Toxicology is a multidisciplinary science in which the adverse effects of toxicants on living organisms are determined both qualitatively and quantitatively. It involves identifying, classifying, and characterizing physical, biological, and chemical substances and their effects on organisms. Toxicology includes determining how an organism could be exposed to an agent, and how the agent will enter the organism, (e.g., through dermal absorption, inhalation, or ingestion), as well as cataloging the adverse effects of exposure to various concentrations of the agent. Toxicologists determine whether effects are at the genetic, cellular, tissue, or organ level, and whether effects are seen in only specific subpopulations of organisms.

Bisphenol A (BPA) is a synthetically produced industrial chemical (formula C15H16O2; Figure 1A) that has been widely used commercially in epoxy resins and polycarbonate plastics since the 1950s. BPA has been used as a component of plastic storage containers of all types, plastic bottles, and food and beverage packaging including the lining of aluminum cans. In addition, BPA is found in thermal paper such as receipts and tickets (Geens et al. 2012 and Figure 1B). Consequently, there are multiple routes for BPA exposure in humans. The majority of exposures are via ingestion, while the next most likely exposure is by dermal absorption (Ma et al. 2019). The highest risk of exposure is through consumption of canned food, where BPA leaches from the can lining, thereby contaminating the interior product (Jalal et al. 2018).

Bisphenol A (BPA) and items that contain it. (A) The chemical structure of BPA. (B) A selection of consumer products that contain BPA and/or its structural analogs. From left to right: a baby bottle, a plastic bottle, an aluminum can, and a thermal paper-based boarding pass.

BPA toxicology has been investigated due to its classification as an endocrine-disrupting chemical, as it interferes with normal hormone function. BPA is composed of two connected phenols (Figure 1A) and is structurally similar to diethylstilbestrol, a synthetic form of the steroid hormone estrogen that is known to cause birth defects and cancer (Reed and Fenton 2013). When naturally produced, estrogen promotes development of secondary sex characteristics in biological females, ensuring proper ovulation, and establishing and maintaining successful pregnancies. The mechanism by which BPA binds to human estrogen receptors has been identified, providing evidence for how BPA interferes with normal sexual function (Li et al. 2015).

Although BPA has been detected in ∼90% of individuals, in most cohort studies, BPA levels have been at or below the limits of 50 µg/kg body weight/day set by the US Environmental Protection Agency and the European Food Safety Authority (Jalal et al. 2018). Several sources of evidence, however, indicate that BPA can contribute to human disease, including cancers, neurobehavioral disorders, and infertility, even at very low doses (Vogel 2009). It is especially concerning that the cohort found to have the highest exposure to BPA is young children (Ma et al. 2019). For children, their increased BPA exposure is likely due to their higher food intake per pound of body mass and potential metabolic differences compared to adults (Braun and Hauser 2011). Additionally, high levels of BPA exposure have also been reported in pregnant women (Ma et al. 2019). Although this is presumed to result from increased dietary exposure to contaminated food (Gorecki et al. 2017; Pacyga et al. 2019), the link between diet and higher urinary BPA concentrations during pregnancy is still being investigated .

BPA’s ability to leach from polycarbonate sources and its effects were initially suggested in the early 1990s. The estrogenic effect of leached BPA was first suggested by Krishnan et al. (1993) when studying estrogen in yeast cells. They discovered that BPA had leached from an autoclaved flask into the culture media used for yeast. Once in the medium, BPA acted as an agonist on artificially introduced mammalian estrogen receptors in the yeast cells, suggesting BPA mimicked estrogen. By the mid-1990s, leaching of BPA into consumer goods intended for human consumption had been detected (Ben-Jonathan and Steinmetz 1998).

