Precision Neoantigen Discovery Using Large-Scale Immunopeptidomes and Composite Modeling of MHC Peptide Presentation.

Major histocompatibility complex (MHC)-bound peptides that originate from tumor-specific genetic alterations, known as neoantigens, are an important class of anticancer therapeutic targets. Accurately predicting peptide presentation by MHC complexes is a key aspect of discovering therapeutically relevant neoantigens. Technological improvements in mass spectrometry-based immunopeptidomics and advanced modeling techniques have vastly improved MHC presentation prediction over the past 2 decades. However, improvement in the accuracy of prediction algorithms is needed for clinical applications like the development of personalized cancer vaccines, the discovery of biomarkers for response to immunotherapies, and the quantification of autoimmune risk in gene therapies. Toward this end, we generated allele-specific immunopeptidomics data using 25 monoallelic cell lines and created Systematic Human Leukocyte Antigen (HLA) Epitope Ranking Pan Algorithm (SHERPA), a pan-allelic MHC-peptide algorithm for predicting MHC-peptide binding and presentation. In contrast to previously published large-scale monoallelic data, we used an HLA-null K562 parental cell line and a stable transfection of HLA allele to better emulate native presentation. Our dataset includes five previously unprofiled alleles that expand MHC diversity in the training data and extend allelic coverage in underprofiled populations. To improve generalizability, SHERPA systematically integrates 128 monoallelic and 384 multiallelic samples with publicly available immunoproteomics data and binding assay data. Using this dataset, we developed two features that empirically estimate the propensities of genes and specific regions within gene bodies to engender immunopeptides to represent antigen processing. Using a composite model constructed with gradient boosting decision trees, multiallelic deconvolution, and 2.15 million peptides encompassing 167 alleles, we achieved a 1.44-fold improvement of positive predictive value compared with existing tools when evaluated on independent monoallelic datasets and a 1.17-fold improvement when evaluating on tumor samples. With a high degree of accuracy, SHERPA has the potential to enable precision neoantigen discovery for future clinical applications.

Precision Neoantigen Discovery Using Large-scale Immunopeptidomes and Composite Modeling of MHC Peptide Presentation.

Major histocompatibility complex (MHC)-bound peptides that originate from tumor-specific genetic alterations, known as neoantigens, are an important class of anticancer therapeutic targets. Accurately predicting peptide presentation by MHC complexes is a key aspect of discovering therapeutically relevant neoantigens. Technological improvements in mass-spectrometry-based immunopeptidomics and advanced modeling techniques have vastly improved MHC presentation prediction over the past two decades. However, improvement in the sensitivity and specificity of prediction algorithms is needed for clinical applications such as the development of personalized cancer vaccines, the discovery of biomarkers for response to checkpoint blockade, and the quantification of autoimmune risk in gene therapies. Toward this end, we generated allele-specific immunopeptidomics data using 25 monoallelic cell lines and created Systematic HLA Epitope Ranking Pan Algorithm (SHERPA), a pan-allelic MHC-peptide algorithm for predicting MHC-peptide binding and presentation. In contrast to previously published large-scale monoallelic data, we used an HLA-null K562 parental cell line and a stable transfection of HLA alleles to better emulate native presentation. Our dataset includes five previously unprofiled alleles that expand MHC-binding pocket diversity in the training data and extend allelic coverage in under profiled populations. To improve generalizability, SHERPA systematically integrates 128 monoallelic and 384 multiallelic samples with publicly available immunoproteomics data and binding assay data. Using this dataset, we developed two features that empirically estimate the propensities of genes and specific regions within gene bodies to engender immunopeptides to represent antigen processing. Using a composite model constructed with gradient boosting decision trees, multiallelic deconvolution, and 2.15 million peptides encompassing 167 alleles, we achieved a 1.44-fold improvement of positive predictive value compared with existing tools when evaluated on independent monoallelic datasets and a 1.15-fold improvement when evaluating on tumor samples. With a high degree of accuracy, SHERPA has the potential to enable precision neoantigen discovery for future clinical applications.

