Trends in Biotechnology
OpinionRational drug design via intrinsically disordered protein
Introduction
The pharmaceutical industry is currently struggling to find promising new drug targets. Although research and development (R&D) spending has increased by an average of 8% (adjusted for inflation) each year for at least the past 35 years, and although the number of completed phase III clinical trials per year match or even exceed the R&D spending trend, the number of new molecule entities (NMEs) approved by the Food and Drug Administration (FDA; http://www.fda.gov) has not kept pace (Figure 1). For example, despite a twofold increase in R&D spending during the past decade (adjusted for inflation), the number of approved NMEs in 2004 increased just 10% compared with 1995, with no upward trend in the intervening years (Figure 1). By the end of 2005, 18 NMEs were approved for the year by the FDA (http://www.fda.gov/cder/rdmt/nmecy2005.htm), showing that 2005 was a particularly poor year for drug discovery – one of the two worst since before 1990 – and further showing the gradual decline in the five-year-average since a peak in the 1997–1998 period.
Current drug therapy uses ∼500 targets, less than 10% of a reasonable potential list [1]; membrane receptors (mostly G-protein coupled) and various enzymes account for ∼70% of the current drug molecule targets [2]. New approaches, including high-throughput screening, structure-based and fragment-based design, proteomics, microarrary analysis and systems biology, have not yielded the hoped-for increases in the discovery rates for NMEs, at least not to date, apparently due in significant measure to the failure to translate the new discoveries and understanding into new targets.
A major potential source of new drug targets is provided by the set of protein–protein interactions within a cell (http://www.nature.com/horizon/chemicalspace/background/protein.html). Upon environmental stimuli, the complex networks of protein–protein interactions generate a coordinated response; therefore, such interactions offer attractive targets for a generation of therapeutic intervention. However, the protein–protein interactions involved in cell signalling are transient and difficult to assay, which inherently hampers drug discovery efforts focused on the modulation of these interactions. Systems biology approaches are mapping out the protein interactome and an understanding of these results might help to identify the protein–protein interactions that appear to be particularly desirable drug targets [3]. A further complication is that kinetics is known to be of central importance in every drug design problem. The recent development of quantitative interaction network mapping is making this approach more powerful by including kinetics information [4]. Although the next step of finding drug molecules that modulate protein–protein interactions has not been generally successful in the past [5], recent promising examples are creating a new optimism in this arena 6, 7.
We at Molecular Kinetics noticed that for several of the protein–protein interactions that are being successfully blocked by small molecules one of the partners in each case undergoes a disorder-to-order transition upon binding to its structured complement. Although flexibility and rearrangement of side chains on the surface of a protein have been observed as a way for binding sites, or ‘hot spots’ to change shape to better fit small molecule ligands [8] through a mechanism that seems analogous to Koshland's induced fit [9], the importance of backbone flexibility coupled with a disorder-to-order transition upon binding has not received a mention in any of the current discussions on the modulation of protein–protein interactions by small molecule ligands. Although our concept of disorder-based modulation of protein–protein interactions currently suffers from a lack of extensive experimental support (because of the novelty of this field), one well-characterized example, the p53–MDM2 interaction, provides a useful illustration (for more discussion see below and 10, 11, 12, 13, 14, 15, 16, 17). Furthermore, the application of our computational tools revealed that the mechanisms where a small molecule can block the protein–protein interaction in which one partner undergoes disorder-to-order transition is likely to be applicable for the modulation of several additional protein–protein interactions such as tubulin polymerization and the formation of the Bcl2–BH3, Bcl-xL–BH3, XIAP–caspase-9, XIAP–caspase-3 and Rac–Trio complexes; however, the probable importance of intrinsic disorder in the druggable interactions was unmentioned [6]. The goals of this article are twofold: first, to show how this intrinsic backbone and side-chain disorder is important, perhaps even crucial, for the druggability of protein–protein interactions; and second, to use disorder knowledge to generalize from the current successes to identify, on a proteome-wide scale, the protein–protein interactions that are the most promising targets for therapeutic intervention. These analyses yield a cornucopia of new opportunities (Molecular Kinetics, patent pending) across a wide spectrum of diseases.
