Understanding signal transmission from an MHC bound peptide via the T-cell receptor to the inside of a T-cell is one of the most important challenges in immunology. Changes in the peptide/MHC interface in reaction to T-cell receptor engagement are of importance as the combined surface of peptide and MHC is the only site where the T-cell physically interacts with the peptide antigen and hence this site is likely to be initial trigger of TCR signalling. We present by far the largest set of Molecular Dynamics simulations investigating the effects of T-cell receptor binding on the dynamics of peptide/MHC. We found that simulations with and without T-cell receptor have conserved differences in H-bonds, flexibility, and solvent accessible surface area indicating that the MHC binding groove can open and close only if no T-cell receptor is present.
The binding between potentially pathogenic peptides and MHCs is one of the most important processes in the human immune system. However, no approach has yet been able to give structural insight into how peptides detach from the MHC. We allow this insight for the first time by combining coarse graining, hierarchical Monte Carlo, temperature modulation, and stochastic conformational optimization. Properties of resulting peptide detachment trajectories show a high agreement with experimental measurements. Understanding peptide/MHC detachment processes has broad implications in the treatment of many human diseases such as allergies or cancer.
H. Zhang, H. Lim, B. Knapp, C. M. Deane, M. Aleksic and P. A. van der MerweC. The contribution of major histocompatibility complex contacts to the affinity and kinetics of T cell receptor binding. Scientific Report, accepted, 2016.
The interaction between the TCR and antigenic peptide in complex with MHCs is a crucial step in T cell activation. The relative contributions of TCR:peptide and TCR:MHC contacts to the overall binding energy remained unclear. This has important implications for our understanding of T cell development and function. In this study we used site directed mutagenesis and computational simulations to estimate the contribution of HLA-A2 side-chains to the binding of four different TCRs. Our results show that these TCRs have very different energetic 'footprints' on HLA-A2.
K. Krawczyk, A. Sim, B. Knapp, C. M. Deane, P. MinaryC. Tertiary element interaction in HIV-1 TAR. Journal of Chemical Information and Modeling. 56 (9), p 1746–1754. 2016.
HIV-1 replication requires binding to occur between Trans-activation Response Element (TAR) RNA and the TAT protein. This TAR-TAT binding depends on the conformation of TAR, and therapeutic development has attempted to exploit this dynamic behavior. Here we hierachical Move Monte Carlo algorithms to simulate TAR dynamics in the context of mutations inhibiting TAR binding. We found that two tertiary elements, the apical loop and the bulge, can interact directly, and this interaction may be linked to the affinity of TAR for TAT.
S. Demharter, B. Knapp, C. M. Deane, P. MinaryC. Modelling Functional Motions of Biological Systems by Customised Natural Moves. Biophysical Journal. 111(4): p710–721. 2016.
In this methodological study we show the application of hierarchical Move Monte Carlo (HMMC) on a protein and DNA system. HMMC simulations are highly efficient conformational sampling method with built-in customization capabilities that allows researchers to design and perform a large number of simulations to investigate functional motions in biological systems within reasonable runtime. Firstly, we investigated the plasticity of an MHC-II in the absence of a bound peptide. Secondly, we study the effects of the epigenetic mark 5-hydroxymethyl on cytosine on the structure of the Dickerson-Drew dodecamer. We show how our customized natural moves protocol can be used to investigate causal relationships of functional motions in biological systems.
R. EsmaielbeikiC, K. Krawczyk, B. Knapp, J. Nebel#, C. M. Deane#. Progress and Challenges in Predicting Protein-Protein Interfaces. Briefings in Bioinformatics. 17 (1): 117-131. 2016. http://dx.doi.org/10.1093/bib/bbv027
The majority of biological processes are mediated via protein-protein interactions. In this work we review the current state of the art in the field of computational protein-protein complex prediction. We categorise, compare, and discuss available computational methods.
S. Wan#, B. Knapp#, D. Wright, C. Deane, P. CoveneyC: Rapid, Precise and Reproducible Prediction of Peptide-MHC Binding Affinities from Molecular Dynamics that Correlate Well with Experiment. Journal of Chemical Theory and Computation. 11 (7), 3346–3356. 2015. http://dx.doi.org/10.1021/acs.jctc.5b00179
Presentation of potentially pathogenic peptides by MHC molecules is one of the most important processes in the adaptive immune defence. Machine learning techniques to predict peptide/MHC binding achieve excellent results but are highly depended the availability of training data. In this study we implemented a training-free Molecular Mechanics Poisson-Boltzmann surface area (MMPBSA) binding free energy calculation. For a diverse set of peptides we achieve a correlation of up to 0.91 with experimental data.
