Introduction

Weight loss and exercise remain the cornerstones of diabetes prevention. Nevertheless, the paradigm that dietary over-consumption and physical inactivity are the single determinants of weight gain and diabetes risk has recently been challenged as overly simplistic. In 2002, Baillie-Hamilton postulated a role for chemical toxins in the etiology of obesity by showing that the obesity epidemic coincided with the marked increase of industrial chemicals in the environment over the past 40 years [1]. The global hypothesis put forth by Baillie-Hamilton was that environmental toxins contribute to obesity through dysregulated feeding behavior. To date, this hypothesis is unproven, but has sparked considerable interest in the role of pesticides, organic pollutants and environmental toxins in human health and disease. Indeed, a recently published post hoc analysis from the 2003–2004 U.S. National Health and Nutrition Examination Survey (NHANES) showed a 39% increase in incident type 2 diabetes for every 1 standard deviation increase in urinary concentration of the environmental toxin, bisphenol A (BPA) [2]. Importantly, the 4 other phenols measured did not correlate with diabetes incidence or any other disease. The authors concluded that BPA conferred particular risk for diabetes which is especially concerning given the widespread human exposure.

Proposed mechanisms linking BPA to the development of diabetes are numerous and include fetal programming [3], sex steroid imbalance [4], altered adipocyte biology [5], disrupted beta cell function [6] and oxidative stress [7]. Nevertheless, because BPA is hepatically metabolized, its toxicity in the liver appears particularly prominent [7]. Further, there is reason to speculate that BPA may be an environmental inhibitor of hepatic glucokinase, specifically. Glucokinase has been definitively shown as the body’s glucose sensor in beta cells, enteroendocrine cells of the gut, hypothalamus and liver. Retinoic acid [8], [9], biotin [10], [11], and sorbitol [12] have all been shown to alter the transcription, translation, or activity of glucokinase, but none of these factors have changed much in the American diet over the past 30 years, making them unlikely candidates as dietary regulators of diabetes risk. In contrast, BPA – new to the American diet over the past 40 or so years – is also known as an “endocrine disruptor” due its estrogenic effects [6], [13]–[15]. The latter is of particular interest as estrogen action may be antagonized, not synergized, by BPA specifically in the liver [16]–[18] where glucokinase has a key glucoregulatory role. Despite the known actions of estrogen on glucokinase, and of BPA on estrogen action, no study to date had examined BPA’s effect on glucokinase, hence was the aim of the current investigation.

Methods

Results

Acute BPA Exposure

Hepatic glucokinase kinetics for the BPA-treated and control mice are described in Table 1. Glucokinase activity was significantly lower at 5 mmol glucose for the BPA exposed animals (p = 0.029), as well as at 10 mmol/l (p = 0.002) and 20 mmol/l (p = 0.002) glucose. As depicted in Figure 1A, 2-hour BPA exposure largely abolished glucokinase activity over the range of physiologic glucose concentrations tested (overall treatment effect p<0.001) through a decline in the Vmax and an increase in the enzyme’s Hill coefficient.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Glucokinase specific activity at different glucose concentrations (range 0–20 mmol/l) in mice 2 hours after acute BPA exposure (A; overall treatment effect p<0.001) or after 2 weeks chronic BPA exposure (B; overall treatment effect p = 0.003).

https://doi.org/10.1371/journal.pone.0069991.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Glucokinase (GCK) kinetics in the control and BPA groups after acute and chronic BPA exposure.

https://doi.org/10.1371/journal.pone.0069991.t001

Discussion

The rise in human exposure to bisphenol A [2,2-bis(p-hydroxylphenyl)propane] (BPA) has largely mirrored the rise in incident type 2 diabetes. BPA is a chemical used widely in the production of epoxy resins and polycarbonate plastics. Because of its pervasive use in the lining of food and beverage containers, medical supplies and dental sealants, human exposure has been deemed virtually ubiquitous. Population biomonitoring studies estimate the current adult exposure to BPA ≤1 µg/kg/day [20], well below the purported “safe dose” of 50 µg/kg/day set by the Environmental Protection Agency in the 1980’s. Yet few assertions in medicine today are as controversial. A recent review of the >80 biomonitoring studies cast serious doubt on whether past estimates are accurate or whether the acceptable level of exposure ensures any degree of safety [21]. Because of BPA’s known anti-estrogenic effects [7] and the known effect of estrogen to stimulate glucokinase [22] – the body’s glucose sensor – we examined the direct effect of BPA on hepatic glucokinase activity. Major findings from this study demonstrate that both acute and chronic BPA exposure – at a dose deemed safe in humans –impair hepatic glucokinase activity and function. These results provide a potential link as to how BPA exposure may increase risk for diabetes and have alarming implications for public health.

To explore the association between BPA exposure and increased risk for diabetes, we examined the effect of acute ingestion of BPA on glucose sensing. Orally administered BPA is metabolized by glucuronidation during first pass metabolism in the liver, with its peak seen at 2 hours, a biological half-life of approximately 6 hours and nearly complete elimination in 24 hours [23]. Based on the above BPA absorption parameters established in humans, we administered the maximum “safe dose” (50 µg/kg) dissolved in water to mice expecting maximum biologic effects 1–3 hours post-ingestion. Results from the current study demonstrate nearly complete suppression in overall hepatic glucokinase activity and significant alteration in intrinsic enzyme kinetics in mice 2 hours after ingestion of BPA. Although human exposure is unlikely to occur at this level in a single dose, these data are the first to contend that this dose is not safe and has physiological ramifications relevant to the development of diabetes, at least in mice.

Due to possible exposure through dermal absorption, inhalation and ingestion, the precise pharmacokinetics of chronic BPA exposure are less clear [24]. What is clear is that BPA exposure is virtually ubiquitous and constant. A recent survey identified BPA in 63 of 105 food samples, including fresh turkey, canned green beans, and infant formula [25], suggesting that human exposure is frequent, but also at low levels, rather than in large boluses. Ad libitum water (spiked with 1.75 mmol BPA) intake by our mice over 14 days also demonstrated a significant negative effect on overall hepatic glucokinase activity and function, albeit less pronounced than the effect of the acute BPA exposure. Importantly, however, these experiments capture the influence of low and intermittent exposure of BPA on hepatic glucose sensing, likely more representative of its true effects in the human population. Furthermore, mathematical models [26] would contend that impaired glucose sensing to the degree observed in the 2-hour and 2-week experiments could lead to clinical diabetes [27].

Several limitations of the current study are worth noting in its interpretation. First, experiments were conducted exclusively in mice, hence results may not be reproducible in humans. Second, all mice were male, hence we cannot comment on the impact of BPA exposure in female mice. Lastly, the effect of BPA exposure on glucokinase activity was presumed through the action of BPA on the estrogen receptor, however, this was not specifically investigated in this brief report.

In conclusion, major findings from this study demonstrate that both acute and chronic BPA exposure – at a dose deemed safe in humans –impair hepatic glucokinase activity and function in mice. These results provide a plausible link between BPA exposure and risk for diabetes. Future research is needed to determine whether BPA impairs glucose sensing in humans and if it can be restored through BPA avoidance. Such research could have a significant impact on food packaging, dietary selections and legislation relevant to public health.