1 Nature Reviews Drug Discovery 2007 Vol: 6(11):891-903. DOI: 10.1038/nrd2410

Target deconvolution strategies in drug discovery

Recognition of some of the limitations of target-based drug discovery has recently led to the renaissance of a more holistic approach in which complex biological systems are investigated for phenotypic changes upon exposure to small molecules. The subsequent identification of the molecular targets that underlie an observed phenotypic response — termed target deconvolution — is an important aspect of current drug discovery, as knowledge of the molecular targets will greatly aid drug development. Here, the broad panel of experimental strategies that can be applied to target deconvolution is critically reviewed.

Mentions
Figures
FIGURE 1 | Phenotype-based versus target-based drug discovery.
The diagram illustrates the early phase of drug discovery, in which the aim is to identify target and lead molecules. In the phenotype-based approach, lead molecules are obtained first, followed by target deconvolution to identify the molecular targets that underlie the observed phenotypic effects. In the target-based approach, molecular targets are identified and validated before lead discovery starts; assays and screens are then used to find a lead. FIGURE 2 | Affinity-chromatography-based methods for target deconvolution.
Affinity chromatography can be applied to target deconvolution in different ways. In brief, the small molecule ligand (L) is tethered to a matrix (shown in gray) and incubated with a protein extract that includes the target protein (T). After all unbound proteins have been removed by a series of washing steps, any ligand-bound proteins are then eluted using buffer conditions that disrupt intermolecular interactions (solid-phase elution) and investigated by SDS–PAGE. To minimize the identification of nonspecifically bound proteins, the protein pattern that is obtained with an inactive ligand analogue (A) is also determined (the 'comparison variant'), and the two outcomes are compared. In the 'competition variant', protein elution is accomplished by an excess of free ligand. In serial affinity chromatography, the matrix is incubated with protein extract which is then incubated with fresh matrix. Most of the proteins that bind specifically are captured by the first matrix, whereas the amounts of nonspecific binding proteins are similar for both matrices. The coloured bands represent the matrix-bound proteins from the extract, following separation by SDS–PAGE. FIGURE 3 | Three-hybrid systems for target deconvolution.
The three-hybrid system comprises one component that consists of a DNA-binding domain fused to a ligand-binding domain (DHFR), one component that consists of a ligand molecule (MTX) linked to a small molecule, and one component that consists of a transcriptional activation domain fused to a protein from a cDNA library (which might be a target protein). The binding of the small molecule to its target protein results in the interaction of the three hybrid components, which form a trimeric complex. This complex then activates the expression of a reporter gene, providing a measure of the interaction. DHFR, dihydrofolate reductase; MTX, methotrexate. FIGURE 4 | Display technologies for target deconvolution.
a | Phage display. A phage population displaying potential target proteins on their surface is exposed to an immobilized small molecule. After affinity selection, the eluted phage population is amplified and subjected to further rounds of affinity enrichment. At the end of the procedure, the monoclonal phage population can be analysed for target identification. b | mRNA display. An mRNA-displayed protein library is created in vitro and exposed to an immobilized small molecule. After affinity selection, cDNA target molecules are amplified by PCR and used in the next selection round to generate a new library that is enriched for drug-binding proteins. Several rounds of reiteration lead to the identification of target molecules. ds, double stranded; ss, single stranded. FIGURE 5 | Microarray technologies for target deconvolution.
a | Protein microarrays. In the illustrated example, proteins (exemplified by P1, P2 and P3) are immobilized on a glass slide through a glutathione S-transferase (GST) tag and exposed to a labelled small molecule (in this example, a biotinylated form of the small molecule). Bound target proteins are detected by adding a fluorescently labelled streptavidin (S) conjugate, and then identified by their positions on the array. As a 'loading' control, the array is probed with GST-specific antibodies. b | 'Reverse transfected' cell microarrays. Cells are transfected with cDNAs that are arrayed on a glass slide to generate a 'living' microarray. A labelled small molecule (in this example, a radiolabelled small molecule) is added to the chip and the proteins that bind to it are detected by autoradiography. B, biotin. FIGURE 6 | Biochemical suppression in target deconvolution.