B. KnappC, R. Bardenet, M. O. Bernabeu, R. Bordas, M. Bruna, B. Calderhead, J. Cooper, A. G. Fletcher, D. Groen, B. Kuijper, J. Lewis, G. McInerny, T. Minssen, J. Osborne, V. Paulitschke, J. Pitt-Francis, J. Todoric, C. A. Yates, D. Gavaghan, C. M. Deane. Ten simple rules for a successful cross-disciplinary collaboration. PLoS Computational Biology. 11(4): e1004214. 2015. http://dx.doi.org/10.1371/journal.pcbi.1004214
In this work we combine the experience of researchers from Oxford, Cambridge, and UCL to describe guidelines for collaborations between scientist from different fields. We outline major benefits/pitfalls, and show how to make the most out of cross-disciplinary settings for the benefit of all partners. The article was reported by the Times Higher Education “The secrets to successful interdisciplinary work” (https://www.timeshighereducation.co.uk/news/people/the-secrets-to-succes...)
V. PaulitschkeC, W. Berger, P. Paulitschke, E. Hofstätter, B. Knapp, R. Dingelmaier-Hovorka, D. Födinger, W. Jäger, T. Szekeres, A. Meshcheryakova, A. Bilek, N. Floimayr, H. Pehamberger, C. Gerner, R. Kunstfeld: Vemurafenib resistance signature by proteome analysis offers new strategies and rational therapeutic concepts. Molecular Cancer Therapeutics. 14(3):757-68, 2015. http://dx.doi.org/10.1158/1535-7163.MCT-14-0701
Vemurafenib is an effective late-stage caner drug for melanoma patients. However, Vemurafenib resistance is common. In this study we characterised the Vemurafenib-resistant M24met melanoma cell line in comparison to the Vemurafenib-sensitive A375 melanoma cell line. Using a wide range of methods we described key features of Vemurafenib resistance.
B. KnappC, S. Demharter, R. Esmaielbeiki, C. M. Deane. Current Status and Future Challenges in (TCR)pMHC Molecular Dynamics Simulations. Briefings in Bioinformatics. 16(6), 1035-1044. 2015. http://dx.doi.org/10.1021/acs.jcim.5b00511
In this review we summarised and compared recent Molecular Dynamics simulations of T cell receptors, peptides, and MHC complexes. We show that studies often disagree. This might be due to a small numbers of simulations per study and the problem of multiple testing. We give recommendations on how to avoid this challenge in future studies (see also the paper below).
B. KnappC, J. Dunbar, C.M. Deane: Large Scale Characterization of the LC13 TCR and HLA-B8 Structural Landscape in Reaction to 172 Altered Peptide Ligands: A Molecular Dynamics Simulation Study. PLoS Computational Biology, 10(8):e1003748, 2014 (paper selected for the front page). http://dx.doi.org/10.1371/journal.pcbi.1003748
The interaction between T cell receptor, peptide, and MHC is one of the most important processes in the adaptive human immune response. In this study we described the by far largest set of Molecular Dynamics simulations of T cell receptor / peptide / MHC interactions. We showed that more and less immunogenic peptides differ in their hydrogen bond networks and interface distances. We further showed that many previously suggested dynamical and structural properties of the TCR/peptide/MHC interaction are unlikely to be casual for peptide immunogenicity. Our findings will be the basis to make peptide immunogenicity predictable for the first time. This is of paramount interest for the treatment of allergies, cancer, and vaccine design.
The paper was selected for the front page of PLoS Comp Biol:
H. Wilman, J. P. Ebejer, J. Shi, C. M. Deane, B. KnappC: Crowdsourcing yields a new standard for kinks in protein helices. Journal of Chemical Information and Modeling, 22; 54(9):2585-93, 2014. http://dx.doi.org/10.1021/ci500403a Kinks in protein helices are known to be important sites for signal transductions. However, current computational methods disagree on the definition of a kink. Using a crowd sourcing approach we coined a new gold standard to make kinks in protein helices comparable. This comparability is of high interest for drug design.
