Energy landscapes of peptide-MHC binding.

Molecules of the Major Histocompatibility Complex (MHC) present short protein fragments on the cell surface, an important step in T cell immune recognition. MHC-I molecules process peptides from intracellular proteins; MHC-II molecules act in antigen-presenting cells and present peptides derived from extracellular proteins. Here we show that the sequence-dependent energy landscapes of MHC-peptide binding encode class-specific nonlinearities (epistasis). MHC-I has a smooth landscape with global epistasis; the binding energy is a simple deformation of an underlying linear trait. This form of epistasis enhances the discrimination between strong-binding peptides. In contrast, MHC-II has a rugged landscape with idiosyncratic epistasis: binding depends on detailed amino acid combinations at multiple positions of the peptide sequence. The form of epistasis affects the learning of energy landscapes from training data. For MHC-I, a low-complexity problem, we derive a simple matrix model of binding energies that outperforms current models trained by machine learning. For MHC-II, higher complexity prevents learning by simple regression methods. Epistasis also affects the energy and fitness effects of mutations in antigen-derived peptides (epitopes). In MHC-I, large-effect mutations occur predominantly in anchor positions of strong-binding epitopes. In MHC-II, large effects depend on the background epitope sequence but are broadly distributed over the epitope, generating a bigger target for escape mutations due to loss of presentation. Together, our analysis shows how an energy landscape of protein-protein binding constrains the target of escape mutations from T cell immunity, linking the complexity of the molecular interactions to the dynamics of adaptive immune response.

Structural basis for mouse LAG3 interactions with the MHC class II molecule I-Ab.

The immune checkpoint protein, Lymphocyte activation gene-3 (LAG3), binds Major Histocompatibility Complex Class II (MHC-II) and suppresses T cell activation. Despite the recent FDA approval of a LAG3 inhibitor for the treatment of melanoma, how LAG3 engages MHC-II on the cell surface remains poorly understood. Here, we determine the 3.84 Å-resolution structure of mouse LAG3 bound to the MHC-II molecule I-Ab, revealing that domain 1 (D1) of LAG3 binds a conserved, membrane-proximal region of MHC-II spanning both the α2 and ß2 subdomains. LAG3 dimerization restricts the intermolecular spacing of MHC-II molecules, which may attenuate T cell activation by enforcing suboptimal signaling geometry. The LAG3-MHC-II interface overlaps with the MHC-II-binding site of the T cell coreceptor CD4, implicating disruption of CD4-MHC-II interactions as a mechanism for LAG3 immunosuppressive function. Lastly, antibody epitope analysis indicates that multiple LAG3 inhibitors do not recognize the MHC-II-binding interface of LAG3, suggesting a role for functionally distinct mechanisms of LAG3 antagonism in therapeutic development.

Redox-dependent plasticity of oxMIF facilitates its interaction with CD74 and therapeutic antibodies.

MIF is a ubiquitous protein involved in proinflammatory processes, which undergoes an oxidation-driven conformational change to oxidized (ox)MIF. We demonstrate that hypochlorous acid, produced by neutrophil-released myeloperoxidase (MPO) under inflammatory conditions, effectively oxidizes MIF into the oxMIF isoform, which is specifically recognized by the anti-oxMIF therapeutic antibody, ON104. NMR investigation of MIF oxidized by the MPO system revealed increased flexibility throughout the MIF structure, including at several catalytic and allosteric sites. Mass spectrometry of MPO-oxMIF revealed methionines as the primary site of oxidation, whereas Pro2 and Tyr99/100 remained almost unmodified. ELISA, SPR and cell-based assays demonstrated that structural changes caused by MPO-driven oxidation promoted binding of oxMIF to its receptor, CD74, which does not occur with native MIF. These data reveal the environment and modifications that facilitate interactions between MIF and its pro-inflammatory receptor, and a route for therapeutic intervention targeting the oxMIF isoform.

Structural insights into human MHC-II association with invariant chain.

