supplementary figures. When MC-LR was docked with AR, the free binding energy was −7.2 kcal/mol. MC-LR has only one amino acid in their cyclic chemical structure. The leucine chain nitrogen of MCLR can interact with PRO-682. CYN (Figure S5B) shows a higher propensity to form stable complexes with AR, establishing a complex with a free binding energy of −8.2 kcal/mol. The interaction between CYN and AR involves six significant interactions. In the docking of AR with CYN, the guanidine group of CYN binds to GLN-798, while the sulfate group binds with LYS-847 and ARG-846. Similarly, the CYN uracil group binds to ARG-855, GLN-858, and LEU-797. 4. Discussion The potential ED activity of the cyanotoxins MC-LR and CYN has gained interest in the last years, although there are still few studies, especially in the case of CYN. Recently, a review focused on this topic revealed that the results obtained from these studies are contradictory, although, in general, both toxins showed ED activity mediated by very different mechanisms (Casas-Rodríguez et al., 2022). However, the potential ED effects of both toxins have not been evaluated by international standardized test guidelines, published by the OECD for the detection of endocrine disrupting chemicals. Stable transfected transactivation assays are a valuable tool for assessing the potential of chemicals to act as EDs in a controlled laboratory setting (Grimaldi et al., 2015). These assays provide information on how chemicals might interact with hormone receptors within cells, potentially influencing the activation or suppression of hormone-regulated genes (Grimaldi et al., 2015; Kenda et al., 2020). It is important to note that these assays have limitations. They may not identify EDs that affect other components of the endocrine system, such as enzymes involved in hormone processing, or processes such as hormone production, breakdown, movement within the body, and elimination (OECD 455, 2021; OECD 458, 2023). Among the in vitro assays at level two of the conceptual framework of the OECD, the stable transfected transactivation assay to detect (anti)agonists of ER α in the Fig. 3. Androgen receptor agonist (A) and antagonist (B) effect of MC-LR and CYN. Data are represented as mean ±SD of at least two independent repeats. DHT: dihydrotestosterone; SC: spike in control; RTA: Relative Transcriptional Activation. A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 6 hER α -HeLa-9903 cell line, derived from human cervical cancer cells (OECD 455, 2021), has been applied to evaluate various potential EDs: persistent organic pollutants and parabens (Kim et al., 2011), preservatives (Kenda et al., 2020), flavonoids (Skledar et al., 2020), diverse alternatives of bis (2-ethylhexyl)phthalate (Park et al., 2019), bisphenol A and S and analogs (Durcik et al., 2022), azole pesticide products (Jung et al., 2023) and even experimental mixtures of ED chemicals (Durcik et al., 2023). Furthermore, the androgenic and anti-androgenic activity assessment using AR-EcoScreen cells has also been applied to chemicals in household applicants (Lee et al., 2018), various compounds (Kenda et al., 2020) or veterinary drugs (Park et al., 2021). In the present work, the potential agonistic and antagonistic effects of MC-LR and CYN on human ER α and AR were investigated. For this purpose, two OECD performance-based test guidelines: No. 455 (2021) STTA and No. 458 (2023) ARTA were applied. The results obtained provide the first evidence that pure MC-LR exhibits weak ER α agonistic activity, while CYN did not show an agonistic response at the concentrations tested (25 pM – 2.5 µM). Furthermore, both toxins displayed ER α antagonistic effects, confirmed by reverse sigmoidal concentrationresponse curves in the ER α STTA without intrinsic cytotoxicity to these cells. Several studies have examined the potential of cyanotoxins to act as EDs, specifically targeting ER and AR, but these studies have not followed standardized OECD guidelines. One such study used a genetically modified cell line containing an estrogen-activated luciferase gene (Oziol and Bouaïcha, 2010). The results showed that low concentrations of MC-LR triggered luciferase activity, suggesting a weak estrogenic effect. This effect might be due to an indirect interaction with the ER. However, when a specific ER antagonist was introduced, luciferase activity