|
Mechanisms of Small-RNA–Guided Gene Silencing in Mammalian Cells
Summary: Thomas Tuschl studies the role of double-stranded RNA in gene silencing, with the aim of understanding its biology as well as its application as a research tool and in treatment of genetic disease.
Biochemical and Reverse Genetic Analysis of RNA Silencing
Double-stranded RNA (dsRNA), an important regulator of gene expression in many eukaryotes, triggers different types of sequence-specific gene silencing that are collectively referred to as RNA silencing or RNA interference (RNAi). A key step in known silencing pathways is the processing of dsRNAs into short RNA duplexes of characteristic size and structure. These short RNAs guide RNA silencing by distinct mechanisms for different degrees of complementarity to their targets (Figure 1).
Recently, we and other laboratories identified a new class of small RNAs that is restricted to mammalian germline cells. These small RNAs interact with Piwi proteins, most of which are specifically expressed in germline cells, and they were therefore termed piRNAs (Figure 2). The piRNAs measure 26 to 31 nt and are 2'-O-methylated at their 3' ends. Sequence analysis of piRNAs indicates that they originate from long single-stranded, noncoding, nonconserved primary transcripts. The piRNAs carry a 5' phosphate and preferentially start with a 5' uridine, but otherwise they appear to be randomly excised from the long primary transcripts. Because Piwi proteins are essential for germ cell development, it is possible that piRNAs and Piwi proteins regulate the timing of meiotic and postmeiotic events through transcriptional and post-transcriptional repression.
We develop biochemical and cell-based assays to dissect the mechanism of small interfering RNA (siRNA)– and microRNA (miRNA)–guided mRNA degradation and translational repression. During this process, we identify new factors involved in RNA silencing and determine their specific function, such as assembly of silencing complexes or mRNA target recognition. We also examine factors required for nuclear maturation of miRNAs.
To address the function and biogenesis of piRNAs, we are establishing a human biochemical system. We are screening germline tumor cell lines for Piwi-protein and piRNA expression. We hope to find a suitable cell line that can be manipulated to provide sufficient material for biochemical analysis and to develop cell-based functional assays.
To understand the RNA-protein interaction network, we have developed biochemical approaches combined with deep sequencing to identify important regulatory RNA sequence segments. We have included in such analyses not only proteins associated with small-RNA–containing RNPs, but also other mRNA-binding proteins, such as FMRP, FXR1 and 2, IGF2BP 1ן, and several others. The ultimate aim is to generate genome-wide post-transcriptional regulatory maps and crack the "mRNP code" that defines mRNA localization, stability, and rate of translation.
First, we isolate mRNPs from various affinity-tagged cell lines and determine the enrichment of target mRNAs in immunoprecipitates through Affymetrix array analysis. If a reproducible mRNA enrichment is observed, we identify the targeted mRNA segments by parallel sequencing the RNA fragments residing in partially nuclease digested and immunopurified complexes. Finally, to assess whether alterations in post-transcriptional control contribute to explaining complex human genetic diseases, we are examining the frequency of sequence alterations in the identified mRNA and miRNA segments by parallel sequencing of large pools of patient and control DNA samples.
To dissect RNA-silencing pathways in living cells, we have established stable cell lines that carry reporter genes under the control of RNA-silencing processes. Our cell-based assays measure activation of reporter gene expression, which occurs upon disruption of the RNA-silencing machinery. In the past, disruption of RNA silencing was recorded as reduction of reporter gene expression, which could be complicated by secondary effects such as reduced cell viability or proliferation. Our positive readout systems are less prone to these types of artifacts and are also easier to analyze because of the absence of reporter background expression in untreated cells. These cell lines are used for high-throughput screening, in a collaboration with Hakim Djaballah (Memorial Sloan-Kettering Cancer Center), to identify small-molecule inhibitors of RNA-silencing pathways and in siRNA-based genome-wide screens to identify new candidate genes required for RNA silencing.