In biochemical suppression, the addition of a small molecule to cytoplasmic protein extracts inhibits an activity of interest, which is measured by means of an activity assay. Fractions of 'uninhibited' extract are added to the 'inhibited' extract to identify the fraction that suppresses the small molecule's inhibitory activity. Subsequent rounds of fractionation and activity-assay measurements are carried out until the 'suppressor' activity is purified.
Altmetric
References
  1. Sneader, W. Drug discovery: a history (John Wiley & Sons, Chichester, UK, 2005) , .
    • . . . At the beginning of the nineteenth century, a key step towards modern drug discovery was made when it became possible to isolate the pharmacologically active substances that are responsible for the observed effects; this led to the discovery of many important drugs1 . . .
  2. International Human Genome Sequencing Consortium. Finishing the euchromatic sequence of the human genome. Nature 431, 931-945 (2004) , .
    • . . . This approach, which was thought to be more rational and efficient, was spurred on by the sequencing of the human genome and the arrival of a new era in which knowledge of the gene sequences of all potential drug targets is available2, 3, 4, 5. . . .
  3. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860-921 (2001) , .
    • . . . This approach, which was thought to be more rational and efficient, was spurred on by the sequencing of the human genome and the arrival of a new era in which knowledge of the gene sequences of all potential drug targets is available2, 3, 4, 5. . . .
  4. Venter, J. C. et al. The sequence of the human genome. Science 291, 1304-1351 (2001) , .
    • . . . This approach, which was thought to be more rational and efficient, was spurred on by the sequencing of the human genome and the arrival of a new era in which knowledge of the gene sequences of all potential drug targets is available2, 3, 4, 5. . . .
  5. Overington, J. P., Al-Lazikani, B. & Hopkins, A. L. How many drug targets are there? Nature Rev. Drug Discov. 5, 993-996 (2006).This study examines the results of a comprehensive survey and proposes a consensus number of current drug targets for all classes of therapeutic drugs , .
    • . . . This approach, which was thought to be more rational and efficient, was spurred on by the sequencing of the human genome and the arrival of a new era in which knowledge of the gene sequences of all potential drug targets is available2, 3, 4, 5. . . .
    • . . . Moreover, owing to the polygenic and multifactorial nature of many important diseases, the therapeutic concept of a 'magic bullet' striking a specific target is questionable, and compounds that have multiple cellular targets (polypharmacologic compounds) might be preferable5 . . .
  6. Sams-Dodd, F. Target-based drug discovery: is something wrong? Drug Discov. Today 10, 139-147 (2005) , .
    • . . . In this postgenomic era, the perceived 'failure' of target-based drug discovery6 has recently led to the renaissance of a more holistic approach that involves the screening of test compounds to determine whether they elicit any phenotypic changes in mammalian cells or in model organisms, such as nematode worms and zebrafish (Box 1) . . .
  7. Butcher, E. C. Can cell systems biology rescue drug discovery? Nature Rev. Drug Discov. 4, 461-467 (2005) , .
    • . . . This approach, which goes beyond individual genes and proteins as it involves the investigation of signalling pathways in a systems-based manner7, is also referred to as chemical genetics8, in analogy to classical forward genetic screens in model organisms . . .
  8. Strausberg, R. L. & Schreiber, S. L. From knowing to controlling: a path from genomics to drugs using small molecule probes. Science 300, 294-295 (2003) , .
    • . . . This approach, which goes beyond individual genes and proteins as it involves the investigation of signalling pathways in a systems-based manner7, is also referred to as chemical genetics8, in analogy to classical forward genetic screens in model organisms . . .
  9. Graziani, F., Aldegheri, L. & Terstappen, G. C. High throughput scintillation proximity assay for the identification of FKBP-12 ligands. J. Biomol. Screen. 4, 3-7 (1999) , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
  10. Caligiuri, M. et al. MASPIT: three-hybrid trap for quantitative proteome fingerprinting of small molecule-protein interactions in mammalian cells. Chem. Biol. 13, 711-722 (2006).This paper describes the application of methotrexate-linked small-molecule ligands to the configuration of a mammalian three-hybrid interaction trap for the proteome-wide identification of small molecule targets , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
    • . . . To circumvent these limitations, the three-hybrid system was recently adapted to mammalian cells in the form of a mammalian small molecule protein interaction trap (MASPIT)10, which is based on the cytokine-receptor-associated JAK (Janus-activated kinase)–STAT (signal transducer and activator of transcription) signal transduction system . . .
