Trending ▼   ResFinder  

college 12

8 pages, 0 questions, 0 questions with responses, 0 total responses,    0    0
ishita
   		+Fave Message
ProfileTimelineUploads
 Home > ishita >

Formatting page ...

REVIEWS FUNCTIONAL GENOMICS TO NEW DRUG TARGETS Richard Kramer and Dalia Cohen Abstract | The completion of the sequencing of the human genome, and those of other organisms, is expected to lead to many potential new drug targets in various diseases, and it is predicted that novel therapeutic agents will be developed against such targets. The role of functional genomics in modern drug discovery is to prioritize these targets and to translate that knowledge into rational and reliable drug discovery. Here, we describe the field of functional genomics and review approaches that have been applied to drug discovery, including RNA profiling, proteomics, antisense and RNA interference, model organisms and high-throughput, genome-wide overexpression or knockdowns, and outline the future directions that are likely to yield new drug targets from genomics. Novartis Institutes for BioMedical Research, 250 Massachusetts Avenue, Cambridge, Massachusetts 02139, USA. Correspondence to D.C. e-mail: dalia.cohen@ pharma.novartis.com doi:10.1038/nrd1552 For the past several decades, drug discovery has focused primarily on a limited number of families of druggable genes against which medicinal chemists could readily develop compounds with a desired biochemical effect. These targets were usually exhaustively investigated with dozens or even hundreds of related publications often available before huge investments in discovery programmes began. Today, that comfort level is often missing with many genomics-based targets. Although the genomics approach will undoubtedly increase the probability of developing novel therapies, the limited knowledge available will (and has) increase(d) the risk and almost certainly the attrition rate for early-stage research projects. To effectively and competitively exploit the completed human genome sequence, the incorporation of technologies capable of identifying, validating and prioritizing thousands of genes to select the most promising as targets will be required. The estimated 35,000 genes in the human genome, as well as multiple splice variants of many mRNAs, mandates that these technologies must be higher in throughput than most current technologies, as it will be impossible to develop the traditional depth of knowledge about each target. Importantly, no single technology will be sufficient to generate all of the necessary information, and the integration of knowledge from several approaches is required to select the best new drug targets for drug development. NATURE REVIEWS | DRUG DISCOVERY This review describes the elements of a multipronged, high-throughput functional genomics approach, as implemented at the Novartis Institutes for BioMedical Research (NIBR), to bridge the gap between raw genomic information and viable drug targets, and presents examples of the use of these approaches for target discovery and validation (FIG. 1). Technological platforms that range from nucleotide chemistry and molecular and cell biology to proteomics, bioinformatics and model organisms are applied in parallel, if possible, to generate a multi-dimensional profile of each gene that will be used to identify the most promising new targets. Future challenges will be to integrate, analyse and interpret large sets of diverse data types to better understand the role of genes in biological pathways that are involved in various diseases and to select the best points of intervention. Gene-family mining and comparative genomics One way to exploit the information encoded in the human genome is to mine the sequence for identity and similarity of known gene families with a range of sequence-search methods. The priority in drug discovery is to mine for additional members of families that are already known to contain drug targets. The advantages of identifying extra members of such druggable gene families are twofold: the potential identification of novel drug targets, and the generation VOLUME 3 | NOVEMBER 2004 | 9 6 5 REVIEWS a Bioinformatics Genome collections RNA interference cDNA Expression profiles RNA Protein High-throughput cell-based assays (for example, cytokine induction) Promoterreporter gene Promoterendogenous gene Potential drug targets Verification/ validation Functional database Protein protein interactions Pathways Disease-relevant cells Model organisms Drosophila melanogaster Zebrafish Drug target b High-throughput screening (HTS) Compounds Drug discovery pipeline Lead optimization Early development support Clinical study Commercialization Disease models/toxicology Figure 1 | Approaches for target discovery and validation. Diagrammatic representation of the parallel, multi-pronged, high-throughput functional genomics approaches used at the Novartis Institutes for BioMedical Research for drug target discovery and validation as described in this review. The various technological platforms are represented in boxes and the arrows indicate the flow of data and information about genes leading to a drug target. Functional genomics activities are shown in panel a, and the conventional drug discovery pipeline (not covered in this review) in panel b. 966 | NOVEMBER 2004 | VOLUME 3 of knowledge of additional members of a gene family that can be used to help design drugs that are selective only for the target under consideration. All sequence-search methods use sequence similarity as a basis for identifying homologues. In general, it is relatively easy to identify homologues that share a sequence identity of 30% or greater using pairwise sequence-search methods such as BLAST (Basic Local Alignment Search Tool)1 and FASTA2. The success rate for identifying homologues with a sequence identity in the range of 20 30% is only approximately 50% when using such methods, and the success of the searches is much lower for identities of less than 20%. In order to identify these remote homologues, it is imperative to use search algorithms such as PSIBLAST (Position Specific Iterative BLAST)3, HMMER (Hidden Markov Model profiler)4 and SAM (Sequence Alignment Modeler)5. We have used Smith Waterman6, PSI-BLAST and HMM algorithms (HMMER, SAM) to detect close and remote homologues of gene families that are of specific interest in target discovery. As each method produces overlapping but non-identical results, all three algorithms were used and the results combined for maximum