Olivier Michielin

The adoptive transfer of T cell receptor–engineered (TCR-engineered) T cells (ACT) targeting the HLA-A2–restricted cancer-testis epitope NY-ESO-1 _157–165 (A2/NY) has yielded favorable clinical responses against several cancers. Two approaches to improve ACT are TCR affinity optimization and T cell coengineering to express immunomodulatory molecules that can exploit endogenous immunity. By computational design we previously developed a panel of binding-enhanced A2/NY-TCRs including A97L, which augmented the in vitro function of gene-modified T cells as compared with WT. Here, we demonstrated higher persistence and improved tumor control by A97L–T cells. In order to harness macrophages in tumors, we further coengineered A97L–T cells to secrete a high-affinity signal regulatory protein α (SiRPα) decoy (CV1) that blocks CD47. While CV1-Fc–coengineered A97L–T cells mediated significantly better control of tumor outgrowth and survival in Winn assays, in subcutaneous xenograft models the T cells, coated by CV1-Fc, were depleted. Importantly, there was no phagocytosis of CV1 monomer–coengineered T cells by human macrophages. Moreover, avelumab and cetuximab enhanced macrophage-mediated phagocytosis of tumor cells in vitro in the presence of CV1 and improved tumor control upon coadministration with A97L–T cells. Taken together, our study indicates important clinical promise for harnessing macrophages by combining CV1-coengineered TCR–T cells with targeted antibodies to direct phagocytosis against tumor cells.

Immune checkpoints are crucial in the maintenance of antitumor immune responses. The activation or blockade of immune checkpoints is dependent on the interactions between receptors and ligands; such interactions can provide inhibitory or stimulatory signals, including the enhancement or suppression of T-cell proliferation, differentiation, and/or cytokine secretion. B-and T-lymphocyte attenuator (BTLA) is a lymphoid-specific cell surface receptor which is present on T-cells and interacts with herpes virus entry mediator (HVEM), which is present on tumor cells. The binding of HVEM to BTLA triggers an inhibitory signal which attenuates the immune response. This feature is interesting for studying the molecular interactions between HVEM and BTLA, as they may be targeted for novel immunotherapies. This work was based on the crystal structure of the BTLA/HVEM complex showing that BTLA binds the N-terminal cysteine-rich domain of HVEM. We investigated the amino acid sequence of HVEM and used molecular modeling methods to develop inhibitors of the BTLA/HVEM interaction. We synthesized novel compounds and determined their ability to interact with the BTLA protein and inhibit the formation of the BTLA/HVEM complex. Our results suggest that the HVEM (14–39) peptide is a potent inhibitor of the formation of the BTLA/HVEM protein complex.

SwissTargetPrediction is a web tool, on-line since 2014, that aims to predict the most probable protein targets of small molecules. Predictions are based on the similarity principle, through reverse screening. Here, we describe the 2019 version, which represents a major update in terms of underlying data, backend and web interface. The bioactivity data were updated, the model retrained and similarity thresholds redefined. In the new version, the predictions are performed by searching for similar molecules, in 2D and 3D, within a larger collection of 376 342 compounds known to be experimentally active on an extended set of 3068 macromolecular targets. An efficient backend implementation allows to speed up the process that returns results for a druglike molecule on human proteins in 15–20 s. The refreshed web interface enhances user experience with new features for easy input and improved analysis. Interoperability capacity enables straightforward submission of any input or output molecule to other on-line computer-aided drug design tools, developed by the SIB Swiss Institute of Bioinformatics. High levels of predictive performance were maintained despite more extended biological and chemical spaces to be explored, e.g. achieving at least one correct human target in the top 15 predictions for >70% of external compounds. The new SwissTargetPrediction is available free of charge (www.swisstargetprediction.ch).