  11. Harding, M. W., Galat, A., Uehling, D. E. & Schreiber, S. L. A receptor for the immunosuppressant FK506 is a cis-trans peptidyl-prolyl isomerase. Nature 341, 758-760 (1989) , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
  12. Licitra, E. J. & Liu, J. O. A three-hybrid system for detecting small ligand-protein receptor interactions. Proc. Natl Acad. Sci. USA 93, 12817-12821 (1996).This study demonstrates that a yeast three-hybrid system can be used to discover receptors for small ligands , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
    • . . . The yeast three-hybrid system12 is an extension of the original yeast two-hybrid system39 in which a third hybrid component — the small molecule, linked to a ligand — is also present (Fig. 3) . . .
  13. MacBeath, G. & Schreiber, S. L. Printing proteins as microarrays for high-throughput function determination. Science 289, 1760-1763 (2000).This study demonstrates that protein microarrays can be used for the identification of protein targets of small molecules , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
  14. McPherson, M., Yang, Y., Hammond, P. W. & Kreider, B. L. Drug receptor identification from multiple tissues using cellular-derived mRNA display libraries. Chem. Biol. 9, 691-698 (2002).This proof-of-concept study demonstrates that mRNA display technology can be used to select proteins that bind to a drug of interest , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
    • . . . To our knowledge, no applications of mRNA display for target deconvolution have been published, apart from a successful proof-of-concept study14 in which a library of mRNA–protein fusion molecules from human liver, kidney and bone marrow transcripts and an immobilized FK506–biotin conjugate were used to identify full-length FKBP12. . . .
  15. Sche, P. P., McKenzie, K. M., White, J. D. & Austin, D. J. Display cloning: functional identification of natural product receptors using cDNA-phage display. Chem. Biol. 6, 707-716 (1999) , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
    • . . . Evaluating results after early selection rounds might be useful for the identification of additional binding proteins15 . . .
  16. Siekierka, J. J., Hung, S. H., Poe, M., Lin, C. S. & Sigal, N. H. A cytosolic binding protein for the immunosuppressant FK506 has peptidyl-prolyl isomerase activity but is distinct from cyclophilin. Nature 341, 755-757 (1989) , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
    • . . . Affinity chromatography, a 'classical' method for the purification of proteins18, has been applied to the identification of cellular target proteins that interact with small organic molecules16, 19, 20 . . .
  17. Ziauddin, J. & Sabatini, D. M. Microarrays of cells expressing defined cDNAs. Nature 411, 107-110 (2001).This paper demonstrates the feasibility of using reverse-transfected cell microarrays for the identification of drug targets , .
    • . . . For proof-of-concept studies, the pair FK506–FKBP12 (FK506 binding protein 12kDa) is often used9, 10, 11, 12, 13, 14, 15, 16, 17, because the high cellular abundance of FKBP12, coupled with its high affinity for FK506, makes it ideally suited for such studies . . .
    • . . . A technology that uses microarrays of cells that express defined cDNAs17 (Fig. 5b) circumvents the need to use large numbers of individually purified proteins (which is a necessity for protein microarrays) . . .
    • . . . Proof of concept of this technology for target deconvolution using HEK293T cells was demonstrated by the identification of cells that expressed FKBP12 when radiolabelled FK506 was added to the culture medium17 . . .
    • . . . Furthermore, by using a radiolabelled form of the dopamine D1 receptor antagonist SCH23390, which displays only a relatively weak D1 receptor affinity (the kD of the interaction is approximately 7 10-6M), the D1 receptor could be identified17. . . .
  18. Cuatrecasas, P. Affinity chromatography and purification of the insulin receptor of liver cell membranes. Proc. Natl Acad. Sci. USA 69, 1277-1281 (1972) , .
    • . . . Affinity chromatography, a 'classical' method for the purification of proteins18, has been applied to the identification of cellular target proteins that interact with small organic molecules16, 19, 20 . . .
  19. Taunton, J., Hassig, C. A. & Schreiber, S. L. A mammalian histone deacetylase related to the yeast transcriptional regulator Rpd3p. Science 272, 408-411 (1996) , .
    • . . . Affinity chromatography, a 'classical' method for the purification of proteins18, has been applied to the identification of cellular target proteins that interact with small organic molecules16, 19, 20 . . .