effectiveness. The databases that we searched include genomic sequence information and various protein translations predicted from proprietary and public sequence information. The list of hits obtained was then processed to remove sequences of low complexity and redundancy on the basis of criteria that were specifically developed for each gene family according to the sequence characteristics of the appropriate family members. A good example is the G-protein-coupled receptors (GPCRs), in which two-thirds of the members have just one exon7, making similarity searches more straightforward. We have identified several GPCRs using the methods described above. The ligand for one of them has subsequently been identified8, whereas several others are still ORPHAN RECEPTORS and are candidates for deorphanization. Although the availability of the human genome sequence has led to the identification of additional members of known gene families, the sequencing of the genomes of several more organisms has contributed to the prediction and understanding of gene function in biological pathways. Comparative genomics helps in the understanding of the biological context by looking at similarities and differences among organisms. Collating all known and predicted protein sequences of each genome and clustering them using different methods9 has several immediate applications in target discovery. For example, if sequences of micro-organisms are included in the analysis, clusters without human/ eukaryotic members should include potential antimicrobial targets for infectious diseases. Clusters with only human members indicate that the genes are unique to the human genome. There have been several publications speculating that approximately 40% of human genes have biological functions that are unknown and that there could be many disease targets in this pool of www.nature.com/reviews/drugdisc REVIEWS Human genome Comparative genomics: cluster sequences across multiple genomes Mining the human genome using iterative homology-based search methods Novel members of known gene families Novel gene families Identification of candidate genes Assessment of suitability as drug target using all available lines of evidence Figure 2 | Computational mining of the human genome. Information flow in two of the computational biology approaches used for drug target discovery. The candidate genes enter the functional genomics pipeline in FIG. 1 as potential drug targets . unknown genes10. The identification of families of genes from this unknown set would be of great scientific interest as well as considerable pharmaceutical importance. FIGURE 2 describes the use of gene-family mining and comparative genomics in the in silico identification of drug target candidates. RNA profiling in drug discovery ORPHAN RECEPTOR A receptor for which the ligand has not been identified. TWO-DIMENSIONAL GEL ELECTROPHORESIS A commonly used method for fractionating proteins, in which proteins are first separated (first dimension) on a polyacrylamide gel according to isoelectric point, then separated at a 90 angle (second dimension) on the basis of molecular mass. One of the highest-throughput methods for functionalizing the genome available today is RNA profiling using high-density arrays of DNA on glass. There are several methods available for RNA profiling (for a review, see REF. 11). Most of these consist of attaching DNA probe sequences to glass, labelling the RNA and hybridizing this to the DNA array. Today, the most commonly used methods attach as many as 500,000 short (25 60 mer) DNA sequences to a <2 cm2 glass surface. RNA extracted from cells or tissues is converted into cDNA and labelled, usually fluorescently, either as cDNA or as antisense RNA. The labelled target is hybridized to the probes and the label bound to each probe is determined. In this way, the expression level of more than 10,000 genes can be routinely determined in a single hybridization from as little as 10 ng RNA. Although the phenotype of a cell is largely determined by the expressed proteins and their interactions with each other and the environment, RNA profiling has several advantages. RNA profiling offers much greater throughput, generally greater coverage of the genome and can provide data from smaller samples. However, although many differences at the RNA level are reflected in differences in the protein level, there is not always a good correlation, and RNA levels cannot reflect protein modifications or interactions. Nevertheless, for new target discovery, the most common applications compare RNA levels in diseased tissues with those from normal tissues, either from humans or animal models or in tissues or cells that were drug NATURE REVIEWS | DRUG DISCOVERY treated. The initial data set from an RNA-profiling experiment typically has 200,000 400,000 data points and there are various tools available for handling these large data sets. The outcome of the application of any or all of these tools is a list of genes, typically tens to hundreds of genes, that are probably modulated in disease or by a drug and which therefore merit further characterization as potential drug targets or biomarkers. One area in which RNA profiling has proved extremely valuable is in the classification of cancers. Recent studies have shown that RNA profiling can subdivide the disease accurately and is informative about patient outcome in breast, brain and prostate cancers, lymphomas and leukaemias12 17. Studies have shown that RNA profiling of tumour-derived cell lines can reflect the identity of the source tissue18 and could be used to predict drug responsiveness19. Another area that has received significant attention for RNA profiling is neurobiology. One of the main problems has been gaining access to good sample material, particularly for diseases for which there are no good animal models, such as schizophrenia. Several studies of RNA profiling of gross brain samples20 25 have been published. These studies have identified novel drug targets to pursue for the development of new therapies for schizophrenia. Proteomics The genome determines its potential for gene and protein expression, but does not specify which proteins are expressed in the various types of cell in an organism or individual, to what level they are expressed or the extent of their post-translational modifications. Proteomics is used to determine differential protein expression, posttranslational modifications and alternative splicing and processed products. TWO-DIMENSIONAL GEL ELECTROPHORESIS is