decreased. This suggests that the weak estrogenic effect might be reversible. At higher MC-LR concentrations, the study reported a decrease in luciferase activity, probably caused by the cytotoxic properties of the toxin. The authors proposed that estrogenic activity of MC-LR could involve signaling pathways similar to those of okadaic acid. Furthermore, they suggest that oxidative stress caused by reactive oxygen species could play a role in modulating signaling proteins. In addition, some phytoplankton species can produce compounds with estrogenic and retinoid-like activity (Jonas et al., 2015). However, it is important to note that these endocrine effects may not be directly related to the cyanotoxins themselves. Another study by Mallia et al. (2020) investigated the potential endocrine activity of entire cyanobacterial cultures, including various species of Microcystis and Planktothrix, rather than isolated cyanotoxins. They used reporter gene assays to detect estrogenic, androgenic and glucocorticoid activity, but the results were not conclusive. Our findings on MC-LR’s estrogenic effects on ER α align with recent research on SGC-7901 cells (Wang et al., 2023). Their study suggests that MC-LR might possess estrogenic activity. The authors proposed that MC-LR could achieve this by reducing a specific cellular process (Hsp90 phosphorylation) and thus causing an overactivation of the estrogen signaling pathway. This leads to the movement of activated ER α into the cell nucleus and to an increase in the production of a protein called Krt16 in stomach cells. This innovative research provides valuable information on the molecular mechanisms underlying the estrogenic potential of MC-LR. For pure CYN, in vitro studies are practically inexistent and, to our knowledge, no works have been published on its potential effects on the ER or AR receptor according to the OECD guidelines. The potential estrogenic activity of CYN was investigated by a yeast estrogen screening (YES) assay (Liu et al., 2018), and the toxin was the agonist in this test, and its binding affinity to the estrogen receptor was related to its intrinsic properties. In contrast, in the present work, CYN did not show agonistic estrogenic activity, only antagonistic estrogenic effects. However, in in silico studies, firstly to confirm the reproducibility of the ER docking analysis, the ER agonist and antagonist activities were compared with previously published data using the same validated methodology (Cascajosa-Lira et al., 2023). Natural ligands for the receptors 3ERD and 1ERE are Diethylstilbestrol (DES) and Estradiol (E2), with binding energies of −6.6 kcal/mol for DES and −10.5 kcal/mol for E2, respectively. The binding energies of the toxins MC-LR and CYN, at −8.2 kcal/mol and −7.0 kcal/mol respectively, fall between the binding energies of DES and E2. These values indicate that MC-LR and CYN have sufficiently low binding energies to form stable interactions with the receptor, suggesting that these toxins can bind effectively and potentially compete with natural ligands for receptor occupancy (Cascajosa-Lira et al., 2023). In an earlier study, Cascajosa-Lira et al. (2023) used an identical molecular model of receptors featuring antagonistic conformations to predict the binding energy of their native ligands. Specifically, they reported a binding energy of −8.1 for raloxifene with the 1ERR receptor and −9.1 for 4-hydroxytamoxifen with the 3ERT receptor. Despite the minor stability of the binding energies associated with MC-LR and CYN, they remain notably low and establish interactions with various amino acids within the receptor structure. Consequently, both toxins could potentially exert antagonistic effects on this receptor. Previous researchers who used in silico models to assess estrogen receptor affinity have categorized the energy range as indicative of an intermediate probability of binding to ER in the context of MC-LR (Skledar et al., 2020). Globally, with respect to the ER, the molecular docking analysis revealed the key functional groups of each cyanotoxin with the ability to interact with the 4 conformations of the ER. In the case of MC-LR the amino acids arginine, methyl aspartic acid and glutamic acid could be involved, whiles the characteristic amino acid of MCs, Adda (3-amino,9-methoxy,10-phenyl,2,6,3-trimethyl-deca-4(E),6(E)-dienoic acid), seems