  20. Guiffant, D. et al. Identification of intracellular targets of small molecular weight chemical compounds using affinity chromatography. Biotechnol. J. 2, 68-75 (2007) , .
    • . . . Affinity chromatography, a 'classical' method for the purification of proteins18, has been applied to the identification of cellular target proteins that interact with small organic molecules16, 19, 20 . . .
  21. Laing, P. Luminescent visualization of antigens on blots. J. Immunol. Methods 92, 161-165 (1986) , .
    • . . . Finally, the protein is typically identified by immunodetection21 or mass spectrometry22, 23, 24. . . .
  22. Bennett, K. L., Brond, J. C., Kristensen, D. B., Podtelejnikov, A. V. & Wisniewski, J. R. Analysis of large-scale MS data sets: the dramas and the delights. Drug Discov. Today: TARGETS 3 (Suppl. 1), 43-49 (2005) , .
    • . . . Finally, the protein is typically identified by immunodetection21 or mass spectrometry22, 23, 24. . . .
  23. Schuchardt, S. & Sickmann, A. Protein identification using mass spectrometry: a method overview. EXS 97, 141-170 (2007) , .
    • . . . Finally, the protein is typically identified by immunodetection21 or mass spectrometry22, 23, 24. . . .
  24. Yates, J. R. III. Mass spectrometry and the age of the proteome. J. Mass Spectrom. 33, 1-19 (1998) , .
    • . . . Finally, the protein is typically identified by immunodetection21 or mass spectrometry22, 23, 24. . . .
  25. Hahn, R., Berger, E., Pflegerl, K. & Jungbauer, A. Directed immobilization of peptide ligands to accessible pore sites by conjugation with a placeholder molecule. Anal. Chem. 75, 543-548 (2003) , .
    • . . . Particular problems might be caused by steric hindrance25 or by the creation of a hydrophobic environment that results in nonspecific binding26, although nonspecific binding can often be avoided by adding linkers that have polar functional groups27 . . .
  26. Shaltiel, S. Hydrophobic chromatography. Methods Enzymol. 104, 69-96 (1984) , .
    • . . . Particular problems might be caused by steric hindrance25 or by the creation of a hydrophobic environment that results in nonspecific binding26, although nonspecific binding can often be avoided by adding linkers that have polar functional groups27 . . .
  27. Shiyama, T., Furuya, M., Yamazaki, A., Terada, T. & Tanaka, A. Design and synthesis of novel hydrophilic spacers for the reduction of nonspecific binding proteins on affinity resins. Bioorg. Med. Chem. 12, 2831-2841 (2004) , .
    • . . . Particular problems might be caused by steric hindrance25 or by the creation of a hydrophobic environment that results in nonspecific binding26, although nonspecific binding can often be avoided by adding linkers that have polar functional groups27 . . .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
  28. Bach, S. et al. Roscovitine targets, protein kinases and pyridoxal kinase. J. Biol. Chem. 280, 31208-31219 (2005) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
    • . . . Affinity chromatography with solid-phase elution (Fig. 2) using an SDS–PAGE sample buffer was performed with the drug immobilized on a sepharose matrix28 . . .
  29. Sato, S. et al. Polyproline-rod approach to isolating protein targets of bioactive small molecules: isolation of a new target of indomethacin. J. Am. Chem. Soc. 129, 873-880 (2007) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
    • . . . In this way, CBP (CREB (cyclic-AMP-responsive-element-binding protein)-binding protein) was identified as a target of the colon cancer lead compound ICG-001 (Ref. 31), and glyoxalase 1 (GLO1) was identified as a new target of indomethacin, a widely used anti-inflammatory drug29. . . .
  30. Godl, K. et al. An efficient proteomics method to identify the cellular targets of protein kinase inhibitors. Proc. Natl Acad. Sci. USA 100, 15434-15439 (2003) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
  31. Emami, K. H. et al. A small molecule inhibitor of -catenin/CREB-binding protein transcription. Proc. Natl Acad. Sci. USA 101, 12682-12687 (2004) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
    • . . . In this way, CBP (CREB (cyclic-AMP-responsive-element-binding protein)-binding protein) was identified as a target of the colon cancer lead compound ICG-001 (Ref. 31), and glyoxalase 1 (GLO1) was identified as a new target of indomethacin, a widely used anti-inflammatory drug29. . . .
