Efficient discovery of anti-inflammatory small-molecule combinations using evolutionary computing

Journal name:
Nature Chemical Biology
Year published:
Published online


The control of biochemical fluxes is distributed, and to perturb complex intracellular networks effectively it is often necessary to modulate several steps simultaneously. However, the number of possible permutations leads to a combinatorial explosion in the number of experiments that would have to be performed in a complete analysis. We used a multiobjective evolutionary algorithm to optimize reagent combinations from a dynamic chemical library of 33 compounds with established or predicted targets in the regulatory network controlling IL-1β expression. The evolutionary algorithm converged on excellent solutions within 11 generations, during which we studied just 550 combinations out of the potential search space of ∼9 billion. The top five reagents with the greatest contribution to combinatorial effects throughout the evolutionary algorithm were then optimized pairwise. A p38 MAPK inhibitor together with either an inhibitor of IκB kinase or a chelator of poorly liganded iron yielded synergistic inhibition of macrophage IL-1β expression. Evolutionary searches provide a powerful and general approach to the discovery of new combinations of pharmacological agents with therapeutic indices potentially greater than those of single drugs.

At a glance


  1. Combinatorial evolutionary inhibition of IL-1β expression.
    Figure 1: Combinatorial evolutionary inhibition of IL-1β expression.

    Known drugs were tested alone before being used at a single concentration (3 μM) in a chemical library (loop 1, clockwise). Initialization of IBEA creates a random selection of combinations that are incubated with stimulated cells before measurement of cell death (LDH release) and IL-1β expression. Evaluation of these data against the number of compounds in the combination (n = 3 for all data) is performed by IBEA before a new generation of combinations is computed and tested. After 11 generations, concentration-dependent optimization (loop 2) of five top-ranked reagents was undertaken. Synergy was detected in new dual combinations. L, low; M, medium; H, high; PBS, phosphate-buffered saline; TMB, 3,3′,5,5′-tetramethyl benzidine; HRP, horseradish peroxidase.

  2. Combinatorial multiobjective optimization of reagents inhibiting IL-1β production.
    Figure 2: Combinatorial multiobjective optimization of reagents inhibiting IL-1β production.

    Analysis of successive generations (generations 1 (initialization), 5 and 10) of reagent combinations reveals their convergence to a subset of highly effective combinations reflecting the inhibition of IL-1β expression with concomitant decreases in LDH release and the number of member reagents. All data presented are the means of three determinations. Data points appearing as zero on the axis labeled “Number of reagents” reflect positive control responses (LPS, 1 μg ml−1; DMSO, 0.5% v/v). n = 3 for all data.

  3. Analysis of the evolutionary multiobjective optimization of the inhibition of IL-1β production.
    Figure 3: Analysis of the evolutionary multiobjective optimization of the inhibition of IL-1β production.

    Population average rank for inhibition of IL-1β expression (top left), number of component reagents in combinations (top right), LDH release (bottom left) and overall IBEA hypervolume (bottom right). The IBEA hypervolume is a composite (described in Methods) of the performance of the different generations with regard to the three objectives, where a smaller number indicates better performance. Error bars, s.e.m.

  4. Rank ordering of component reagents.
    Figure 4: Rank ordering of component reagents.

    Analysis of all evolutionary algorithm generations (1–11) in the presence (top) and absence (bottom) of SB203580 yielded a rank order for the fitness contribution (described in Methods) of each reagent within the library. Only five top-ranked reagents are displayed here either alone (single) or in double or triple combinations in the presence and absence of SB203580.

  5. Concentration-dependent adaptive dose-matrix optimization of paired reagent combinations.
    Figure 5: Concentration-dependent adaptive dose-matrix optimization of paired reagent combinations.

    Concentration-dependent adaptive dose matrix optimization of paired reagent combinations was achieved by adaptively changing the concentrations of reagents after assessing the inhibition of IL-1β expression. (a) A p38 MAPK inhibitor (SB203580) and an Iκ kinase inhibitor (IKKi) were assessed alone and as a paired combination. (b) Simple additive effects of the SB203580-IKKi combination. (c) Potential synergy of the SB203580-IKKi combination (as in a) revealed by subtraction of simple additive effects of the experimental data in b, calculated from single-reagent data in the absence of the other reagent. (d) A p38 MAPK inhibitor (SB203580) and an iron chelator (SIH) were assessed alone and as a paired combination. (e) Simple additive effects of the SB203580-SIH combination. (f) Potential synergy of the SB203580-SIH combination (as in d) revealed by subtraction of simple additive effects of the experimental data in e, calculated from single-reagent data in the absence of the other reagent. Synergistic inhibition of IL-1β expression was revealed with the combinations of SB203580 with IKKi in c or with SIH in f; synergistic inhibition is considered as 'extra' inhibition relative to the additive inhibitions measured for the individual reagents.