SwissSimilarity is a new web tool for rapid ligand-based virtual screening of small to unprecedented ultralarge libraries of small molecules. Screenable compounds include drugs, bioactive and commercial molecules, as well as 205 million of virtual compounds readily synthesizable from commercially available synthetic reagents. Predictions can be carried out on-the-fly using six different screening approaches, including 2D molecular fingerprints as well as superpositional and fast nonsuperpositional 3D similarity methodologies. SwissSimilarity is part of a large initiative of the SIB Swiss Institute of Bioinformatics to provide online tools for computer-aided drug design, such as SwissDock, SwissBioisostere or SwissTargetPrediction with which it can interoperate, and is linked to other well-established online tools and databases. User interface and backend have been designed for simplicity and ease of use, to provide proficient virtual screening capabilities to specialists and nonexperts in the field. SwissSimilarity is accessible free of charge or login at http://www.swisssimilarity.ch.

Molecular docking is a computational approach for predicting the most probable position of ligands in the binding sites of macromolecules and constitutes the cornerstone of structure‐based computer‐aided drug design. Here, we present a new algorithm called Attracting Cavities that allows molecular docking to be performed by simple energy minimizations only. The approach consists in transiently replacing the rough potential energy hypersurface of the protein by a smooth attracting potential driving the ligands into protein cavities. The actual protein energy landscape is reintroduced in a second step to refine the ligand position. The scoring function of Attracting Cavities is based on the CHARMM force field and the FACTS solvation model. The approach was tested on the 85 experimental ligand–protein structures included in the Astex diverse set and achieved a success rate of 80% in reproducing the experimental binding mode starting from a completely randomized ligand conformer. The algorithm thus compares favorably with current state‐of‐the‐art docking programs. © 2015 The Authors. Journal of Computational Chemistry Published by Wiley Periodicals, Inc.

Motivation: Most bioactive molecules perform their action by interacting with proteins or other macromolecules. However, for a significant fraction of them, the primary target remains unknown. In addition, the majority of bioactive molecules have more than one target, many of which are poorly characterized. Computational predictions of bioactive molecule targets based on similarity with known ligands are powerful to narrow down the number of potential targets and to rationalize side effects of known molecules. Results: Using a reference set of 224 412 molecules active on 1700 human proteins, we show that accurate target prediction can be achieved by combining different measures of chemical similarity based on both chemical structure and molecular shape. Our results indicate that the combined approach is especially efficient when no ligand with the same scaffold or from the same chemical series has yet been discovered. We also observe that different combinations of similarity measures are optimal for different molecular properties, such as the number of heavy atoms. This further highlights the importance of considering different classes of similarity measures between new molecules and known ligands to accurately predict their targets. Contact:olivier.michielin@unil.ch or vincent.zoete@unil.ch Supplementary information: Supplementary data are available at Bioinformatics online.

In the Gac/Rsm signal transduction pathway of Pseudomonas fluorescens CHA0, the dimeric RNA-binding proteins RsmA and RsmE, which belong to the vast bacterial RsmA/CsrA family, effectively repress translation of target mRNAs containing a typical recognition sequence near the translation start site. Three small RNAs (RsmX, RsmY, RsmZ) with clustered recognition sequences can sequester RsmA and RsmE and thereby relieve translational repression. According to a previously established structural model, the RsmE protein makes optimal contacts with an RNA sequence 5′-A/UCANGGANGU/A-3′, in which the central ribonucleotides form a hexaloop. Here, we questioned the relevance of the hexaloop structure in target RNAs. We found that two predicted pentaloop structures, AGGGA (in pltA mRNA encoding a pyoluteorin biosynthetic enzyme) and AAGGA (in mutated pltA mRNA), allowed effective interaction with the RsmE protein in vivo. By contrast, ACGGA and AUGGA were poor targets. Isothermal titration calorimetry measurements confirmed the strong binding of RsmE to the AGGGA pentaloop structure in an RNA oligomer. Modeling studies highlighted the crucial role of the second ribonucleotide in the loop structure. In conclusion, a refined structural model of RsmE-RNA interaction accommodates certain pentaloop RNAs among the preferred hexaloop RNAs.