  32. Snyder, J. R. et al. Dissection of melanogenesis with small molecules identifies prohibitin as a regulator. Chem. Biol. 12, 477-484 (2005) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
    • . . . This experimental refinement led to the identification of prohibitin as a target protein of melanogenin (an inducer of pigmentation)32, and also to the identification of the primary molecular target (cytosolic malate dehydrogenase) for E7070, an anticancer drug that has reached phase 2 clinical trials37 . . .
  33. Wang, G., Shang, L., Burgett, A. W., Harran, P. G. & Wang, X. Diazonamide toxins reveal an unexpected function for ornithine -amino transferase in mitotic cell division. Proc. Natl Acad. Sci. USA 104, 2068-2073 (2007) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
    • . . . Bound proteins were eluted from the matrices and separated by SDS-PAGE and then the resultant protein patterns (obtained after the gels had been stained) were compared, resulting in the identification of ornithine -aminotransferase (OAT), which may have an unanticipated role in tumourigenesis33. . . .
  34. Shimizu, N. et al. High-performance affinity beads for identifying drug receptors. Nature Biotech. 18, 877-881 (2000) , .
    • . . . Biotin and polyethylene glycol have been successfully used as linkers for applications in target deconvolution14, 28, 29, 30, and agarose beads comprise the most frequently used matrix31, 32, 33, although other matrices have been applied27, 34. . . .
  35. Labrou, N. & Clonis, Y. D. The affinity technology in downstream processing. J. Biotechnol. 36, 95-119 (1994) , .
    • . . . High-affinity ligands typically display dissociation constant (kD) values that range from 10-7 to 10-15 M, whereas so-called general-affinity ligands exhibit weaker binding (with kD values in the range of 10-4 to 10-6 M)35 . . .
  36. Graves, P. R. et al. Discovery of novel targets of quinoline drugs in the human purine binding proteome. Mol. Pharmacol. 62, 1364-1372 (2002) , .
    • . . . To improve the detection of low-abundance proteins in a cell extract, a reduction of sample complexity, before affinity chromatography is carried out, can be helpful, as was recently demonstrated by the identification of novel targets for quinoline drugs in the fraction of the human purine-binding proteome36. . . .
  37. Oda, Y. et al. Quantitative chemical proteomics for identifying candidate drug targets. Anal. Chem. 75, 2159-2165 (2003) , .
    • . . . This experimental refinement led to the identification of prohibitin as a target protein of melanogenin (an inducer of pigmentation)32, and also to the identification of the primary molecular target (cytosolic malate dehydrogenase) for E7070, an anticancer drug that has reached phase 2 clinical trials37 . . .
  38. Yamamoto, K., Yamazaki, A., Takeuchi, M. & Tanaka, A. A versatile method of identifying specific binding proteins on affinity resins. Anal. Biochem. 352, 15-23 (2006) , .
    • . . . In order to circumvent problems that might be caused by slow kinetic dissociation or low compound solubility, a serial affinity chromatography approach was introduced38 (Fig. 2) . . .
    • . . . In proof-of-concept experiments, the protein targets of FK506 (FKBP12), benzenesulphonamide (carbonic anhydrase II) and methotrexate (dihydrofolate reductase (DHFR)) were identified38. . . .
  39. Fields, S. & Song, O. A novel genetic system to detect protein-protein interactions. Nature 340, 245-246 (1989) , .
    • . . . The yeast three-hybrid system12 is an extension of the original yeast two-hybrid system39 in which a third hybrid component — the small molecule, linked to a ligand — is also present (Fig. 3) . . .
  40. Becker, F. et al. A three-hybrid approach to scanning the proteome for targets of small molecule kinase inhibitors. Chem. Biol. 11, 211-223 (2004) , .
    • . . . The yeast three-hybrid system has been successfully used to identify the targets of small molecule kinase inhibitors40 . . .
  41. Gray, N. S. et al. Exploiting chemical libraries, structure, and genomics in the search for kinase inhibitors. Science 281, 533-538 (1998) , .
    • . . . A methotrexate ligand was fused through a polyethylene glycol linker to various kinase inhibitors, including purvalanol B, which is a well-characterized CDK inhibitor41, 42 . . .
