Tumor suppressor genes and proto-oncogenes play critical roles in cell cycle regulation, apoptosis, and cell senescence. Moreover, p16, p53, and p21 have important functions in the DNA-damage repair pathways which are among the most frequently compromised pathways in pathological conditions such as tumor growth (57). In this sense, most cancers have inactivating mutations in one or more proteins that normally function to restrict progression through the G₁ stage of the cell cycle (e.g.: p16), and in proteins such as p53 that normally function at crucial cell-cycle checkpoints, stopping the cycle if a previous step has occurred incorrectly or if DNA has been damaged (58). In fact, the hypermethylation of p16 and p53 promoter regions is an epigenetic pattern frequently observed in human cancer development, and this condition is generally associated with reduced methylation level of global genomic DNA (59).

The effect of glyphosate on DNA methylation was first reported by Kwiatkowska et al. (33) in vitro ( Table 1 ). These authors showed that high concentrations of glyphosate (from 84.54 to 1690 μg/ml) induce DNA lesions in peripheral blood mononuclear cells (PBMCs), decreased global 5mC percentage, and increased methylation of p53 promoter. Recently, similar results were reported, even at lower doses of glyphosate and AMPA (100–1,000 times lesser), in PBMC cells (34, 60) ( Table 1 ). Importantly, they found that the hypermethylation of p16, p53, and p21 genes was able to downregulate their mRNA expression and activate proto-oncogenes, which could lead to genomic alterations, downstream function dysregulation, and cancer development risk. Supporting these possible effects, it was later reported by Santovito et al. (61) that human lymphocytes exposed to lower glyphosate concentrations (0.025–0.500 μg/ml) increased the frequency of chromosomal aberration and micronuclei. Later, Woźniak et al. (36) found that glyphosate changes the expression of DNMT1, DMNT3A, and histone deacetylase (HDAC) 3, while AMPA changes the expression of DNMT1 and HDAC3 in PBMCs. These enzymes are involved in the regulation of chromatin architecture and, thus, could affect methylation patterns and histone modification, leading to changes in gene expression. On the other hand, Duforestel et al. (35) found that glyphosate triggered a significant reduction in DNA methylation and increased ten-eleven translocation (TET) 3 activity in MCF10A cells. TET enzymes oxidize 5-methylcytosines and reverse methylation. Combining glyphosate with enhanced expression of miRNA 182-5p (associated with breast cancer) induced tumor development in mice, suggesting that DNA hypomethylation occurring via the TET pathway primes cells for oncogenic response in the presence of another potential risk factor (35).

Although controversies have grown about the carcinogenicity and toxicity consequence of glyphosate, effects have been shown on skin cancer promotion in mice and proliferation of human breast cells (62, 63). Taking into account the role of this herbicide as a “probable human carcinogen”, it would be interesting to analyze if the epigenetic changes of tumor suppressor genes observed in vitro could be replicated in in vivo models and if these molecular alterations could explain, at least in part, some of the adverse effects produce by glyphosate.

Estrogens, a class of steroid hormones, regulate the growth, development, and physiology of the human reproductive system (64). They are produced principally by the gonads and placenta, but have multiple physiological functions on target organs such as the uterus, hypothalamus, pituitary, bone, mammary tissue, and liver (65). Estrogen signaling is mainly mediated through the classic nuclear receptor, estrogen receptor (ER) α. The expression of ERα occurs through different promoters depending on the tissue and physiological or developmental stages. In rats, five promoters have been described that result in transcripts with different 5′ untranslated regions derived from exons OS, ON, O, OT, and E1 (66). Importantly, the loss of expression, which is frequently observed in breast cancer, and the presence of triple negative tumors are often associated with hypermethylation of ER regulatory regions (37).

Several works report how the exposure to glyphosate alter the ER expression in vivo (67, 68) and in vitro (69–72), although the regulation mechanisms involved are still under study. Recently, Gomez et al. (37) and Lorenz et al. (38) found that developmental exposure to GBH (Filial 0, F0) from gestational day (GD) 9 until weaning induces epigenetic changes in ERα of F1 rats ( Table 1 ). The first group reported that standard diet supplemented with a GBH in two doses, 3.50 and 350 mg of glyphosate/kg bw/day, decreased the expression of OS, O, OT, and E1 transcripts in male mammary glands at postnatal day (PND) 60. These changes were accompanied by an increased in the DNA methylation of their promoter regions (73). Along the same line, Lorenz et al. (38) found, in a similar model, that perinatal exposure to GBH in a dose of 350 mg of glyphosate/kg bw/day, upregulated the expression of ERα mRNA in the pregnant rat uterus (F1) at GD5 (preimplantation period). This change was associated with an increase in the abundance of the O transcript variant and a decrease in DNA methylation of its promoter. Supporting these transcriptional changes, histone H4 acetylation and H3K9me3 were enriched in the O promoter in GBH-exposed rats, whereas H3K27me3 was decreased.