often used to fractionate the numerous proteins from a cell or tissue and to identify differentially expressed or modified proteins. This is followed by mass spectrometry to identify the individual protein spots of interest from the gels. This approach has lower throughput than RNA-expression profiling, but the resulting differentially expressed or modified proteins identified by proteomics represent the actual potential target or disease-associated molecules. A benefit of proteomic analyses studies is that they can simplify the sample being studied by examining only a fraction of the proteome, thereby focusing the study to address more specific enquiries. Generally, this is achieved by subcellular fractionation or by affinity methods such as so-called activity-based probes for binding substrates to a specific class of proteases26. Comparative image analyses of the two-dimensional PAGE protein patterns of multiple proteomes allow the quantitation of changes in protein expression. A powerful alternative to classic two-dimensional PAGE approaches is differential in-gel electrophoresis (DIGE). In this method, the two samples of proteins are differentially labelled with nonspecific Cy3 and Cy5 fluorescent dyes. The samples are mixed together, VOLUME 3 | NOVEMBER 2004 | 9 6 7 REVIEWS Box 1 | Proteomic identification of a molecular drug target In order to determine the biochemical effects33 of the cytostatic agent bengamide, twodimensional PAGE was used to analyse differences in H1299 cells following treatment with the natural products bengamide, bengamide E and a synthetic bengamide analogue, LAF389. Proteins that were reproducibly altered in mobility following bengamide treatment were selected for identification by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). Thorough analysis of one of the proteins, 14-3-3, on the basis of two separate proteolytic digests, followed by complete nanoelectrospray/ mass spectrometry-mass spectrometry (nano-ES/MS-MS) sequencing of the amino-terminal peptides, showed that LAF389 caused a change in processing of the amino terminus of 14-3-3. The normal form of the protein has its N-terminal methionine cleaved by methionine aminopeptidase (MetAp) and is further processed by acetylation of the proximal valine. In cells treated with LAF389, N-terminal processing is blocked and a new form of 14-3-3 emerges that retains its initiator methionine (FIG. 3). The two forms of 14-3-3 were resolved by charge on isoelectric focusing gels as, at physiological pH, the unprocessed, methionine-containing form is positively charged relative to the blocked, acetylated form. The change in 14-3-3 processing is an example of a cellular marker for pharmacological inhibition of MetAp activity, and this response has the potential to serve as a biomarker for in vivo responses to MetAp inhibition. WATSON CRICK BINDING The binding of one strand of a nucleic acid with another strand through the pairing of complementary bases (A T and G C) to form a doublehelix structure. 968 separated simultaneously on the same gel and imaged by overlaying the two colour channels, which allows direct comparisons within the same gel27. An alternative approach to gels that attempts to separate every protein involves proteolytic digestion of all proteins to generate peptide components before fractionation. Although high-performance liquid chromatography (HPLC) is a mature technology, bottlenecks might emerge in the daunting task of sorting the millions of peptides that are generated from a complex mixture of all proteins. This complex problem can be made more tractable by affinity selection for peptides that contain a specific amino acid or modification. This method commonly uses the isotope-coded affinity tags (ICAT) approach, which allows for both enrichment of cysteinecontaining peptides and mass-spectrometric-based quantitation through mixing two sets of proteins, one with an addition of a light tag (hydrogen labelled) sample and one with a heavy tag (deuterium-labelled) sample28. The relative difficulty in specifically detecting and identifying proteins compared with nucleic acids has caused protein-array technologies to lag behind those for detecting and measuring gene expression at the RNA level. The field of protein arrays can be roughly divided into three technologies that differ in how they work and, therefore, in what problems they can be applied to. One approach uses chips with various activated surfaces to capture target proteins from a crude lysate or other mix, and then identifies the captured proteins with mass spectrometry29. Another approach uses chips spotted with capture molecules, such as antibodies or compounds, for binding the target proteins, which are then detected by fluorescent probes that are either attached directly to the target proteins or directed to tags on the target proteins30. The third approach consists of the target proteins themselves spotted onto chips, either as individual protein spots or in lysates of cells or tissues. As the target, rather than the probe, is immobilized in this last approach, the chips are sometimes referred to as reverse phase arrays31,32. | NOVEMBER 2004 | VOLUME 3 Proteomics approaches can identify post-translational modifications that can provide an insight into a biological pathway or the action of biologically active compounds. Proteomics has been successfully used (BOX 1) to identify the modulation of post-translational modification induced by the cytostatic agent bengamide, which has led to the development of a biomarker and has provided indications towards the identification of the molecular target (FIG. 3) for this compound class33. Proteomics can also be used to identify protein protein interactions. High-throughput-interaction analysis has resulted in the mapping of networks in yeast34,35, worms36, fruit fly37 and mammals38,39. Oligonucleotides in drug discovery Successfully bringing a new medicine to the market depends on many factors, including having a compound with the right pharmacology and targeting the right molecular target. To confirm that the selected molecular target is an appropriate one for the disease, a potent, selective low-molecular-mass inhibitor compound must be identified either the drug itself or a lead compound that is subsequently modified to become the drug. This process is resource intensive and has a high rate of failure. The value of oligonucleotides in the target/drug discovery process is based on the premise that they can be rapidly and specifically used to simulate