Small-RNA Profiling of Human Cell Types in Normal and Disease States and Analysis of miRNA Function
Approximately 100 of the 400 known mammalian miRNAs are recognized on average per cell or tissue type at a small-RNA cloning depth of several thousand clones. Recently, several links between the RNA-silencing factors and inherited or acquired genetic disorders were recognized. To understand the biology of RNA silencing, we identify all active dsRNA expression units in all cell types as a function of development and disease state, including viral infection. A subset of these miRNA precursor clusters show interesting tissue and cell-type specificity (Figure 3). To further clarify the expression patterns in specific tissues with heterogeneous cell types, we developed a highly sensitive and specific in situ hybridization method (Figure 4).
In collaboration with Mihaela Zavolan (University of Basel, Switzerland), Chris Sander (Memorial Sloan-Kettering Cancer Center), and James Russo (Columbia University Genome Center), we prepared, sequenced, and analyzed more than 250 small-RNA libraries. The distinct miRNA expression patterns have been collected in a miRNA gene expression atlas. We are evaluating whether dysregulation of miRNA expression contributes to disease states.
Our extensive efforts to sequence small RNAs have yielded an almost complete collection of mammalian miRNA sequences. To allow for miRNA profiling of large collections of patient material, we will be using miRNA microarrays and methods we are developing in our laboratory and in collaboration with Miltenyi Biotec. We developed a new solid-phase synthesis protocol for preadenylated mono- and polynucleotide fluorescent conjugates, which can be effectively ligated to the 3' end of miRNAs for labeling. We also prepared a synthetic miRNA reference standard that contains equimolar amounts of the predominantly cloned sequence of each nonredundant rodent, human, and human viral miRNA. This universal standard is used to correct for sequence-dependent variations in miRNA labeling and array hybridization efficiency. With this collection of molecular tools, we are in a strong position to evaluate the role of miRNAs in normal and pathological conditions.
In animals, miRNAs regulate many different biological processes, including cell-lineage specification, apoptosis, neuronal development, cholesterol metabolism, and hormonal secretion. The vast majority of the several hundred mammalian miRNAs have no known function.
We use several approaches to identify and validate targets of miRNAs. The first approach relies on genome-wide predictions of miRNA targets based on sequence similarity and evolutionary cross-species conservation (in collaboration with the Zavolan and Sander laboratories). As our mechanistic understanding of miRNA-guided repression and its sequence specificity improves, we will refine rules for the target predictions. To identify miRNA function and targets in mice, we inhibit specific miRNAs by injecting nuclease-resistant cholesterol-conjugated 2'-O-methyl antisense oligoribonucleotides (antagomirs) in collaboration with Markus Stoffel (Swiss Federal Institute of Technology [ETH], Zürich). Antagomirs are also used to block miRNAs expressed in cultured cells. Following miRNA inhibition, we examine cells or tissues by cDNA array analysis to reveal regulated mRNAs.
As an alternative to the bioinformatic-driven approach, we develop biochemical means to isolate miRNA–target mRNA complexes directly. Together with the determination of the spatial and temporal expression of miRNAs and their target mRNAs, these approaches will contribute to the elucidation of the biological function of miRNA regulatory networks.
Single-nucleotide changes as well as micro- and macrodeletions and/or insertions can abrogate, strengthen or weaken miRNA–target RNA interactions. As our understanding of miRNA regulation and function progresses, we will be able to initiate genotyping studies to examine the association of sequence variations in miRNA- and mRNA-targeting sites with heritable diseases or predisposition to diseases. The targeting sites for miRNAs are predominantly located outside the protein-coding region in the 3'-untranslated region of target mRNAs, a region that has received little attention in disease-related mutational analysis. Perturbation of the conserved network of miRNA and target mRNA interactions may offer new explanations concerning idiopathic diseases.
Development of RNA-Silencing Tools for Knockdown of Mammalian Gene Expression
Molecules that can specifically silence gene expression are powerful research tools. Our laboratory pioneered the application of siRNAs for silencing genes expressed in mammalian somatic cells. The robustness of this approach has motivated numerous biotechnology organizations and academic institutions to develop siRNA libraries for high-throughput genome-wide screening in mammalian cells. The largest remaining challenge for siRNA applications is the development of methods for effective delivery of siRNAs to all cell types and in vivo.
Last updated April 02, 2008
|
