Bisphenols Come in Different Flavors: Is “S” Better Than “A”?

In this issue, Boucher et al (1) describe the stimulatory effects of bisphenol S (BPS) on the differentiation and lipid accumulation in human adipocytes. Human exposure to bisphenol A (BPA), and very recently to its close analog BPS, has been associated with widespread health concerns for over 2 decades. Consequently, this subject deserves a brief perspective with respect to the following: 1) the prevalence of the bisphenols in the environment and the extent of human exposure; 2) their metabolic activities and putative mechanisms of action; and 3) an overview of ongoing conflicts between research scientists, environmental groups, the chemical industry, and regulatory agencies in the battle over the recognition of bisphenols as hazardous endocrine-disrupting chemicals (EDCs).

BPA is a synthetic small molecule composed of 2 phenol groups. It is produced in very large quantities (over 5.4 million tons worldwide in 2015) in the manufacture of polycarbonate plastics and epoxy resins, both of which are made of repeating BPA monomers (Figure 1). BPS is an analog of BPA with a similar structure of 2 phenol groups on each side of a sulfonyl group (2). Polymers made of repeating BPS units are called polyethersulfone (PES). BPA-containing consumer products include plastic bottles, food utensils, lining of beverage and food cans, as well as dental cements, thermal receipt papers, iv medical devices, and water pipes (3). Although the carbonate linkages are rather stable and the polymers are chemically inert, BPA can leach out as a result of incomplete polymerization and/or because of some degradation of the polymer by elevated temperatures or acidic conditions (4, 5). Humans can be exposed to BPA through ingestion, skin absorption, inhalation, and through iv catheters. BPA at nanomolar concentrations has been measured in blood or urine of most individuals tested and has also been detected in the amniotic fluid and fetal plasma, indicating its passage across the placental during pregnancy, and underlying its potential to affect human fetal development (6). Given the worldwide use of polycarbonate plastics, epidemiological studies have had difficult times finding control groups that have not been exposed to BPA.

Figure 1.

Chemical structures of BPA and BPS and the corresponding polymers: polycarbonate and PES.

BPA was first synthesized in 1891, with the first evidence for its estrogenicity coming from studies with ovariectomized rats in the 1930s. A landmark study in 1993 found an estrogen-like compound in water autoclaved in polycarbonate bottles for use in tissue culture, which was subsequently identified as BPA (7). Although BPA-based plastics have been in commercial use since the 1950s, reports on its estrogen-like properties with an environmental impact began to appear only in the early 1990s (8). Figure 2 shows an exponential increase in publications on BPA in PubMed, reaching 3000 between 2011 and 2015 and totaling over 6000. In response to the increasing pressure by concerned scientists and environmental groups to ban the use of BPA in consumer products, the chemical industry began to introduce BPS as a “safe substitute” for BPA in the mid-2000s. Structural/chemical data show that PESs have excellent thermal, optical, and mechanical properties as plastics. However, soon thereafter, reports on EDC properties of BPS began to appear (Figure 2) (2), as well as initial findings on its detection in human urine (9).

Figure 2.

Number of publications (in PubMed) on BPA (left axis) vs BPS (right axis), per period from 1991 to the end of 2015.

By now, thousands of in vitro and in vivo studies have provided evidence that BPA is a prototypical EDC that affects reproduction, neural development, behavior, cardiovascular functions, and metabolism, as well as the promotion of several types of cancer. Extensive research has revealed that there is no single mechanism by which BPA acts as an estrogen agonist. Proposed targets include classical nuclear estrogen receptors (ERs) α and β, membrane-associated ERs, estrogen-responsive G protein-coupled receptor 30, and nuclear estrogen-related receptor-γ (10). Actions via other nuclear and nonnuclear receptors have also been suggested. Given that BPA has a significantly lower binding affinity than natural estrogens to most of these receptors, while often showing equipotency with estradiol, the issue of its mechanism of action is far from being settled. Some scientists even argue for existence of a specific BPA receptor, reminiscent of the discovery of opioid receptors and endogenous endorphins decades after the neurological disruptive effects of opiate drugs were recognized. Another confounding issue is the nonmonotonic (nonlinear) mode of BPA action, whereby an endocrine-like activity at low nanomolar doses is followed by a nadir, and then by toxic-like effects at micromolar doses (11). Moreover, there is an ongoing debate as to the clearance rate of BPA from the circulation, the formation of biologically active vs inactive metabolites, and the accumulation and retention of BPA in adipose tissue, given its lipophilic nature (12).