the biological and pharmacological effects of target inhibition by lowmolecular-mass compounds in cellular assays, in animal models of disease and even in humans. Inhibition is initiated by WATSON CRICK BINDING of the so-called antisense oligonucleotide to its target mRNA. Normal translation of the mRNA is subsequently prevented by one of several mechanisms, including induced degradation of the message, interference with the splicing process or a physical blocking of the translational machinery. Another approach uses single-stranded RNA known as a ribozyme with sequences complementary to the target flanking a sequence with RNA-cleaving catalytic activity. Several types of modified oligonucleotide with distinct properties have been used for antisense gene knockdown studies. Four main types of modification have been used methoxyethyl (MOE)40, locked nucleic acids (LNA)41, peptide nucleic acids (PNA)42 and morpholino43 depending on the application. Oligonucleotides partially modified with, for example, MOE and LNA groups induce an RNaseH-driven degradation of the target mRNA, whereas fully modified PNA and morpholino oligonucleotides have been used to cause translational arrest. Single-stranded antisense oligonucleotides are thought to operate both in the nucleus and the cytoplasm, and there are elegant examples of oligonucleotides targeted against splice sites of pre-mRNA to achieve changes in pre-mRNA splicing44. The use of chimeric oligonucleotides in the nucleus has been reported to downregulate specific splice variants, such as members of the interleukin receptor family45. Double-stranded RNA reagents that operate by the RNA interference (RNAi) mechanism, such as small interfering RNA (siRNA) and the vector-driven www.nature.com/reviews/drugdisc REVIEWS Acetyl Endogenous 14-3-3- 14-3-3- accumulates following treatment with bengamide E V M D V R D E R Q E Q L V L Q V Q K K Additional spot on two-dimensional gel 14-3-3 14-3-3- Ac-Val <-> Met-Val- .. amino terminus Methionine aminopeptidase inhibition Bengamide E target: methionine aminopeptidase Figure 3 | Novel drug target discovery by proteomics. As described in BOX 1, two-dimensional gel electrophoresis revealed proteins with altered mobility on treatment of cells with the cytostatic agent bengamide. Mass-spectrometry separation and sequencing of peptides showed that the differences resulted from altered amino-terminal modification, indicating that the target of bengamide is methionine aminopeptidase and that the change in the processing of the protein 14-3-3 could serve as a biomarker for in vivo responses to bengamide. expression of short hairpin RNA (shRNA), have recently become the predominant gene knockdown reagents46. They also induce an enzyme-driven degradation of mRNA by an as yet partially characterized ribonuclease complex known as the RNA-induced silencing complex (RISC). An important difference between the mechanistic action of siRNAs and antisense oligonucleotide is that these reagents are reported to operate in the cytoplasm only47. Intracellular expression of shRNAs that function by the RNAi mechanism are reported to show long-term target suppression48. A further advantage of RNAi is the possibility of delivering the reagents by viral vectors into cell types that have typically been difficult to transfect49. There are several notable examples in which oligonucleotides have been targeted against genes with already well-characterized roles in disease states and these examples have served as a proof of concept for the strategy50,51. An example of how oligonucleotide technologies were used to filter a large list of genes to highlight those of prime importance in a disease indication is that of chronic neuropathic pain. A list of potential targets was developed from RNA-expression profiling in several rat models of chronic pain; this was filtered, and priority was assigned to those genes that were upregulated, that had no previous association with pain, and those that were novel and came from protein families with a history of successful drug discovery. Even with such stringent filtering, the remaining genes were still too numerous to initiate drug discovery programmes without further target validation. To this end, a process was established by which oligonucleotides were used to explore the effects of downregulating the genes in the various animal models of neuropathic pain. P2X3, a purinergic receptor that is specific for sensory ganglia with reported links to pain, was selected for proof of concept. The potential of both antisense and siRNA to NATURE REVIEWS | DRUG DISCOVERY target P2X3 was carefully characterized in vitro for their ability to downregulate P2X3 mRNA and protein52. A functional assay established that P2X3 responds to ATP and forms ion channels after homodimerization or heterodimerization with P2X2, through which a current could be detected and shut off by gene inhibition. MOE-bearing oligonucleotides and siRNAs were synthesized and delivered into the animal models53,54 and showed that a dose-dependent, target-specific inhibition of P2X3 would inhibit pain in various rat models. The results demonstrated the potential value of inhibiting the function of this channel in pain indications and allowed a prioritization of P2X3. Since these results were obtained, low-molecular-mass inhibitors of P2X3 have indeed shown therapeutic relief in rat models of chronic pain55. Systematic analysis of gene function Genomics approaches, such as genome and cDNA sequencing, systematically determine gene structure, whereas gene-expression analyses allow comprehensive characterization and comparison of gene-expression behaviour. Such data can often associate a gene with a biological process but alone cannot define any essential role for the encoded protein in a disease. Analysis of gene function in mammalian cell culture is an ideal approach to identify regulators of diseaserelevant biological processes in a high-throughput and systematic manner. In many cases, molecular signaltransduction pathways and disease phenotypes can be accurately modelled in cell-culture systems. Most genes have now been identified to some level of accuracy, and so it should be possible to define the role of genes in cell-based phenotypic assays by systematic overexpression or inhibition of expression. Furthermore, by reiteratively testing defined genes in many phenotypic assays, it should be possible to build databases of gene function, analogous to those for gene expression and protein interactions, which would