J. DunbarC, B. Knapp, A. Fuchs, J. Shi; C.M. Deane: Examining variable domain orientations in antigen receptors gives insight into TCR-like antibody design. PLoS Computational Biology, 18;10(9):e1003852., 2014. http://dx.doi.org/10.1371/journal.pcbi.1003852
Antibody and T cell receptor structures are highly similar. In this study we presented the first systematic comparison between those two structure types. By comparing the relative chain orientation between VH/VL and Vβ/Vα we found that antibodies and T cell receptors, while being highly similar, bind in distinct ways. We found the packing of the long Vα CDR3 as casual for the difference in binding mode. Side-chain substitutions in this area can make antibodies bind in a TCR like mode. This is of impact for the engineering of therapeutic TCR-like antibodies.
H. Dien, C.M. Deane, B. KnappC: Gro2mat: A package to efficiently read Gromacs output in Matlab: Journal of Computational Chemistry, 35(20) 1528-1531, 2014. http://dx.doi.org/10.1002/jcc.23650
In this paper we presented the first library that connects the Molecular Dynamics package Gromacs with the statistics package Matlab. Our package speeds up the data transfer between the two packages by several orders of magnitude up. This allows for analysis methods that have not been possible before.
J. OsborneC, M. Bernabeu, M. Bruna, B. Calderhead, J. Cooper, N. Dalchau, S. Dunn, A. Fletcher, R. Freeman, D. Groen, B. Knapp, G. McInerny, G. Mirams, J. Pitt-Francis, B. Sengupta, D. Wright, C. Yates, D. Gavaghan, S. Emmott, C. M. Deane: Ten Simple Rules for Effective Computational Research. PLoS Computational Biology, 10(3): e1003506, 2014. http://dx.doi.org/10.1371/journal.pcbi.1003506
In this work we combined the broad experience of bioinformaticians and computational biologists from Oxford, Cambridge, and UCL to provide guidelines for computer-based research in computational biology.
J. Schwaiger, J. H. Aberle, K. Stiasny, B. Knapp, W. Schreiner, I. Fae, G. Fischer, O. Scheinost, V. Chmelik, F. X. HeinzC: Specificities of Human CD4 + T cell Responses to an Inactivated Flavivirus Vaccine and Infection: Correlation with Structure and Epitope Prediction. Journal of Virology, 88(14):7828-42, 2014. http://dx.doi.org/10.1128/JVI.00196-14
Tick-borne encephalitis virus is endemic in many parts of the world. In this study we combined experimental and computational methods to investigate the immunogenicity of viral proteins involved in tick-borne encephalitis. We show that computational tools have a high accuracy but also limits when predicting the outcome of patient specific immune reaction data.
B. Hischenhuber, H. Havlicek, J. Todoric, S. Hoellriegl-Binder, W. Schreiner, B. KnappC: Differential geometric analysis of alterations in MH alpha-helices. Journal of Computational Chemistry, 34(21): 1862–1879, 2013. (paper selected for the cover). http://dx.doi.org/10.1002/jcc.23328
Helices are an important secondary structure in proteins. In this study we developed a software package to characterize the alpha helices flanking the MHC binding groove by means of differential geometric parameters such as curvature, torsion, ruled surface area, conical curvature and distribution parameter. This is the first package that allows for the systematic and reliable comparison of MHC structures. This is of high impact to the in silico design of specific T cell receptors as well as peptide vaccines. The paper was selected for the cover of JCC:
J. Todoric, A. Handisurya, B. Knapp, A. Tura, G. Pacini, A. Kautzky-WillerC: Relationship of pentraxin 3 with insulin sensitivity in gestational diabetes. European Journal of Clinical Investigation, 43(4):341-9, 2013. http://dx.doi.org/10.1111/eci.12051
Pentraxin 3 is a cytokine-inducible molecule involved in a variety of inflammatory conditions. In this study we showed for the first time that the serum concentrations are associated with glucose levels in pregnant women and that this association is stronger in patients with gestational diabetes mellitus.
B. KnappC, G. Dorffner, W. Schreiner: Early relaxation dynamics in the LC 13 T cell receptor in reaction to 172 altered peptide ligands: A molecular dynamics simulations study. PLoS ONE, 6;8(6):e64464, 2013. http://dx.doi.org/10.1371/journal.pone.0064464
The interaction between T cell receptor, peptide, and MHC is one of the most important processes in the adaptive human immune response. In this study we showed for the first time that the initial relaxation dynamics of of more and less immunogenic peptides differ statistically significant from each other during Molecular Dynamics simulations. This study is the proof of concept that peptide immunogenicity can be predicted based on the structural dynamics of a T-cell receptor, peptide, MHC complex.
