P03.01 High immunogenic VLP-based vaccines elicit new T cell specificities against melanoma neoantigens in mice

BackgroundNeoantigens’ (neoAg) identification, which determines T-cell responses against tumors, has fostered the development of personalized vaccines with promising results. While the ranking of the most immunogenic neoAg can be addressed using predictive techniques, their formulation as vaccines needs to be improved. To maximize their therapeutic potential, optimal neoAg-based vaccines should be manufactured in a superb delivery platform that enhances robust new immune responses, able to bypass thymic tolerance and the humoral immunosuppressive microenvironment. These novel T cell responses generated at the periphery will not be exhausted, opposite to TILs. We aim to develop a highly immunogenic vaccine platform, based on engineered HIV-derived Virus-Like Particles (VLP) expressing approximately 2500 copies of each selected neoAg. We tested different neoAgs loaded VLPs (neoVLP) in a melanoma mouse model to evaluate their capability to generate new immunogenic specificities.Material and MethodsSpecific non-synonymous mutations from B16F10 cells were identified, selected and used to generate a list of prioritized peptides. NeoAgs were classified as: Tier1, acquiring a mutation that creates an anchor residue to the MHC-I, not present in the WT peptide; Tier2, acquiring a mutation in a position that largely impacts contact with the TCR respect to WT; and Tier3, acquiring a mutation in the TCR contact region but inducing a less drastic change than in Tier2. Frame shift (FS) mutations, expected to be highly immunogenic, were also included.Thirteen to fifteen selected neoAgs from each group were loaded on highly immunogenic neoVLPs. Their immunogenicity was evaluated in C57bl/6 mice by immunization with a neoVLP-coding plasmid DNA (prime) and purified neoVLPs as soluble particles (boost). Splenocytes were used to evaluate neoAg-specific T cell responses.ResultsWe have successfully generated and purified neoVLPs, exposing neoAgs from all groups by transient transfection of Expi293 cells. Protein integrity and VLP morphology were confirmed by western blot and cryo-EM. When used for immunization assays, neoVLPs, containing neoAgs from Tier2, Tier3 and FS groups, were capable of generating humoral responses against viral proteins and T cell responses against neoAgs present in the neoVLP. B16F10 inoculated animals, but not vaccinated, did not develop detectable T cell responses against neoAgs present in any tested neoVLP, suggesting that the vaccination with neoVLPs promoted new specificities against selected neoAgs that might contribute to tumor control and eradication.ConclusionOur data show that the neoVLPs promote the generation of new antitumor-specific immune responses against selected neoepitopes, suggesting that neoVLPs vaccination could be an alternative to current therapeutic vaccine approaches and a promising candidate for future personalized immunotherapy.Disclosure InformationC. Aguilar-Gurrieri: None. A. Barajas: None. A. Pons-Grifols: None. I. Varela: None. P. Amengual-Rigo: None. R. Farriol: None. M. Vazquez: None. R. Lepore: None. C. Anjos-Souza: None. J. Blanco-Heredia: None. L. de Mattos-Arruda: None. V. Guallar: None. A. Valencia: None. B. Clotet: F. Consultant/Advisory Board; Significant; Albajuna Therapeutics. J. Carrillo: Other; Significant; Albajuna Therapeutics. J. Blanco: F. Consultant/Advisory Board; Significant; Albajuna Therapeutics.

Towards guided mutagenesis: Gaussian process regression predicts MHC class II antigen mutant binding

Antigen-specific immunotherapies (ASI) require successful loading and presentation of antigen peptide into the major histocompatibility complex (MHC) binding cleft. One route of ASI design is to mutate native antigens for either stronger or weaker binding interaction to MHC. Exploring all possible mutations is costly both experimentally and computationally. To reduce experimental and computational expense, here we investigate the minimal amount of prior data required to accurately predict the relative binding affinity of point mutations for peptide-MHC class II (pMHCII) binding. Using data from different residue subsets, we interpolate pMHCII mutant binding affinities by Gaussian process (GP) regression of residue volume and hydrophobicity. We apply GP regression to an experimental dataset from the Immune Epitope Database, and theoretical datasets from NetMHCIIpan and Free Energy Perturbation calculations. We find that GP regression can predict binding affinities of 9 neutral residues from a 6-residue subset with an average R2 coefficient of determination value of 0.62 ± 0.04 (± 95% CI), average error of 0.09 ± 0.01 kcal/mol (± 95% CI), and with an ROC AUC value of 0.92 for binary classification of enhanced or diminished binding affinity. Similarly, metrics increase to an R2 value of 0.69 ± 0.04, average error of 0.07 ± 0.01 kcal/mol, and an ROC AUC value of 0.94 for predicting 7 neutral residues from an 8-residue subset. Our work finds that prediction is most accurate for neutral residues at anchor residue sites without register shift. This work holds relevance to predicting pMHCII binding and accelerating ASI design.