The loading of processed peptides on to major histocompatibility complex II (MHC-II) molecules for recognition by T cells is vital to cell-mediated adaptive immunity. As part of this process, MHC-II associates with the invariant chain (Ii) during biosynthesis in the endoplasmic reticulum to prevent premature peptide loading and to serve as a scaffold for subsequent proteolytic processing into MHC-II-CLIP. Cryo-electron microscopy structures of full-length Human Leukocyte Antigen-DR (HLA-DR) and HLA-DQ complexes associated with Ii, resolved at 3.0 to 3.1 Å, elucidate the trimeric assembly of the HLA/Ii complex and define atomic-level interactions between HLA, Ii transmembrane domains, loop domains, and class II-associated invariant chain peptides (CLIP). Together with previous structures of MHC-II peptide loading intermediates DO and DM, our findings complete the structural path governing class II antigen presentation.

De novo identification of CD4+ T cell epitopes.

CD4+ T cells recognize peptide antigens presented on class II major histocompatibility complex (MHC-II) molecules to carry out their function. The remarkable diversity of T cell receptor sequences and lack of antigen discovery approaches for MHC-II make profiling the specificities of CD4+ T cells challenging. We have expanded our platform of signaling and antigen-presenting bifunctional receptors to encode MHC-II molecules presenting covalently linked peptides (SABR-IIs) for CD4+ T cell antigen discovery. SABR-IIs can present epitopes to CD4+ T cells and induce signaling upon their recognition, allowing a readable output. Furthermore, the SABR-II design is modular in signaling and deployment to T cells and B cells. Here, we demonstrate that SABR-IIs libraries presenting endogenous and non-contiguous epitopes can be used for antigen discovery in the context of type 1 diabetes. SABR-II libraries provide a rapid, flexible, scalable and versatile approach for de novo identification of CD4+ T cell ligands from single-cell RNA sequencing data using experimental and computational approaches.

TripHLApan: predicting HLA molecules binding peptides based on triple coding matrix and transfer learning.

Human leukocyte antigen (HLA) recognizes foreign threats and triggers immune responses by presenting peptides to T cells. Computationally modeling the binding patterns between peptide and HLA is very important for the development of tumor vaccines. However, it is still a big challenge to accurately predict HLA molecules binding peptides. In this paper, we develop a new model TripHLApan for predicting HLA molecules binding peptides by integrating triple coding matrix, BiGRU + Attention models, and transfer learning strategy. We have found the main interaction site regions between HLA molecules and peptides, as well as the correlation between HLA encoding and binding motifs. Based on the discovery, we make the preprocessing and coding closer to the natural biological process. Besides, due to the input being based on multiple types of features and the attention module focused on the BiGRU hidden layer, TripHLApan has learned more sequence level binding information. The application of transfer learning strategies ensures the accuracy of prediction results under special lengths (peptides in length 8) and model scalability with the data explosion. Compared with the current optimal models, TripHLApan exhibits strong predictive performance in various prediction environments with different positive and negative sample ratios. In addition, we validate the superiority and scalability of TripHLApan's predictive performance using additional latest data sets, ablation experiments and binding reconstitution ability in the samples of a melanoma patient. The results show that TripHLApan is a powerful tool for predicting the binding of HLA-I and HLA-II molecular peptides for the synthesis of tumor vaccines. TripHLApan is publicly available at https://github.com/CSUBioGroup/TripHLApan.git.

Graph-pMHC: graph neural network approach to MHC class II peptide presentation and antibody immunogenicity.