allow the prediction of disease relevance, specificity and potential efficacy of a given gene as a drug target. In this regard, several technological achievements have made genome-scale analyses of gene function in mammalian systems possible. First, efforts during the past few years have focused on producing full-length, fully sequenced cDNAs for all mouse and human genes56. Such cDNA collections can be used to examine gene function by systematic overexpression. In addition, recombination-based methods of cDNA cloning57 allow high-throughput, sequence-independent manipulation of open reading frames. Therefore, hundreds or thousands of cDNAs can be placed into appropriate expression vectors in an efficient, industrialized manner. Conversely, the use of RNAi approaches for mammalian systems using small oligonucleotides allows the production of gene-specific inhibitors in a reliable genome-scale manner. In the past year, a few reports have emerged that illustrate the tremendous potential for systematic screening of gene function in cell culture. All of these studies used overexpression of large numbers of VOLUME 3 | NOVEMBER 2004 | 9 6 9 REVIEWS Figure 4 | Optical sections of adult Drosophila melanogaster eyes. The eye on the left is wild type and the eye on the right is from a transgenic fly that expresses A . The fluorescent immunostaining shows A deposition in the photoreceptor layer (green). The outline of the photoreceptor cells is shown in red (actin stain). Progressive accumulation of A is indicated by the fact that A is detectable in the more mature posterior photoreceptors (bottom) but not in the less mature anterior photoreceptors (top). cDNAs using miniaturized cell-based assays. The premise behind such experiments is essentially the same for expression cloning. It assumes that overexpression of an exogenous cDNA that is missing or limited in a cell type and that uniquely induces a phenotype, such as receptor binding, viability or induction of specific gene products can be found. The main difference is that rather than testing large numbers of genes in large pools, individual genes are directly tested, a technique that is referred to here as the gene-by-gene approach. The first study of this kind demonstrated a high-throughput approach for determining the cellular localization of approximately 100 proteins of unknown function58. The ability of a large set of genes to regulate specific biological processes has been demonstrated in two studies. Michiels et al.59 reported screening 13,000 random, individual cDNAs in an adenoviral expression vector for the ability to induced various phenotypes in cell culture, including induction of osteoblast differentiation, loss of epithelial cell morphology and production of soluble factors that can induce endothelial tube formation. A similar approach was used to screen >150,000 randomly picked cDNAs in small cDNA pools for activation of a nuclear factor- B-encoding reporter gene60. Although neither of these studies demonstrated the biological relevance of any of the active cDNAs found in their screens, each study was effective at identifying many known regulators of the processes under study. 970 | NOVEMBER 2004 | VOLUME 3 The true promise of the gene-by-gene approach is the systematic and quantitative cataloguing of activity for specific genes, which requires screening of individual, defined cDNAs in a reiterative manner. Recently, two reports described the evaluation of approximately 8,000 fully sequenced cDNAs predicted to encode secreted proteins in several disease-relevant biological assays61,62. In both studies, recovered media from cells transfected with the individual cDNAs were tested for the presence of factors that were either immunomodulatory61 or that could modify several metabolic responses centred on glucose uptake, gluconeogenesis or insulin signalling62. In the second study, a novel role for the previously known protein bone morphogenetic protein-9 in glucose homeostasis was discovered. Finally, the gene-by-gene approach was used in an unbiased manner to screen >20,000 characterized cDNAs for the ability to activate various signal-transduction end points63. This study described a search for genes that can activate the promoter for the inflammatory mediator interleukin-8 (IL-8) using a curated set of predicted full-length human cDNAs. In addition, the same cDNAs were examined for their ability to activate expression driven by various specific response elements. A protein of previously unknown function that strongly induced the IL-8 promoter was identified. This protein, referred to as TORC1, was predicted to be a specific activator of the cAMP response element (CRE) on the basis of its activity in the battery of signal-transduction assays used to test the cDNA collection. Indeed, it was shown that TORC1 is a specific co-activator of CRE-binding protein-1 (CREB1) and that the IL-8 promoter contained a previously unrecognized non-canonical CRE. The same set of clones was also used to interrogate p53 and activator protein-1 responses64. Large-scale gene-knockdown studies using RNAi have been reported recently in mammalian systems65,66. The efficacy of transfected synthetic siRNAs and RNA duplexes expressed in situ using RNA polymerase III promoters has prompted several groups to begin producing and testing genome-scale sets of such genespecific inhibitors in high-throughput cell-based assays in mammalian cells. These results indicate that a phenocopy approach using RNAi in mammalian cells will also provide an efficient high-throughput approach for analysis of gene function. The use of model organisms Genetics and genomics research has led to the discovery of the widespread conservation of DNA and protein sequences, gene function and signalling pathways across diverse organisms. This conservation of biochemical and signalling pathways across evolutionary time allows the use of model organisms as surrogates for human patients because experimental results and knowledge obtained using these model systems can often be applied to understanding equivalent processes, pathways and mechanisms in the human situation. Two higher eukaryotic model organisms that are gaining a foothold in the pharma environment are the www.nature.com/reviews/drugdisc REVIEWS Box 2 | Genetic modifier screens in a model of Alzheimer s disease Drosophila melanogaster has been used to generate a model of Alzheimer s disease (AD) that can be used to address the mechanisms of secretion, toxicity and turnover of the amyloid peptide A 