Introduction of Non-natural Amino Acids Into T-Cell Epitopes to Mitigate Peptide-Specific T-Cell Responses

Non-natural modifications are widely introduced into peptides to improve their therapeutic efficacy, but their impact on immunogenicity remains largely unknown. As the CD4 T-cell response is a key factor in triggering immunogenicity, we investigated the effect of introducing D-amino acids (Daa), amino isobutyric acid (Aib), N-methylation, Cα-methylation, reduced amide, and peptoid bonds into an immunoprevalent T-cell epitope on binding to a set of HLA-DR molecules, recognition, and priming of human T cells. Modifications are differentially accepted at multiple positions, but are all tolerated in the flanking regions. Introduction of Aib and Daa in the binding core had the most deleterious effect on binding to HLA-DR molecules and T-cell activation. Their introduction at the positions close to the P1 anchor residue abolished T-cell priming, suggesting they might be sufficient to dampen peptide immunogenicity. Other modifications led to variable effects on binding to HLA-DR molecules and T-cell reactivity, but none exhibited an increased ability to stimulate T cells. Altogether, non-natural modifications appear generally to diminish binding to HLA-DR molecules and hence T-cell stimulation. These data might guide the design of therapeutic peptides to make them less immunogenic.

Machine learning optimization of peptides for presentation by class II MHCs

T cells play a critical role in cellular immune responses to pathogens and cancer and can be activated and expanded by MHC-presented antigens contained in peptide vaccines. We present a machine learning method to optimize the presentation of peptides by class II MHCs by modifying their anchor residues. Our method first learns a model of peptide affinity for a class II MHC using an ensemble of deep residual networks, and then uses the model to propose anchor residue changes to improve peptide affinity. We use a high throughput yeast display assay to show that anchor residue optimization improves peptide binding. Supplementary information: Supplementary data are available at Bioinformatics online.

Peptidomes and Structures Illustrate Two Distinguishing Mechanisms of Alternating the Peptide Plasticity Caused by Swine MHC Class I Micropolymorphism

The micropolymorphism of major histocompatibility complex class I (MHC-I) can greatly alter the plasticity of peptide presentation, but elucidating the underlying mechanism remains a challenge. Here we investigated the impact of the micropolymorphism on peptide presentation of swine MHC-I (termed swine leukocyte antigen class I, SLA-I) molecules via immunopeptidomes that were determined by our newly developed random peptide library combined with the mass spectrometry (MS) de novo sequencing method (termed RPLD–MS) and the corresponding crystal structures. The immunopeptidomes of SLA-1*04:01, SLA-1*13:01, and their mutants showed that mutations of residues 156 and 99 could expand and narrow the ranges of peptides presented by SLA-I molecules, respectively. R156A mutation of SLA-1*04:01 altered the charge properties and enlarged the volume size of pocket D, which eliminated the harsh restriction to accommodate the third (P3) anchor residue of the peptide and expanded the peptide binding scope. Compared with 99Tyr of SLA-1*0401, 99Phe of SLA-1*13:01 could not form a conservative hydrogen bond with the backbone of the P3 residues, leading to fewer changes in the pocket properties but a significant decrease in quantitative of immunopeptidomes. This absent force could be compensated by the salt bridge formed by P1-E and 170Arg. These data illustrate two distinguishing manners that show how micropolymorphism alters the peptide-binding plasticity of SLA-I alleles, verifying the sensitivity and accuracy of the RPLD-MS method for determining the peptide binding characteristics of MHC-I in vitro and helping to more accurately predict and identify MHC-I restricted epitopes.

HLA binding of self-peptides is biased towards proteins with specific molecular functions

Human leukocyte antigen (HLA) is highly polymorphic and plays a key role in guiding adaptive immune responses by presenting foreign and self peptides to T cells. Each HLA variant selects a minor fraction of peptides that match a certain motif required for optimal interaction with the peptide-binding groove. These restriction rules define the landscape of peptides presented to T cells. Given these limitations, one might suggest that the choice of peptides presented by HLA is non-random and there is preferential presentation of an array of peptides that is optimal for distinguishing self and foreign proteins. In this study we explore these preferences with a comparative analysis of self peptides enriched and depleted in HLA ligands. We show that HLAs exhibit preferences towards presenting peptides from certain proteins while disfavoring others with specific functions, and highlight differences between various HLA genes and alleles in those preferences. We link those differences to HLA anchor residue propensities and amino acid composition of preferentially presented proteins. The set of proteins that peptides presented by a given HLA are most likely to be derived from can be used to distinguish between class I and class II HLAs and HLA alleles. Our observations can be extrapolated to explain the protective effect of certain HLA alleles in infectious diseases, and we hypothesize that they can also explain susceptibility to certain autoimmune diseases and cancers. We demonstrate that these differences lead to differential presentation of HIV, influenza virus, SARS-CoV-1 and SARS-CoV-2 proteins by various HLA alleles. Finally, we show that the reported self peptidome preferences of distinct HLA variants can be compensated by combinations of HLA-A/HLA-B and HLA-A/HLA-C alleles in frequent haplotypes.

