Potential Endocrine Disruption of Cyanobacterial Toxins, Microcystins and Cylindrospermopsin: A Review

Microcystins (MCs) and cylindrospermopsin (CYN), although classified as hepatotoxins and cytotoxins, respectively, have been shown to also induce toxic effects in many other systems and organs. Among them, their potential endocrine disruption (ED) activity has been scarcely investigated. Considering the increasing relevance of ED on humans, mammals, and aquatic organisms, this work aimed to review the state-of-the-art regarding the toxic effects of MCs and CYN at this level. It has been evidenced that MCs have been more extensively investigated than CYN. Reported results are contradictory, with the presence or absence of effects, but experimental conditions also vary to a great extent. In general, both toxins have shown ED activity mediated by very different mechanisms, such as estrogenic responses via a binding estrogen receptor (ER), pathological changes in several organs and cells (testis, ovarian cells), and a decreased gonad-somatic index. Moreover, toxic effects mediated by reactive oxygen species (ROS), changes in transcriptional responses on several endocrine axes and steroidogenesis-related genes, and changes in hormone levels have also been reported. Further research is required in a risk assessment frame because official protocols for assessment of endocrine disrupters have not been used. Moreover, the use of advanced techniques would aid in deciphering cyanotoxins dose-response relationships in relation to their ED potential.


Introduction
The endocrine system is a collection of numerous glands that directly secrete different hormones into blood circulation and includes different axes, such as hypothalamicpituitary-adrenal/interrenal (for fish) (HPA/HPI), hypothalamic-pituitary-gonad (HPG), and hypothalamic-pituitary-thyroid (HPT) [1]. Each axis regulates different functions of the body, including metabolism, growth, and development, via axis-specific hormones. The HPA axis mainly controls bodily responses to stress by glucocorticoids, cortisol, or corticosterone, while the HPG axis coordinates reproduction by steroid hormones, synthesized from cholesterol in a process called steroidogenesis, and the HPT axis regulates energy metabolism and development by thyroid hormones [2,3].
This system can be altered by different substances called endocrine disruptors (EDs) [4]. In the literature, other authors have defined ED as an exogenous substance that interferes with the function of hormonal systems and produces a range of developmental, reproductive, neurological, immune, or metabolic diseases in humans and wildlife [5]. According to the World Health Organization (WHO), an ED is "an exogenous substance or mixture that alters function(s) of the endocrine system and consequently causes adverse health effects in an intact organism, or its progeny, or (sub)population" [6]. Moreover, for the European increased body length of embryos was also observed after exposure to biomass extracts. All of these findings highlight the importance of investigating the activity of estrogens, androgens, retinoids, and other bioactive compounds present in cyanobacterial blooms. In contrast, more recently, Mallia et al. [52] investigated the in vitro endocrine activities of 26 cyanobacterial cultures of Microcystis or Planktothrix species by screening them in estrogen-, androgen-, and glucocorticoid-responsive reporter gene assay (RGA), and the results were not conclusive.
In zebrafish, one of the first studies suggested that VTG abundance increased in MC-LR-exposed zebrafish brains, and this was explained because the toxin might mimic the effects of endocrine-disrupting chemicals (EDCs) [56]. In zebrafish larvae exposed to purified MC-LR or lyophilized M. aeruginosa cells containing MC-LR, an upregulation of VTG genes was observed in the case of exposed Microcystis larvae, although not in the group exposed to MC-LR [19]. VTG are a group of lipoproteins produced in the liver in response to estrogens, and are transported through the blood and deposited in the developing oocytes of female fish [69]. The fact that only Microcystis cells showed induction of VTG could be explained by the presence of substances produced by these cells, such as phytoestrogens.
Because of the endocrine disorders induced by MC-LR, the toxin exhibits reproductive toxicity, as has been reviewed [33][34][35], and zebrafish have been chosen as experimental models to understand the effects of MCs on fish. Male and female zebrafish exposed by subchronic immersion (30 days (d)) to MC-LR suffered from toxic effects in terms of gonads, hatchability, and hormone levels; whole body VTG levels significantly increased in females, while they decreased in males [57]. Moreover, the VTG1 transcriptional level was significantly reduced in the livers of both sexes. Marked histological lesions were observed in the livers, ovaries, and testes of treated fish, and a significantly elevated apoptosis rate was observed in ovaries instead of testes. These results confirmed that female zebrafish were more vulnerable than males, and that MC-LR did not exert any estrogenic effects on adult zebrafish [57].
The sex-dependence effects of MC-LR exposure on fish reproduction was extensively investigated by Liu et al. [41], and some important molecular biomarkers related to gametogenesis in zebrafish were altered after MC-LR exposure (1-20 µg/L for 30 d). Thus, levels of E2, T and 11-keto testosterone , and FSH were increased in serum from all of the treated females, while T, FSH, and LH levels changed in all of the treated males. Histopathological changes were observed, with retarded oogenesis and spermatogenesis. Transcriptional changes in 22 genes on the HPG axis exhibited sex-specific responses.
Zhao et al. [58] investigated the effects of MC-LR on zebrafish reproduction by affecting oogenesis, as well as the effects on sex hormones and the transcription of genes on the HPG axis. Decreased egg production after exposure to MC-LR (>10 µg/L) and increased concentration of E2 and VTG at 10 µg/L MC-LR was reported, whereas at higher concentrations of 50 µg MC-LR/L, concentrations of E2, VTG, and T declined. Moreover, changes in the transcription of steroidogenic pathway genes, and in numerous intra-and extra-ovarian factors, which regulate oogenesis, were detected. Female zebrafish exposed to MC-LR experienced decreased egg production, fertilization, and hatching rates, which may reflect a defect in folliculogenesis and possible poor oocyte quality. Consequently, the changes in these endpoints could affect unexposed F1 or F2 generations, leading to potential multigenerational and transgenerational effects, respectively. Later, Hou et al. [44] studied the effects of MC-LR during life cycle exposure in zebrafish and hatchlings (5 d post fertilization), and demonstrated for the first time that MC-LR induced growth inhibition and decreased ovary weight and ovarian ultra-pathological lesions. Decreased ovarian T levels indicated that MC-LR disrupted sex steroid hormone balance. Moreover, the significantly upregulated transcription of brain FSHβ and LHβ along the ovarian ERα, FSHR, and LHR suggested that positive feedback regulation in the hypothalamic-pituitary-gonadal-liver (HPGL) axis was induced as a compensatory mechanism for MC-LR damage [44]. In contrast, ovarian VTG content and hepatic ERα and VGT1 expression were downregulated, which agreed with reduced vitellus storage found in the histopathological study. In summary, MC-LR impairs the development and reproduction of female zebrafish by disrupting the transcription of related HLPG-axis genes.
The same authors [59] evidenced that MC-LR caused gonadal development retardation through disrupting the growth hormone/insulin-like growth factors (GH/IGFs) system, which plays an important role in the endocrine regulation of fish growth. Thus, zebrafish hatchlings (5 d post fertilization) exposed to MC-LR for 90 days until sexual maturity showed delayed ovarian maturation and sperm development, as well as lesions in the brain and liver. The retarded gonadal development was parallel to an inhibition of the GH/IGFs system, which was characterized by significant decreases in some mRNA expression profiles of the genes in the brain, hepatic, and gonadal levels in males [59]. In this work, the gonadal development of males was more vulnerable than that of females after MC-LR exposure.
Nevertheless, Quiao et al. [57] reported that adult zebrafish females exhibited more susceptibility than males, and Liu et al. [41] showed that exposure to MC-LR could disrupt the function of the HPG endocrine system and indicated sex-dependent effects of MC-LR on fish reproduction. These discrepancies indicate that more studies are needed to investigate if the sex-dependent effects of MC-LR could depend on the tested life stage, or on the general conditions assayed.
