Interfering with the brain: Use of RNA interference for understanding the pathophysiology of psychiatric and neurological disorders

Abstract

Psychiatric and neurological disorders are among the most complex, poorly understood, and debilitating diseases in medicine. The burgeoning advances in functional genomic technologies have led to the identification of a vast number of novel genes that are potentially implicated in the pathophysiology of such disorders. However, many of these candidate genes have not yet been functionalized and require validation in vivo. Traditionally, abrogating gene function is one of the primary means of examining the physiological significance of a given gene product. Several methods have been developed for gene ablation or knockdown, however, with limited levels of success. The recent discovery of RNA interference (RNAi), as a highly efficient method for gene knockdown, has been one of the major breakthroughs in molecular medicine. In vivo application of RNAi is further demonstrating the promise of this technology. Recent efforts have focused on applying RNAi-based knockdown to understand the genes implicated in neuropsychiatric disorders. However, the greatest challenge with this approach is translating the success of RNAi from mammalian cell cultures to the brain in animal models of disease and, subsequently, in patients. In this review, we describe the various methods that are being developed to deliver RNAi into the brain for down-regulating gene expression and subsequent phenotyping of genes in vivo. We illustrate the utility of various approaches with a few successful examples and also discuss the potential benefits and pitfalls associated with the use of each delivery approach. Appropriate tailoring of tools that deliver RNAi in the brain may not only aid our understanding of the complex pathophysiology of neuropsychiatric disorders, but may also serve as a valuable therapy for disorders, where there is an immense unmet medical need.

Introduction

Perowne: “Best hope now, apparently, is RNA interference.” Baxter: “Yeah. Gene silencing. One day, perhaps…” From “Saturday” by Ian McEwan, Jonathan Cape Press, 2005.

Since its recent discovery, first in the worm, Caenorhabditis elegans (Fire et al., 1998), and subsequently in mammalian cells (Bitko & Barik, 2001, Caplen et al., 2001, Elbashir et al., 2001a), RNA interference (RNAi) has emerged as one of the most important innovations in modern molecular medicine. This is reflected by the fact that the discovery of small RNAs was Science magazine's “Scientific Breakthrough of the Year” in 2002 (Couzin, 2002). A number of biotechnology companies have already been established, with strategies specifically dedicated to develop RNAi-based medicines in order to cure diseases that have been traditionally difficult to approach (Pollack, 2004). This also led the Fortune magazine to label RNAi as “Biotech's Billion Dollar Breakthrough” in 2003 (Stipp, 2003). The ascent of RNAi is further revealed in that it has already permeated the realms of modern literary fiction, as in Ian McEwan's latest novel. In this review, we focus on the potential utility of RNAi to impact our understanding of the brain function. It is hoped that RNAi-based in vivo methods will improve the target validation process for neuropsychiatric disorders. Further, we describe how such efforts may lead to the development of better therapeutic strategies for psychiatric and neurodegenerative diseases where there is an immense unmet medical need.

Post-genomic biomedical research has paved our way from mere clinical observations of neuropsychiatric disorders to modern microarray strategies that help uncover disease mechanisms and detect potentially target genes relevant to the disease pathology (Mirnics et al., 2000, Dickey et al., 2003, McClung & Nestler, 2003, Newton et al., 2003, Wong & Licinio, 2004, Yao et al., 2004). Microarray-based genomic analyses are further complimented by proteomic platforms that enable the identification of proteins with altered expression in a disease state (Denslow et al., 2003, Rohlff & Hollis, 2003, Freeman & Hemby, 2004, Kim et al., 2004b, Vercauteren et al., 2004, Williams et al., 2004, Zhang & Goodlett, 2004). Consequently, the high-throughput gene and protein expression profiling permit the systematic identification of a large number of targets, or so-called “hits”, with pathophysiological implications or relevance. Targets identified from the use of such high-throughput technologies warrant an in vivo validation of their etiological role in the disease.

In vivo validation of targets is currently accomplished using tools that employ either pharmacological, immunological, or genetic means of manipulating the target function (Fig. 1). Pharmacological tools directly activate (e.g., agonists, positive allosteric modulators) or inhibit (e.g., antagonists, negative allosteric modulators) the protein product of the gene and provide a rapid means of in vivo target validation. Similarly, antibodies serve as immunological tools that readily inhibit the target protein to validate its function in vivo (Salfeld, 2004). However, the rapidity of these tools in revealing a target-based phenotype is largely superseded by both the difficulty and time-consuming process of finding small molecule compounds or generating antibodies that are selective for the target protein. Too often, we lack highly selective and potent ligands for even well-known targets. Another disadvantage of the pharmacological approach for initial target validation, following gene microarray studies, is that the gene product must be identified and reasonably well characterized. This is not always the case with genetic studies, where many of the hits are genes, which have not been functionalized or are not full-length sequences but just expressed sequence tags (ESTs).

The genetic approach of producing target mutants offers an alternative far selective to the pharmacological or immunological means of target validation in vivo. It includes genetic manipulation by homologous recombination in the embryonic stem (ES) cells, such that the gene of interest is either overexpressed or knocked out to produce stable transgenic animals (Capecchi, 1989, Koller & Smithies, 1992, Shastry, 1998). However, a major disadvantage to the use of this methodology is the restricted feasibility of homologous recombination to rodent ES cells, which is further compounded by the phenotypic differences observed among different strains following mutation of the same gene (Crawley, 1996, Gerlai, 1996, Lathe, 1996, Crabbe et al., 1999, Nadeau, 2001). In addition to the enormous amount of time involved in generating stably expressing mutant offsprings, this approach is not applicable to the phenotyping of genes that are critically involved in the early developmental stages of life. Furthermore, the most significant and commonly encountered drawback of producing mutant animals is the genetic compensation as well as developmental adaptations that may mask the establishment of a clear phenotype. Generation of inducible transgenics evades several limitations associated with the use of traditional mutant animals (van der Neut, 1997, Lewandoski, 2001, Misra & Duncan, 2002, Williams et al., 2003, Beglopoulos & Shen, 2004). Nonetheless, this process still remains time and labor intensive, which provides further impetus for research efforts into the development of other more efficient ways of manipulating gene expression during adulthood.

Manipulating the gene expression in somatic systems circumvents some of the problems encountered with transgenic mutants. In this case, gene knockdown is achieved using antisense oligodeoxynucleotides (Robinson et al., 1997, Weiss et al., 1997, Szklarczyk & Kaczmarek, 1999, Godfray & Estibeiro, 2003) or ribozymes (Welch et al., 1998, Cairns et al., 2002); however, at unpredictable frequencies and occasional acquaintance of non-specific events. Clearly, the shortcomings of existing methodologies demand an effective alternative that enables a selective and rapid validation of novel targets in vivo. More recently, the endogenous regulatory mechanism of RNAi has emerged as a potentially superior alternative to the traditional approaches for target validation in adult animals.