to play an unimportant role in possible interactions with this receptor. In the case of CYN, the groups potentially involved are the guanidine ring nitrogens and the sulphate group. While molecular docking can provide valuable insights into ligandreceptor interactions, it is essential to validate these predictions with experimental data. As reported by the OECD in the Conceptual Framework for endocrine disruption (2018), a combination of computational approaches (level 1) and in vitro experiments (level 2) can provide a more comprehensive understanding of the biological effects of compounds such as MC-LR and CYN on ER α and AR. In that sense, all the data obtained by in silico studies agree with the in vitro results obtained with ER but not with those obtained with AR. Molecular docking showed the ability of both toxins to form stable complex with the different conformations of ER and the in vitro studies showed that MC-LR has agonist activity and both toxins could have antagonistic activities on the ER. Nevertheless, molecular docking predicted the ability of both toxins to form complex with the AR, while in vitro no interactions were found. There are also some discrepancies due to computational limitations, as molecular docking relies on assumptions about protein structure and ligand flexibility that may not always accurately capture the complexity of biological systems, and experimental variability in in vitro experiments can be influenced by factors such as cell type, culture conditions and assay sensitivity. Taking all into account, these findings remark the necessity of complete the in silico studies with experimental procedures that can confirm or not the predictions obtained in docking assays. However, some studies suggest that the chemical structure of MCs makes it unlikely that they directly bind to the ER like other EDs (Hou et al., 2018). These researchers indicated that the estrogenic effects of MC-LR might be caused by stimulating steroidogenesis, a process in which the body produces estrogen, rather than directly interacting with ER. Our current research using molecular docking simulations shows that MC-LR can bind to the ER with a relatively stable interaction energy. This finding seems to contradict the idea that MC-LR does not bind to ER. It is important to point out that in vitro studies, while valuable, may not always predict the full impact of chemicals on living organisms (Kenda et al., 2020). This is because in vitro assays often lack the complex breakdown, absorption, and excretion processes (pharmacokinetics) that occur in the body. For example, Grenet et al. (2019) found no clear connection between in vitro endocrine disruption assays and A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 7 long-term effects in rats. However, other studies using cell lines have shown some promise in predicting in vivo outcomes. For instance, Henneberg et al. (2014) observed consistent endocrine disruption effects in both cell lines and living fish exposed to river samples. Recently, the potential estrogenic effects of MC-LR and CYN in vivo using rats, following established OECD guidelines, have been investigated (Casas-Rodriguez et al., 2023). The results did not show changes in the weight of the uterus (wet or blotted) or its microscopic structure. Interestingly, among the measured serum hormones, progesterone levels increased in a dose-dependent manner in rats exposed to MC-LR. These in vivo findings suggest that MC-LR and CYN may not act as estrogens under the conditions tested in the ovariectomized rat uterotrophic assay. This contrasts with the estrogenic activity observed in our current in vitro experiments. The differences observed between our in vitro and in vivo findings are not uncommon. There are several factors that can contribute to these discrepancies (Kenda et al., 2020). One factor is the use of different test organisms. In vitro studies often use isolated cells or human cell lines that express specific receptors, while in vivo studies involve whole living organisms such as rats. Moreover, other factors that can contribute to the different results obtained between in vitro and in vivo experiments are the different biological complexity of the models, the exposure routes, the metabolism and elimination (in vitro limited ability to simulate metabolic processes vs in vivo liver enzymes and