Antigen presentation on MHC class II (pMHCII presentation) plays an essential role in the adaptive immune response to extracellular pathogens and cancerous cells. But it can also reduce the efficacy of large-molecule drugs by triggering an anti-drug response. Significant progress has been made in pMHCII presentation modeling due to the collection of large-scale pMHC mass spectrometry datasets (ligandomes) and advances in machine learning. Here, we develop graph-pMHC, a graph neural network approach to predict pMHCII presentation. We derive adjacency matrices for pMHCII using Alphafold2-multimer and address the peptide-MHC binding groove alignment problem with a simple graph enumeration strategy. We demonstrate that graph-pMHC dramatically outperforms methods with suboptimal inductive biases, such as the multilayer-perceptron-based NetMHCIIpan-4.0 (+20.17% absolute average precision). Finally, we create an antibody drug immunogenicity dataset from clinical trial data and develop a method for measuring anti-antibody immunogenicity risk using pMHCII presentation models. Our model increases receiver operating characteristic curve (ROC)-area under the ROC curve (AUC) by 2.57% compared to just filtering peptides by hits in OASis alone for predicting antibody drug immunogenicity.

Identification of mouse CD4+ T cell epitopes in SARS-CoV-2 BA.1 spike and nucleocapsid for use in peptide:MHCII tetramers.

Understanding adaptive immunity against SARS-CoV-2 is a major requisite for the development of effective vaccines and treatments for COVID-19. CD4+ T cells play an integral role in this process primarily by generating antiviral cytokines and providing help to antibody-producing B cells. To empower detailed studies of SARS-CoV-2-specific CD4+ T cell responses in mouse models, we comprehensively mapped I-Ab-restricted epitopes for the spike and nucleocapsid proteins of the BA.1 variant of concern via IFNγ ELISpot assay. This was followed by the generation of corresponding peptide:MHCII tetramer reagents to directly stain epitope-specific T cells. Using this rigorous validation strategy, we identified 6 immunogenic epitopes in spike and 3 in nucleocapsid, all of which are conserved in the ancestral Wuhan strain. We also validated a previously identified epitope from Wuhan that is absent in BA.1. These epitopes and tetramers will be invaluable tools for SARS-CoV-2 antigen-specific CD4+ T cell studies in mice.

Accurate modeling of peptide-MHC structures with AlphaFold.

Major histocompatibility complex (MHC) proteins present peptides on the cell surface for T cell surveillance. Reliable in silico prediction of which peptides would be presented and which T cell receptors would recognize them is an important problem in structural immunology. Here, we introduce an AlphaFold-based pipeline for predicting the three-dimensional structures of peptide-MHC complexes for class I and class II MHC molecules. Our method demonstrates high accuracy, outperforming existing tools in class I modeling accuracy and class II peptide register prediction. We validate its performance and utility with new experimental data on a recently described cancer neoantigen/wild-type peptide pair and explore applications toward improving peptide-MHC binding prediction.

Combining Sequence Similarity with Physicochemical Properties to Predict Binders for MHC-II Molecules.

Prediction of binders of the major histocompatibility complex class II (MHC-II) molecules is critical for T cell immunogenicity. As protein-protein interaction also relies on physicochemical properties, we aim to build a novel model combining sequence information and the physicochemical properties of proteins. Our research used data from the NetMHCIIpan 3.2 study. Features include BLOSUM50 and the physicochemical properties from iFeature Python package. We created a hybrid model of recurrent neural layers and feedforward layers. The final Area Under the Receiver Operating Characteristics (AUROC) on the test data was 0.755.

Loading dynamics of one SARS-CoV-2-derived peptide into MHC-II revealed by kinetic models.

Major histocompatibility complex class II (MHC-II) plays an indispensable role in activating CD4+ T cell immune responses by presenting antigenic peptides on the cell surface for recognition by T cell receptors. The assembly of MHC-II and antigenic peptide is therefore a prerequisite for the antigen presentation. To date, however, the atomic-level mechanism underlying the peptide-loading dynamics for MHC-II is still elusive. Here, by constructing Markov state models based on extensive all-atom molecular dynamics simulations, we reveal the complete peptide-loading dynamics into MHC-II for one SARS-CoV-2 S-protein-derived antigenic peptide (235ITRFQTLLALHRSYL249). Our Markov state model identifies six metastable states (S1-S6) during the peptide-loading process and determines two dominant loading pathways. The peptide could potentially approach the antigen-binding groove via either its N- or C-terminus. Then, the consecutive insertion of several anchor residues into the binding pockets profoundly dictates the peptide-loading dynamics. Notably, the MHC-II αA52-E55 motif could guide the peptide loading into the antigen-binding groove via forming ß-sheets conformation with the incoming peptide. The rate-limiting step, namely S5→S6, is mainly attributed to a considerable desolvation penalty triggered by the binding of the peptide C-terminus. Moreover, we further examined the conformational changes associated with the peptide exchange process catalyzed by the chaperon protein HLA-DM. A flipped-out conformation of MHC-II αW43 captured in S1-S3 is considered a critical anchor point for HLA-DM to modulate the structural dynamics. Our work provides deep structural insights into the key regulatory factors in MHC-II responsible for peptide recognition and guides future design for peptide vaccines against SARS-CoV-2.

