In 2008, Kozul-Horvath and Hamilton found that mice fed a grain-based chow (Purina 5001) had no changes in gene expression when exposed to arsenic while they were dramatic when fed purified diet AIN-76A (Click here for pdf). They found out that "background" arsenic in Purina 5001 masked their ability to study gene expression changes in the liver and lung of mice exposed to low doses (10 and 100 ppb) of arsenic while they found clear changes with the AIN-76A. In fact, this chow was loaded with arsenic ~ 390 ppb, 4 to 40 fold higher than the levels they were studying. This initial work suggested that arsenic in this chow affected expression of genes in phase I and phase II enzymes involved in metabolism of xenobiotics in liver and also those of the lung involved in immune signaling. In their follow-up work to that study in 2009 (Click here for pdf), they used the AIN-76A to do more low dose arsenic work and again found that mice fed arsenic in water (10 - 100 ppb) had a strong effect on expression of genes and proteins important to the immune response and confirmed there is indeed an effect on the phenotype of these mice with the 100 ppb level of arsenic using an in vivo model of influenza A (H1N1 or Swine flu) (Click here for pdf). They next wanted to do a follow-up study which was published in 2012 (Click here for pdf) studying 10 ppb arsenic in water again to see whether there were any effects on development when exposing mice in utero, postnatally, or when exposed during both periods. They found dramatic and adverse effects on growth in offspring regardless of when exposure occurred which was likely due to reduced energy from dam's breast milk (i.e. lower fat content). Their interest in looking at lower doses of arsenic in their studies is based on the current recommendations of the EPA to reduce arsenic levels in water from 50 to 10 ppb. Please also find press articles below.
Their work has highlighted the importance of choosing a well-defined purified diet for examining how lower doses of potentially toxic substances like arsenic influence an animal's toxicological phenotype. Given the presence of heavy metals and other dietary constituents present in grain-based chows, this work provides further support for the importance of choosing the diet wisely in all in vivo research.