  42. Knockaert, M. et al. Intracellular targets of cyclin-dependent kinase inhibitors: identification by affinity chromatography using immobilised inhibitors. Chem. Biol. 7, 411-422 (2000) , .
    • . . . A methotrexate ligand was fused through a polyethylene glycol linker to various kinase inhibitors, including purvalanol B, which is a well-characterized CDK inhibitor41, 42 . . .
  43. Nagar, B. et al. Crystal structures of the kinase domain of c-Abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res. 62, 4236-4243 (2002) , .
    • . . . Using the ABL tyrosine kinase inhibitor PD173955 (Refs 43,44), the expected targets LYN, FYN (both of which are SRC kinases) and fibroblast growth factor receptor 1 (FGFR1) were identified, in addition to several other kinases, such as ephrin receptor tyrosine kinases and cyclin G-associated kinase, that had not previously been reported to be targets of this compound . . .
  44. Wisniewski, D. et al. Characterization of potent inhibitors of the Bcr-Abl and the c-kit receptor tyrosine kinases. Cancer Res. 62, 4244-4255 (2002) , .
    • . . . Using the ABL tyrosine kinase inhibitor PD173955 (Refs 43,44), the expected targets LYN, FYN (both of which are SRC kinases) and fibroblast growth factor receptor 1 (FGFR1) were identified, in addition to several other kinases, such as ephrin receptor tyrosine kinases and cyclin G-associated kinase, that had not previously been reported to be targets of this compound . . .
  45. Smith, G. P. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science 228, 1315-1317 (1985) , .
    • . . . Phage display was originally developed for the generation of antibody libraries and the selection of antibodies that strongly bind to a specific antigen45 . . .
  46. Rossenu, S., Dewitte, D., Vandekerckhove, J. & Ampe, C. A phage display technique for a fast, sensitive, and systematic investigation of protein-protein interactions. J. Protein Chem. 16, 499-503 (1997) , .
    • . . . Later it was used to identify protein interaction partners46, and it can also be configured to identify drug-binding proteins (Fig. 4a) . . .
  47. Shim, J. S., Lee, J., Park, H. J., Park, S. J. & Kwon, H. J. A new curcumin derivative, HBC, interferes with the cell cycle progression of colon cancer cells via antagonization of the Ca2+/calmodulin function. Chem. Biol. 11, 1455-1463 (2004) , .
    • . . . Recently, phage display was successfully used to identify bound Ca2+–calmodulin as a target of the curcumin derivative HBC47, a molecule which inhibits the proliferation of several human cancer cells. . . .
  48. Mattheakis, L. C., Bhatt, R. R. & Dower, W. J. An in vitro polysome display system for identifying ligands from very large peptide libraries. Proc. Natl Acad. Sci. USA 91, 9022-9026 (1994).This paper describes an in vitro protein synthesis system for the construction of large libraries of peptides displayed on polysomes (mRNA display) , .
    • . . . mRNA display. mRNA display, an in vitro technique that was originally developed to increase the number of displayed peptides as compared with phage display48, was later applied to the identification of protein–protein interactions49 . . .
  49. Hammond, P. W., Alpin, J., Rise, C. E., Wright, M. & Kreider, B. L. In vitro selection and characterization of Bcl-XL-binding proteins from a mix of tissue-specific mRNA display libraries. J. Biol. Chem. 276, 20898-20906 (2001) , .
    • . . . mRNA display. mRNA display, an in vitro technique that was originally developed to increase the number of displayed peptides as compared with phage display48, was later applied to the identification of protein–protein interactions49 . . .
  50. Zhu, H. & Snyder, M. Protein chip technology. Curr. Opin. Chem. Biol. 7, 55-63 (2003) , .
    • . . . To this end, the proteins to be analysed are purified and immobilized (for example, by an amino-terminal glutathione S-transferase tag) on a glass microscope slide or on another derivatized surface50 . . .
  51. Jacinto, E. & Hall, M. N. Tor signalling in bugs, brain and brawn. Nature Rev. Mol. Cell Biol. 4, 117-126 (2003) , .
    • . . . In a study that aimed to find new components of the target of rapamycin (TOR) signalling network (a central nutrient-response network that controls cell growth51), small molecule inhibitors of rapamycin (SMIRs) were identified . . .