The zinc ions stabilize the three-dimensional structure and are required for the binding of staphylococcal enterotoxin-like protein P (SEIP) with MHC-II receptors.

Staphylococcus aureus is a common human and animal pathogen. These bacteria have various pathogenicity factors, including enterotoxin-like proteins. SElP (staphylococcal enterotoxin-like protein P) has potential zinc ion-binding sites and is able to interact with major histocompatibility complex class II (MHCII) and T-cell receptor (TCR). A method for the expression and isolation of the enterotoxin-like protein of Staphylococcus aureus (SElP) was developed. The expression was carried out in E. coli cells, and the protein was isolated by affinity chromatography on a NiNTA column. The endotoxins were separated by affinity chromatography on Affi-Prep® polymyxin. It was shown by gel filtration that the resulting protein had a monomeric form. The protein in zinc-bound and zinc-free forms was characterized by protein melting using fluorescence method and it was shown that zinc stabilizes the spatial structure of SElP. The functional activity of SElP was investigated by the ability to interact with the histocompatibility antigen class II receptor (MHC-II) exposed on the B cell line Raji by flow cytofluorometry. The zinc-bound and zinc-free forms were shown to differ in their interaction with MHC-II. The localization of the zinc-binding site was confirmed by the introduction of the H225 and D227 mutations. The mutant protein was characterized by melting, and its propensity to form aggregates was shown.

CRESSP: a comprehensive pipeline for prediction of immunopathogenic SARS-CoV-2 epitopes using structural properties of proteins.

The development of autoimmune diseases following SARS-CoV-2 infection, including multisystem inflammatory syndrome, has been reported, and several mechanisms have been suggested, including molecular mimicry. We developed a scalable, comparative immunoinformatics pipeline called cross-reactive-epitope-search-using-structural-properties-of-proteins (CRESSP) to identify cross-reactive epitopes between a collection of SARS-CoV-2 proteomes and the human proteome using the structural properties of the proteins. Overall, by searching 4 911 245 proteins from 196 352 SARS-CoV-2 genomes, we identified 133 and 648 human proteins harboring potential cross-reactive B-cell and CD8+ T-cell epitopes, respectively. To demonstrate the robustness of our pipeline, we predicted the cross-reactive epitopes of coronavirus spike proteins, which were recognized by known cross-neutralizing antibodies. Using single-cell expression data, we identified PARP14 as a potential target of intermolecular epitope spreading between the virus and human proteins. Finally, we developed a web application (https://ahs2202.github.io/3M/) to interactively visualize our results. We also made our pipeline available as an open-source CRESSP package (https://pypi.org/project/cressp/), which can analyze any two proteomes of interest to identify potentially cross-reactive epitopes between the proteomes. Overall, our immunoinformatic resources provide a foundation for the investigation of molecular mimicry in the pathogenesis of autoimmune and chronic inflammatory diseases following COVID-19.

A previously unappreciated polymorphism in the beta chain of I-As expressed in autoimmunity-prone SJL mice: Combined impact on antibody, CD4 T cell recognition and MHC class II dimer structural stability.