Section snippets
Intrinsically disordered proteins
In contrast to structured (ordered) proteins, where the 3D structures exhibit only small oscillatory backbone excursions about their equilibrium positions, intrinsically disordered proteins exist as dynamic ensembles in which the atom positions and backbone Ramachandran angles vary significantly during time with no specific equilibrium values. Thus, intrinsically disordered proteins and regions differ from their structured counterparts by their dynamics and not necessarily by the temporal
Disorder in genomes, functional classes and diseases
The number of proteins experimentally characterized as disordered, as well as the number of documented instances of disorder involvement in cell signalling and regulation, are both growing rapidly 19, 25, 31, 32. In agreement with these experimental results, intrinsically disordered proteins are predicted to be widespread 33, 34, 35. The application of PONDR®s to several different proteomes indicates that eukaryotic proteins contain long regions of disorder in 35–51% of their sequences, whereas
Intrinsic disorder and drug discovery
The ultimate goal of drug discovery is to synthesize or discover new molecules to treat disease. Although not generally realized, there has been an indirect focus on intrinsically disordered proteins in drug discovery. For example, kinase and protein phosphorylation sites (which reside primarily within the intrinsically disordered regions 19, 38, 39, 40) have been the focus of drug discovery efforts during the past decade or longer. However, strategies for structure-based drug design (SBDD)
The p53–MDM2 interaction
Currently, many promising drug targets involve protein–protein interactions within signalling pathways [3]. Several small molecules have been recently discovered to inhibit protein–protein interactions that are essential to cancer cell proliferation 6, 7, 10, 11. These small molecules appear to act by mimicking the shape and physicochemical characteristics of a disordered protein fragment after it has transitioned to an ordered helical structure as a result of interaction with its binding
Concluding remarks
Application of computational methodologies based on protein intrinsic disorder is herein suggested to provide a novel pathway for drug discovery. Protein intrinsic disorder is abundant in human and other eukaryotic genomes; such disorder has been found to have important roles in cell signalling and regulation, and occurs in the proteins associated with several major diseases. Protein–protein interactions based on disorder-to-order transition of one partner can make ideal druggable targets.
Acknowledgements
We thank the NIH Small Business Innovation Research (SBIR) grant program for support (NIH SBIR Grant numbers 1R43CA97629-01, 1R43GM066412-01A1, 5R43CA099053-02, 1R43CA097568-01A1 and 1R43CA119429-01), and the Indiana 21st Century Fund for providing Molecular Kinetics with matching funds for the last four NIH SBIR Awards.
References (53)
Antagonists of protein–protein interactions
Chem. Biol.
(2000)Protein–protein interactions and cancer: small molecules going in for the kill
Curr. Opin. Chem. Biol.
(2005)A small synthetic peptide, which inhibits the p53–hmd2 interaction, stimulates the p53 pathway in tumour cell lines
J. Mol. Biol.
(2000)Design of a synthetic Mdm2-binding mini protein that activates the p53 response in vivo
Curr. Biol.
(1997)Ca2+ binding to alpha-synuclein regulates ligand binding and oligomerization
J. Biol. Chem.
(2001)Intrinsically unstructured proteins
Trends Biochem. Sci.
(2002)Intrinsic disorder in cell-signalling and cancer-associated proteins
J. Mol. Biol.
(2002)Prediction and functional analysis of native disorder in proteins from the three kingdoms of life
J. Mol. Biol.
(2004)Preformed structural elements feature in partner recognition by intrinsically unstructured proteins
J. Mol. Biol.
(2004)Local structural elements in the mostly unstructured transcriptional activation domain of human p53
J. Biol. Chem.
(2000)
Studies of the RNA degradosome-organizing domain of the Escherichia coli ribonuclease RNase E
J. Mol. Biol.
The C-terminal domain of measles virus nucleoprotein belongs to the class of intrinsically disordered proteins that fold upon binding to their physiological partner
Virus Res.
The role of innovation in drug development
Nat. Biotechnol.
Drug discovery: a historical perspective
Science
Virtual identification of essential proteins within the protein interaction network of yeast
Proteomics
A quantitative protein interaction network for the ErbB receptors using protein microarrays
Nature
Targeting protein–protein interactions for cancer therapy
J. Mol. Med.
Small-molecule inhibitors of protein–protein interactions: progressing towards the dream
Nat. Rev. Drug Discov.
Comparison of experimental binding data and theoretical models in proteins containing subunits
Biochemistry
Inhibition of the p53–MDM2 interaction: targeting a protein–protein interface
Mol. Cancer Res.
In vivo activation of the p53 pathway by small-molecule antagonists of MDM2
Science
Signalling to the p53 tumour suppressor through pathways activated by genotoxic and nongenotoxic stress
p53-mediated death of cells overexpressing MDM2 by an inhibitor of MDM2 interaction with p53
Oncogene
Structure of the MDM2 oncoprotein bound to the p53 tumour suppressor transactivation domain
Science
Small-molecule antagonists of p53–MDM2 binding – research tools and potential therapeutics
Cell Cycle
Predicting intrinsic disorder from amino acid sequence
Proteins
Cited by (213)
Flanking regions, amyloid cores, and polymorphism: the potential interplay underlying structural diversity
2023, Journal of Biological ChemistryTargeting disordered-structured domain interactions in Galectin-3 based on NMR and enhanced MD
2022, Biophysical JournalFunctional inhibition of c-Myc using novel inhibitors identified through “hot spot” targeting
2022, Journal of Biological ChemistryIntrinsically disordered proteins/regions and insight into their biomolecular interactions
2022, Biophysical ChemistryDesign and Development of Hydrophobicity and Net charge Based Artificial Neural Network Model for IDP/IDPR Prediction
2022, Procedia Computer ScienceDisaggregation mechanism of prion amyloid for tweezer inhibitor
2021, International Journal of Biological Macromolecules