  52. Huang, J. et al. Finding new components of the target of rapamycin (TOR) signaling network through chemical genetics and proteome chips. Proc. Natl Acad. Sci. USA 101, 16594-16599 (2004) , .
    • . . . To gain insights into their molecular mechanisms, biotinylated SMIRs were synthesized and used to probe yeast proteome chips for SMIR-binding activities52 . . .
  53. Peterson, J. R., Lebensohn, A. M., Pelish, H. E. & Kirschner, M. W. Biochemical suppression of small-molecule inhibitors: a strategy to identify inhibitor targets and signaling pathway components. Chem. Biol. 13, 443-452 (2006).This paper introduces a target deconvolution method that is based on functional suppression of chemical inhibition in vitro , .
    • . . . In addition to functional methods employed in chemical genetics research53 such as haploinsufficiency profiling54, which do not require physical interaction, very recently 'biochemical suppression' was introduced as an experimental strategy for target deconvolution55 . . .
    • . . . Two distinct suppressor activities were identified, with the CDC42–RhoGDI (Rho-GDP dissociation inhibitor) complex shown to be a direct target of pirl1, and the Arp (actin and related proteins) 2–3 complex being shown to be a downstream component of the actin assembly pathway that is capable of relieving upstream inhibition of CDC42–RhoGDI when it is added at high concentrations53 . . .
  54. Mayer, T. U. Chemical genetics: tailoring tools for cell biology. Trends Cell Biol. 13, 270-277 (2003) , .
    • . . . In addition to functional methods employed in chemical genetics research53 such as haploinsufficiency profiling54, which do not require physical interaction, very recently 'biochemical suppression' was introduced as an experimental strategy for target deconvolution55 . . .
  55. Giaever, G. et al. Chemogenomic profiling: identifying the functional interactions of small molecules in yeast. Proc. Natl Acad. Sci. USA 101, 793-798 (2004) , .
    • . . . In addition to functional methods employed in chemical genetics research53 such as haploinsufficiency profiling54, which do not require physical interaction, very recently 'biochemical suppression' was introduced as an experimental strategy for target deconvolution55 . . .
  56. Li, X. et al. Multicopy suppressors for novel antibacterial compounds reveal targets and drug efflux susceptibility. Chem. Biol. 11, 1423-1430 (2004) , .
    • . . . In conceptual analogy with genetic high-copy suppressor screens56, 57, small molecules that act as inhibitors in phenotypic screens are used as 'mutations', and the addition of partially purified protein mixtures is used as a way of 'over-expressing' potential suppressors . . .
  57. Luesch, H. et al. A genome-wide overexpression screen in yeast for small-molecule target identification. Chem. Biol. 12, 55-63 (2005) , .
    • . . . In conceptual analogy with genetic high-copy suppressor screens56, 57, small molecules that act as inhibitors in phenotypic screens are used as 'mutations', and the addition of partially purified protein mixtures is used as a way of 'over-expressing' potential suppressors . . .
  58. Gonzalez-Couto, E. et al. Huntington's disease: from experimental results to interaction networks, patho-pathway construction and disease hypothesis. BMC Syst. Biol. 1, 45-47 (2007) , .
    • . . . To discriminate between positive and false-positive interactions, an initial bioinformatics analysis of the identified target protein is useful to investigate whether this protein maps into the pathway that was targeted by the phenotypic assay58, 59 . . .
  59. Ekins, S., Nikolsky, Y., Bugrim, A., Kirillov, E. & Nikolskaya, T. Pathway mapping tools for analysis of high content data. Methods Mol. Biol. 356, 319-350 (2007) , .
    • . . . To discriminate between positive and false-positive interactions, an initial bioinformatics analysis of the identified target protein is useful to investigate whether this protein maps into the pathway that was targeted by the phenotypic assay58, 59 . . .
  60. Macchiarulo, A., Nobeli, I. & Thornton, J. M. Ligand selectivity and competition between enzymes in silico. Nature Biotech. 22, 1039-1045 (2004) , .
    • . . . If three-dimensional protein structures are available, then a structural analysis of the compound–protein interaction can be used to evaluate the interaction profile60, 61 . . .
  61. Mueller, M., Martens, L. & Apweiler, R. Annotating the human proteome: beyond establishing a parts list. Biochim. Biophys. Acta 1774, 175-191 (2007) , .