In the process of structure-function studies on the MHC class II molecule expressed in autoimmunity prone SJL mice, I-As, we discovered a disparity from the reported sequence of the MHC class II beta chain. The variant is localized at a highly conserved site of the beta chain, at residue 58. Our studies revealed that this single amino acid substitution of Pro for Ala at this residue, found in I-As, changes the structure of the MHC class II molecule, as evidenced by a loss of recognition by two monoclonal antibodies, and elements of MHC class II conformational stability identified through molecular dynamics simulation. Two other rare polymorphisms in I-As involved in hydrogen bonding potential between the alpha chain and the peptide main chain are located at the same end of the MHC class II binding pocket, studied in parallel may impact the consequences of the ß chain variant. Despite striking changes in MHC class II structure, CD4 T cell recognition of influenza-derived peptides was preserved. These disparate findings were reconciled by discovering, through monoclonal antibody blocking approaches, that CD4 T cell recognition by I-As restricted CD4 T cells focused more on the region of MHC class II at the peptide's amino terminus. These studies argue that the conformational variability or flexibility of the MHC class II molecule in that region of I-As select a CD4 T cell repertoire that deviates from the prototypical docking mode onto MHC class II peptide complexes. Overall, our results are consistent with the view that naturally occurring MHC class II molecules can possess polymorphisms that destabilize prototypical features of the MHC class II molecule but that can maintain T cell recognition of the MHC class II:peptide ligand via alternate docking modes.

Blocking IAg7 class II major histocompatibility complex by drug-like small molecules alleviated Sjögren's syndrome in NOD mice.

BACKGROUND: Sjögren's syndrome (SjS) is an autoimmune disease with a strong genetic association. To date, no vaccine or therapeutic agent exists to cure SjS, and patients must rely on lifelong therapies to treat symptoms. Human leukocyte antigens (HLA) are primary susceptibility loci that form the genetic basis for many autoimmune diseases, including SjS. In this study, we sought to determine whether blocking MHC class II IAg7 antigen presentation in the NOD mouse would alleviate SjS by preventing the recognition of autoantigens by pathogenic T cells. METHODS: Mapping of the antigenic epitopes of Ro60 autoantigen to IAg7 of the NOD mice was performed using structural modeling and in-vitro stimulation. Tetraazatricyclo-dodecane (TATD) and 8-Azaguanine (8-Aza) were previously identified as potential binders to IAg7 of the NOD mice using in silico drug screening. Mice were treated with 20mgs/kg via IP every day five days/week for 23 weeks. Disease profiling was conducted. FINDINGS: Specific peptides of Ro60 autoantigen were identified to bind to IAg7 and stimulated splenocytes of the NOD mice. Treating NOD mice with TATD or 8-Azaguanine alleviated SjS symptoms by improving salivary and lacrimal gland secretory function, decreasing the levels of autoantibodies, and reducing the severity of lymphocytic infiltration in the salivary and lacrimal glands. INTERPRETATION: This study presents a novel therapeutic approach for SjS by identifying small molecules capable of inhibiting T cell response via antigen-specific presentation. FUNDING: CQN is supported financially in part by PHS grants AI130561, DE026450, and DE028544 from the National Institutes of Health.

Discovery of an ancient MHC category with both class I and class II features.

Two classes of major histocompatibility complex (MHC) molecules, MHC class I and class II, play important roles in our immune system, presenting antigens to functionally distinct T lymphocyte populations. However, the origin of this essential MHC class divergence is poorly understood. Here, we discovered a category of MHC molecules (W-category) in the most primitive jawed vertebrates, cartilaginous fish, and also in bony fish and tetrapods. W-category, surprisingly, possesses class II-type α- and ß-chain organization together with class I-specific sequence motifs for interdomain binding, and the W-category α2 domain shows unprecedented, phylogenetic similarity with ß2-microglobulin of class I. Based on the results, we propose a model in which the ancestral MHC class I molecule evolved from class II-type W-category. The discovery of the ancient MHC group, W-category, sheds a light on the long-standing critical question of the MHC class divergence and suggests that class II type came first.

Suppression of Plasmodium MIF-CD74 signaling protects against severe malaria.