  1. Nathan, C. & Ding, A. Nonresolving inflammation. Cell 140, 871882 (2010).
  2. Dinarello, C.A. Interleukin-1 in the pathogenesis and treatment of inflammatory diseases. Blood 117, 37203732 (2011).
  3. Dinarello, C.A. Biologic basis for interleukin-1 in disease. Blood 87, 20952147 (1996).
  4. Luheshi, N.M., Rothwell, N.J. & Brough, D. Dual functionality of interleukin-1 family cytokines: implications for anti-interleukin-1 therapy. Br. J. Pharmacol. 157, 13181329 (2009).
  5. Relton, J.K. & Rothwell, N.J. Interleukin-1 receptor antagonist inhibits ischaemic and excitotoxic neuronal damage in the rat. Brain Res. Bull. 29, 243246 (1992).
  6. Lu, Y.-C., Yeh, W.-C. & Ohashi, P.S. LPS/TLR4 signal transduction pathway. Cytokine 42, 145151 (2008).
  7. Hopkins, A.L. Network pharmacology: the next paradigm in drug discovery. Nat. Chem. Biol. 4, 682690 (2008).
  8. Fell, D.A. & Thomas, S. Physiological control of metabolic flux: the requirement for multisite modulation. Biochem. J. 311, 3539 (1995).
  9. Jeon, Y.J. et al. Dexamethasone inhibits IL-1beta gene expression in LPS-stimulated RAW 264.7 cells by blocking NF-κB/Rel and AP-1 activation. Immunopharmacology 48, 173183 (2000).
  10. Lehár, J., Stockwell, B.R., Giaever, G. & Nislow, C. Combination chemical genetics. Nat. Chem. Biol. 4, 674681 (2008).
  11. Feng, Y., Mitchison, T.J., Bender, A., Young, D.W. & Tallarico, J.A. Multi-parameter phenotypic profiling: using cellular effects to characterize small-molecule compounds. Nat. Rev. Drug Discov. 8, 567578 (2009).
  12. Paolini, G.V., Shapland, R.H., van Hoorn, W.P., Mason, J.S. & Hopkins, A.L. Global mapping of pharmacological space. Nat. Biotechnol. 24, 805815 (2006).
  13. Coello, C.A.C., Lamont, G.B. & Veldhuizen, D.A.V. Evolutionary Algorithms for Solving Multi-objective Problems (Springer, 2007).
  14. Handl, J., Kell, D.B. & Knowles, J. Multiobjective optimization in bioinformatics and computational biology. IEEE/ACM Trans. Comput. Biol. Bioinform. 4, 279292 (2007).
  15. Knowles, J., Corne, D. & Deb, K. Multiobjective Problem Solving from Nature: From Concepts to Applications (Springer, 2008).
  16. Bäck, T., Fogel, D.B. & Michalewicz, Z. Handbook of Evolutionary Computation (Institute of Physics Pub., 1997).
  17. Goldberg, D.E. The Design of Innovation: Lessons From and For Competent Genetic Algorithms (Kluwer Academic Publishers, 2002).
  18. Knight, C.G. et al. Array-based evolution of DNA aptamers allows modelling of an explicit sequence-fitness landscape. Nucleic Acids Res. 37, e6 (2008).
  19. O'Hagan, S., Dunn, W.B., Brown, M., Knowles, J.D. & Kell, D.B. Closed-loop, multiobjective optimization of analytical instrumentation: gas chromatography/time-of-flight mass spectrometry of the metabolomes of human serum and of yeast fermentations. Anal. Chem. 77, 290303 (2005).
  20. Knowles, J. closed-loop evolutionary multiobjective optimization. IEEE Comput. Intell. Mag. 4, 7791 (2009).
  21. Zitzler, E. & Künzli, S. Indicator-based selection in multi-objective search. in Parallel Problem Solving from Nature—PPSN VIII (eds. Yao, X. et al.) 832842 (Springer, 2004).
  22. Allmendinger, R. & Knowles, J. Analysis of Several Evolutionary Algorithms on the Noisy Three-Objective Chemical Mixture Optimization Problem. Technical Report MLO-12009, 1–9 〈 http://www.cs.manchester.ac.uk/~allmendr/publications.html〉 (2009).
  23. Allmendinger, R. & Knowles, J. Evolutionary optimization on problems subject to changes of variables. in Parallel Problem Solving from Nature—PPSN XI (eds. Schaefer, R., Cotta, C., Kolodziej, J. & Rudolph, G.) 151160 (Springer, 2010).
  24. Kell, D.B. Iron behaving badly: inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases. BMC Med. Genomics 2, 2 (2009).
  25. Kell, D.B. Towards a unifying, systems biology understanding of large-scale cellular death and destruction caused by poorly liganded iron: Parkinson's, Huntington's, Alzheimer's, prions, bactericides, chemical toxicology and others as examples. Arch. Toxicol. 84, 825889 (2010).