These studies proposed that the adverse observed effects of GBH on mammary gland growth (37, 73–75), embryo implantation (76), uterine development and reproduction (67, 68, 77), could be mediated, at least in part, by aberrant DNA methylation and histone acetylation/methylation of ERα gene. In this sense, the disruption of ER was previously related to male and female outcomes, including infertility, abnormal uterine and sperm maturation, atypical ovarian functions, and implantation deficits (78). These results require particular attention since all the epigenetic alterations mentioned above were observed after the GBH exposure has ended, suggesting that this exposure during sensitive periods of development (gestation) could perturb epigenetic programming and could have a long-lasting impact later in life. So far, it is necessary to clarify whether these effects could be due to the active principle (glyphosate), the co-formulants, or a combination of both, since previous studies have shown that commercial formulations are more toxic than glyphosate alone (79). In addition, it would be interesting to consider different administration routes, timings of exposure, and time points as conditioning factors.

Epigenetic transgenerational inheritance is a non-genetic form of inheritance that allows environmental factors to produce epigenetic alterations in the germline (sperm or egg) at critical periods of development that could be passed to subsequent generations, leading to pathologies or phenotypic variation in the absence of continued direct exposures (41). DNA methylation reprogramming could occur in the early embryo following fertilization (80), in the primordial germ cells in early gonadal development (81) or, even, during adult spermatogenesis in the testis (1).

Recently, Deepika Kubsad et al. (41) studied the transgenerational effect of glyphosate exposure (25 mg/kg bw/day) on pregnant rats (F0) during GD8 to 14. They found that this exposure produced in F1, F2, and F3 differential DNA methylation regions (DMRs) in the sperm ( Table 1 ). DMR associated gene categories were mainly related to transcription, signaling, metabolism, receptors, and cytokines and include metabolic and cancer pathways. Negligible pathology was observed in the F0 and F1 generations, while a significant increase in prostate, kidney and ovarian diseases, obesity, and parturition (birth) abnormalities was observed in the F2 generation grand-offspring and F3 generation great-grand-offspring. Tumor development was also monitored in males and females and found to increase in the F2 generation glyphosate female lineage; the most predominant were mammary adenomas. In another work from the same group, sperm from F3 generation was studied for DMRs and differential histone retention sites (DHRs), that were correlated with known pathology specific-associated genes (42) ( Table 1 ). Interestingly, overlapping sets of DMRs and DHRs were identified that were common for all the pathologies. These results support previous works that found adverse effects related to fetal parameters and structural congenital anomalies after perinatal exposure of GBH (F0) in second-generation of rats (F2) (76, 82, 83). However, Deepika Kubsad et al. (41) and Maamar et al. (42) reported for the first time that transgenerational inheritance of disease in rodents could be produced by DMRs and DHRs in the male germline, and these sites could potentially act as a biomarkers for specific diseases. Based on these findings, further studies are needed to deepen on the generational toxicology of glyphosate, in the disease etiology of the future generations.

The effect of GBH on miRNAs was recently reported by Hua ji et al. (39) who studied the effect of glyphosate exposure (1% Roundup; equivalent to 50 mg/kg bw/day) during pregnancy and lactation (GD14 to PND7) in the offspring brain ( Table 1 ). A miRNA microarray detected 55 upregulated (i.e: miR-711, miR-27b-3p, miR-142a-3p) and 19 downregulated (i.e: miR-34b-5p) miRNAs in the prefrontal cortex of mice at PND28 after maternal exposure. In addition, they reported abnormalities of the Wnt/β-catenin and Notch pathways in these animals that correspond with the dysregulation found in miRNA patterns. This support previous works that showed a disruption of Wnt proteins by neonatal exposure to GBH (2 mg/kg bw/day) from PND1 to PND7, in rat uterus (PND21) and implantation sites (GD9) (68, 77). In addition, exposure to glyphosate also produced a downregulation of these pathways in neuron cultures (84). In a second work and using the same model of exposure, these authors also found that circular RNA (circRNA) profile was significantly altered in the hippocampus of perinatal glyphosate exposure group (40) ( Table 1 ). circRNAs are a special class of non-coding RNAs which may interact with miRNAs to regulate gene expression. The altered miRNA and circRNAs were related to biological functions, including neurogenesis, neuron differentiation, brain development, stress-associated steroid metabolism pathways, among others.

Some of the miRNAs and their target genes disrupted by glyphosate exposure were also reported to be involved in pathological conditions, such as neurological disorders, prostate cancer and breast cancer (40, 85–88). Particularly, miR-34b-5p affects Numbl and Notch1 genes, which are involved in the Notch signaling pathway. In addition, the 3′-untranslated regions of β-catenin and Lef-1, which are involved in the Wnt signaling, contain miR-34 binding sites and are sensitive to miR-34b-dependent regulation. Abnormal activation of the Wnt/β-catenin or Notch pathways may serve an important role in the pathogenesis of various reproductive outcomes, including preeclampsia (85), embryo implantation (86), endometriosis (87), and ovarian tumors (88). Taking all together, these findings provide a new basis for identifying the mechanism of action of glyphosate-induced neurotoxicity in the developing brain and could serve as a beginning for elucidating the more general mechanisms of GBH toxicity in human and animal models. In addition, more investigations are needed to clarify the interaction between circRNAs, miRNAs, and genes as possible target of glyphosate exposure.