Neoantigen peptides are tumor-specific epitopes which are encoded by somatic mutations, frameshift insertion mutations or post-transcriptional modification and are expressed by cancer cells but not by normal tissue of the patient. In theory, neoantigens can be recognized as foreign by the immune system, enabling strong and patient-specific T-cell responses without any potential for autoimmunity. Neoantigens are an attractive target for personalized cancer immunotherapy, in contrast to conventional "off-the-shelf" cancer vaccines. Neoantigens have been loaded on to adjuvanted peptide vaccines as well as nucleic-acid-based vaccines. The current generation of neoantigen vaccines can be theoretically combined with immune checkpoint inhibitors to further augment patient's endogenous anti-tumor T-cell responses and thus can generate long-lasting clinical benefits without off-target toxicities commonly observed with standard-of-care chemoradiotherapy. Neoantigen-directed cancer vaccines are in early-stage clinical trials for several solid tumor types. The goal of neoantigen identification algorithms is to design a personalized vaccine for each patient that contains the optimal set of neoantigens for each individual.
Neoantigens are peptides consisting of 8 to 25 amino acids that are produced by somatic mutations that have been acquired over the course of tumorigenesis, and are presented in the peptide-binding cleft of HLA class I or class II molecules on the surface of tumor cells. In contrast to tumor-associated antigens which are overexpressed or aberrantly expressed self-proteins, neoepitopes are truly non-self. As a result, high affinity T-cell clones that can potentially strongly react to these antigens are not deleted during central tolerance. As such, neoantigens are potentially immunogenic as a result of their sequence being different than the germline proteome and can be discovered through either analysis of the T-cell repertoire or in silico prediction from mass-spectrometric immunopeptidomic data. The production of a neoantigen starts with a mutation in the DNA, which is transcribed into mRNA, translated into a mutant protein, processed by the immunoproteasome, translocated into the endoplasmic reticulum or endosomal compartments, trimmed to an appropriate length, loaded onto newly synthesized HLA molecules, and trafficked to the cell surface, where the pHLA complex will act as a ligand for TCRs. As each tumor has a distinct set of acquired mutations as a result of carcinogen exposures, DNA-repair pathway mutations or mutations in oncogenic drivers, the resulting set of neoantigens that a tumor presents is nearly individual-specific, creating a molecular fingerprint. The peptide nature of neoantigens makes them easily reproducible in peptide libraries, and vaccines can be created which train the immune system to proliferate T cells with specificity against these neoantigens. This specificity will allow for the T cells to migrate to tumor sites, and carry out a cytolytic function.
Fig. 1 Classification of neoantigen-based therapies.1,5
The somatic mutation profile of a patient's tumor thus serves as a historical record of the evolutionary pressures, DNA-damage burden and microenvironmental exposures that have been encountered by that cancer. Neoepitopes that are generated by passenger mutations are of particular interest, since they are unlikely to have been subject to negative selection in the thymus due to their absence from the germline. After performing matched whole tumor and normal sequencing, followed by rigorous filtering and variant calling to ensure that only true somatic mutations are used for prediction, various computational algorithms use additional information such as clonality, mutation allele burden, and expression of the corresponding gene at the RNA level to nominate mutations which are predicted to be presented as true HLA ligands. However, even when a mutation is predicted to be immunogenic by a given algorithm, this still only informs on its potential, since a neoepitope will only be targeted if a T cell with the appropriate specificity exists within a patient. Thus, orthogonal validation assays to demonstrate recognition of a mutation by a patient's T cells such as co-culture of autologous dendritic cells pulsed with a mutant peptide, or tetramer staining of peripheral blood or tumor-infiltrating T cells, are still needed. Additionally, due to the processes of subclonal selection and immunoediting, the mutational landscape of a tumor will evolve over time. Thus, new biopsies or liquid biopsies may be obtained at a later time-point and reanalyzed to update a patient's neoantigen profile.