detoxification pathways that can alter the bioavailability and toxicity of the compounds) and the individual variability (in vitro standardized conditions and cell lines vs in vivo differences in genetic makeup, health status and other factors that can influence susceptibility). These differences justify the need to perform assays at different levels, as the OECD recommends in its conceptual framework for testing and evaluating EDs, with five different levels (OECD 150, 2018). In this sense, in order to clarify the potential androgenic/antiandrogenic effects of these cyanotoxins, further in vivo studies could be performed using the OECD Hershberger assay (OECD 441, 2009). Finally, recapping all the information exposed, the estrogenic and androgenic effects of MC-LR and CYN may vary depending on the target tissue and the specific hormonal environment. These toxins might also indirectly influence endocrine function by affecting steroid hormone synthesis, metabolism, or transport. In that sense, as reviewed by Casas-Rodríguez et al., (2022), some studies have evidenced endocrine disruption caused by MCs, resulting in damage in related cells such as, ovarian and leydig cells or GnRH neurons. Moreover, additional effects/mechanisms implicated are changes in hormone levels and steroidogenesis-related genes; interference with steroidogenic enzymes; transcriptional responses of HPG-, HPIand HPT-axis related genes; activation of the ERK1/2 signalling pathway or changes in the activity of GnRH transcription factors. On the other hand, although studies dealing with CYN are very limited, and the results obtained were also variable, it showed ED effects mediated by changes in the transcript levels of related hormones, apoptosis induction, changes in thyroid-hormone-related genes, oxidative stress, or alterations in hormone levels. It’s important to note that while these molecular mechanisms have been proposed, the full extent of the endocrine-disrupting effects of MC-LR and CYN remains to be explored. Further research is needed to understand the long-term consequences of exposure to these toxins and to develop strategies for prevention and mitigation. 5. Conclusions In this study, the ER α and AR agonistic/antagonistic effects by two pure cyanotoxins, MC-LR and CYN, were investigated for the first time. The results of both experiments, in silico predictions and in vitro assays, indicated that MC-LR has ER α -mediated endocrine-disrupting potential acting as both, agonist and antagonist. In the case of CYN, predictions suggest that the toxin can act as an agonist or antagonist, but in vitro experiments only reflect the capacity of CYN to act as ER α antagonist. Moreover, despite the in silico predictions showing that both toxins could have the potential to form complexes with AR, no agonist or antagonist effects were found in the in vitro model. These results suggested that these toxins could have adverse effects on the endocrine system mediated by interactions with the ER α . However, further studies should be conducted to determine the (anti)androgenic effects of these cyanotoxins. Funding This work was supported by project PID 2019-104890RB-I00 funded by MICIU/AEI/10.13039/501100011033, project PID2023-147444OBI00 funded by MICIU/AEI /10.13039/501100011033 and FEDER, UE. The FPI grant number PRE2020-094412 awarded to Antonio CasasRodríguez was funded by Spanish Ministerio de Ciencia e Innovaci´ on. The FPU grant number FPU2019-01247 awarded to Antonio CascajosaLira was funded by the Spanish Ministerio de Universidades. CRediT authorship contribution statement Antonio Casas-Rodríguez: Writing – review & editing, Writing – original draft, Visualization, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Antonio CascajosaLira: Writing – original draft, Software, Formal analysis, Data curation, Conceptualization. María Puerto: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Ana María Came´ an: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Funding acquisition. Angeles Jos: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Declaration of Competing Interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Angeles Jos reports financial support was provided by SPANISH MINISTERIO DE CIENCIA E INNOVACI´ ON. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments The authors would acknowledge to the MICIU/AEI/10.13039/ 501100011033, project number PID 2019-104890RB-I00, the MICIU/ AEI/10.13039/501100011033 and FEDER, project number PID2023147444OB-I00. A.C.