42, which has been implicated in AD. A 42 was ectopically expressed in the nervous system of flies, causing a visible phenotype77 (FIG. 4) that can be readily scored. This model has been used in genetic modifier screens in which suppressors or enhancers of the phenotype are identified as potential components of the A pathway. fruitfly D. melanogaster, and the zebrafish, Danio rerio. All share essential features that are necessary for a good genetic model system, including ease of culture, short generation time (two weeks and three months, respectively) and the production of large numbers of progeny. Equally important is the availability and continual improvement of methods for experimental and genetic manipulation; these methods allow researchers to carry out genome-scale genetic screens and analyse the identified genes and pathways at a level of sophistication that is not typically practical in other organisms. Many conserved signalling and developmental pathways were first discovered and characterized using this genetic approach, and similar analyses in mammalian systems have provided further evidence for the functional conservation of genes, pathways and processes between these model organisms and humans. Furthermore, the widespread use of model organisms in the research community, combined with data obtained using whole-genome gene-expression analyses and comparative genomics, has created a large increase in the amount of available data. Learning how best to use this knowledge base and how to make this information accessible and valuable to others in the research community is one of the key challenges facing those working with model organisms in the pharma environment today. Each model organism has specific advantages for drug discovery research. D. melanogaster in particular has been a mainstay of genetics and genomics research for almost 100 years, and, as a result, there is a large body of accumulated knowledge concerning the biology, anatomy, genetics and physiology of the Box 3 | Compound screening by mutation suppression in the zebrafish Peterson et al.76 demonstrated the use of zebrafish for whole-organism, small-molecule screening to identify compounds that reversed the phenotype of a genetic mutation. They used the zebrafish mutation gridlock (grl), which leads to disruption of aortic blood flow in a manner similar to that of the human disease aortic coarctation. They arrayed embryos in 96-well plates and treated them with 5,000 compounds from a structurally diverse chemical library. Two structurally related compounds were observed to restore circulation to the tail and cause resolution of the morphological defect in the aorta in a dose-dependent manner. Gene expression was analysed in fish treated with the more potent of the two compounds, and increased expression of vascular endothelial growth factor, which is important in the formation of the aorta, was seen. This study demonstrates that phenotype-based compound screening in zebrafish models has the potential for identifying drug leads. A challenge of this approach will be to further develop the leads with little or no knowledge of the molecular target. Technologies such as the one described in BOX 2 will become more important as the use of cell- and wholeorganism-based screens increases. NATURE REVIEWS | DRUG DISCOVERY fruitfly. There is an impressive array of well-developed research reagents and methods for D. melanogaster, including a large set of publicly available insertional mutations, site-specific recombination systems for clonal analysis, tissue- and stage-specific regulation of gene expression and gene replacement through homologous recombination. These advantages have led to the extensive use of D. melanogaster for genetic modifier screens, in which particular phenotypes are created using the available tools and then genetic modifiers of the phenotype are identified using chemical mutagenesis or available mutations. Such screens have been used to define and identify new components of various signalling pathways with clear disease associations, such as the insulin signalling pathway67. Recently, we have used D. melanogaster to identify new components in the A processing pathway68, which indicates that D. melanogaster might serve as a model for identifying new targets in the Alzheimer s disease pathway. BOX 2 describes an example of the use of genetic modifier screens in a fruitfly model of Alzheimer s disease. The zebrafish is a relatively new figure in drug development. The small vertebrate is a genetically tractable model organism that is also amenable to compound screening. Zebrafish embryos are initially transparent, they develop in simple salt solution and the primordial form of many of their organs appears within the first 24 hours of development. Although used mainly as a developmental genetics tool, the zebrafish offers several characteristics that make it attractive for the drug discovery process69. For example, drugs known to affect the QT interval in humans have also been shown, for the most part, to affect heart rate in zebrafish embryos70. Many drugs have been pulled from the clinic at a relatively late stage because of this problem71. Measuring heart rate as a means to predict QT-interval prolongation or torsade de pointes is a relatively simple assay and can be enhanced through automation. Similarly, angiogenesis inhibitors induce specific vascular defects in the patterning of the embryonic vasculature72. Embryonic vascular phenotypes could be used in both target validation and lead optimization phases of the development pipeline. Other areas of therapeutic interest that can be addressed using zebrafish include lipid metabolism, applications that involve phospholipase-dependent fluorescent dyes73 and drug abuse74,75. BOX 3 describes the use of zebrafish for whole-organism compound screening with small molecules76. Conclusions Modern pharmaceutical discovery is emerging as a new branch of science, thanks in large part to the technological advances that are allowing us to truly functionalize the genome. The investment made in sequencing the human (and other species ) genome was made in reaction to the promise that this information would revolutionize medicine. We have every reason to believe, given the tools at our disposal, that this promise will be fulfilled. VOLUME 3 | NOVEMBER 2004 | 9 7 1 REVIEWS 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 972 Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 215, 403 410 (1990). Pearson, W. R. Rapid and sensitive sequence comparison with FASTP and FASTA. Methods Enzymol. 