In 2003, two back-to-back publications showed that, on its own, BPA could function as a weak estrogen and disrupt normal reproductive processes in mammals. Howdeshell et al. (2003) reported that damage to mice cages and bottles made from polycarbonate and polysulfone led to the leaching of BPA, resulting in oral and dermal absorption of BPA by the caged mice. Corroborating the earlier findings in yeast, the leached BPA produced estrogenic activity in the exposed female mice. At the same time, Hunt et al. (2003) reported a similar scenario in their mouse populations. They uncovered a role for BPA in the disruption of meiosis, leading to infertility. Their initial observations were not from a planned experiment, rather they had observed loss of fertility in their mouse population for ∼5 years and were perplexed by this problem. Unexpectedly, the eggs of even their wild-type mice had a dramatic spike in chromosomal abnormalities, indicating severe defects in meiosis. To test their hypothesis that the meiosis abnormalities were due to caging material, Hunt and colleagues set up an experiment with three groups of cages and bottles: (1) new (i.e., nondamaged), (2) previously used/damaged, and (3) newly damaged. They found that mice using damaged cages and bottles showed an increased rate of defects in meiotic chromosome structure and number when compared to animals in nondamaged cages and bottles. In comparison, no defects were observed in animals given glass bottles.

Nearly 20 years later, it is now known that BPA interferes with reproduction by mechanisms that affect cellular and genomic integrity. For example, BPA-induced genotoxicity, the ability of chemical agents to damage DNA, is associated with the formation of multiple DNA lesions. These include adducts (chemical modification of DNA), and single- and double-strand DNA breaks (Jalal et al. 2018). At the molecular level, it is known that BPA acts as an endocrine disruptor via binding to at least five distinct nuclear receptors, including the estrogen receptor (Li et al. 2015). Strikingly, despite having a binding affinity ∼1000 times lower than that of estradiol, BPA is still able to produce estrogenic effects.

Finally, mounting evidence indicates that meiotic defects may also derive from BPA substitutes. Since the 2003 report by Hunt et al. that BPA disrupts chromosome, kinetochore and spindle alignment, Yang et al. (2020) found that chromosome and spindle-related defects occur in mouse oocytes treated with bisphenol F, a structural analog of BPA. Similarly, Chen et al. (2016) showed that exposure to bisphenol S alters germline function in C. elegans. At the cellular level, like BPA, bisphenols F and S cause meiotic defects in a dose-dependent manner.

There have been many studies linking endocrine-disrupting chemicals to developmental and reproductive disorders in humans (Marques-Pinto and Carvalho 2013). In the years following the 2003 reports on BPA’s effects in mice (Howdeshell et al. 2003; Hunt et al. 2003), the potential for BPA to limit human reproduction became a topic of considerable attention. In one study, human oocytes derived from patients undergoing in vitro fertilization (IVF) were cultured in petri dishes, treated with BPA, and observed to exhibit a dose-dependent increase in meiotic arrest (Machtinger et al. 2013). At doses equivalent to the amount of BPA found in the fluid surrounding the ovum in the ovarian follicle, human oocytes exhibited defects in chromosome alignment and spindle formation in prophase and metaphase I of meiosis, respectively (Machtinger et al. 2013). Also, Ehrlich et al. (2012) found a positive association between the levels of BPA in urine and increased odds of implantation failure in women undergoing IVF treatments.

Findings from preclinical and clinical studies have now revealed ways in which prenatal, perinatal, and postnatal exposure to BPA affects human females (Pivonello et al. 2020). For example, BPA has been shown to accumulate in maternal blood, urine, amniotic fluid, placental tissue, and follicular fluid (Benachour and Aris 2009). Multiple lines of evidence have now uncovered potential mechanisms by which BPA accumulation interferes with fertilization and causes adverse pregnancy and birth outcomes (Cantonwine et al. 2013). For women undergoing reproductive therapies such as IVF, accumulation of BPA in follicular fluid was positively associated with a higher number of degenerated oocytes (Poormoosavi et al. 2019). Other studies have reported an association between BPA exposure and a decrease in the number of oocytes produced, as well as a reduction in the ability of oocytes to mature and be fertilized (Peretz et al. 2014). These effects are particularly concerning given that females are born with a finite number of eggs whose quality normally sharply declines at ages >35 years. (Nagaoka et al. 2012).

It is important to note that BPA-induced reproductive toxicity is not limited to females (Cariati et al. 2019). Thus far, fewer clinical studies have been conducted on the rate of male infertility due to BPA exposure. Nonetheless, studies have linked BPA exposure and decreased semen quality via effects on sperm concentration, total count, and vitality, which lessen the chance that sperm will successfully reach and fertilize a viable egg (Li et al. 2011; Rahman et al. 2015). Additionally, it is thought that BPA can inhibit fertilization by downregulating fertility-related proteins in the spermatozoa, such as the cytoskeletal component actin (Rahman et al. 2015), a protein required for sperm::egg interactions (Brener et al. 2003). Another study found that BPA exposure in adult rats leads to a decrease in follicle-stimulating hormone (Sadowski et al. 2014), a hormone that promotes the production of sperm and the growth and release of eggs.