  26. Puneet, P. et al. SphK1 regulates proinflammatory responses associated with endotoxin and polymicrobial sepsis. Science 328, 12901294 (2010).
  27. Greco, W.R., Bravo, G. & Parsons, J.C. The search for synergy: a critical review from a response surface perspective. Pharmacol. Rev. 47, 331385 (1995).
  28. Lehár, J. et al. Synergistic drug combinations tend to improve therapeutically relevant selectivity. Nat. Biotechnol. 27, 659666 (2009).
  29. Zimmermann, G.R., Léhar, J. & Keith, C.T. Multi-target therapeutics: when the whole is greater than the sum of the parts. Drug Discov. Today 12, 3442 (2007).
  30. Ejim, L. et al. Combinations of antibiotics and nonantibiotic drugs enhance antimicrobial efficacy. Nat. Chem. Biol. 7, 348350 (2011).
  31. Knight, Z.A., Lin, H. & Shokat, K.M. Targeting the cancer kinome through polypharmacology. Nat. Rev. Cancer 10, 130137 (2010).
  32. Oda, K. & Kitano, H. A comprehensive map of the toll-like receptor signaling network. Mol. Syst. Biol. 2, 2006.0015 (2006).
  33. Bornkamp, B. et al. Innovative approaches for designing and analyzing adaptive dose-ranging trials. J. Biopharm. Stat. 17, 965995 (2007).
  34. Jia, J. et al. Mechanisms of drug combinations: interaction and network perspectives. Nat. Rev. Drug Discov. 8, 111128 (2009); erratum 8, 516 (2009).
  35. Fitzgerald, J.B., Schoeberl, B., Nielsen, U.B. & Sorger, P.K. Systems biology and combination therapy in the quest for clinical efficacy. Nat. Chem. Biol. 2, 458466 (2006).
  36. Sun, C.-P. et al. Integrative systems control approach for reactivating Kaposi's sarcoma-associated herpesvirus (KSHV) with combinatory drugs. Integr. Biol. (Camb) 1, 123130 (2009).
  37. Wong, P.K. et al. Closed-loop control of cellular functions using combinatory drugs guided by a stochastic search algorithm. Proc. Natl. Acad. Sci. USA 105, 51055110 (2008).
  38. Feala, J.D. et al. Systems approaches and algorithms for discovery of combinatorial therapies. Wiley Interdiscip. Rev. Syst. Biol. Med. 2, 181193 (2010).
  39. Chatterjee, M.S., Purvis, J.E., Brass, L.F. & Diamond, S.L. Pairwise agonist scanning predicts cellular signaling responses to combinatorial stimuli. Nat. Biotechnol. 28, 727732 (2010).
  40. Geva-Zatorsky, N. et al. Protein dynamics in drug combinations: a linear superposition of individual-drug responses. Cell 140, 643651 (2010).
  41. Horackova, M., Ponka, P. & Byczko, Z. The antioxidant effects of a novel iron chelator salicylaldehyde isonicotinoyl hydrazone in the prevention of H2O2 injury in adult cardiomyocytes. Cardiovasc. Res. 47, 529536 (2000).
  42. Sindrilaru, A. et al. An unrestrained proinflammatory M1 macrophage population induced by iron impairs wound healing in humans and mice. J. Clin. Invest. 121, 985997 (2011).
  43. Hammaker, D. & Firestein, G.S. “Go upstream, young man”: lessons learned from the p38 saga. Ann. Rheum. Dis. 69 (suppl. 1), i77i82 (2010).
  44. Dobson, P.D. & Kell, D.B. Carrier-mediated cellular uptake of pharmaceutical drugs: an exception or the rule? Nat. Rev. Drug Discov. 7, 205220 (2008).
  45. Kell, D.B. & Oliver, S.G. Here is the evidence, now what is the hypothesis? The complementary roles of inductive and hypothesis-driven science in the post-genomic era. Bioessays 26, 99105 (2004).
  46. Scully, P. et al. Plasma cytokine profiles in females with irritable bowel syndrome and extra-intestinal co-morbidity. Am. J. Gastroenterol. 105, 22352243 (2010).
  47. Löfdahl, C.-G. COPD and co-morbidities, with special emphasis on cardiovascular conditions. Clin. Respir. J. 2 (suppl. 1), 5963 (2008).
  48. Akella, L.B. & DeCaprio, D. Cheminformatics approaches to analyze diversity in compound screening libraries. Curr. Opin. Chem. Biol. 14, 325330 (2010).
  49. Tobinick, E.L. The value of drug repositioning in the current pharmaceutical market. Drug News Perspect. 22, 119125 (2009).
  50. Hughes, J. Evolutionary multi-objective ranking with uncertainty and noise. in Evolutionary Multi-Criterion Optimization, Vol. 1993 (eds. Zitzler, E. et al.) 329343 (Springer, 2001).