Su et al. [60] also demonstrated that life cycle exposure to MC-LR caused endocrine disruption in male zebrafish, with growth inhibition and organic and functional damage of the testis. A significant decrease in the T/E2 ratio, a sensitive biomarker of abnormal sex hormone levels in fish [70], indicated that MC-LR disrupted sex steroid hormones balance. The changes in transcriptional responses of HPG-axis-related genes revealed that the toxin promoted the conversion of T to E2 in circulating blood, and the upregulation of vgt1 mRNA expression in the liver indicated that MC-LR-induced estrogenic-like effects [60].
More recently, a study carried out on female zebrafish subchronically exposed to MC-LR (30 d) investigated the reproductive toxicity of the toxin by affecting oocyte development and fertilized eggs [64]. Pathological changes were observed in the ovaries after MC-LR exposure, and a significant increase in the rate of the zebrafish oocytes germinal vesicle breakdown (GVBD), malformations of the offspring, and decreased cAMP and VTG levels were detected at the highest concentration assayed. Moreover, the phosphorylation levels of the extracellular signal-regulated kinases (ERK) were elevated in the ovaries, as well as phosphorylated cyclin B levels. The authors suggested that MC-LR promotes oocyte maturation by activating the ERK1/2 (the most extensively studied members of the ERK family) and maturation-promoting factor (MPF) signaling pathways, and that cAMP is involved in this process.
In addition to the studies that demonstrated the endocrine-disrupting activity of MC-LR pure standard in zebrafish, the effects of M. aeruginosa on the HPGL axis in female zebrafish after short-term exposure (96 h) were investigated by Liu et al. [62]. Not only did the cyanobacteria cause histological lesions in the liver and gonads, but the fertilization rate and hatchability of eggs spawned in treated groups were also decreased, and transgenerational effects could appear. Moreover, M. aeruginosa decreased E2 and T plasma levels, and the vtg1 transcriptional level was decreased in the liver, whereas plasma VTG protein levels increased. The downregulation of the hepatic vtg1 gene could result from decreased E2 levels in the plasma, and from liver lesions caused by M. aeruginosa. In contrast, plasma VTG levels increased in a concentration-dependent manner, and this fact could be caused by the impairment of the ovaries, making it difficult for plasma VTG to incorporate into the oocytes. Similarly, Quiao et al. [57] found that MC-LR decreased hepatic vtg1 gene expression and increased VTG levels in zebrafish, as previously mentioned.
In the same model, synergism after combined exposure of MC-LR and nitrite causing reproductive dysfunction by interferences with the HPGL axis was also demonstrated in male zebrafish [61]. Both MC-LR and nitrite caused concentration-dependent effects, including testicular pathological changes, but the effects were more consistent in comparison to the single exposure groups. Exposure to MC-LR or nitrite alone significantly decreased T levels by the downregulation of gene expression in the HPLG axis, and interaction between them was significant. In contrast, E2 levels and the transcriptional levels of cyp19a1b, cyp19a1a, and vtg1 were increased with MC-LR concentrations, confirming the estrogen-like effects of MC-LR, as previously reported [60].
Moreover, the recovery mechanism of the reproductive function of adult zebrafish exposed to MC-LR exposure (0-50 µg/L for 21 d) and later transferred to MC free water for another 21 d was demonstrated by Kawan et al. [47]. In this work, after MC-LR exposure, several changes were reported: histopathological lesions in the gonads; a decreased percentage of mature oocytes and number of spawned eggs; decreased fertilization and hatching rates; increased concentrations of E2, T, and VTG in females; some gene transcriptions of the HPG axis were changed; and ovarian protein levels were increased. However, after 21 d of depuration (in water MCs free) the reproductive changes were also reversible.
In mammals, MC-LR induced female reproductive toxicity in mice after 28 d of exposure (i.p.), with reduced relative ovary weight and pathological changes in ovaries. MC-LR induced decreases in progesterone (P4), but no FSH or LH effects were reported. The alterations of the estrus cycle could be explained as a result of direct impact on the ovaries rather than indirect actions from the hypothalamus or pituitary. The toxin was detected in the ovaries of treated mice [66]. The same authors investigated the effects of MC-LR in mice after oral exposure (3 or 6 months) and the toxin stimulated follicle atresia, decreased developmental follicles, reduced the gonadosomatic index (GSI), and confirmed the changes in estrus cycles [67].
Finally, an extensive study was performed in rats, providing data from which to understand the endocrine-disrupting effects of MC-LR [3]. In this interesting and novel work, after a single i.p. injection of MC-LR (median lethal dose), the histopathology of several organs (hypothalamus, pituitary, adrenal, ovary, and thyroid) was analyzed, and concentrations of hormones in serum and gene expression of the HPA, HPG, and HPT axes were examined. The authors suggested that MC-LR affected the three axes. The changes in concentrations of hormones were not dose-dependent, and they hypothesized that the threshold for endocrine-disrupting effects of MC-LR might be less than 36.5 µg/kg b.w. In addition, they reported that non-monotonic dose responses (NMDRs) of hormones and genes were observed. NMDRs are relatively common in studies of endocrine-disrupting chemicals (EDCs), and it would be suitable to investigate endocrine disorders induced by smaller concentrations of MC-LR. Further studies are needed to confirm the effects of MC-LR and other congeners on the three axes.
In conclusion, the estrogenic effects of MC-LR have been demonstrated in different models (mainly in zebrafish), and several experiments have demonstrated that the toxin is able to disrupt the transcription of related HPGL axes through several mechanisms. In mammals, the toxin affected concentrations of hormones in the serum and gene expressions of the HPA, HPG and HPT axes, and the threshold level for endocrine-disrupting effects of MC-LR proposed in rats still needs to be confirmed. More studies are needed in several directions: (1) to elucidate if MC-LR effects are sex-dependent; (2) to investigate the main action of pure MC-LR in comparison to other substances present in cyanobacterial blooms (which may be phytoestrogens); (3) in vitro (level 2) and in vivo studies (level 3) following OECD guidelines with several pure MCs congeners (MC-LR, MC-RR, MC-YR) are needed in order to know if they are able to interact with ER, e.g., their effects in uterine weight or uterotrophic response in mammals have not been investigated yet; 4) further studies to explore the combination of cyanotoxins and other pollutants (metals, pesticides, plastics), some of them considered as EDs. 0.3, 1, 3 and 10 g dw/L for 120 hpf Teratogenic effects in zebrafish embryos exposed to extract of A. gracile at 1 g dw/L. Deformities of the tip of the tail and spine. Edema of the heart and trunk, small head and yolk retention, at high concentrations.
Estrogenic effects were not significant. Significant differences in locomotion compared to control for 0.3 g dw/L extracts of M. aeruginosa and P. agardhii. The assay revealed concentration-dependent retinoid-receptor-mediated activity in the biomass extracts of all tested cyanobacteria.
Estrogenic potency was detected in all samples. No androgenic, glucocorticoid, or antiandrogenic activities were detected in any of the tested biomass extracts. A potential estrogenic activity was indicated for the extracts of A. gracile, P. agardhii, and M. aeruginosa at 0.3 g dw/L.

MC-LR standard Male and female Zebrafish
Histological examination of ovaries. Determination of E2, T, and VTG plasma levels by ELISA. Expression of different genes (gnrh, lh, fsh) in brain, liver, and ovary by RT-PCR.
2, 10 and 50 µg/L for 21 d 10 µg/L MC-LR caused stronger effect on increasing sex hormones and related regulating genes and inducing maturation, whereas 50 µg/L MC-LR induced a greater impact on the inhibition of maturation and ovulation in fish. MC-LR strongly impaired follicular development by influencing intra-and extra-ovarian factors involved in these processes. [58] MC-LR standard Male and female Zebrafish Determination of E2, T and VTG levels by ELISA. Expression in brain, ovary, and liver of gnrh, lh, fsh, and vtg genes by RT-PCR.   Loss of contact between follicular cells and oocytes in ovary and cellular deterioration in testes. Decrease in both vitellogenic and post-vitellogenic oocytes and in the number of spermatozoa. In all fish serum levels of E2, T, 11-KT, and E2/11-KT ratio increased. E2/T ratio decreased in females.