A. HaschemiC, P. Kosma, L. Gille, C. R. Evans, C. F. Burant, P. Starkl, B.Knapp, R. Haas, J. A. Schmid, C. Jandl, S. Amir, G. Lubec, J. Park, H. Esterbauer, M. Bilban, L. Brizuela, J. A. PospisilikC, L. E. Otterbein, O.Wagner. The Sedoheptulose Kinase CARKL Directs Macrophage Polarization Through Control of Glucose Metabolism. Cell Metabolism, 15(6), 813–826, 2012. http://dx.doi.org/10.1016/j.cmet.2012.04.023
In this study we screened for novel regulators of macrophage activation. We found nonprotein kinases of glucose metabolism among the most enriched classes of candidate immune modulators. We also found that one of these, the carbohydrate kinase-like protein CARKL, is rapidly down regulated in vitro and in vivo upon LPS stimulation. We showed that CARKL-dependent metabolic reprogramming is required for proper M1- and M2-like macrophage polarization and uncover a rate-limiting requirement for appropriate glucose flux in macrophage polarization.
B. Knapp, G. Fischer, D. Van,Hemelen, I. Fae, B. Maillere, C. Ebner, W. Schreiner, B. Bohle, B. Jahn-SchmidC: Association of HLA-DR1 with the allergic response to the major mugwort pollen allergen: molecular background. BMC Immunology, 8;13(1):43, 2012. Flagged by the journal as “Highly accessed”. http://dx.doi.org/10.1186/1471-2172-13-43
Mug pollen contains major allergens often resulting in allergic symptoms in humans. The frequency of HLA-DR1 is highly increased in mugwort-allergic individuals. However, the major mug pollen allergen also binds strongly to HLA-DR4 but this allele is not increased in allergic individuals. We investigated the background for this behaviour by a combination of experimental and computational techniques. We demonstrated that differences in peptide mobility within the HLA-DR complex might be the reason why HLA-DR1, but not -DR4 permits successful T cell activation.
J. TodoricC, A. Handisurya, T. Perkmann, B. Knapp, O. Wagner, A. Tura, G. Pacini, H. Esterbauer, A. Kautzky-Willer: Circulating Progranulin Levels in Women with Gestational Diabetes Mellitus and Healthy Controls During and After Pregnancy, European Journal of Endocrinology, 167(4):561-7, 2012. http://dx.doi.org/10.1530/EJE-12-0060
Progranulin is a marker of chronic inflammatory response in obesity and type 2 diabetes which is capable of directly affecting the insulin signaling pathway. In this study we showed that Progranulin concentrations were significantly higher in pregnant women compared to post partum levels. However, the gestational glucose tolerance state (gestational diabetes mellitus vs normal glucose tolerance) did not influence the Progranulin concentrations.
B. Hischenhuber, F. Frommlet, W. Schreiner, B. KnappC: MH²c: Characterization of major histocompatibility α-helices - An Information Criterion Approach. Computer Physics Communications, 183 (7), 1481–1490, 2012. (paper was ranked 1st in the list of the most downloaded papers of Computer Physics Communications within the last 90 days (June 2012). http://dx.doi.org/10.1016/j.cpc.2012.02.008
Helices are important secondary structures in proteins. In this study we developed a software package to characterize and compare these helices. Using information criteria we compared different curve and spline models and thereby provide a package to reliably characterize helices for a broad range of applications.
The convergence and non-convergence of Molecular Dynamics simulations is a frequently discussed topic in the literature. Convergence of a simulation can never be proofed. Here we presented a new method that determines non-convergence. For this purpose we used a matrix of RMSD comparisons of each structure in the trajectory with respect to each other structure. Unless a lagged comparison of these values has not reached stationary shape a simulation is unlikely to have converged.
Genetic Algorithms are a heuristic methods to solve complex problems. In this study we developed a software package that employs Genetic Algorithms to create sets of high affinity binders for MHC. Starting from random peptides our algorithm is able to create high affinity binders for various different scoring functions within 10 generations. This tool allows for the high throughput design of high affinity binders.