Alterations in the HLA-B*57:01 Immunopeptidome by Flucloxacillin and Immunogenicity of Drug-Haptenated Peptides

Neoantigen formation due to the interaction of drug molecules with human leukocyte antigen (HLA)-peptide complexes can lead to severe hypersensitivity reactions. Flucloxacillin (FLX), a β-lactam antibiotic for narrow-spectrum gram-positive bacterial infections, has been associated with severe immune-mediated drug-induced liver injury caused by an influx of T-lymphocytes targeting liver cells potentially recognizing drug-haptenated peptides in the context of HLA-B*57:01. To identify immunopeptidome changes that could lead to drug-driven immunogenicity, we used mass spectrometry to characterize the proteome and immunopeptidome of B-lymphoblastoid cells solely expressing HLA-B*57:01 as MHC-I molecules. Selected drug-conjugated peptides identified in these cells were synthesized and tested for their immunogenicity in HLA-B*57:01-transgenic mice. T cell responses were evaluated in vitro by immune assays. The immunopeptidome of FLX-treated cells was more diverse than that of untreated cells, enriched with peptides containing carboxy-terminal tryptophan and FLX-haptenated lysine residues on peptides. Selected FLX-modified peptides with drug on P4 and P6 induced drug-specific CD8+ T cells in vivo. FLX was also found directly linked to the HLA K146 that could interfere with KIR-3DL or peptide interactions. These studies identify a novel effect of antibiotics to alter anchor residue frequencies in HLA-presented peptides which may impact drug-induced inflammation. Covalent FLX-modified lysines on peptides mapped drug-specific immunogenicity primarily at P4 and P6 suggesting these peptide sites as drivers of off-target adverse reactions mediated by FLX. FLX modifications on HLA-B*57:01-exposed lysines may also impact interactions with KIR or TCR and subsequent NK and T cell function.

HLA class I genotypes customize vaccination strategies in immune simulation to combat COVID-19

Memory CD8+ T cells are associated with a better outcome in Coronavirus Disease 2019 (COVID-19) and recognized as promising vaccine targets against viral infections. This study determined the efficacy of population-dominant and infection-relevant human leukocyte antigens (HLA) class I proteins to present severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) peptides through calculating binding affinities and simulating CD8+ T cell responses. As a result, HLA class I proteins distinguished or shared various viral peptides derived from viruses. HLA class I supertypes clustered viral peptides through recognizing anchor and preferred residues. SARS-CoV-2 peptides overlapped significantly with SARS but minimally with common human coronaviruses. Immune simulation of CD8+ T cell activation using predicted SARS-CoV-2 peptide antigens depended on high-affinity peptide binding, anchor residue interaction, and synergistic presentation of HLA class I proteins in individuals. Results demonstrated that multi-epitope vaccination, employing a strong binding affinity, viral adjuvants, and heterozygous HLA class I genes, induced potent immune responses. Therefore, optimal CD8+ T cell responses can be achieved and customized contingent on HLA class I genotypes in human populations, supporting a precise vaccination strategy to combat COVID-19.

Machine learning optimization of peptides for presentation by class II MHCs

T cells play a critical role in normal immune responses to pathogens and cancer and can be targeted to MHC-presented antigens via interventions such as peptide vaccines. Here, we present a machine learning method to optimize the presentation of peptides by class II MHCs by modifying the peptide’s anchor residues. Our method first learns a model of peptide affinity for a class II MHC using an ensemble of deep residual networks, and then uses the model to propose anchor residue changes to improve peptide affinity. We use a high throughput yeast display assay to show that anchor residue optimization successfully improved peptide binding.

Structural studies of geranylgeranylglyceryl phosphate synthase, a prenyltransferase found in thermophilic Euryarchaeota

Archaea are uniquely adapted to thrive in harsh environments, and one of these adaptations involves the archaeal membrane lipids, which are characterized by their isoprenoid alkyl chains connected via ether linkages to glycerol 1-phosphate. The membrane lipids of the thermophilic and acidophilic euryarchaeota Thermoplasma volcanium are exclusively glycerol dibiphytanyl glycerol tetraethers. The first committed step in the biosynthetic pathway of these archaeal lipids is the formation of the ether linkage between glycerol 1-phosphate and geranylgeranyl diphosphate, and is catalyzed by the enzyme geranylgeranylglyceryl phosphate synthase (GGGPS). The 1.72 Å resolution crystal structure of GGGPS from T. volcanium (TvGGGPS) in complex with glycerol and sulfate is reported here. The crystal structure reveals TvGGGPS to be a dimer, which is consistent with the absence of the aromatic anchor residue in helix α5a that is required for hexamerization in other GGGPS homologs; the hexameric quaternary structure in GGGPS is thought to provide thermostability. A phylogenetic analysis of the Euryarchaeota and a parallel ancestral state reconstruction investigated the relationship between optimal growth temperature and the ancestral sequences. The presence of an aromatic anchor residue is not explained by temperature as an ecological parameter. An examination of the active site of the TvGGGPS dimer revealed that it may be able to accommodate longer isoprenoid substrates, supporting an alternative pathway of isoprenoid membrane-lipid synthesis.