Both serum levels of FSH and LH increased in females and decreased in males. In females, mRNA levels of cyp19b, erβ, and fshr increased, while gnrh2 expression decreased. In males, significant downregulation of gnrh2, erβ, and fshr, and significant up-regulation of 17βhsd. In both males and females, protein levels of 17βhsd and CYP19a increased.  Ultrastructural changes in the liver of male zebrafish were highly consistent with those in females. In gonads of males and females, ultrastructural changes were observed at 3 and 30 µg/L. In both sexes, MC-LR induced a marked concentration-dependent decrease in the expression of brain ghrh and pacap1. The gene expression in brain gh showed a similar downward tendency. In the liver, MC-LR increased ghrb expression but decreased the mRNA expression of igf2a and igf2b. In terms of the transcription of genes in the gonad, there were significant downregulations in the mRNA expression of ghra, igf3, and igf2r. Apparently, alterations in the transcription of related GH/IGFs axis genes were more significant in males than in females. [59]

MC-LR standard Male Zebrafish
Pathological analysis and morphometry of testis. Determination of E2 and T testis levels by ELISA. Expression of steroidogenic genes (CYP, StAR, 17βhsd) in brain, liver, and testis by RT-PCR. Determination of Erα levels in liver by Western blot.
0.3, 1, 3, 10 and 30 µg/L for 30 d Dose-dependent testicular damage, characterized by spermatogenesis suppression and the degeneration of interstitial and sustentacular cells. These pathological changes, as symptoms of male reproductive impairment, were consistent with extensive downregulation of brain genes involved in upstream regulation along the HPGL axis. Sex hormone levels and steroidogenesis gene (CYP, StAR, 17βhsd) expression in testis exhibited concurrent remarkable increases. The expression levels of testicular genes exhibited stronger positive correlation with E2 and T contents than those in the brain and liver, indicating that MC-LR induced steroidogenesis disruption by primarily affecting the synthesis and autocrine in the testis.
[48] opposite tendency with a significant concentration-dependent increase in MC-LR exposure groups, with a decrease in nitrite exposure groups. In co-exposure groups, E2 levels presented an increasing trend. Both MC-LR and nitrite significantly decreased T levels; the serum. In the brain: mRNA levels of gnrh2, gnrhr3, fshβ, and lhβ decreased in a concentration-dependent manner; CYP19a1b gene expression increased. In the testis: a significant downregulation of ar, fshr, and lhr, with a concentration-dependent upregulation in CYP19a1a. Significant interactive suppressions between MC-LR and nitrite on the transcriptional levels of testis ar and fshr. [61]

Microcystis aeruginosa Female Zebrafish
Histopathological analysis of brain and ovary tissue sections. Determination of E2, T, and VTG plasma levels. Expression in brain, gonads, and liver of gnrh, lh, and fsh genes by RT-PCR.
Both toxins with ER-antagonist activity caused significant estrogenic activity. At low concentrations (not cytotoxicity), the induction of the luciferase activity was significantly higher with MC-LR than with NOD-R, indicating that the dose-response effects of MC-LR appear more rapidly than NOD-R. Only MC-LR at high concentrations significantly decreased cell viability. Estrogenic potency of some extracts of cyanobacteria was demonstrated, being higher than extracts of single cyanobacterial species. There was no dioxin-like, glucocorticoid, or anti/retinoic activities for any of the extracts studied. Except for Aphanizomenon flos-aquae, exudates from all tested pure strains of cyanobacteria were estrogenic. The greatest induction of the luciferase reporter gene was caused by extracts of Microcystis aeruginosa.
The greatest tested concentrations of exudates from Aphanizomenon gracile induced a maximum response of luciferase expression, while the other exudates did not cause a maximum response. [49] MC-LR from a purified extract of The major estradiol oxidation products in the HLM assay were 2-hydroxyestradiol and estrone, with only traces of 4-hydroxyestradiol and estriol. The concentrations of 2-hydroxyestradiol following co-incubation of estradiol were significantly lower compared to when estradiol was incubated with HLM alone. In contrast, the concentrations of estrone were relatively higher. [52]

MC-LR standard
Female SD rats The reduction of GSI indicated a detrimental effect to the female reproductive system. The composition ratio of the number of primordial follicles in the high-dose group decreased. The duration of the proestrus and estrous stage decreased: mice exposed to MC-LR exhibited an abnormal estrous cycle. P4 level decreased without evident changes in FSH, LH, or E2. Occasional shrinkage of neurons and darkened staining of cells in hypothalamus. In the pituitary gland, parenchymal cells lost cytoplasm, with fragmentation and lysis of nuclei. In ovaries, rats exposed to 54.75 or 73 µg MC-LR/kg exhibited hyperaemia, cytoplasmic loss, abnormal nuclear change, and nuclear dissolution. Necrosis of local granular cells at 73 µg MC-LR/kg. Broken nuclei, necrosis of follicular epithelial cells, and reduced intracellular colloid in thyroid gland. Serum concentrations of CRH and ACTH decreased. Less concentrations of GnRH and E2, but greater concentrations of LH, FSH, and T. TRH, fT4, and fT3 decreased, but TSH concentration increased. mRNA for crh and gr were downregulated compared with the control. Expression of gr was downregulated while expression of pomc was up regulated. In the adrenal gland, levels of mRNA for cyp11a1 and 3βhsd were significantly lower. In the hypothalamus, mRNA expressions of gnrh1, erα, and erβ were downregulated.

Androgenic/Antiandrogenic Effects of MCs and Consequences on Reproductive Toxicity
The androgenic/antiandrogenic effects of MCs and their potential influence on the reproduction of males are shown in Table 2. Most of the studies were carried out in vivo on mammals, while in vitro studies were less frequent, highlighting the research performed with a double strategy, using in vivo and parallel in vitro assays [45,46,[71][72][73][74][75][76][77]. Moreover, only a few studies have been performed using fish [78] and, occasionally, the effects of MC-LR have been explored in amphibians [79,80] and crustaceans [81,82]. Almost all of the studies were carried out with the pure standard MC-LR, apart from Chen et al. [78], and, consequently, all of the endocrine disrupting effects observed are due to the toxin and could not be attributed to other active compounds present in purified Microcystis extracts.
The first studies demonstrated MC-LR-induced endocrine effects and male reproductive toxicity via both acute and subchronic exposures in mammals [71,83]. MC-LR lowered the T, LH, and FSH serum levels and decreased testis weight and sperm concentration in rats after 28 days of exposure [71], and, similarly, decreased serum T concentrations, impaired sperm quality, and led to testicular injury in mice chronically exposed to low-dose exposures of MC-LR (3-6 months) [83]. Numerous studies have reported a decrease in T levels after exposure to MC-LR in mammals [46,71,72,[75][76][77][81][82][83][84]. However, the response in the case of FSH and LH levels or the expression of FSH and LH genes after toxin exposure was more variable: sometimes it decreased [71,72]; at times it increased [46,80,83,84]; and even initial increases were followed by decreases in mice exposed to MC-LR [45,46]. All these changes suggested the potential endocrine toxicity of MC-LR.
In relation to the distribution of MC-LR in male gonads, the toxin was found in aquatic organisms [85,86] and in fish [87]. In mammals, diverse studies have confirmed the localization of MC-LR in testes [45], and the target cells in testes were explored, such as cultured spermatogonia, Sertoli cells, and Leydig cells (LCs) [45,71,76]. Li et al. [71] investigated the toxicity of MC-LR to cultured LCs, and demonstrated decreased T production and increased reactive oxygen species (ROS) and lipid peroxidation (LPO), showing that oxidative stress could play a role in cell apoptosis of these cells.
