(RG9) Molecular Interactions Research Group (MIRG) Aaron P. Yamniuk, John Newitt, Fumio Arisaka, Anthony Giannetti, Preston Hensley, David Myszka, Fred Schwarz, Ed Eisenstein, Michael Doyle ABRF 2013 Palm Springs CA March 4, 2013 Session Organizer: Aaron Yamniuk, Bristol-Myers Squibb 1
Session Outline Session 1: Introduction to a Protein Interaction System used for Quantitative Evaluation of Biomolecular Interactions Aaron Yamniuk, Bristol-Myers Squibb Session 2: Results of the MIRG 2012 Benchmark Study on Quantitative Evaluation of a Protein-Protein Interaction Satya Yadav, Cleveland Clinic Lerner Research Institute 2
MIRG Mission To show how biophysical tools (biosensor, ITC and AUC) are used to quantitatively characterize molecular interactions To show how these tools work in a resource facility environment To educate the ABRF members in the methodologies of the three core technologies To provide test systems to be used to compare the capabilities of individual laboratories with each other, either within the three disciplines or among the disciplines To compile the results of the analyses of these systems and publish the results in the Journal of Biomolecular Techniques or other appropriate publications To meet regularly to organize and run workshops or other activities at the annual ABRF meetings to accomplish this mission A + B AB AUC ITC SPR 3
Past MIRG Benchmark Studies MIRG 2002 Benchmark Study: Quantitative analysis of Enzyme/Inhibitor interaction 4-Carboxybenzenesulfonamide (201.2 Da) MIRG 2010 Benchmark Study: Molecular Interactions in a Three- Component System protein B protein C kinetics & affinity protein A?? Carbonic Anhydrase II (~30,000 Da) PDB:112CA Ternary complex? Competitive binding? For more information see: www.abrf.org/mirg 4
MIRG Objective: Develop a protein-protein p interaction system to test limits of core molecular interaction technologies Use system to: Test limits of MIRG core interaction technologies Evaluate new molecular interaction technologies Possible future use as a reference standard for core laboratories? Desired properties: Easy production with good yield High solubility and stability Desired affinity range of K D = ~10-9 10-5 M can we use one molecule pair to test high and low end? High resolution structures for computational analysis 5
MIRG Plan: Evaluate the use of the (mutant) Barnase/Barstar protein interaction system Barnase (or mutant Barnase) K ~10-3 -14 D 10 M Barstar (or mutant t Barstar) Well characterized proteins (structure, t folding, binding) (Hartley, Fersht, others) Easy production with high yield Favorable stability and solubility properties Large collection of mutants available (K D 10-3 10-12 M) 6
Construct design to generate bivalent interaction with desired affinities his-binase1 his-binase2 R59A R59A K d ~10-8 10-9 M K d ~10-5 10-6 M d Also produce wild type for comparison (~10x higher affinity) Y29A his-barstar*-y29a (Y29A, C40A, C82A) wild type his-barstar* (C40A, C82A) 7
Production of barstar and binase proteins Protein Predicted Mass (g/mol) a Theor. pi ESI/TOF Mass (g/mol) DLS Rh (nm) b SEC-MALS (g/mol) His-Barstar 12,031 5.3 12,030 1.8 ± 0.0 12,300 His-Barstar-Y29A 11,939 5.3 11,939 1.9 ± 0.1 11,800 His-Binase1 26,283 8.5 26,284 2.6 ± 0.1 26,700 His-Binase2 26,283 8.5 26,284 2.6 ± 0.1 27,300 0.5, 2.0, 10ug His-binase1 Barstar 10,205 4.9 10,205 1.6 ± 0.1 10,100 Barstar-Y29A 10,112 4.9 10,112 1.6 ± 0.1 9,790 Binase1 24,434 8.8 24,434 2.2 ± 0.2 24,900 Binase2 24,434 8.8 24,434 2.3 ± 0.1 24,800 0.5, 2.0, 10ug His-barstar-Y29A Protocols developed for efficient expression, purification, and his-tag cleavage (E. Eisenstein, J. Newitt) Proteins are high purity, monomeric in solution with expected mass 8
ITC: Barstar*-Y29A vs Binase1 or 2 5.00 Time (min) 0 30 60 90 120 1E-4 1E-5 K o d values (25 C) Av ± SD 4.80 µcal/sec 4.60 4.40 4.20 Kd (M) 1E-6 1E-7 1E-8 High affinity site Low affinity site 0.00 1E-9 KCal/Mole of Inje ectant -2.00-4.00-6.00-8.00 R59A, -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Molar Ratio - Tested multiple concentrations, temperatures (18 measurements, 3 laboratories) - N = 2.0, highly reproducible (full activity) - his-tagged was equivalent to untagged - Binase1 was equivalent to binase2 - Reproducible K D determination possible 9 but challenging
AUC: Barstar*-Y29A vs Binase1 or 2 Data at low [protein] Protein Kd1 (nm) Kd2 (um) domain Binase 1 10-40 46 Binase 1 5-15 16-20 Binase 2 10-40 Additional measurements in progress Data at high [protein] R59A, domain - Tested multiple loading concentrations, 2 rotor speeds (sedimentation ti equilibrium) i -K D values in agreement with ITC 10
SPR: Barstar*-Y29A vs Binase1 or 2 1E-3 1E-4 Av ± SD Kd (M) 1E-5 Triplicate measurements 1E-6 1E-7 High affinity site Low affinity site R59A, 1E-8 - Tested multiple analyte concentrations, & surface densities (20 measurements, 3 labs) - his-tagged was equivalent to untagged - Binase1 was equivalent to binase2 - Reproducible K D determination ti possible but challenging 11
Comparison: ITC vs AUC vs SPR Barstar*-Y29A vs Binase1 or 2 High affinity site 1E-3 Low affinity site 1E-6 1E-7 1E-4 Av ± SD Kd d (M) 1E-8 Kd (M) 1E-5 1E-9 ITC AUC SPR (amine) 1E-6 ITC AUC SPR (amine) Significantly higher affinity determined by ITC compared to SPR for each site AUC agrees well with ITC for higher affinity site, intermediate between ITC and SPR values for lower affinity site Why do different techniques give different K D values for this protein-protein interaction? 12 Solution vs surface effects?