This brings us back to the article by Boucher et al (1) and the implications of an inadvertent exposure to BPS to the obesity epidemics. Obesity results from 2 complementary processes: hyperplasia, or increased recruitment of mature adipocytes from a pool of preadipocytes by the process of adipogenesis, and hypertrophy, or increased lipid accretion by each adipocyte. In this study, the authors incubated human sc preadipocytes with BPS at doses ranging from 0.1nM to 25μM. They found that BPS induced expression of several key adipogenic markers and augmented lipid accumulation. This is an important paper for several reasons. First, the authors used human adipocytes, unlike previous studies that used 3T3-L1 murine adipocytes (13). Second, BPS was used at a full range of doses, revealing a nonmonotonic action. Third, coincubation with the ER antagonist ICI-182780 and glucocorticoid receptor antagonist RU486 pointed towards BPS actions via ERs, although such effects were not uniform across all genes examined. Collectively, these data suggest a significant heterogeneity of BPS actions, and its potential interactions with peroxisome proliferator-activated receptor-γ, the master regulator of adipocyte differentiation. This study adds support to previous reports, using human adipose tissue and adipocytes, on adverse effects of BPA on the release of adipokines, such as adiponectin, IL-6, and TNFα, which exacerbate the metabolic syndrome (14, 15). However, the present study also leaves some unresolved questions. Among these is the relatively low potency of BPS at environmentally relevant low nanomolar doses and the need for a side-by-side comparison of the effects of BPS and BPA, to clarify whether or not they act in a similar fashion.

Given the overwhelming scientific evidence on the endocrine-disrupting properties of bisphenols, a relevant question is why it has been so difficult to reach regulatory decisions whether they pose hazards to human health. Undoubtedly, replacement of the ubiquitous bisphenols is a very costly endeavor to the chemical and polymer industries. Second, among the thousands of studies, some have used very large concentrations of bisphenols in vitro or in vivo, providing fodder to contrarians. Third, a legitimate concern has been lack of standardization among animal studies in terms of routes of administration, pharmacokinetics across species, differences between sexes, sensitive windows of exposure, and specific disease endpoints.

To address some of these concerns, 3 agencies: the National Institute of Environmental Health Sciences, the National Toxicology Program, and the Food and Drug Administration, joined forces in 2012 and formed a consortium, named CLARITY-BPA (16). Under the CLARITY consortium, thousands of male and female rats were housed and treated at the National Center for Toxicological Research at Jefferson, AR. Rats were given a daily gavage with one of 5 equally spaced BPA doses as well as 2 doses of estrogen as a reference. Treatments started at early pregnancy, continued during early neonatal life, and lasted up to 2 years of age. Twelve laboratories across the United States, who submitted competitive grant applications, were selected to cover different endpoints, including behavior, brain development, reproduction, cardiovascular, metabolism, and cancer. The university-based researchers were provided animals and/or tissues or serum samples according to their specifications from the National Center for Toxicological Research facilities and did the analysis in their own laboratories. Studies are ongoing and await peer-reviewed publications.

The real issue is that the industry is replacing a toxic chemical with another, yet untested chemical, which will require large investments of research funds to carry out applicable studies, such as those by Boucher et al (1). Notably, in addition to BPA and BPS, other bisphenols, including bisphenol F and BPA diglycidyl ether, also have endocrine-disrupting properties (12). Does all this mean that the vast amount of work already invested in characterizing BPA is now being disregarded by its replacement with BPS by the industry? Moreover, to what extent should regulators take all bisphenols into consideration upon making future decisions?