Chen et al. [83] suggested that MCs in mice are transported to the testis via the blood, and that the Leydig cells were the initial target. The LCs have an important role in the synthesis and secretion of androgens, mainly T, which are essential for spermatogenesis and sperm maturation. Moreover, these authors indicated that serum LH and FSH increased in response to MC-LR treatment, especially after 6 months of exposure, and this could be due to the regulation of secretion of T by LH in a negative feedback manner. Consequently, the suppression of T secretion stimulated the release of LH to maintain homeostasis. In addition, the apoptosis of cells, increased in a dose-and time-dependent manner to MC-LR, disrupted the ability of Sertoli cells to secrete inhibin leading to an increased secretion of FSH [83].
Chen et al. [84], in rats exposed to i.p. with MC-LR for a long time (50 d), found decreased T levels, while FSH and LH were increased, and morphological alterations of the testes were reported. Thus, the testes index (calculated by the formula: (testes weight/body weight) × 100%) in the group of rats exposed to the highest dose decreased in comparison to the controls, and the space between the seminiferous tubules was increased (light microscope). Ultrastructural observations showed cytoplasmic shrinkage, cell membrane blebbing, swollen mitochondria, and a deformed nucleus. The authors suggested that the cytoskeleton disruption and mitochondria dysfunction induced by MC-LR could interact through ROS formation and resulted in an impairment of the reproductive system. The effects on LCs cells induced by MC-LR were confirmed by Chen et al. [73] after chronic exposure to low concentrations of MC-LR in mice. In this work, the number of LCs in the testes of mice was decreased, while macrophages were significantly increased. By using a co-culture system, they studied the interaction between macrophages and LCs in the presence of MC-LR: stimulation of macrophages to produce TNF-α, and secreted TNF-α induced LCs apoptosis by binding to the tumor necrosis factor receptor 1 (TNFR1) on these cells, and by activating the ROS-p38MAPK signaling pathway. Furthermore, they reported that GAS6 (a protein that binds apoptotic cells and macrophages, creating a bridge) mediated phagocytosis of apoptotic LCs by binding to the Axl receptor on macrophages and phosphatidylserine (PtdSer) on apoptotic LCs. In summary, reduced serum T levels could be associated with decreased LCs due to LCs apoptosis by immune cells after MC-LR exposure in mice.
In contrast to the suggestion that LCs are a potential cellular target of MC-LR, Wang et al. [88] reported that MC-LR could enter testicular tissues in mammals, and spermatogonia and Sertoli cells are critical target cells of MC-LR, whereas MC-LR was not detected in LCs and had no cytotoxicity on these cells [45,88]. These authors indicated that MC-LR could exert an indirect dysfunction of these cells, a secondary effect due to the damage to the HPG axis caused by the toxin. This hypothesis was supported by the decrease induced by MC-LR in the hypothalamic gonadotropin-releasing hormone (GnRH) expression in a dose-and duration-dependent manner [45].
Similarly, Xiong et al. [72] evaluated the effects of MC-LR on the HPG axis in mice exposed to different concentrations of MC-LR (1-14 d). In this work, the toxin impaired the spermatogenesis of mice, perhaps through the direct or indirect inhibition of GnRH synthesis at the hypothalamic level, which resulted in decreased levels of LH and a suppression of T production in the testes. The in vitro experiments carried out by these authors on LCs confirmed that MC-LR did not affect T synthesis by direct damage of these cells.
As the LCs (the main producers of T) were not injured by MC-LR in some experiments, Wang et al. [46] demonstrated that the GnRH neurons were the target of MC-LR in rats. In this work, serum levels of GnRH, LH, FSH, and T showed a similar pattern of earlystage increases and late-stage declines. The authors indicated that there might be two mechanisms by which MC-LR affected the hypothalamic-pituitary axis: (1) MC-LR could attack multiple targets along the axis, and, consequently, some of the targeting led to decreased transcription of GnRH while others led to disrupted pituitary activity; (2) MC-LR only affected a single target, such as the GnRH neurons, affecting the expression and secretion of GnRH. Later, the same group investigated the toxic effects of MC-LR on GnRH neurons in the hypothalamus [75] and demonstrated that the toxin could enter GnRH neurons and inhibit GnRH synthesis, resulting in the decrease of serum GnRH and T levels in mice. In vitro, in GT1-7 cells, they also associated this inhibitory effect on GnRH synthesis with the activation of the cyclic adenosine monophosphate (cAMP/protein kinase A (PKA)/cAMP response element-binding protein (CREB)/c-Fos. Furthermore, they found that miR-329-3p was significantly reduced in GT1-7 cells after MC-LR exposure, and the toxin induced PKA activation by regulating the expression of PRKAR1A and PRKACB at the post-transcriptional level. These data contribute to understanding, at the molecular level, male infertility induced by MC-LR because of dysfunction of the HPG axis.
With the aim of studying the specific mechanisms of the uptake of MC-LR by GnRHsecreting neurons, Ding et al. [89] demonstrated in vitro, in GT1-7 cells, that at least four organic anion transporting polypeptides (Oatp1a4, Oatp1a5, Oatp5a1, Oatp2b1) were expressed in GnRH neurons at the mRNA level, but only Oatp1a5 was expressed at the protein level. They also reported that MC-LR could not be transported into Oatp1a5-deficient GT1-t7 cells, which were protected against the effects of the toxin. Later, the same authors confirmed in vivo that, after MC-LR exposure, mice exhibited decreased GnRH levels [74]. Furthermore, in GT1-7 cells, the toxin stimulated intracellular Ca 2+ and cAMP to activate the protein kinase c (PKC), PKA, and MAPK signaling pathways in GnRH neurons. Then, the toxin altered some protein levels and changed the activity of GnRH transcription factors, such as Pbx1a, Oct-1, Dlx-2, Otx-2, c-Jun, and c-Fos. All these data support the conclusion that MC-LR can decline the synthesis of GnRH in GnRH neurons via the Ca 2+ -PKC-NF-kB-cAMP-PCA-Creb and Erk/P38-MAPK signaling pathways. Recently, Jin et al. [77], using a combination of in vitro (the same GT1-7 cells) and in vivo experiments (male mice), demonstrated that MC-LR could activate the endoplasmic reticulum stress (ERs), leading to apoptosis of GnRH neurons and decrease GnRH synthesis, and reduce the secretion of T. Moreover, pre-treatment with a ERs inhibitor reduced the apoptotic rate on cells, whereas apoptotic cell death was increased by pre-treatment of GT1-7 cells with an autophagy inhibitor. Autophagy is a protective cellular process, activated by ERs stress and could potentially protect cells from apoptosis induced by MC-LR.
In contrast to previous studies, which suggested that LCs are an indirect target of MC-LR [45,74], a recent study carried out in vivo in mouse testes and in vitro in TM3 cells [76] indicated that MC-LR can enter and accumulate into mouse testes and in TM3 cells, and provided results indicative that LCs are a direct target of MC-LR. The main findings of this work were: (1) MC-LR reduced serum and testicular T levels; (2) MC-LR downregulated steroidogenic proteins and synthesis in both models; (3) MC-LR evoked GCN2/elF2α signaling in TM3 cells; (4) GCN2iB, a specific inhibitor of GCN2 signaling, attenuated MC-LR-induced eIF2α phosphorylation and subsequent downregulation of steroidogenic proteins; (5) pre-treatment with N-Tert-Butyl-α-Phenylnitrone (PBN), a free radical scavenger, reduced the activation of GCN2/elF2α and the downregulation of steroidogenic proteins in TM3. All these results indicated that MC-LR at least partially inhibited T synthesis via the ROS-mediated GCN2/elF2α pathway.
