Laboratory diet profoundly alters gene expression and confounds genomic analysis in mouse liver and lung

Abstract

Nutritional studies in laboratory animals have long shown that various dietary components can contribute to altered gene expression and metabolism, but diet alone has not been considered in whole animal genomic studies. In this study, global gene expression changes in mice fed either a non-purified chow or a purified diet were investigated and background metal levels in the two diets were measured by ICP-MS. C57BL/6J mice were raised for 5 weeks on either the cereal-based, non-purified LRD-5001 diet or the purified, casein-based AIN-76A diet, as part of a larger study examining the effects of low dose arsenic (As) in the diet or drinking water. Affymetrix Mouse Whole Genome 430 2.0 microarrays were used to assess gene expression changes in the liver and lung. Microarray analysis revealed that animals fed the LRD-5001 diet displayed a significantly higher hepatic expression of Phase I and II metabolism genes as well as other metabolic genes. The LRD-5001 diet masked the As-induced gene expression changes that were clearly seen in the animals fed the AIN-76A diet when each dietary group was exposed to 100 ppb As in drinking water. Trace metal analysis revealed that the LRD-5001 diet contained a mixture of inorganic and organic As at a total concentration of 390 ppb, while the AIN-76A diet contained approximately 20 ppb. These findings indicate that the use of non-purified diets may profoundly alter observable patterns of change induced by arsenic and, likely, by other experimental treatments, particularly, altering gene and protein expression.

Introduction

Nutritional studies in laboratory animals have long shown that various dietary components can substantially influence expression of individual genes and proteins and also alter metabolism, but to date the effects of diet alone have not been carefully considered in whole animal genomic studies. Gene arrays are a powerful tool for examining global patterns of gene expression and can greatly aid in understanding underlying mechanisms of action, such as pharmacological or toxicological investigations of the effects of drugs and toxicants in vivo [1]. However, many other variables, which are often not controlled for in experimental designs, can also influence gene expression, confounding the effects of the experimental treatment. A number of these factors, including housing, handling, diurnal cycles, euthanasia, and necropsy procedures, have previously been shown to alter gene expression patterns [2], [3].

Laboratory animal diets fall into three main categories: cereal-based (non-purified), purified, and chemically defined [4]. This study compared the global effects of a purified (AIN-76A) and non-purified diet (LRD-5001) on gene expression using whole genome microarrays. Recent insights have indicated the importance of a carefully controlled laboratory diet, as a number of publications have reported that variability in estrogenic activity in lab diets, particularly from isoflavones, can greatly affect experimental results [5], [6], [7], [8], while others have reported the variable contamination of chow with toxicants of interest [9], [10]. In addition, considerable lot-to-lot variability has also been reported for normal constituents [8], [11]. Laboratory diets can be either an open formula, in which the exact composition is readily available to consumers, or a closed formula, in which the exact composition is known only by the manufacturer [12]. In many closed diets, the protein, fat, and carbohydrate sources vary from lot-to-lot, as do the proportions of these ingredients, based on product availability and cost.

The AIN-76A diet is a purified diet introduced in 1977 by the American Institute of Nutrition (AIN) for use with experimental rodents. The major source of protein is casein and the diet contains purified and refined sources of minerals and vitamins in standard proportions with minimal variability from lot-to-lot [4], [13]. While the importance of laboratory diet has become clear, the type of diet in not commonly reported in the published toxicogenomics literature or is listed as standard lab chow, and many animal housing facilities continue to use such cereal-based, non-purified, non-refined diets. One of the more commonly used is the Purina Laboratory Rodent diet 5001 (LRD-5001). The LRD-5001 chow is a closed formula consisting of variable sources of protein, one of which is fish meal (a probable source of arsenic and other contaminants) and the ingredients and proportions in this diet vary according to raw material availability and price. It was recently reported that this diet contained varying and often high concentrations of methylmercury, likely due to the fishmeal, which was shown to have an impact on experimental results [9].

The goal of in vivo gene array studies is to reveal gene expression patterns associated with a specific treatment variable. However, a diet with high and variable background levels of the toxicant under investigation can potentially skew or mask results. In addition, it is well known that specific dietary components can be potent inducers or inhibitors of specific gene or protein expression or inhibit certain metabolic enzymes. The composition of the LRD-5001 diet and results of previous studies led us to examine the background levels of metals in this chow, particularly since our laboratory investigates the mechanism underlying the toxicity of arsenic (As) and other toxic metals of concern. Arsenic is classified as a human carcinogen by the U.S. EPA and the World Health Organization [14], [15], and chronic ingestion has been associated with increased risk of a number of serious diseases including various cancers, type II diabetes, cardiovascular disease and reproductive and developmental abnormalities [15], [16], [17], [18], [19], [20]. In numerous published studies, As has been shown to interfere with a wide variety of physiological processes, including DNA repair [21], [22], [23], endocrine signaling [24], [25], [26], [27], and immune function [28], [29]. This study was conducted as part of a larger experiment investigating altered patterns of gene expression in the liver and lung in response to low dose As exposure administered through the diet or in drinking water. As part of this larger study, two groups of animals were fed either the non-purified LRD-5001 lab chow or the purified AIN-76A diet. Sub-groups from each diet background were exposed to As at 10 or 100 ppb in the drinking water (ad lib) or 10 ppb added to each diet for 5 weeks. This study revealed a significant alteration in the gene expression pattern in the livers and lungs of the mice as a direct result of diet. Moreover, while the animals on the purified diet showed a striking and consistent pattern of altered gene expression as a result of As exposure, these effects were largely masked by the non-purified diet, due to the dominant pattern of effects from the diet itself on many of these same pathways.