Targeted gene mutation approaches to the study of anxiety-like behavior in mice

Abstract

Studying the behavioral phenotypes of transgenic and gene knockout mice is a powerful means to better understand the pathophysiology of neuropsychiatric disorders and ultimately improve their treatment. This paper provides an overview of the methods and findings of studies that have tested for anxiety-related behavioral phenotypes in gene mutant mice. In the context of improving the side effect burden of benzodiazepines, gene targeting has been valuable for dissociating the functional roles (i.e., anxiolytic, sedative, amnestic) of individual GABA_A receptor subunits. Supporting the link between abnormalities in CRH function and anxiety, CRH overexpressing transgenic mice and CRH-R2 receptor knockout mutants have displayed significantly increased anxiety-like behavior, while CRH-R1 receptor knockout mice have shown an anxiolytic-like phenotype. Consistent with an important role for the serotonergic system in anxiety, 5-HT1_A receptor deficient mice have consistently exhibited heightened anxiety-like behavior, while the evidence from 5-HT1_B and 5-HT2_C deficient mice remains somewhat equivocal. Mutant mice lacking either of the monoamine degrading enzymes, MAOA or COMT, have shown a number of behavioral and neurological effects, including alterations in anxiety-like behavior. With enhanced spatial and temporal control over gene mutations, in combination with an improved battery of behavioral tests, gene mutant mice will provide an increasingly valuable tool for understanding the neural substrates of anxiety.

Introduction

The emergence of molecular techniques that allow the alteration of genes and gene function in the intact animal is fashioning a new era in behavioral neuroscience. Protocols used in the generation of transgenic and gene knockout mice are now established as routine, and their widespread availability is producing proliferating numbers of gene mutant mice with direct relevance to the study of neural processes [1]. Fortunately for the behavioral neuroscientist, the mouse represents a good subject for studying the effects of genetic manipulations on behavior. Once generated, a mutation can be maintained in numbers of mice large enough to facilitate sound behavioral studies, with relatively little expense. Moreover, while there has been a long history of using the rat in behavioral neuroscience, many laboratories now have behavioral tests that are specifically designed, or can be successfully adapted, for use with mice.

The technology behind the generation of transgenic and gene knockout mice has been extensively described elsewhere, and the reader is referred to one of many excellent sources for a fuller discussion [2], [3], [4]. A transgenic mouse is often generated in order to study the functional consequences of gene overexpression. Briefly, this is achieved by microinjecting foreign DNA containing copies of a given gene into developing mouse embryos, where they have a chance to integrate into the host genome. The overall level of transgene expression is largely determined by the location at which the transgene inserts into the genome, which is random. However, patterns of expression can be directed to specific areas of the brain by using a region or cell specific promoter. With a gene knockout mouse, one can study the effects of removing a gene. First a DNA construct is designed that causes a functional disruption in the gene of interest. The DNA is then integrated into pluripotent embryonic stem cells (via homologous recombination) and inserted into foster embryos. Progeny are examined to see whether they have incorporated the null mutation (e.g., via coat color), and those chimeric offspring carrying the mutation in germline cells (thereby allowing the mutation to be transferred across generations) can be interbred to produce mice that are heterozygous for the missing gene. Assuming that the absence of the gene does not impact prenatal survival, interbreeding heterozygote offspring will normally give a complement of heterozygous null mutant, homozygous null mutant and ‘wild type’ littermate controls, in the ratio of 2:1:1.

Where a genetic alteration does impact prenatal survival, studying the behavioral consequences of that mutation is, of course, impossible. The ‘classical’ methods for generating gene knockout and transgenic mice produce genetic mutations in all cells where the gene is expressed, throughout ontogeny and adulthood. Therefore, even in viable animals, the presence of the mutation during development can complicate interpretation of behavioral phenotypes (observed or absent) in a mutant mouse due to indirect or multiple effects of the gene mutation. Thus, additional gene product in a transgenic mouse may lead to a cascade of molecular and neurochemical effects, confusing the causal link between phenotype and targeted gene. Conversely, where a gene has been inactivated, developmental adaptations may compensate for the deletion, thereby masking that gene's normal function. On one level, developmental effects can provide insight into the plasticity of neural systems [5]. Moreover, rendering a genetic alteration that is both global and chronic is a powerful means to model genetic contributions to a neurological or a neuropsychiatric disorder, when that is the goal. However, these same factors become undesirable when chronic, global expression impacts survival, or when studying the function of a gene in the normal adult brain [6]. In this context, techniques which allow for greater control over the temporal and spatial characteristics of a gene mutation represent an important advance [7].

One technique which produces regional restriction of a gene deletion works by flanking the gene of interest with loxP sequences that act as recognition sites for the bacteriophage enzyme Cre recombinase [8]. Intercrossing animals carrying the loxP flanking regions with Cre-expressing transgenic mice will result in excision of the ‘floxed’ site containing the target gene. Therefore, when Cre is driven by a promoter that is specific for a given region, inactivation of the gene is restricted to that region. In order to add temporal control over the genetic mutation, Cre expression can be linked to a regulating system, such as tetracycline or interferon [9]. The ability to determine where and when a gene is inactivated or overexpressed is already providing important insights into the link between genes and learning and memory [10], [11], and it is hoped that it will have a similar impact for studying anxiety-related processes. Of course, understanding the effects of increasingly sophisticated genetic manipulations necessitates sounds methods for behavioral phenotyping. The next section provides and introduction and overview of some of the behavioral methods used to assess anxiety-related phenotypes in gene mutant mice.