-R. acknowledges the Spanish Ministerio de Ciencia e Innovaci´ on for the predoctoral grant awarded (PRE-2020-094412). A.C.-L. thanks the Spanish Ministerio de Universidades for the funding FPU grant (FPU2019-01247). The authors would like to acknowledge the Grants of the VII Plan Propio de Investigaci´ on of the University of Sevilla for the use of General Research Services (I.3 - Anualidad 2024). The authors also thank the biology service of CITIUS (University of Seville) for the technical assistance offered. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ecoenv.2024.117456. Data Availability Data will be made available on request. A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 8 References Bouai¨cha, N., Miles, C., Beach, D., Labidi, Z., Djabri, A., Benayache, N., NguyenQuang, T., 2019. Structural diversity, characterization, and toxicology of microcystins. Toxins 11, 714. https://doi.org/10.3390/toxins11120714. Buratti, F.M., Manganelli, M., Vichi, S., Stefanelli, M., Scardala, S., Testai, E., Funari, E., 2017. Cyanotoxins: producing organisms, occurrence, toxicity, mechanism of action and human health toxicological risk evaluation. Arch. Toxicol. 91, 1049–1130. https://doi.org/10.1007/s00204-016-1913-6. Casas-Rodriguez, A., Moyano, R., Molina-Hern´ andez, V., Came´ an, A.M., Jos, A., 2023. Potential oestrogenic effects (following the OECD test guideline 440) and thyroid dysfunction induced by pure cyanotoxins (micrtocystin-LR, cylindropermopsin) in rats. Environ. Res. 226, 115671. https://doi.org/10.1016/j.envres.2023.115671. Casas-Rodríguez, A., Came´ an, A.M., Jos, A., 2022. Potential endrocrine disruption of cyanobacterial toxins, microcystins and cylindrospermopsin. A Rev. Toxins 14, 882. https://doi.org/10.3390/toxins14120882. Cascajosa-Lira, A., Medrano-Padial, C., Pichardo, S., de la Torre, J.M., Ba˜ nos, A., Jos, ´ A., Came´ an, A.M., 2023. Identification of in vitro metabolites of an Allium organosulfur compound and environmental toxicity prediction as part of its risk assessment. Environ. Res. 229, 116001. https://doi.org/10.1016/j.envres.2023.116001. Chen, J., Liu, J., Han, S., et al., 2023. Nontraditional biomanipulation: a powerful ecotechnology to combat cyanobacterial blooms in eutrophic freshwaters. Innov. Life 1 (3), 100038. https://doi.org/10.59717/j.xinn-life.2023.100038. Chen, L., Shi, T., Wang, Y., He, J., Zhao, X., Wang, Y., Giesy, J., Chen, F., Chen, Y., Tuo, X., et al., 2021. Effects of acute exposure to microcystins on hypothalamicpituitary-adrenal (HPA), -gonad (HPG) and -thyroid (HPT) axes of female rats. Sci. Total Environ. 778, 145196. https://doi.org/10.1016/j.scitotenv.2021.145196. Durcik, M., Grobin, A., Roskar, R., Trontelij, J., Masic, L.P., 2023. Estrogenic potency of endrocrine disrupting chemicals and their mixtures detected in environmental waters and wastewaters. Chemosphere 330, 138712. https://doi.org/10.1016/j. chemosphere.2023.138712. Durcik, M., Hiti, L., Tomasic, T., Masic, L.P., 2022. New bisphenol A and bisphenol S analogs: evaluation of their hER α agonistic and antagonistic activities using the OECD 455 in-vitro assay and molecular modeling. Chem. Biol. Interact. 354, 109820. https://doi.org/10.1016/j.cbi.2022.109820. European Chemical Agency (ECHA) and European Food Safety Authority (EFSA) with the technical support of the Joint Research Centre (JRC, Andersson, N., Arena, M., Auteri, D., Barmaz, S., Grignard, E., Kienzler, A., Lepper, P., Lostia, A.M., Munn, S., et al., 2018. Guidance for the identification of endocrine disruptors in the context of Regulations (EU) No 528/2012 and (EC) No 1107/2009. EFSA J. 16, e05311. https://doi.org/10.2903/j.efsa.2018.5311. Froscio, S.M., Humpage, A.R., Burcham, P.C., Falconer, I.R., 2003. Cylindrospermopsininduced protein synthesis inhibition and its dissociation from acute toxicity in mouse hepatocytes. Environ. Toxicol. 18, 243–251. https://doi.org/10.1002/tox.10121. Gonçalves da Silva, C., Dutra Duque, M., Freire Nordi, C.S., Viana-Niero, C., 2021. New insights into toxicity of microcystins produced by cianobacteria using in silico ADMET prediction. Toxicon 204, 64–71. https://doi.org/10.1016/j. toxicon.2021.11.002. Grenet, I., Comet, J.P., Schorsch, F., Ryan, N., Wichard, J., Rouqui´ e, D., 2019. Chemical in vitro bioactivity profiles are not informative about the long-term in vivo endocrine mediated toxicity. Comput. Toxicol. 