The deadliest complication of infection by Plasmodium parasites, cerebral malaria, accounts for the majority of malarial fatalities. Although our understanding of the cellular and molecular mechanisms underlying the pathology remains incomplete, recent studies support the contribution of systemic and neuroinflammation as the cause of cerebral edema and blood-brain barrier (BBB) dysfunction. All Plasmodium species encode an orthologue of the innate cytokine, Macrophage Migration Inhibitory Factor (MIF), which functions in mammalian biology to regulate innate responses. Plasmodium MIF (PMIF) similarly signals through the host MIF receptor CD74, leading to an enhanced inflammatory response. We investigated the PMIF-CD74 interaction in the onset of experimental cerebral malaria (ECM) and liver stage Plasmodium development by using a combination of CD74 deficient (Cd74-/- ) hosts and PMIF deficient parasites. Cd74-/- mice were found to be protected from ECM and the protection was associated with the inability of brain microvessels to present parasite antigen to sequestered and pathogenic Plasmodium-specific CD8+ T cells. Infection of WT hosts with PMIF-deficient sporozoites or infection of Cd74-/- hosts with WT sporozoites impacted the survival of infected hepatocytes and subsequently reduced blood-stage associated inflammation, contributing to protection from ECM. We recapitulated these finding with a novel pharmacologic PMIF-selective antagonist that reduced PMIF/CD74 signaling and fully protected mice from ECM. These findings reveal a conserved mechanism for Plasmodium usurpation of host CD74 signaling and suggest a tractable approach for new pharmacologic intervention.

A simple pan-specific RNN model for predicting HLA-II binding peptides.

The prediction of human leukocyte antigen (HLA) class II binding peptides plays important roles in understanding the mechanism of immune recognition and developing effective epitope-based vaccines. In this work, gated recurrent unit (GRU)-based recurrent neural network (RNN) was successfully employed to establish a pan-specific prediction model of HLA-II-binding peptides by using only the HLA and peptide sequence information. In comparison with the existing pan-specific models of HLA-II-binding peptides, the GRU-based RNN model covered a broad spectrum of HLA-II molecules including 50 HLA-DR, 47 HLA-DQ, and 19 HLA-DP molecules with peptide lengths varying from 8 to 43 mers. The results demonstrated strong discriminant capabilities of the GRU-based RNN model, of which the AUC values were 0.92, 0.88, and 0.88 for the training, validation, and test sets, respectively. Also, the GRU-based model showed state-of-the-art performances in predicting the binding peptides with the length ranging from 8-32 mers, which provides an efficient method for predicting HLA-II-binding peptides of longer lengths in comparison with the available methods. Overall, taking the advantages of the RNN architecture, the established pan-specific GRU model can be used for predicting accurately the HLA-II-binding peptides in a simple and direct manner.

HLA class II peptide-binding-region analysis reveals funneling of polymorphism in action.

BACKGROUND: HLA-class II proteins hold important roles in key physiological processes. The purpose of this study was to compile all class II alleles reported in human population and investigate patterns in pocket variants and their combinations, focusing on the peptide-binding region (PBR). METHODS: For this purpose, all protein sequences of DPA1, DQA1, DPB1, DQB1 and DRB1 were selected and filtered, in order to have full PBR sequences. Proportional representation was used for pocket variants while population data were also used. RESULTS: All pocket variants and PBR sequences were retrieved and analyzed based on the preference of amino acids and their properties in all pocket positions. The observed number of pocket variants combinations was much lower than the possible inferred, suggesting that PBR formation is under strict funneling. Also, although class II proteins are very polymorphic, in the majority of the reported alleles in all populations, a significantly less polymorphic pocket core was found. CONCLUSIONS: Pocket variability of five HLA class II proteins was studied revealing favorable properties of each protein. The actual PBR sequences of HLA class II proteins appear to be governed by restrictions that lead to the establishment of only a fraction of the possible combinations and the polymorphism recorded is the result of intense funneling based on function.