Human iPSC-Derived 2D and 3D Platforms for Rapidly Assessing Developmental, Functional, and Terminal Toxicities in Neural Cells.

With increasing global health threats has come an urgent need to rapidly develop and deploy safe and effective therapies. A common practice to fast track clinical adoption of compounds for new indications is to repurpose already approved therapeutics; however, many compounds considered safe to a specific application or population may elicit undesirable side effects when the dosage, usage directives, and/or clinical context are changed. For example, progenitor and developing cells may have different susceptibilities than mature dormant cells, which may yet be different than mature active cells. Thus, in vitro test systems should reflect the cellular context of the native cell: developing, nascent, or functionally active. To that end, we have developed high-throughput, two- and three-dimensional human induced pluripotent stem cell (hiPSC)-derived neural screening platforms that reflect different neurodevelopmental stages. As a proof of concept, we implemented this in vitro human system to swiftly identify the potential neurotoxicity profiles of 29 therapeutic compounds that could be repurposed as anti-virals. Interestingly, many compounds displayed high toxicity on early-stage neural tissues but not on later stages. Compounds with the safest overall viability profiles were further evaluated for functional assessment in a high-throughput calcium flux assay. Of the 29 drugs tested, only four did not modulate or have other potentially toxic effects on the developing or mature neurospheroids across all the tested dosages. These results highlight the importance of employing human neural cultures at different stages of development to fully understand the neurotoxicity profile of potential therapeutics across normal ontogeny.

A machine learning algorithm with subclonal sensitivity reveals widespread pan-cancer human leukocyte antigen loss of heterozygosity.

Human leukocyte antigen loss of heterozygosity (HLA LOH) allows cancer cells to escape immune recognition by deleting HLA alleles, causing the suppressed presentation of tumor neoantigens. Despite its importance in immunotherapy response, few methods exist to detect HLA LOH, and their accuracy is not well understood. Here, we develop DASH (Deletion of Allele-Specific HLAs), a machine learning-based algorithm to detect HLA LOH from paired tumor-normal sequencing data. With cell line mixtures, we demonstrate increased sensitivity compared to previously published tools. Moreover, our patient-specific digital PCR validation approach provides a sensitive, robust orthogonal approach that could be used for clinical validation. Using DASH on 610 patients across 15 tumor types, we find that 18% of patients have HLA LOH. Moreover, we show inflated HLA LOH rates compared to genome-wide LOH and correlations between CD274 (encodes PD-L1) expression and microsatellite instability status, suggesting the HLA LOH is a key immune resistance strategy.

Functional and Mechanistic Neurotoxicity Profiling Using Human iPSC-Derived Neural 3D Cultures.

Neurological disorders affect millions of people worldwide and appear to be on the rise. Whereas the reason for this increase remains unknown, environmental factors are a suspected contributor. Hence, there is an urgent need to develop more complex, biologically relevant, and predictive in vitro assays to screen larger sets of compounds with the potential for neurotoxicity. Here, we employed a human induced pluripotent stem cell (iPSC)-based 3D neural platform composed of mature cortical neurons and astrocytes as a model for this purpose. The iPSC-derived human 3D cortical neuron/astrocyte co-cultures (3D neural cultures) present spontaneous synchronized, readily detectable calcium oscillations. This advanced neural platform was optimized for high-throughput screening in 384-well plates and displays highly consistent, functional performance across different wells and plates. Characterization of oscillation profiles in 3D neural cultures was performed through multi-parametric analysis that included the calcium oscillation rate and peak width, amplitude, and waveform irregularities. Cellular and mitochondrial toxicity were assessed by high-content imaging. For assay characterization, we used a set of neuromodulators with known mechanisms of action. We then explored the neurotoxic profile of a library of 87 compounds that included pharmaceutical drugs, pesticides, flame retardants, and other chemicals. Our results demonstrated that 57% of the tested compounds exhibited effects in the assay. The compounds were then ranked according to their effective concentrations based on in vitro activity. Our results show that a human iPSC-derived 3D neural culture assay platform is a promising biologically relevant tool to assess the neurotoxic potential of drugs and environmental toxicants.