The primary advantage of targeting non-self antigens is that it circumvents tolerance mechanisms which limit therapies that target self-antigens. Neoantigen peptides are a prime example of this; their mutant amino-acid sequence is different from the sequence in the HLA groove presented during thymic selection. Thus, there is a naïve, high-avidity T-cell clone in the periphery that can be activated without competing regulatory T cells that normally suppress the response to over-expressed self-antigens. Peptide-based delivery methods take advantage of this; they recapitulate the natural presentation of the mutant pHLA complex and thus provide a stable complex to prime T cells. Synthetic long peptides, in particular when combined with nanoparticulate or liposomal adjuvants, are effectively phagocytosed by professional antigen presenting cells, directed to MHC class I and class II processing pathways, and can induce multifunctional CD8+ effector and CD4+ helper responses. Because the epitope is non-self, on-target off-tissue toxicities are limited as normal cells will not have the mutation and thus will not be able to present the target peptide. Therefore, dose escalation regimens which would be contraindicated for shared self-antigens may be possible. In addition, the clarity of concept in a binary "mutant" versus "wild-type" antigen allows for easier pharmacodynamic evaluation: tetramer or dextramer reagents can be made to detect mutation-specific T cells in blood and tumor and multiplexed serum assays can be used to measure cytokine polarization profiles that are associated with clinical response. In the end, the non-self nature of neoantigen peptides make every tumor mutation a potential Achilles heel that can be selectively exploited without damaging healthy tissues.
Beyond personalized medicines with molecularly targeted small molecules, neoantigen peptides represent the next step of personalization to the de novo programming of an adaptive immune response, written in the alphabets of an individual tumor. Neoantigen peptides are thus the letters that construct the language, translating individual somatic DNA mutations into the syntax of immune signals that are intelligible to T cells. Whole-exome and transcriptome sequencing to identify mutations, HLA haplotyping to predict peptide presentation, and proteomic immunopeptidomics to confirm the in vivo display of candidate peptides on tumor cells now collectively paint an integrated multi-omics picture. Machine-learning ensembles trained on large-scale T-cell reactivity data then prioritize candidates based on putative binding affinity and other physiochemical determinants of immunogenicity such as hydrophobic anchor residues, proteasomal cleavage motifs, and homology to pathogen-derived epitopes, which could have already primed cross-reactive memory pools. The selected peptide candidates are produced using good-manufacturing-practice procedures and pooled into multivalent formulations to maximize epitope spreading. The vaccines are then administered in repeated prime-boost schedules combined with and often synchronized to immune checkpoint blockade to avoid early T-cell exhaustion. This results in an adaptive immune response that is then tracked at different molecular levels: clonal T-cell expansions in the blood, tissue-resident memory phenotypes at metastatic locations, changes in tumor microenvironment composition towards interferon-γ–dominated inflammation, and epitope spreading to unvaccinated neoantigens as a marker of endogenous antigen cascade. Neoantigen peptides can also be repurposed to guide ex vivo enrichment of mutation-reactive T-cell receptors for autologous or allogeneic cellular therapies that combine specificity with the proliferative potential of T cells engineered for persistent expansion.
The development of a personalised peptide vaccine requires the translation of a unique somatic mutanome into an injectable and clinically graded peptide drug product. In practice, this involves the extraction of tumour and matched normal tissue or blood from the patient, sequencing of the tumour sample to high depth (single nucleotide variants, insertions and deletions (indels), fusions, splice site changes), processing of raw sequencing reads through an antigen discovery pipeline to infer the patient's HLA type, and predicting proteasomal cleavage, peptide–HLA binding affinity and then ranking based on expression, clonal fraction and predicted immunogenicity to identify a shortlist of peptides of interest. The shortlisted peptides are then synthesised under Good Manufacturing Practice (GMP) conditions (often as long synthetic constructs including both the core peptide and flanking regions to aid proteasomal processing), purified, subjected to quality control and a series of pre-clinical validation experiments (solubility, stability, non-cytotoxicity, immunogenicity) and finally formulated (often with an adjuvant or encapsulated in nanoparticles) and filled into vials under GMP manufacturing conditions before being released for administration to the patient. The whole process requires rigorous and auditable documentation and a version-controlled algorithmic pipeline to allow each peptide batch to be traceable back to its original computationally derived neoantigen target.