Does amine-coupling impact affinity? SPR using his-tag capture: Barstar*-Y29A vs Binase1 or 2 R59A, Anti-his Fab R59A, Density (RU) K D 1 (nm) K D 2 (um) his-binase2 315 57 33 his-binase2 450 48 29 his-binase2 640 54 31 - Oriented his-binase surface - Both sites = higher affinity compared to amine-coupling. 13
Summary of all data 1E-6 High affinity site 1E-3 Low affinity site 1E-7 1E-4 Av ± SD Kd K (M) 1E-8 Kd (M) 1E-5 1E-9 ITC AUC SPR (amine) SPR (anti-his) 1E-6 ITC AUC SPR (amine) SPR (anti-his) Different affinities determined using solution and surface-based measurements amine-coupling appears to influence affinity because his-capture experiments = higher affinity Surface electrostatic effects may also contribute to lower SPR affinity compared to solution-based measurements See Yamniuk et al, manuscript in preparation 14
Conclusions: Protocols in place for high yield production of barstar and binase proteins Barstar/binase proteins have suitable biophysical properties for use in benchmarking (high solubility, stability, fractional activity) Barstar*-Y29A / binase interaction has affinities in the desired range to test the limits of label-free interaction technologies Affinities appear Challenging to obtain Different for surface (SPR) vs solution (ITC/AUC) binding, but reproducibly obtained using a given technique System t suitable for benchmark study (see MIRG2012 Benchmark study results)? Possible future use as a reference standard for core laboratories? 15
Acknowledgements MIRG Members Simon Bergqvist, Pfizer Mike Doyle, Bristol-Myers Squibb Satya Yadav, Cleveland Clinic Aaron Yamniuk (Chair), Bristol- Myers Squibb Thomas Neubert (EB liaison), New York University School of Medicine Collaborators Ed Eisenstein, University Maryland John Newitt, Bristol-Myers Squibb Fumio Arisaka, Tokyo Institute of Technology Anthony Giannetti, Genentech Inc Preston Hensley, Lotus Translational Medicine David Myszka, University of Utah Fred Schwarz, National Institute of Standards and Technology Interested in joining the MIRG? Please contact Aaron Yamniuk (MIRG Chair) at aaron.yamniuk@bms.com
ABSTRACT Session 1 Introduction to a Protein Interaction System used for Quantitative Evaluation of Biomolecular Interactions - Aaron Yamniuk, Bristol-Myers Squibb A central goal of molecular biology is the determination of biomolecular function. This comes largely from a knowledge of the non-covalent interactions that biological small and macromolecules experience. The fundamental mission of the Molecular Interactions Research Group (MIRG) of the ABRF is to show how solution biophysical tools are used to quantitatively characterize molecular interactions, and to educate the ABRF members and scientific community on the utility and limitations of core technologies such as biosensors, microcalorimetry or analytical ultracentrifugation, for evalution of such interactions. To this end, the MIRG developed a model system for the quantitative analysis of protein-protein interactions, and validated the system using several orthogonal techniques. In this presentation (MIRG Session 1), the details of the protein-protein interaction system will be described, along with an summary of experimental data obtained in MIRG laboratories. The quantitation of the binding affinity was found to be reproducible in multiple MIRG laboratories and using several techniques such as surface plasmon resonance, isothermal titration calorimetry, and analytical ultracentrifugation. In addition, the proteins were shown to be highly stable and soluble, suggesting that the system was suitable for use in benchmarking. Therefore, in late 2012 the MIRG conducted a benchmarking study to evaluate the capabilities of different laboratories and instrumentation to quantitatively characterize the binding affinity and stiochiometry for the model MIRG system, the results of which will be summarized in the subsequent presentation (MIRG Session 2). 17