183, 63 98 (1990). Altschul, S. F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389 3402 (1997). Eddy, S. R. HMMER: profile hidden Markov models for protein sequence analysis [online], <http://hmmer.wustl.edu/> (2001). Karplus, L. et al. What is the value added by human intervention in protein structure prediction? Proteins 45 (Suppl. 5) 86 91 (2001). Smith, T. F. & Waterman, M. S. Identification of common molecular subsequences, J. Mol. Biol. 147, 195 197 (1981). Lismaa, T. P. et al. G Protein-Coupled Receptors (Springer, New York, 1995). Jones, C. E. et al. Expression and characterization of 5-oxo-6E, 8Z, 11Z, 14Z-eicosatetraenoic acid receptor highly expressed on human eosinophils and neutrophils. Mol. Pharmacol. 63, 471 477 (2003). Tatusov, R. L., Koonin, E. V. & Lipman, D. J. A genomic perspective on protein families. Science 278, 631 637 (1997). Venter, J. C. et al. The sequence of the human genome. Science 291, 1304 1351 (2001). Scheel, J. et al. Yellow pages to the transcriptome. Pharmacogenomics 3, 791 807 (2002). Perou, C. M. et al. Distinctive gene expression patterns in human mammary epithelial cells and breast cancers. Proc. Natl Acad. Sci. USA 96, 9212 9217 (1999). Golub, T. R. et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286, 531 537 (1999). Ash, A. et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403, 503 511 (2000). References 12 14 were among the earliest works to look at gene-expression patterns as a way to classify cancers and to distinguish among phenotypically similar cancers. Such classification is often crucial in determining the outlook and most effective treatment for the cancer. Scott, L. et al. Prediction of central nervous system embryonal tumor outcome based on gene expression. Nature 415, 436 442 (2002). Yeoh, E. J. et al. Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic lukemia by gene expression profiling. Cancer Cell 1, 133 143 (2002). Singh, D. et al. Gene expression correlates of clinical prostate cancer behavior. Cancer Cell 1, 203 209 (2002). Ross, D. T. et al. Systematic variation in gene expression in patterns in human cancer cell lines. Nature Genet. 24, 227 235 (2000). Scherf, U. et al. A gene expression database for the molecular pharmacology of cancer. Nature Genet. 24, 236 244 (2000). Mirnics, K., Middleton, F. A., Marquez, A., Lewis, D. A. & Levitt, P. Molecular characterization of schizophrenia viewed by microarray analysis of gene expression in prefrontal cortex. Neuron 28, 53 67 (2000). Hakak, Y. et al. Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia. Proc. Natl Acad. Sci. USA 98, 4746 4751 (2001). Middleton, F. A., Mirnics, K., Pierri, J. N., Lewis, D. A. & Levitt, P. Gene expression profiling reveals alterations of specific metabolic pathways in schizophrenia. J. Neurosci. 22, 2718 2729 (2002). Vawter, P. M. et al. Microarray analysis of gene expression in the prefrontal cortex in schizophrenia: a preliminary study. Schizophr. Res. 58, 11 20 (2002). Hemby, S. E. et al. Gene expression profile for schizophrenia: discrete neuron transcription patterns in the entorhinal cortex. Arch. Gen. Psychiatry 59, 631 640 (2002). Buxton, F. et al. RNA profiling in the anterior cingulated in schizophrenia. (Abstracts of the 9th International Congress on Schizophrenia Research). Schizophr. Res. 60, 95 (2003). Saghatelian, A., Jessani, M., Joseph, A., Humphrey, M. & Cravatt, B. F. Activity-based probes for the proteomic profiling of metalloproteases. Proc. Natl Acad. Sci. USA 101, 10000 10005 (2004). Unlu, M., Morgan, M. E. & Minden, J. S. Difference gel electrophoresis: a single gel method for detecting changes in protein extracts. Electrophoresis 18, 2071 2077 (1997). Gygi, S. P. et al. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nature Biotechnol. 17, 994 999 (1999). Merchant, M. & Weinberger, S. R. Recent advancements in surface-enhanced laser desorption/ionization-time of flightmass spectrometry. Electrophoresis 6, 1164 1177 (2000). | NOVEMBER 2004 | VOLUME 3 30. Taunton, J., Hassig, C. A. & Schreiber, S. L. A mammalian histone deacetylase related to the yeast transcripitonal regulator Rpd3p. Science 272, 408 411 (1996). 31. Petricoin, E. F. et al. Use of proteomic patterns in serum to identify ovarian cancer. Lancet 359, 572 577 (2002). 32. Ramachandran, N. et al. Self-assembling protein microarrays. Science 305, 86 90 (2004). 33. Towbin, H. et al. Proteomics-based target identification: bengamides as a new class of methionine aminopeptidase inhibitors. J. Biol. Chem. 278, 52964 52971 (2003). 34. Gavin, A.-C. et al. Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141 147 (2002). 35. Ho, Y. et al. Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415, 180 183 (2002). 36. Li, S. et al. A map of the interactome network of the metazoan C. elegans. Science 303, 540 543 (2004). 37. Giot, L. et al. A protein interaction map of Drosophila melanogaster. Science 302, 1727 1736 (2003). 38. Colland, F. et al. Functional proteomic mapping of a human signaling pathway. Genome Res. 14, 1324 1332 (2004). 39. Bouwmeester, T. et al. A physical and functional map of the human TNF- /NF- B signal transduction pathway. Nature Cell Biol. 6, 97 105 (2004). 40. Dorn, G. et al. Specific inhibition of the rat ligand-gated ion channel P2X3 function via methoxyethoxy-modified phosphorothioated antisense oligonucleotides. Antisense Nucleic Acid Drug Dev. 11, 165 174 (2001). 41. Wahlestedt, C. et al. Potent and nontoxic antisense oligonucleotides containing locked nucleic acids. Proc. Natl Acad. Sci. USA 97, 5633 5638 (2000). 42. Good, L., Awasthi, S. K., Dryselius, R., Larsson, O. & Nielsen, P. E. Bactericidal antisense effects of peptide PNA conjugates. Nature Biotechnol. 19, 360 364 (2001). 43. McCaffrey, A. P., Meuse, L., Karimi, M., Contag, C. H. & Kay, M. A. A potent and specific morpholino antisense inhibitor of hepatitis C translation in mice. Hepatology 38, 503 508 (2003). 44. Mercatante, D., Mohler, J. & Kole, R. Cellular response to an antisense-mediated shift of bcl-x mRNA splicing and antineoplastic agents. J. Biol. Chem. 277, 49374 49382 (2002). 45. Karras, J. G., Maier, M. A., Lu, T., Watt, A. & Manohran, M. P. Peptide nucleic acids are potent modulators of endogenous pre-mRNA splicing of the murine interleukin-5 receptor- chain. Biochemistry 40, 7853 7859 (2001). 46. Hall, J. Unraveling the general properties of siRNAs: strength in numbers and lessons from the past. Nature Rev. Genet. 5, 552 227 (2004). 