Raw genomic data, by themselves, cannot yield a list of candidate T-cell epitopes. At the center of all current pipelines, they must first be computationally processed to identify somatic variants, typically through alignment of tumor and normal paired DNA reads to a reference genome, probabilistic variant callers that call somatic mutations while filtering out sequencing errors, and functional annotation of each mutation. This variant calling pipeline is usually supplemented with RNA-sequencing data to measure allelic expression and to filter out non-expressed mutations (e.g. below threshold expression levels) to avoid synthesizing peptides that will not be naturally processed and presented. HLA typing is also performed on the same sequencing libraries, typically by alignment to the hyperpolymorphic exons that encode the peptide-binding cleft of class I and class II HLA molecules; a four-digit resolution is usually required since peptide-binding motifs vary significantly across the thousands of HLA alleles. Each mutation-containing peptide is then computationally docked into the antigen-binding cleft of all HLA molecules found in each patient, and predicted binding is usually performed with either a collection of neural networks or with structure-based scoring functions that are trained on eluted ligand datasets. Downstream filters also include proteasomal cleavage prediction, transporter associated with antigen processing (TAP) binding, and gene essentiality to downrank epitopes that may be lost under immuno-selective pressure. Immunogenicity predictors also consider features including amino-acid physiochemistry, similarity to microbial proteomes, and predicted T-cell receptor contact residues in an attempt to further enrich for epitopes that are more likely to elicit de-novo immune responses as opposed to merely binding HLA without being recognized by T cells. Computational prediction of tumor subclonality may also be incorporated to prioritize driver mutations that are present in the majority of tumor cells instead of subclonal passenger mutations that would make a smaller fraction of cancer cells visible to T cells. Visualization dashboards are usually used to collect mutation-centric, peptide-centric, and HLA-centric evidence so that each candidate epitope can be inspected in its genomic, transcriptomic, and immunologic context by a human curator prior to experimental validation by chemical synthesis.
Fig. 2 The flow chart of neoantigen vaccine generation.2,6
Bioinformatic triage of candidate mutant epitopes feeds the algorithmic workflow to the chemistry laboratory, where peptide sequences are translated to physical entities by solid-phase peptide synthesis (SPPS). Iterative chain elongation with amino-acid residues anchored to a polymeric matrix is the standard mode of synthesis. The key steps of each cycle are deprotection of the N-terminal amine group, activation and coupling of the next protected amino acid, and capping of the free amines to suppress deletion products. For many neoantigen targets, the synthetic reaction is scaled to the microgram-to-milligram range and carried out under specifications that will meet both the research-grade purity needs and the anticipated regulatory requirements. In the majority of studies, long peptides of fifteen to thirty-five residues are synthesized, as they are likely to need intracellular processing by professional antigen-presenting cells, thus providing class I and class II epitopes in a single polypeptide chain. Post-translational modifications are sometimes added to improve the chemical or biological stability of the peptides. Terminal D-amino acids can be incorporated to protect the peptides from exopeptidases, and lipidation or cholesterylation can be used to tether them to albumin or cell membranes to increase their half-life and enhance lymph-node delivery. Cysteine side chains are also sometimes added to the sequence to provide sites for disulfide conjugation to carrier proteins or nanoparticle scaffolds for multivalent presentation that can recapitulate the geometric periodicity of pathogen surfaces. Crude peptides cleaved from the resin and deprotected in the global sense are purified by reversed-phase high-performance liquid chromatography, and their molecular weight and degree of aggregation under physiologic buffers are validated by mass spectrometry. Endotoxin levels and sterility are also determined, as even background levels of contaminating proteins or nucleic acids can have a confounding effect on later immunologic assays. Lyophilized aliquots are typically stored under nitrogen in sealed glass vials at controlled humidity, and are characterized by accelerated stability studies to support their expiration dating. Finally, complete documentation of the synthetic history of the peptide—resin batch, coupling reagents, cleavage cocktail, chromatographic gradient, quality-control spectra—is recorded to ensure that the peptide, if needed again, can be reproduced with identical physicochemical characteristics for booster vaccinations or companion diagnostics.