12, 100098. https://doi.org/10.1016/j. comtox.2019.100098. Grimaldi, M., Boulahtouf, A., Delfosse, V., Thouennon, E., Bourguet, W., Balaguer, P., 2015. Reporter cell lines for the characterization of the interactions between human nuclear receptors and endocrine disruptors. PMID: 26029163. Front. Endocrinol. 6, 1–12. https://doi.org/10.3389/fendo.2015.00062. Henneberg, A., Bender, K., Blaha, L., Giebner, S., Kuch, B., K¨ ohler, H.R., Maier, D., Oehlmann, J., Richter, D., Scheurer, M., Schulte-Oehlmann, U., Sieratowicz, A., Ziebart, S., Triebskorn, R., 2014. Are in vitro methods for the detection of endocrine potentials in the aquatic environment predictive for in vivo effects? Outcomes of the Projects SchussenAktiv and SchussenAktivplus in the Lake Constance Area, Germany, 9 PLoS One 5 (6), e98307. https://doi.org/10.1371/journal. pone.0098307. Hou, J., Su, Y., Lin, W., Guo, H., Li, L., Anderson, D., Li, D., Tang, R., Chi, W., Zhang, X., 2018. Estrogenic potency of MC-LR is induced via stimulating steroidogenesis: In vitro and in vivo evidence. Environ. Pollut. 240, 615–622. https://doi.org/10.1016/j. envpol.2018.04.140. Huisman, J., Codd, G.A., Paerl, H.W., Ibelings, B.W., Verspagen, J.M.H., Visser, P.M., 2018. Cyanobacterial blooms. Nat. Rev. Microbiol. 16 (8), 471–483. https://doi.org/ 10.1038/s41579-018-0040-1. Jaeger-Honz, S., Nitschke, K., Altaner, S., Klein, K., Dietrich, D.R., Schreiber, F., 2022. Investigation of microcystin conformation and binding towards PPP1 by molecular dynamics simulation. Chem. Biol. Interact. 351, 109766. https://doi.org/10.1016/j. cbi.2021.109766. Jonas, A., Scholz, S., Fetter, E., Sychrova, E., Novakova, K., Ortmann, J., 2015. Endocrine, teratogenic and neurotoxic effects of cyanobacteria detected by cellular in vitro and zebrafish embryos assays. Chemosphere 120, 321–327. https://doi.org/ 10.1016/j.chemosphere.2014.07.074. Jung, D.-W., Jeong, D.-H., Lee, H.-S., 2023. Azole pesticide products and their hepatic metabolites cause endocrine disrupting potential by suppressing the homodimerization of human estrogen receptor alpha. Environ. Pollut. 318, 120894. https://doi.org/10.1016/j.envpol.2022.120894. Kenda, M., Kuzelicki, N.K., Iida, M., Kojima, H., Dolenc, M.S., Triclocarban, Triclosan, 2020. Bromochlorophene, Chlorophene, and Climbazole effects on nuclear receptors: an in Silico and in vitro study. Environ. Health Perspect. 128, 107005. Kim, T.S., Kim, C., Lee, H.K., Kang, I.H., Kim, M.G., Jung, K.K., Kwon, Y.K., Nam, H.-S., Hong, S.K., Kim, H.S., Yoon, H.J., Rhee, G.S., 2011. Estrogenic activity of persistent organic pollutants and parabens based on the stably transfected human estrogen receptorα transcriptional activation assay (OECD TG 455). Toxicol. Res. 27, 181–184. https://doi.org/10.5487/TR.2011.27.3.181. Lee, H.S., Park, E.J., Han, S., Oh, G.Y., Kang, H.S., Suh, J.H., Shin, M.K., Oh, H.S., Hwang, M.S., Moon, G., Koh, Y.H., Park, Y., Hong, J.H., Koo, Y.E., 2018. Assessment of androgen receptor agonistic/antagonistic effects on 25 chemicals in household applicants by OECD in vitro stably transfected transcriptional activation assays. Chemosphere 191, 589–596. https://doi.org/10.1016/j.chemosphere.2017.10.084. Liu, J., Hern´ andez, S.E., Swift, S., Singhal, N., 2018. Estrogenic activity of cylindrospermopsin and anatoxin-a and their oxidative products by Fe III -B*/H 2 O 2 . Water Res. 132, 309–319. https://doi.org/10.1016/j.watres.2018.01.018. Mallia, V., Ivanova, L., Eriksen, G., Harper, E., Connolly, L., Uhlig, S., 2020. Investigation of in vitro endocrine activities of Microcystis and Planktothrix cyanobacterial strains. Toxins 12, 228. https://doi.org/10.3390/toxins12040228. OECD, 2009. Test No. 441: Hershberger Bioassay in Rats: A Short-term Screening Assay for (Anti)Androgenic Properties. OECD Guidelines for the Testing of Chemicals, Section 4. OECD Publishing, Paris. https://doi.org/10.1787/9789264076334-en. OECD, 2018. Revised Guidance Document 150 on Standardised Test Guidelines for Evaluating Chemicals for Endocrine Disruption. OECD Series on Testing and Assessment. OECD Publishing, Paris. https://doi.org/10.1787/9789264304741-en. OECD, 2021. Test No. 455: Performance-Based Test Guideline for Stably Transfected Transactivation In