In the case of fish, only one study performed on tilapias (Nile tilapia) demonstrated the effects of MCs extracted from lyophilized cells of Microcystis aeruginosa and purified MC-LR on endocrine endpoints and reproduction. After 28 d of exposure, T levels increased in Microcystis and MC-LR-exposed fish, and interferences with the expression of genes involved in the HPLG axis and balance of sex hormones were also induced [78].
Endocrine-disrupting effects induced by MC-LR have also been investigated in male frogs (R. nigromaculata) [79,80]. MC-LR induced toxic effects on the reproductive system of frogs, decreased T levels, and increased E2 contents. After prolonged exposure, the relative expression levels of P450 aromatase (CYP19A1) and steroidogenic factor 1 (SF-1), important factors in reproductive toxicity in males by mediating molecular regulatory mechanisms, were enhanced [79]. Later, these authors [80] confirmed decreased T and increased E2 concentrations after exposure to 1 and 10 µg/L MC-LR for 14 d. The toxicity mechanisms proposed in amphibians were: (1) the downregulation of HSD17B3 gene expression caused the disruption of T synthesis; (2) the upregulation of CYP19A1 gene expression directly stimulated conversion of T to E2; (3) this upregulation of CYP19A1 stimulated conversion of androstenedione to estrone (E1) and the upregulation of HSD17B1 gene expression stimulated conversion of E1 to E2. Moreover, the decrease of T and abnormal gene expression of the AR and oestrogen receptors (ESR1) caused interferences in spermatogenesis in frogs [80]. In the case of the giant freshwater prawn, M. rosenbergii [81], results showed that MC-LR could disrupt the testicular development of M. rosenbergii, perhaps by affecting T levels (decreased) and the expression of gonadal developmental-related genes in the testes and eyestalk. Moreover, testicular germ cells, mitochondria, and cell junctions were also damaged. In aquatic crustaceans (M. nipponense), the parental transference of MC-LR induced testicular dysfunction and reproductive and offspring immune changes in exposed prawns [82]. MC-LR-damaged testicular germ cells, reduced T levels in the hemolymph, and inhibited the development of the testes. The F1 offspring showed a downregulation of immunity molecules and antioxidant enzymes, and higher expression levels of apoptoticrelated genes. The mechanisms involved could be the induction of mitochondrial apoptosis of the F1 offspring after continuous exposure to MC-LR.
Moreover, changes in androgenic biomarkers have also been observed in females. Thus, some studies, as included in Table 1, demonstrated reproductive effects in female zebrafish after exposure to MC-LR, with changes in several hormones (e.g., T levels) which modulate reproductive toxicity, as well as changes in the transcription of HPG-axis genes [41,44,47,58]. In female rats, acute exposure to MC-LR affected histopathology and induced changes in several hormones in the HPG axis (GnRH, LH, FSH, T, and E2), and in the transcription of genes for hormone synthesis (as well as in the case of the HPA and HPT axes) suggesting endocrine-disrupting effects [3].
In summary, reports have shown that MC-LR can show androgenic effects mediated by different mechanisms, although cellular targets need to be still defined (LCs, GnRH neurons, etc.). Moreover, in this case, differential responses between pure MC-LR, and other congeners have not been explored. Furthermore, further cyanobacterial blooms/extracts investigations would be of interest, as they represent a more real exposure scenario.     Pre-treatment with an endoplasmic reticulum stress inhibitor, such as 4-phynyl butyric acid, reduced apoptosis rate in the cells. The pre-treatment with the autophagy inhibitor 3-MA could increase the apoptosis caused by the toxin.  Gnrh transcription levels decreased with increasing MC-LR concentrations. Oatp1a4, Oatp1a5, Oatp5a1, and Oatp2b1 genes were expressed in GT1-7 cells. Only Oatp1a5 protein was expressed. [89]

Effects of MCs on Steroidogenesis
Steroidogenesis is the process by which steroid hormones are synthesized from cholesterol. There are two principal pathways of synthesis, the adrenal and the gonadal pathways, and, in both cases, numerous enzymes are involved [2,80]. Thus, the steroidogenic acute regulatory protein (StAR), cytochrome P450 cholesterol side chain cleavage enzyme (CYP11A1), cytochrome p450 17 α-hydroxylase/17,20hydroxylase (CYP17A1), 3ß-hydroxysteroid dehydrogenase type 2 (HSD3B2), HSD17B3, and cytochrome P450 aromatase (CYP19A1) are important steroidogenic enzyme profiles along the HPG axes in testes [2,80]. Pavagadhi et al. [90] found, in zebrafish exposed to MCs for 30 d, that cholesterol synthesis was significantly affected, suggesting that MCs can cause endocrine disruption even under sublethal conditions; however, separately from cholesterol synthesis, various metabolic pathways, such as essential fatty acids and lipid oxidation, were also perturbed. Several studies included in Table 1 [48,58] and Table 2 [63,76] show the capacity of MC-LR to interfere on the expression of genes involved in steroidogenesis.
As mentioned above (Section 2.2.), a combined in vitro an in vivo approach is a useful way by which to understand the activities of endocrine disruptors, and both methods were employed by Hou et al. [48] to understand the estrogenic effects of MC-LR via stimulating the steroidogenesis. On the one hand, the effects of MC-LR (1-5000 µg/L) on steroidogenesis were assessed in vitro, in the H295R cells after 48 h of exposure, and the levels of E2 and T increased in a non-dose dependent manner, which showed positive correlations with the expression of steroidogenic genes. Induction of E2 via interference with steroidogenic enzymes, such as an aromatase, the rate-limiting enzyme converting T to E2, is the mechanism suggested for estrogenicity of MC-LR, separately from binding ER [91]. On the other hand, the authors investigated the in vivo effects of MC-LR on male zebrafish exposed to the toxin (0.3-30 µg/L MC-LR) for 30 d. Similar to the in vitro results, sex hormone levels and steroidogenesis gene expression in zebrafish increased, and the critical estrogen threshold was established between 1 and 30 µg/L MC-LR, which is considered environmentally relevant. In this study, E2 and T contents in the testis increased, and upregulation of some steroidogenic genes, especially CYP19a, was reported. In the liver, the vtg1 gene was upregulated, while the transcriptional and protein levels of the ER were decreased. All these results indicated that the estrogenic effects of MC-LR were non-dose dependent, which may result from steroidogenesis stimulation via a non-ER-mediated pathway [48]. These authors indicated that the cyclic chemical structure of MCs is not like other EDCs (nonyl phenol, bisphenol A), and its hydrophilic feature reduces the possibility of it functioning as an ER ligand. Thus, MC-LR is unlikely to cause estrogenic responses via binding ER, and they suggested that MC-LR induces E2 increases via stimulating steroidogenesis rather via the ER-mediated pathway.
Moreover, in vitro MC-LR disturbed steroidogenesis of primary cultured ovarian granulosa cells, responsible for the production of E2 and P4. These cells could uptake MC-LR and should be the target of the toxin, which induces oxidative stress [92]. The ovarian granulosa cells (GCs), essential for the growth and development of follicles, were also exposed to MC-LR for 48 h, and microRNAs (miRNAs) and mRNAs microarray technologies were applied. Numerous miRNAs and mRNAs significantly changed in the treated MC-LR group, and functional analysis indicated that they were involved in proliferation, apoptosis, immunity, metabolism, etc., and, consequently, interferences with the normal function of the cells were reported. Disorders of estrogen and progesterone, and premature ovarian failure (POF) are important factors of infertility, while changes in GCs function are crucial for the induction of infertility [92].
The effects of MC-LR on steroidogenic proteins and GCn2/eIF2α signaling in Leydig cells [76] have been previously commented on (Section 2.2.). In this study, ROS-mediated GCn2/eIF2α activation could explain, in part, the downregulation caused by MC-LR on steroidogenic proteins and synthases.