Download references

Author information

  1. These authors contributed equally to this work.

    • Ben G Small &
    • Barry W McColl


  1. Doctoral Training Centre, Integrative Systems Biology Molecules to Life, Manchester Centre for Integrative Systems Biology, Manchester Interdisciplinary Biocentre, The University of Manchester, Manchester, UK.

    • Ben G Small
  2. School of Chemical Engineering and Analytical Science, The University of Manchester, Manchester, UK.

    • Ben G Small
  3. The Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush, Midlothian, UK.

    • Barry W McColl
  4. School of Computer Science, The University of Manchester, Manchester, UK.

    • Richard Allmendinger,
    • Jürgen Pahle,
    • Joshua Knowles &
    • Pedro Mendes
  5. NeuroSystems Section, Faculty of Life Sciences, The University of Manchester, Manchester, UK.

    • Gloria López-Castejón,
    • Nancy J Rothwell &
    • David Brough
  6. Manchester Centre for Integrative Systems Biology, Manchester Interdisciplinary Biocentre, The University of Manchester, Manchester, UK.

    • Pedro Mendes &
    • Douglas B Kell
  7. Virginia Bioinformatics Institute, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA.

    • Pedro Mendes
  8. School of Chemistry, The University of Manchester, Manchester, UK.

    • Douglas B Kell


B.G.S. designed, performed and analyzed the data from the single-reagent, combinatorial, evolution-directed and adaptive dose-matrix studies and wrote the first draft of the manuscript. B.W.M. designed, performed and analyzed the data from single-reagent and evolution-directed combinatorial studies. J.P. suggested and was involved in the design of the concentration-dependent adaptive dose-matrix search protocol of paired combinations and wrote the code in R to produce the combination response-shape plots. G.L.-C. conducted adaptive dose-matrix studies. D.B.K., D.B. and N.J.R. conceived the idea of applying an evolutionary algorithm to select combinations of reagents for inhibition of IL-1β expression; wrote the grant application; investigated and supervised, together with B.G.S.; and advised on experimental design. J.K. supervised R.A., generally advised on evolutionary algorithms and multiobjective optimization and, with R.A., evaluated and selected the evolutionary algorithm (IBEA) applied here. R.A. wrote the IBEA code in Java, helped implement its use and generated post hoc reagent-fitness measurements. P.M. supervised B.G.S. and advised on experimental design and data analysis. All authors contributed to the writing of the manuscript and approved its final form.

Competing financial interests

D.B.K. transferred his responsibility as principal investigator of Biotechnology and Biological Sciences Research Council (BBSRC) Grant BB/F018398/1 to N.J.R. and P.M. prior to taking up his appointment as chief executive of the BBSRC on 1 October 2008. N.J.R. is a nonexecutive director of AstraZeneca. B.G.S. is a former employee of AstraZeneca and Eli Lilly and Company and is a shareholder of stock from both companies.

Corresponding author

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Text and Figures (2M)

    Supplementary Methods and Supplementary Results

Excel files

  1. Supplementary Spreadsheet 1 (152K)


Zip files

  1. Supplementary Source Code File 1 (20K)

    Cocktail Generator IBEA

  2. Supplementary Source Code File 2 (8K)


Additional data