Adoptive cell transfer using engineered T cells is emerging as a promising treatment for metastatic melanoma. Such an approach allows one to introduce TCR modifications that, while maintaining the specificity for the targeted antigen, can enhance the binding and kinetic parameters for the interaction pMHC. Using the well-characterized 2C TCR/SIYR/H-2K(b) structure as a model system, we demonstrated that a binding free energy decomposition based on the MM-GBSA approach provides a detailed and reliable description of the TCR/pMHC interactions at the structural and thermodynamic levels. Starting from this result, we developed a new structure-based approach, to rationally design new TCR sequences, and applied it to the BC1 TCR targeting the HLA-A2 restricted NY-ESO-1157-165 cancer-testis epitope. 54% of the designed sequence replacements exhibited improved pMHC-binding as compared to the native TCR, with up to 150 fold increase in affinity, while preserving specificity. Genetically-engineered CD8+ T cells expressing these modified TCRs showed an improved functional activity compared to those expressing BC1 TCR. We measured maximum levels of activities for TCRs within the upper limit of natural affinity. Beyond the affinity threshold at KD < 1 μM we observed an attenuation in cellular function. We have also developed a homology modeling-based approach, TCRep 3D, to obtain accurate structural models of any TCR-pMHC complexes. We have complemented the approach with a simplified rigid method to predict the TCR orientation over pMHC. These methods potentially extend the use of our TCR engineering method to entire TCR repertoires for which no X-ray structure is available. We have also performed a steered molecular dynamics study of the unbinding of the TCR-pMHC complex to get a better understanding of how TCRs interact with pMHCs. This entire rational TCR design pipeline is now being used to produce rationally optimized TCRs for adoptive cell therapies of stage IV melanoma.

Crystallographic data about T-Cell Receptor – peptide – major histocompatibility complex class I (TCRpMHC) interaction have revealed extremely diverse TCR binding modes triggering antigen recognition. Understanding the molecular basis that governs TCR orientation over pMHC is still a considerable challenge. We present a simplified rigid approach applied on all non-redundant TCRpMHC crystal structures available. The CHARMM force field in combination with the FACTS implicit solvation model is used to study the role of long-distance interactions between the TCR and pMHC. We demonstrate that the sum of the coulomb interactions and the electrostatic solvation energies is sufficient to identify two orientations corresponding to energetic minima at 0° and 180° from the native orientation. Interestingly, these results are shown to be robust upon small structural variations of the TCR such as changes induced by Molecular Dynamics simulations, suggesting that shape complementarity is not required to obtain a reliable signal. Accurate energy minima are also identified by confronting unbound TCR crystal structures to pMHC. Furthermore, we decompose the electrostatic energy into residue contributions to estimate their role in the overall orientation. Results show that most of the driving force leading to the formation of the complex is defined by CDR1,2/MHC interactions. This long-distance contribution appears to be independent from the binding process itself, since it is reliably identified without considering neither short-range energy terms nor CDR induced fit upon binding. Ultimately, we present an attempt to predict the TCR/pMHC binding mode for a TCR structure obtained by homology modeling. The simplicity of the approach and the absence of any fitted parameters make it also easily applicable to other types of macromolecular protein complexes.

Amino acids form the building blocks of all proteins. Naturally occurring amino acids are restricted to a few tens of sidechains, even when considering post-translational modifications and rare amino acids such as selenocysteine and pyrrolysine. However, the potential chemical diversity of amino acid sidechains is nearly infinite. Exploiting this diversity by using non-natural sidechains to expand the building blocks of proteins and peptides has recently found widespread applications in biochemistry, protein engineering and drug design. Despite these applications, there is currently no unified online bioinformatics resource for non-natural sidechains. With the SwissSidechain database (http://www.swisssidechain.ch), we offer a central and curated platform about non-natural sidechains for researchers in biochemistry, medicinal chemistry, protein engineering and molecular modeling. SwissSidechain provides biophysical, structural and molecular data for hundreds of commercially available non-natural amino acid sidechains, both in l- and d-configurations. The database can be easily browsed by sidechain names, families or physico-chemical properties. We also provide plugins to seamlessly insert non-natural sidechains into peptides and proteins using molecular visualization software, as well as topologies and parameters compatible with molecular mechanics software.