An optimized vitrification protocol method to cryopreserve murine immature follicles has been developed and it has been used as a screening method by which to investigate the ovotoxic response to different MCs congeners [68]. The dose-response ovotoxicity screening results revealed that MC-LF was the most ovotoxic MC variant. Moreover, MC-LR, compared to MC-LA, MC-LF, or -LT, showed the least adverse impacts on follicle survival and development, indicative that these MCs variants could cause follicle damage through other molecular mechanisms, different to the inhibition of PP2A. For example, the expression profile and function of OATPs in ovaries could be involved. This screening also demonstrated that MC-LF at 0.1 µM significantly inhibited E2 secretion, indicating that the human-relevant exposure of MC-LF may interfere with ovarian steroidogenesis and exert endocrine disruptive effects [68].
In fish, the endocrine system consists of three main pathways, including the HPG, HPT axis, and HPI axis. Several studies have indicated that the effects of EDs on an endocrine axis pathway could indirectly influence the other axes in fish. Thus, it has been demonstrated that MC-LR induced endocrine disruption, through endocrine axes, perhaps due to interference with steroidogenesis, including cholesterol, cortisol, and estrogens, and, subsequently, resulted in a disruption of glucose homeostasis [63]. Furthermore, zebrafish have also been chosen as a good experimental model in which to investigate the effects of pure MC-LR standard on the HPI axis during early embryonic development (embryos/larvae), 4-168 hpf [93]. At the highest concentration assayed, MC-LR induced higher cortisol concentrations, and the expression of genes along the HPI axis and mineralocorticoid receptor (MR-) and glucocorticoid receptor (GR-) were also changed. This was the first study at the molecular level focused on the endocrine-disrupting effects of MC-LR, including the neurocrine, steroidogenic, and receptor signaling pathways.
Further research is required to confirm the mechanisms involved in the effects of MC-LR on the production of E2, T, and all the key enzymes for steroidogenesis, by using the double strategy of in vitro and in vivo assays, following international guidelines. In the case of steroidogenesis, the use of the in vitro assay OECD TG 456, not applied so far for cyanotoxins, is encouraged. The same would apply for other minority congeners (as variant-specific effects have been reported) and cyanobacterial blooms.

Thyroid Endocrine Disruption of MCs
The different studies focused on the induction of thyroid endocrine disruption of MCs are shown in Table 3. All of them have been performed using the pure standard MC-LR congener, except for one study, that investigated the effects of MC-RR on zebrafish (0.3-3.0 mg/L for 96 hpf) [94].
Several studies have reported that subchronic exposure to high MC-LR concentrations can cause thyroid dysfunction in fish [95,96] and in mice [97,98]. Previous studies demonstrated that MCs induced changes in thyroid hormones (THs) production in fish [71], and, since this work, most experiments have been performed using zebrafish embryos or juvenile zebrafish as a model to study the thyroid-disrupting toxicity of MCs [94,96,[99][100][101]; although, in other studies, adult zebrafish have been employed [95,102,103]. In mammals, studies are very scarce [3,97,98]. Moreover, in Xenopus laevis tadpoles exposed to an ecyanobacterial biomass containing MC-LR, no changes in thyroid-stimulating hormone (TSH) were detected after 21 days of exposure [104].
The thyroid hormone receptor-α is an inducible ligand-activated transcription factor and thyroid hormones exert their physiological effects by binding to them on the HPT axis [105]. In zebrafish larvae exposed to 100-500 µg MC-LR/L, body growth retardation associated with significant decreases in mRNA expression of iodothyronine deiodinase 2 (Dio2), as well as decreased T4 and T3 levels, were reported [99]. These findings indicated that MC-LR could alter gene expression in the HPT axis, which might contribute to MC-LRinduced thyroid disruption [99].
Similarly, Liu et al. [95] indicated that MC-LR, at relevant environmental concentrations, resulted in the disturbance of TH homeostasis by disrupting the synthesis and conversion of THs: altered Dio activities and decreased levels of T3. Iodothyronine deiodinases play crucial roles in the mechanisms of thyroid hormone biotransformation and metabolism in peripheral tissues, and are important regulators of circulating and intracellular THs levels in fish [106]. Moreover, hyper-stimulated THs synthesis and secretion, elevated T4 levels, as well as an upregulation of genes involved in THs synthesis could be considered as negative feedback from the hypothalamus and pituitary due to the decreased levels of T3. All these lead to a hypothyroidism state after exposure to this toxin.
Similar effects were also reported in vivo in juvenile Chinese rare minnows exposed to MC-LR for 7 d [107], with alterations in Dio activities that resulted in decreased T3 production. The authors indicated that this decrease could be due to a reduction in the rate of conversion of T4 to T3 because of the reduced activity of Dio2. Moreover, the reduction of Dio3 activity associated with the downregulation of mRNA expression in Dio3 could reflect a negative feedback compensatory mechanism against decreases in T3 concentrations in the exposed fish. The sodium/iodide symporter (NIS) gene is known to be involved in thyroid hormone synthesis, and changes in its expression may alter thyroid hormone production. The down-regulation of NIS expression indicates that MC-LR could induce thyroid disruption. Moreover, the transport protein for thyroid hormones transthyretin (TTR), mainly secreted by the liver, regulates the supply of the various thyroid hormones in target tissues. The authors concluded that the changes in transcription of the NIS, TTR, thyroid hormone receptor-α, and iodothyronine deiodinase genes indicated disturbances in thyroid hormone synthesis, transport, and metabolism of fish, leading to a decline in either T4 or T3 production. In agreement with previous work [95], the hypertrophy of thyroid follicle epithelial cells could be secondary to the negative feedback regulation because of the lowered thyroid hormones, also resulting in hypothyroidism in this juvenile fish species.
In adult zebrafish, the effects of MC-LR on the THs homeostasis were also investigated after longer exposure (1-28 d) to environmental concentrations (1-25 µg MC-LR/L) [102]. In contrast to previous experiments in juvenile zebrafish [95], in adult zebrafish, no differ-ences were found in the histopathology of thyroid follicles and T4 levels were unchanged, confirming that exposure to MC-LR did not inhibit the production of THs in adults. However, in agreement with juvenile zebrafish exposed to the same MC-LR concentrations, a significant decrease in T3 levels associated with decreased Dio2 activity in male zebrafish was observed, suggesting the important role of Dio2 in the disturbances of thyroid hormones. Moreover, the mRNA expression of TSH, TTR, and TRS appeared to be a dynamic process, as expression first decreased and then increased with continued exposure. The authors suggested that although MC-LR exposure can alter the metabolism of THs, fish can trigger compensatory mechanisms to maintain THs homeostasis [103].
This fact has also been confirmed in adult zebrafish exposed to acutely higher concentrations of MC-LR (50-400 µg MC-LR/L) during 24, 48, 72, or 96 h, which disrupted the thyroid hormone metabolism by altering Dio activity and gene expression of the HPT axis [103]. These authors confirmed that MC-LR induced a negative feedback regulation of the HPT axis in adult zebrafish, as the females were more sensitive than the males. Notably, it was reported that these changes may affect the complement system through the regulation of c9 mRNA synthesis, although the relationship between thyroid hormone receptors (TRs) and C9 requires further research. In fact, MC-LR caused immunotoxicity in fish and mammals by several toxicity mechanisms [20], and previous studies have demonstrated that MC-LR may affect the function of the complement system [108], regarded as an essential humoral system.
The negative feedback regulatory response in fish, was also reported in juvenile zebrafish after acute MC-LR exposure (50-400 µg MC-LR/L) [100], in the same experimental conditions that the previous work of Gao et al. [103]. MC-LR exposure led to significant reductions of T4 and T3, and the toxin affected the transcription levels of genes involved in TH synthesis transport and metabolism, and the normal function of the thyroid. All the results confirmed that fish can trigger a compensatory mechanism to maintain TH homeostasis after continual exposure to the toxin.