    • . . . If three-dimensional protein structures are available, then a structural analysis of the compound–protein interaction can be used to evaluate the interaction profile60, 61 . . .
  62. Kramer, R. & Cohen, D. Functional genomics to new drug targets. Nature Rev. Drug Discov. 3, 965-972 (2004) , .
    • . . . Many more experimental strategies exist62, 63, however, they are not described here because a detailed analysis of these methods is beyond the scope of this Review. . . .
  63. Flordellis, C. S., Manolis, A. S., Paris, H. & Karabinis, A. Rethinking target discovery in polygenic diseases. Curr. Top. Med. Chem. 6, 1791-1798 (2006) , .
    • . . . Many more experimental strategies exist62, 63, however, they are not described here because a detailed analysis of these methods is beyond the scope of this Review. . . .
  64. Kola, I. & Landis, J. Can the pharmaceutical industry reduce attrition rates? Nature Rev. Drug Discov. 3, 711-715 (2004) , .
    • . . . There is low probability that a small molecule that modulates such a validated target will result in an efficacious medicine64 . . .
  65. Hu, L., Xu, S., Pan, C., Zou, H. & Jiang, G. Preparation of a biochip on porous silicon and application for label-free detection of small molecule-protein interactions. Rapid Commun. Mass Spectrom. 21, 1277-1281 (2007) , .
    • . . . For instance, mass spectrometry methods could be used in combination with protein microarrays, an approach that was recently demonstrated to be feasible65. . . .
  66. Zhang, Q. et al. Small-molecule synergist of the Wnt/-catenin signaling pathway. Proc. Natl Acad. Sci. USA 104, 7444-7448 (2007) , .
    • . . . One example of this is the recent identification of the purine derivative QS11 as a Wnt signalling 'synergist' that acts through inhibition of the GTPase activating protein of ADP-ribosylation factor 1 (ARFGAP1) and is the only inhibitor of this protein that is known to date66 . . .
  67. Carroll, P. M. & Fitzgerald, K. Model Organisms in Drug Discovery (Culinary and Hospitality Industry Publications Services, Texas, 2003) , .
  68. Clemons, P. A. Complex phenotypic assays in high-throughput screening. Curr. Opin. Chem. Biol. 8, 334-338 (2004) , .
  69. Gangadhar, N. M. & Stockwell, B. R. Chemical genetic approaches to probing cell death. Curr. Opin. Chem. Biol. 11, 83-87 (2007) , .
  70. Caricasole, A. et al. Induction of Dickkopf-1, a negative modulator of the Wnt pathway, is associated with neuronal degeneration in Alzheimer's brain. J. Neurosci. 24, 6021-6027 (2004) , .
  71. Barker, N. & Clevers, H. Mining the Wnt pathway for cancer therapeutics. Nature Rev. Drug Discov. 5, 997-1014 (2006) , .
  72. Lang, P., Yeow, K., Nichols, A. & Scheer, A. Cellular imaging in drug discovery. Nature Rev. Drug Discov. 5, 343-356 (2006) , .
  73. Ignatenko, N. A. et al. Pharmacogenomics of the polyamine analog 3,8,13,18-tetraaza-10,11-[(E)-1,2-cyclopropyl]eicosane tetrahydrochloride, CGC-11093, in the colon adenocarcinoma cell line HCT1161. Technol. Cancer Res. Treat. 5, 553-564 (2006) , .
  74. Lamb, J. et al. The connectivity map: using gene-expression signatures to connect small molecules, genes, and disease. Science 313, 1929-1935 (2006).This paper is the first installment of a reference collection of gene-expression profiles from cultured human cells that were treated with 164 bioactive small molecules , .
  75. Nature Insight Proteomics. [online] (2003) , .
  76. Raggiaschi, R. & Terstappen G. C. Proteomics technologies. Biosci. Rep. 25, 1-2 (2005) , .
  77. Kremer, A., Schneider, R. & Terstappen, G. C. A bioinformatics perspective on proteomics: data storage, analysis, and integration. Biosci. Rep. 25, 95-106 (2005) , .
  78. Towbin, H. et al. Proteomics-based target identification: bengamides as a new class of methionine aminopeptidase inhibitors. J. Biol. Chem. 278, 52964-52971 (2003) , .
Expand