Scientists studying toxicity work with compounds that have the potential to cause harm. Thus, experiments with human subjects presents significant ethical, as well as practical, limitations. To study the effects of these compounds, biomedical researchers have adopted the use of model organisms. Because of their similarities to human development and physiology, the predominant models for studying developmental and reproductive toxicity are rodents (Martin et al. 2009). Of the nonmammalian models, the nematode Caenorhabditis elegans is a popular choice as it is inexpensive, requires minimal upkeep, and shares multiple developmental and reproductive pathways with humans (Williams et al. 2017). Unlike rodent models, C. elegans has a life cycle of only 3–4 days from embryo to reproductive adult, and an enormous reproductive capacity, producing over 500 offspring in a life span when mated (Kimble and Ward 1988). Interestingly, most C. elegans are hermaphroditic. Thus, a single worm can produce both egg and sperm, and it can produce ∼300 viable offspring without mating.

When selecting a model organism to study reproductive toxicity, it is important that its genes and biochemical pathways are relatively well conserved so that findings may be applicable to humans. For example, Hornos Carneiro et al. (2020) showed that a compound called coenzyme Q10 (CoQ10) partially reversed DNA damage resulting from oxidative stress-induced repair defects caused by BPA exposure in C. elegans. C. elegans is useful for identifying drug-based therapies such as CoQ10 because the majority of the genetic pathways found in worms, such as those controlling DNA repair in meiosis, operate similarly in humans (Kaletta and Hengartner 2006). Remarkably, >83% of the >15,000 protein-coding sequences in C. elegans are also found in humans (Lai 2000). This subset accounts for an estimated 42% of genes linked to human diseases (Baumeister and Ge 2002). Because the genes and biochemical pathways controlling DNA repair are conserved between C. elegans and mammals, it is reasonable to hypothesize that in humans, CoQ10 may also reverse BPA reproductive toxicity caused by oxidative damage to DNA.

It is also advantageous for a model organism to be genetically tractable, that is, when it is easy to change, add, or delete a gene or multiple genes within it. For example, the strain of C. elegans tested by Hornos Carneiro et al. was col-121, which has a mutation in a collagen gene known to increase cuticle permeability and thus hypersensitivity to chemicals (Watanabe et al. 2005). Their use of the col-121 mutant allowed for lower, but relevant, doses of BPA to be tested.

Lastly, for researchers studying fertility, the transparent cuticle of C. elegans allows scientists to study meiosis using a standard dissecting light microscope (Corsi et al. 2015). The gonads of an adult worm are quite large with respect the animal’s size. When dissected onto glass slides, researchers can easily examine the entire reproductive system of a single animal. Furthermore, the nuclei in the C. elegans gonad are ordered in a spatiotemporal gradient, meaning every stage of sperm or egg production can be visualized and quantified (Figure 2).

Spatiotemporal organization of C. elegans germ line with respect to DNA double-stranded break (DSB) repair in the presence and absence of DMSO, BPA, and/or CoQ10. The graphic represents an adult C. elegans hermaphrodite containing one bilobed gonad (in purple), organized as a syncytium (i.e., nuclei sharing a common cytoplasm). Nuclei at the distal end (i.e., the proliferative zone) are undergoing mitosis and then enter meiotic prophase I. Other stages of prophase I are labeled in the top gonad, separated by blue dashed lines. In the transition zone (TZ, which corresponds to the leptotene/zygotene stages), DSBs start to form on all chromosomes. The peak of DSB formation happens in mid-pachytene, when homologous chromosomes recombine. By diplotene/diakinesis, all DSBs are repaired, and the paired, homologous chromosomes become condensed prior to being segregated. The inset is a graphical representation of bivalent chromosomes in late pachytene from worms exposed to DMSO only (left), BPA and DMSO (middle), and BPA and CoQ10 (right). During meiosis, the protein RAD-51, represented by dark blue circles, self-assembles on DNA following DSBs. For simplicity, the synaptonemal complex between paired homologs is not shown.