Moreover, after parental exposure to MC-LR, thyroid disruption in F1 larvae zebrafish was demonstrated for the first time and highlighted the transgenerational toxicity of MC-LR [96]. The changes in the F1 offspring were decreased hatching and growth retardation, correlated with reduced THs levels. The decreased THs levels in the progeny may be due to abnormal gene transcription along the HPT axis in F1 larvae and thyroid dysfunction in adult female zebrafish, whereas the T4 and T3 levels were unchanged in males. The different expression pattern in both sexes of fish, could contribute to the sex related of THs levels, although further studies are needed to clarify the mechanisms.
Recently, the effects of the coexistence of MC-LR and polystyrene nanoplastics (NSNPs) on the early growth of F1 zebrafish has been studied, and the combined exposure increased the parental transfer of MC-LR of the offspring and increased growth inhibition [102]. The decreased THs levels and the significant changes in the HPT axis gene expression indicated the thyroid disturbance caused by both contaminants could be the main cause of growth inhibition of F1 larvae. More studies are needed to explore the combination of cyanotoxins and other pollutants through the introduction of technologies, such as transcriptomics and proteomics [101].
Considering that Dios are key regulators of THs, and that the liver is the main organ that expresses the outer-ring deiodination (ORD) activity, the hepatic cell line of grass carp has been also chosen as an experimental model to investigate the effects of MC-LR on the activities of these enzymes [109]. Differences in the mRNA expression and activities of Dios were found after MC-LR exposure, decreasing Dio1 and Dio2 and increasing Dio3, and they could be responsible for the changes in THs levels and disturb the normal THs metabolic processes in fish.
Only one study investigated the thyroid endocrine disruption in developing zebrafish larvae exposed to MC-RR [94], and this MC-congener was able to change the transcription pattern of HPT-axis-related genes, except in the case of the thyroglobulin (TG) gene. Moreover, protein synthesis of TG was not affected, whereas NIS was significantly upregulated, in agreement with gene expression. In this study, the mRNA level of Dio1 decreased, but the transcription of Dio2 gene increased, in contrast to previous results (as mentioned above) obtained after MC-LR exposure by Yan et al. [99]: increased Dio1 and decreased Dio2 mRNA expressions. These facts could indicate that diverse MC-congeners exert thyroid toxicity by different mechanisms.
In addition, the thyroid toxicity of MCs in mammals has been poorly studied [3,97,98]. In the first work, in mice exposed to i.p. with MC-LR for 4 weeks, typical symptoms of hyperphagia, polydipsia, and weight loss with thyroid disfunction were found. Moreover, the animals showed glucose, triglyceride, and cholesterol metabolism disruptions. In this experiment, the high FT3 level found in exposed mice, together with a low FT4 level and a normal TSH concentration, was not consistent with the diagnosis of hyperthyroidism in the animals exposed to MC-LR. Increased expression of the Dio2 protein, increased expression of the THs receptor (TR α), and mTOR expression in the brain after exposure to the highest dose of 20 µg MC-LR/kg were reported. The increased plasma FT3 content regulating mTOR signaling was at first considered responsible for the increased food consumption and energy expenditure in the exposed mice. Furthermore, the glucose and lipid metabolic disorders during thyroid dysfunction could be due to the perturbed genetic expression of several genes of the key enzymes involved, which have been identified as being affected by thyroid hormone levels, after MC-LR exposure [97].
The chronic and low-dose effects of MC-LR on mouse thyroid tissues, exposed via the oral route, have also been investigated, as well as the toxin-induced apoptosis, lymphocyte infiltration, thyroid structural disorders, and changes in THs levels [98]. In this study, the expression of Dio3 increased in thyroid and peripheral tissues, and the phosphorylation level of the extracellular signal-regulated kinase (ERK), p38, and Mitogen-activated-protein kinase (MEK) was induced. This confirms that Dio3 and the p38/MAPK and MEK/ERK signaling pathways may play an important role in thyroid injury induced by MC-LR [110]. Moreover, the effects on thyroid hormones metabolism were also investigated in vitro by the same authors on thyroid follicular cells, Nthy-ori 3-1 cells. These cells accumulated MC-LR and, in agreement with the in vivo study, only the expression of Dio3 was increased, and activation of the p38/MAPK and MEK/ERK signal pathways was detected.
Finally, in an extensive study performed by Chen et al. [3] in female rats injected with a single dose of MC-LR (median lethal dose), several parameters were examined: histopathology of several organs, including thyroid tissue (thyroid follicular cells), concentrations of hormones in serum and gene expression of the HPA, HPG, and HPT axes. These authors suggested that MC-LR affected the HPA, HPG, and HPT axes. For the HPT axis, MC-LR induced higher TSH concentrations, although lower levels of TRH, FT3, and FT4 were reported. Globally, significant positive/negative correlations of concentrations of hormones were reported among the three axes, HPA, HPG, and HPT, and profiles of transcription of genes for synthesis of hormones and nuclear hormone receptors in thyroid, adrenal, and ovaries were altered.
All these experiments open the door to new research by which to clarify the functional changes and mechanism of thyroid dysfunction in mammals (rats) after MC-LR exposure; to understand the potential risks of endocrine toxicity of cyanobacterial toxins, focused mainly on the mechanisms of effects of MC-LR on the different endocrine axes and hormonal functions. Thus, additional studies with mammalian models exposed through human relevant pathways would be welcome, as well as to investigate the effects of cyanobacterial blooms.   the accumulation of MC-LR in the F1 generation. MC-LR significantly reduced T4 levels in F1 larvae. After co-exposure, the concentration of T4 decreased further. T3 level of F1 larvae was also significantly reduced in the high-concentration MC-LR group. T3 content of larvae significantly decreased in all PSNPs + MCLR groups. Parental exposure to PSNPs and MC-LR had more pronounced effects on thyroid hormone levels in larvae. When exposed to 4.5 µg/L MC-LR treatment group, the transcription of tr-α, tr-β, and Dio2 were significantly downregulated, and the mRNA levels of TG, TTR, and Dio1 were significantly increased. When parents were exposed to 22.5 µg/L MC-LR, the mRNA levels of CRH, tr-α, tr-β, and Dio2 were significantly decreased, and the transcriptions of TG, TTR, and Dio1 were significantly multiplied in F1 larvae.
[101]  In ovaries, rats exposed to 54.75 or 73 µg MC-LR/kg exhibited obvious hyperemia, cytoplasmic loss, abnormal nuclear change, and nuclear dissolution. Necrosis of local granular cells in rats exposed to 73 µg MC-LR/kg. Broken nuclei, necrosis of follicular epithelial cells, and reduced intracellular colloid in thyroid gland. Serum concentrations of CRH, ACTH, and CORT significantly decreased. Lower concentrations of GnRH and E2, but higher concentrations of LH, FSH, and T. Concentrations of TRH, FT4, and FT3 significantly decreased in treated rats, but high TSH concentration was recorded. Expressions of genes among HPA, HPG, and HPT axes were diverse. In the hypothalamus, transcripts of thrβ were lower in rats exposed to 73 µg MC-LR/kg, while there were no significant changes in expression of trh, tshr, or thrα. In the pituitary, there were no significant effects on mRNA expressions of trhr, tshβ, thrα, or thrβ. In the thyroid, there were no significant alterations in abundances of transcripts of tshr, thrβ, tg, or tpo. When exposed to 36

Cylindrospermopsin
Studies focused on the potential endocrine-disrupting effects of CYN are very scarce (Table 4). Some estrogenic effects have been detected with cyanobacterial extracts containing CYN [49], but information on the estrogenic potency of CYN is very limited [50].