Vitro Assays to Detect Estrogen Receptor Agonists and Antagonists. OECD Guidelines for the Testing of Chemicals, Section 4. OECD Publishing, Paris. https://doi.org/10.1787/9789264265295-en. OECD, 2023. Test No. 458: Stably Transfected Human Androgen Receptor Transcriptional Activation Assay for Detection of Androgenic Agonist and Antagonist Activity of Chemicals. OECD Guidelines for the Testing of Chemicals, Section 4. OECD Publishing, Paris. https://doi.org/10.1787/9789264264366-en. Oziol, L., Bouaïcha, N., 2010. First evidence of estrogenic potential of the cyanobacterial hepatotoxins the nodularin-R and the microcystin-LR in cultured mammalian cells. J. Hazard. Mater. 174 (1–3), 610–615. https://doi.org/10.1016/j. jhazmat.2009.09.095. Park, J., Parck, C., Gye, M.C., Lee, Y., 2019. Assessment of endocrine-disrupting activities of alternative chemicals for bis (2ethylhexyl)phthalate. Environ. Res. 172, 10–17. https://doi.org/10.1016/j.envres.2019.02.001. Park, Y., Park, J., Lee, H.S., 2021. Endocrine disrupting potential of veterinary drugs by in vitro stably transfected human androgen receptor transcriptional activation assays. Environ. Pollut. 286, 117201. https://doi.org/10.1016/j.envpol.2021.117201. Pichardo, S., Camean, A.M., Jos, A., 2017. In vitro toxicological assessment of Cylindrospermopsin: a review. Toxins 9, 402. https://doi.org/10.3390/ toxins9120402. Puerto, M., Gutierrez-Praena, D., Prieto, A.I., Pichardo, S., Jos, A., Miguel-Carrasco, J.L., Vazquez, C.M., Camean, A.M., 2011. Sub-chronic effects of cyanobacterial cells on the transcription of antioxidant enzyme genes in tilapia (Oreochromis niloticus). Ecotoxicology 20, 479–490. https://doi.org/10.1007/s10646-011-0600-x. Puerto, M., Prieto, A.I., Maisanaba, S., Guti´ errez-Praena, D., Mellado-García, P., Jos, ´ A., Came´ an, A.M., 2018. Mutagenic and genotoxic potential of pure Cylindrospermopsin by a battery of in vitro tests. Food Chem. Toxicol. 121, 413–422. https://doi.org/ 10.1016/j.fct.2018.09.013. Ricciardelli, A., Pollio, A., Costantini, M., Zupo, V., 2023. Harmful and beneficial properties of cyanotoxins: Two sides of the same coin. Biotechnol. Adv. 68, 108235. https://doi.org/10.1016/j.biotechadv.2023.108235. Rogers, E.D., Henry, T.B., Twiner, M.J., Gouffon, J.S., McPherson, J.T., Boyer, G.L., Sayler, G.S., Wilhelm, S.W., 2011. Global gene expression profiling in larval zebrafish exposed to microcystin-LR and microcystis reveals endocrine disrupting effects of Cyanobacteria. Environ. Sci. Technol. 45 (5), 1962–1969. https://doi.org/ 10.1021/es103538b. Runnegar, M.T., Kong, S.M., Zhong, Y.Z., Lu, S.C., 1995. Inhibition of reduced glutathione synthesis by cyanobacterial alkaloid cylindrospermopsin in cultured rat hepatocytes. Biochem. Pharmacol. 49, 219–225. https://doi.org/10.1016/s00062952(94)00466-8. Skledar, D.G., Tvrdy, V., Kenda, M., Zega, A., Por, M., Horky, P., Mladenka, P., Dolenc, M.S., Masic, L.P., 2020. Applicability of the OECD 455 in vitro assay for determination of hER α agonistic activity of isoflavonoids. Toxicol. Appl. Pharmacol. 386, 114831. https://doi.org/10.1016/j.taap.2019.114831. Testai, E., Buratti, F.M., Funari, E., Manganelli, M., Vichi, S., Arnich, A., Bire, R., Fessard, V., Sialehaamoa, A., 2016. Review and analysis of occurrence, exposure and toxicity of cyanobacteria toxins in food. EFSA Support. Publ. 13, 998E. https://doi. org/10.2903/sp.efsa.2016.EN-998. Wang, X., Ding, J., Xiang, Z., Jiang, P., Du, J., Han, X., 2016. Microcystin-LR causes sexual hormone disturbance in male rat by targeting gonadotropin-releasing hormone neurons. Toxicon 123, 45–55. https://doi.org/10.1016/j. toxicon.2016.10.011. Wang, X., Ying, F., Chen, Y., Han, X., 2012. Microcystin (-LR) affects hormones level of male mice by damaging hypothalamic-pituitary system. Toxicon 59 (2), 205–214. https://doi.org/10.1016/j.toxicon.2011.12.001. WHO, (2020b). Cyanobacterial toxins: cylindrospermopsins. Background document for development of WHO Guidelines for drinking-water quality and Guidelines for safe recreational water environments. WHO/HEP/ECH/WSH/2020.4.. WHO (2020a). Cyanobacterial toxins: microcystins. Background document for development of WHO Guidelines for drinking-water quality and Guidelines for safe recreational water environments WHO/HEP/ECH/WSH/2020.6. A. Casas-Rodríguez et al. Ecotoxicology and Environmental Safety 289 (2025) 117456 9