B. KnappC, S. Frantal, M. Cibena, W. Schreiner, P. Bauer. Is an intuitive convergence definition of Molecular Dynamics simulations solely based on the Root Mean Square Deviation possible? Journal of Computational Biology, 18(8):997-1005, 2011. http://dx.crossref.org/10.1089/cmb.2010.0237
The convergence and non-convergence of Molecular Dynamics simulations is a frequently discussed topic in the literature. In this study we showed that the commonly used RMSD is incapable of giving an indication for convergence of Molecular Dynamics simulations.
Myelin basic protein is believed to be important in the process of myelination of nerves and therefore of high relevance in multiple sclerosis. In this study we investigated the molecular background of different peptides originating from myelin basic protein. We combined in vitro, in vivo, and in silico data. We found that irrespective of peptide binding affinity, MHC deformation appears to influence costimulation, which then leads to effective T cell priming and disease induction.
B. KnappC, N. Lederer, U. Omasits, W. Schreiner. vmdICE: A plug-in for rapid evaluation of molecular dynamics simulations using VMD. Journal of Computational Chemistry, 31:2868–2873, 2010. http://dx.doi.org/10.1002/jcc.21581
In this paper we presented a software tool that allows to display measuremetns obtained from Molecular Dynamics simulation analysis directly on a 3D representation of a molecule. The package was written as a plug-in for the popular software package VMD. By this way we provide and easy to use and handy way to investigate Molecular Dynamics simulation trajectories interactively.
S. RoopraC#, B. Knapp#, U. Omasits, W. Schreiner: jSimMacs for GROMACS: a Java application for advanced Molecular Dynamics simulations with remote access capability. Journal of Chemical Information and Modeling, 49 (10), pp 2412–2417, 2009. http://dx.doi.org/10.1021/ci900248f
In this paper we presented a graphical user interface for the Molecular Dynamics package Gromacs. Our user interface offers an easy to use way to setup a simulation. The user interfaces provides an interactive 3D representation of the molecule and the possibility to submit the simulation to a cluster. This program makes Molecular Dynamics simulations easy to handle for inexperienced users.
B. KnappC, U. Omasits, B. Bohle, B. Maillere, C. Ebner, W. Schreiner, B. Jahn-Schmid: 3-Layer-based analysis of peptide-MHC-interaction: in silico prediction, peptide binding affinity and T cell activation in a relevant allergen-specific model. Molecular Immunology, 46, 1839-1844. 2009. http://dx.doi.org/10.1016/j.molimm.2009.01.009
In contrast to MHCI the binding groove of MHCII is not closed at its ends. The effect of the peptide flanking regions sticking out of the MHCII binding groove is controversially discussed in the literature. In this study we used a combination of experimental and computational methods to investigate peptide flanking regions. We found that the binding affinity, the immunogenicity and the structural dynamics of the major mug pollen allergen presented by HLA-DR1 is highly sensitive to the length of its peptide flanking regions.
B. KnappC, U. Omasits, S. Frantal, W. Schreiner: A critical cross-validation of high throughput structural binding prediction methods for pMHC. Journal of Computer Aided Molecular Design, 23, 301-307. 2009. http://dx.doi.org/10.1007/s10822-009-9259-2
The prediction of peptide/MHC binding affinity is an important topic in immune-informatics. In this study we investigated combinations of structural modelling and docking tools for their performance in predicting peptide/MHC binding affinity. We found that SCATD in combination with XSCORE archives the highest accuracy. Our results allow for fully automated structural binding predictions between peptides and MHCs.
In this article we discussed the benefits and caveats of the development of graphical user interfaces for Molecular Dynamics simulations. We outline in which respects a broader access of MD via graphical user interfaces may help to increase the usability of Molecular Dynamics simulations while maintaining their quality.
The accurate modelling of peptide/MHC complexes is of high interest in immunology as experimental X-ray structures are not available for all peptide/MHC combinations. In this study we benchmarked different methods for their applicability in the peptide/MHC interface. We showed that SCWRL yields the highest accuracy when benchmarked against X-ray structures. These results allow for high-throughput modelling for peptide/MHC complexes. The paper was selected for the cover of Protein Science:
U. Omasits, B. KnappC, M. Neumann, O. Steinhauser, H. Stockinger, R. Kobler, W. Schreiner: Analysis of key parameters for molecular dynamics of pMHC molecules. Molecular Simulation, 34 (8), 781-793. 2008. http://dx.doi.org/10.1080/08927020802256298
Reliable Molecular Dynamics studies require a large set of carefully chosen parameters. In this study we tested the effects of many such parameters in the peptide/MHC system. On the basis of this study further simulations can be run straightforwardly.