Regarding the scientific literature, the first study in which pure CYN exhibited this activity was reported in primary human-IVF-derived granulosa cells exposed to low concentrations (0-1 µg CYN/mL) for 2-6 h or for 24-72 h. In this model, the human chorionic gonadotropin (hCG)-stimulated progesterone production was inhibited when exposed to a non-cytotoxic concentration of CYN (1 µg/mL) for 24 h [111]. They also found that although CYN was not cytotoxic and it did not affect hCG-stimulated estrogen production after 6 h, the toxin did abolish hCG-stimulated progesterone production. However, in a second experiment, using IVF-derived cells obtained selectively from women with normal fertility status, CYN up to 3 µM (1.25 µg/mL for 6 h) was not cytotoxic, and did not alter progesterone or estrogen production, with or without hCG stimulation. Protein synthesis was significantly inhibited by 3 µM CYN alone, and CYN in the presence of hCG caused a concentration-dependent decrease in global protein synthesis, but had no effect on the synthesis of steroid hormones. This suggests that StAR and CYP450 aromatase (CYP450arom) protein synthesis was not inhibited by CYN [112].
Some aqueous extracts of cyanobacteria, such as Aphanizomenon flos-aquae (a CYN productor), have been shown to be active in the human breast carcinoma cell line MVLN [49]. The only extract that showed clear dose-dependent antiestrogenic activity in co-exposure with E2 was Aph. flos-aquae PCC, containing 3100 µg CYN/g d.w. Moreover, the potency was not correlated with CYN concentrations, suggesting that the activity was also due to other compounds.
Liu et al. [113] investigated the estrogenic activity of CYN (2.4 × 10 −7 M to 2.4 × 10 −12 M) by a yeast estrogen screen (YES) assay and after oxidation of the toxin. CYN was anagonist in the YES assay, and its binding affinity to the estrogen receptor was linked to its intrinsic properties. The toxin modulated E2 estrogenic activity, resulting in non-monotonic responses, and this behavior is common to xenobiotics, known as estrogen-actived chemicals (EACs). After treatment with oxidative compounds, the by-products obtained had reduced binding affinity to estrogen receptors in comparison to the parent toxin [113].
Recently, the direct role of CYN on testicular function and spermatogenesis in aquatic organisms has been studied using ex vivo cultures of zebrafish testes as experimental models [114]. After exposure to CYN (250-1000 µg/L CYN for 24 h and 7d), the toxin reduced the basal-and gonadotropin-induced process of spermatogenesis, and it may have contributed to a decrease in fertility. Moreover, CYN inhibited all stages of spermatogenesis in zebrafish testes; this disruption of the spermatogenesis process could indicate that CYN can impair male reproduction. Significant changes in fshr, lhr, and insulin-like grown factor 3 (igf ) transcript levels were also found, and T secretion was increased as well. These findings contributed to a better understanding of the mechanisms involved on the effects of CYN on male reproduction by inducing apoptosis and altering gonadotropins, as well as changing production of T and igf3. The last fact, inhibition of igf3 by CYN, could be a contributing factor in the mechanisms of CYN-induced impairment of differentiated spermatogonia development [114].
Thus, considering the relevance of CYN, knowledge of its potential endocrine disruption properties should be further investigated. Reports so far, although scarce, have suggested that CYN is a potent EDs on different models, as well as CYN-containing extracts. hCG-stimulated P production was inhibited at 24 h. CYN was not cytotoxic, and it did not affect hCG-stimulated estrogen production after 6 h. The toxin did abolish hCG-stimulated P production. [ After treatment with oxidative compounds, the byproducts obtained reduced binding affinity to estrogen receptors in comparison to the parent toxin. [113] E2: estradiol; hGC: human corionic gonadotropin; hpf: hours post fertilization; MTT: 3-(4, 5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide); P: progesterone; RIA: radioimmunoassay; T: testosterone; YES: yeast estrogen screen.
The same authors investigated the adverse effects of CYN (0-2000 µg/L CYN) on zebrafish embryonic development [116], and numerous changes were observed: decreased growth, increased developmental abnormalities, such as pericardial and yolk sac edema, as well as swim bladder absence. In addition, CYN induced changes in thyroid-hormonerelated genes (tr1a, Dio1, Dio3) which could be involved in the induction of thyroid disrup-tion. Globally, the authors indicated that CYN exposure alters functions related to thyroid hormones, oxidative stress, and osmoregulation, all of them key components for normal embryo development [116]. These results confirmed the adverse effects on embryonic development induced by CYN in zebrafish exposed to the toxin (2-2000 nM) previously reported by Wang et al. [117], who indicated that they may be associated with oxidative stress and apoptosis.
Finally, the disruption of the estrous cycle and its effects on spermatogenesis in vivo in female and male mice was recently investigated [115]. A single i.p. injection of 64 µg CYN/kg b.w. in females induced an impairment in the estrous cycle and a decrease in progesterone levels. In males, weekly i.p. doses of 20 µg CYN/kg b.w given to several groups (1, 2 or 4 doses) increased T levels in groups administered with 1-2 doses and induced increases in spermatozoa cells. All of these results indicate that CYN interferes in the mammalian reproductive system, and may cause infertility, although further studies should be carried out to confirm this dysfunction in mammals and the mechanisms involved.

General Discussion and Final Remarks
Endocrine disruption effects have gained relevance from a toxicological point of view due to the wide number of xenobiotics exerting disruptive properties and their contribution to different illnesses and health effects, in both humans and the environment. Moreover, very different mechanisms can be involved (i.e., interaction with receptors, changes in metabolism of endogenous hormones, interference with feedback regulation and neuroendocrine cells, genomic instability by interference with the spindle figure can play a role, etc.), as reported by De Coster and van Larebeke [9].
MCs and CYN are cyanobacterial toxins that, although targeting the liver, have been shown to also induce endocrine disruption effects, as evidenced in the present review manuscript.
The endocrine disruption activity of MCs has been more extensively investigated in comparison to CYN. Moreover, some of the studies have been performed with cyanobacterial cultures that contain other kinds of compounds that could be responsible of the effects observed (i.e., [49,65,78,104], etc.). If we focus on pure toxins, the available results are highly variable. Thus, Hou et al. [48] reported that MC-LR is unlikely to cause estrogenic responses via binding ER, as its chemical structure is different to other known EDCs, but there are many studies that have evidenced MCs ED effects by other pathways. These include, among others, pathological damage in related organs and cells such as the testis (i.e., [61]), ovarian cells (i.e., [44,66]), Leydig cells (i.e., [83]), or GnRH neurons [46]. Moreover, these cells can suffer from apoptosis and toxic effects mediated by ROS [84] or immune cells [73]. Additional effects/mechanisms include: decreased gonad-somatic index [41], changes in transcriptional responses of HPG-axis related genes [43,47,60], HPI-axis-related genes [93], HPT-axis-related genes [99], steroidogenesis-related genes [76]; disruption of the GH/IGFs system [59]; activation of the ERK1/2 signaling pathway [64], changes in the activity of GnRH transcription factors [74], interference with steroidogenic enzymes [91], changes in hormone levels (i.e., [66,84,96]), etc.
Conversely, pure CYN did show binding affinity to the estrogen receptor [113]. 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 [114], changes in thyroid-hormone-related genes [116], oxidative stress [117], or alterations in hormone levels [114,115].
In any case, none of the studies reviewed, for either MCs or CYN, were performed following official OCDE guidelines, and the experimental designs and exposure scenarios have been very different. This makes it difficult to establish certainties regarding the ED activity of cyanotoxins. Moreover, there was a study conducted with MC-LR [3] and a study conducted on CYN [114] that reported non-monotonic responses, a phenomenon frequently described for endocrine disruptors, but without standardized approaches in a risk assessment context [118]. Therefore, and based on the scientific literature reviewed, the cyanotoxins MCs and CYN are potential endocrine disruptors, but further research is required, particularly for CYN, but also for MCs, considering: the use of OCDE guidelines; the use of advanced technologies such as transcriptomics and proteomics; combined exposures of cyanotoxins with other pollutants; and potential non-monotonic dose-response relationships.