Invited Perspective: Humanized Mice for Arsenic Metabolism—A Better Model for Investigating Arsenic-Induced Diseases?

In recent years, arsenic exposure has been linked to cardiovascular disease,¹ diabetes mellitus,² neurological dysfunction,³ and cancer, all of which may be exacerbated by early life exposure.^4–6 In a study published in this issue of Environmental Health Perspectives, Sethupathy et al. used the humanized AS3MT mouse model to address important questions in arsenic-promoted disease etiology and obtained interesting results indicating sex differences in transgenerational transmission of susceptibility to diabetes.⁷ In addition, Sethupathy et al. explored critical contemporary questions of the roles of transgenerational transmission and sex in enhancing susceptibility to chronic disease.⁷

Although these studies were conducted with an exposure that would be considered high for humans, the human equivalence is still uncertain if allometric scaling is used. Using U.S. Food and Drug Administration (FDA) guidelines, the $400\text{-ppb}$ exposure to the mice would be equivalent to a human exposure to $32\text{ppb}$ of arsenic in drinking water. This $32\text{-ppb}$ exposure would be below the threshold of human exposures that were found to promote diabetes in a human cohort cited by the authors.⁸ Although this exposure would suggest that arsenic is more diabetogenic in the humanized AS3MT mouse than in humans, it might also speak to the uncertainty of using allometric scaling to estimate dose equivalence.

Determining arsenic exposures that are equivalent between cultured human cells, mice, and humans has been the subject of much debate for decades. We proposed in a 2011 workshop the use of phenotypic markers to calibrate exposures and doses in the different systems,⁹ but the recommendations were not widely implemented. The difficulty in extrapolating a human arsenic exposure expressed as parts per billion in drinking water to a concentration to be used in an in vitro cell culture experiment is complicated by a paucity of data on human tissue and bodily fluid levels correlated with exposure levels. Too often, the design of cell culture experiments ignores the basic toxicological principles of absorption, distribution, metabolism, and elimination; instead, dosing is based on drinking water concentrations common for human exposures rather than the expected internal steady-state blood or tissue concentrations.

The difference in approaches to dose calculation is important because epidemiological studies indicate that the concentration of arsenic in the blood of exposed humans is $\sim 1/15\text{th}\text{to}1/10\text{th}$ that of the water they are drinking.^3,10–13 Thus, an exposure to $100\text{ppb}$ arsenic in drinking water results in a blood arsenic level of $\sim 90–130\text{nM}$. Furthermore, translating a drinking water concentration from human to mouse is complicated not only by the usual issues of body surface area or body mass conversions, but also by differences between mice and humans in volume of water consumed per day and arsenic distribution, metabolism, and excretion.¹⁴ The FDA recommends using allometric scaling to determine human equivalence of doses given to different species.¹⁵ Using their guidelines, a dose to a mouse should be divided by $\sim 12$ to give the human equivalent dose. However, this calculation is based on assumptions that are intended only to reduce uncertainty in approximating the equivalence in mice and humans.¹⁵

The AS3MT humanized mouse model, which was developed by researchers at the University of North Carolina at Chapel Hill,¹⁶ is a big step toward equalizing arsenic distribution, metabolism, and excretion between mice and humans and may be a great advance toward resolving mouse-to-human comparisons. However, there may still be species- and strain-specific differences in cellular arsenic transport mechanisms. Investigators must also consider the more nuanced potential impact on expression of protein-coding and long noncoding RNA genes adjacent to the insertion site and how they might influence the phenotype.

In Sethupathy et al.,⁷ and earlier work characterizing the humanized AS3MT mouse,^10,17 arsenic and metabolite excretion was measured to demonstrate the equivalence of arsenic metabolism in the mice. Measuring blood levels and demonstrating that the mice have the same ratio of drinking water to blood levels as humans would provide greater certainty that the mice are not a more sensitive species. Notwithstanding these limitations, the humanized AS3MT mouse model will be useful to many researchers investigating the wide array of diseases induced by chronic arsenic exposure.

Refers to https://doi.org/10.1289/EHP12785