47. Kisielow, M., Kleiner, S., Magasawa, M., Faisal, A. & Nagamine, M. Isoform-specific knockdown and expression of adaptor protein shc A using small interfering RNA. Biochem. J. 363, 1 5 (2002). 48. Brummelkamp, T. R., Bernards, R. & Agami, R. A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550 553 (2002). 49. Rubinson, D. A. et al. A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nature Genet. 33, 401 406 (2003). 50. Rushworth, S. A. et al. Inhibition of porcine DC40 expression and function by in vitro transfection with antisense oligonucleotides. Transplantation 73, 635 642 (2002). 51. Rijcken, E. et al. ICAM-1 and VCAM-1 antisense oligonucleotides attenuate in vivo leukocyte adherence and inflammation in rat inflammatory bowel disease. Gut 51, 529 535 (2002). 52. Hemmings-Mieszczak, M., Dorn, G., Natt, F., Hall, J. & Wishart, W. Independent combinatorial effect of antisense oligonucleotides and RNAi-mediated specific inhibition of the recombinant rat P2X3 receptor. Nucleic Acids Res. 31, 2117 2126 (2003). 53. Barclay, J. et al. Functional downregulation of P2X3 receptor subunit in rat sensory neurons reveals a significant role in chronic neuropathic and inflammatory pain. J. Neurosci. 22, 8139 8147 (2002). 54. Dorn, G. et al. siRNA relieves chronic neuropathic pain. Nucleic Acids Res. 32, e49 (2004). References 52 and 54 demonstrated the use of antisense and small interfering RNA, respectively, in animal models of disease. 55. McGraughty, S. et al. Effects of A-317491, a novel and selective P2X3/P2X2/3 receptor antagonist, on neuropathic, inflammatory and chemogenic nociception following intrathecal and intraplantar administration. Br. J. Pharmacol. 140, 1381 1388 (2003). 56. Strausberg, R. L. et al. Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proc. Natl Acad. Sci. USA 99, 16899 16903 (2002). Describes the creation of the Mammalian Gene Collection, an important resource for functional genomics. 57. Walhout, A. J. et al. GATEWAY recombinational cloning: application to the cloning of large numbers of open reading 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. frames or ORFeomes. Methods Enzymol. 328, 575 592 (2000). Simpson, J. C., Wellenreuther, R., Poustka, A., Pepperkok, R. & Wiemann, S. Systematic subcellular localization of novel proteins identified by large scale cDNA sequencing. EMBO Rep. 3, 287 292 (2003). Michiels, F. et al. Arrayed adenoviral expression libraries for functional screening. Nature Biotechnol. 20, 1154 1157 (2002). Matsuda, A. et al. Large-scale identification and characterization of human genes that activate NF- B and MAPK signaling pathways. Oncogene 22, 2307 2318 (2003). Fiscella, M. et al. TIP, a T-cell factor identified using highthroughput screening increases survival in a graft-versus-host disease model. Nature Biotechnol. 21, 302 307 (2003). Chen, C. et al. An integrated functional genomics screening program reveals a role for BMP-9 in glucose homeostasis. Nature Biotechnol. 21, 294 301 (2003). Iourgenko, V. et al. Identification of family of cAMP response element-binding protein co-activators by genome-scale functional analysis in mammalian cells. Proc. Natl Acad. Sci. USA 100, 12147 12152 (2003). Describes one of the first high-throughput genomic screens using overexpression from a full-length cDNA collection. Chanda, S. K. et al. Genome-scale functional profiling of the mammalian AP-1 signaling pathway. Proc. Natl Acad. Sci. USA 100, 12153 12158 (2003). Paddison, P. J. et al. A resource for large-scale RNAinterference-based screens in mammals. Nature 428, 427 431 (2004). Berns, K. et al. A large-scale RNAi screen in human cells identifies new components of the p53 pathway. Nature 428, 431 437 (2004). References 65 and 66 demonstrated that large-scale RNA interference screens are of potential use for gene analysis and discovery. Li, H. & Garza, D. in Model Organisms in Drug Discovery (eds Carroll, P. M. & Fitzgerald, K.) 81 117 (Wiley, Chichester, 2003). Konsolaki, M. & Cohen, D. Targets for Alzheimer s disease: lessons learnt from flies. Drug Discov. Today 3, 64 70 (2004). Langheinrich, U. Zebrafish: a new model on the pharmaceutical catwalk. Bioessays 25, 904 912 (2003). Milan, D. J., Peterson, T. A., Ruskin, J. N., Peterson, R. T. & MacRae, C. A. Drugs that induce repolarization abnormalities cause bradycardia in zebrafish. Circulation 107, 1355 1358 (2003). Fermini, B. & Fossa, A. A. The impact of drug-induced QT interval prolongation on drug discovery and development. Nature Rev. Drug Discov. 2, 439 447 (2003). Chan, J., Bayliss, P. E., Wood, J. M. & Roberts, T. M. Dissection of angiogenic signaling in zebrafish using a chemical genetic approach. Cancer Cell 1, 257 267 (2002). Farber, S. A. et al. Genetic analysis of digestive physiology using fluorescent phospholipids reporters. Science 292, 1385 1388 (2001). Darland, T. & Dowling, J. E. Behavioral screening for cocaine sensitivity in mutagenized zebrafish. Proc. Natl Acad. Sci. USA 98, 11691 11696 (2001). Lockwood, B., Bjerke, S., Kobayashi, K. & Guo, S. Acute effects of alcohol on larval zebrafish: a genetic systems for large-scale screening. Pharmaco. Biochem. Behav. 77, 647 654 (2004). Peterson, R. T. et al. Chemical suppression of a genetic mutation in a zebrafish model of aortic coarctation. Nature Biotechnol. 22, 595 599 (2004). Finelli, A., Kelkar, A., Song, H., Yang, H. & Konsolaki, M. A model for Alzheimer s A 42-induced toxicity in Drosophila melanogaster. Mol. Cell. Neurosci. 26, 365 375 (2004). Demonstrates the use of a genetically tractable model organism to screen for novel elements in disease-related pathways. Acknowledgements The authors would like to thank F. Steele, N. Nanguneri, Q. Ma, F. Buxton, J. van Oostrum, J. Hall, M. Labow, D. Garza, M. Konsolaki and F. Serluca for their contributions to this review. Competing interests statement The authors declare competing financial interests: see Web version for details. Online links DATABASES The following terms in this article are linked online to: Entrez: http://www.ncbi.nih.gov/Entrez/ AP1 | BPM9 | CREB1 | grl | IL-8 | P2X3 | P2X3 | p53 | TORC1 OMIM: http://www.ncbi.nlm.nih.gov/Omim/ Alzheimer s disease Access to this interactive links box is free online. www.nature.com/reviews/drugdisc

Formatting page ...

Formatting page ...

Formatting page ...

Formatting page ...

Formatting page ...

Formatting page ...

Formatting page ...

 

  Print intermediate debugging step

Show debugging info


 

 


© 2010 - 2026 ResPaper. Terms of ServiceContact Us Advertise with us

 

ishita chat