Validation assays for candidate neoantigen peptides before clinical administration will not only require chemical stability, but also biological activity in an immunologic setting that closely resembles the clinical scenario. A first hurdle is to ensure solubility and lack of apparent cytotoxicity in dendritic cells or lymphocytes in culture. Subsequently, autologous dendritic cells differentiated from the patient's own CD14⁺ monocytes are loaded with serial dilutions of the peptide pool and co-cultured with antigen-naïve T cells from the same patient's blood or spleen. Responses can be measured by multi-parameter flow cytometry for T-cell activation, cytokine secretion (via enzyme-linked immunospot or bead-based multiplex assays) and enumeration of antigen-specific CD8⁺ and CD4⁺ T cells (via tetramer or dextramer staining). Avidity can be determined by peptide titration to identify the concentration required for half-maximal interferon-γ secretion, which can identify potential epitopes that may be preserved even if the peptide levels within the tumor decrease. To better recapitulate the environment of the existing tumor, some of these assays can be performed in the presence of T regulatory cells, myeloid-derived suppressor cells, or under hypoxic conditions to determine if the peptide-induced effector cells maintain their cytolytic function. Cytotoxicity can be validated by testing in an engineered target cell line which has been modified to express the mutant minigene via lentiviral transduction so that the antigen undergoes natural processing and presentation. Peptide validation in vivo can be tested using humanized mouse models in which patient-derived tumor tissue and autologous immune cells have been engrafted and are subsequently vaccinated with the peptide-adjuvant mix; tumor growth, T-cell infiltration, and systemic cytokine profiles can be monitored over time. For all platforms, it is important that wild-type control peptides are tested alongside mutant peptides to ensure reactivity is mutation-specific and recall antigens are included as an internal control for immune competence. Results from these multi-layered assays can then be fed back into the bioinformatic pipeline to fine-tune the machine-learning algorithms and aid in iterative peptide selection until a final set of peptides with sufficient immunogenicity, safety, and manufacturing fidelity is taken to first-in-human testing.
The era of immunotherapy has ushered in an oncologic revolution by moving the treatment landscape from non-specific killing to specific, adaptive, and long-term tumor suppression. Therapies utilizing the host immune system, including monoclonal antibodies, adoptive cell therapy, or tumor vaccines rely on the exquisite specificity of T and B cell receptor recognition that can discern minor molecular variations between tumor and normal cells. Not only has the response rates increased as a result of these therapies, in the metastatic setting many patients are now surviving longer than previously seen with conventional chemotherapy or radiation. The key advantage of immunotherapy over traditional therapies is the precision targeting of immune checkpoints, tumor-specific antigens, or cytokine signaling with the use of biologic drugs. In addition, organ function and quality of life is often maintained with these therapies, due to the lower side effect profile and greater tissue specificity of these targeted therapies. In contrast to traditional chemotherapy or radiation, many aspects of immunotherapy are modifiable on a patient-by-patient basis depending on the response, including dose, schedule, and other drug combinations. Immunotherapy is therefore a dynamic treatment that utilizes genomic, transcriptomic, and immune biomarkers to deliver precise therapeutic force to the tumor while also creating long-term memory immune surveillance.
In contrast to the effects of cytotoxic chemotherapy, which is not cell cycle specific, and thus any proliferating cell will be affected, immunotherapy works through mechanisms that are specific to the cancer cell, and thus has a different toxicity profile and leaves non-dividing stem cells unharmed. For example, antibody therapy can be specific to antigens that are not found on other cells, such as antigens found on the surface of tumor cells, and bind to tumor cells and lead to their phagocytosis. T cells used in adoptive cell transfer can also be selected to only bind to targets that are not found on normal tissues, such as mutations not found in normal cells, providing a mechanism that cannot be matched by small-molecule chemotherapy agents. These T cells are also not expected to target normal cells due to the loss of high affinity T cell receptors that cross-react with self antigens during development of the immune system in the thymus. As a result, the majority of targets are already not available for immune-mediated recognition in a normal patient without cancer. As a result of these attributes, the U.S. FDA and other drug regulatory agencies have been more likely to grant accelerated approval to immunotherapeutic agents, with monitoring for rare immune-mediated side effects post-approval, rather than requiring additional toxicology testing required of small-molecule drugs. Additionally, in large cohorts of patients followed long-term after treatment, most responders have not had late toxicities or organ toxicities that continued to build up over time.
Reduced off-target toxicity. Toxicity is often caused by off-target interaction of the drug with non-tumor cells, and is less of a concern if the epitope used is either mutation-specific or is expressed at substantially higher levels in tumor cells than in normal cells. In addition, with immune interventions, the inflammatory effects are less likely to occur outside the tumor microenvironment, because: (a) checkpoint inhibition only activates T cells already present and already activated in the tumor microenvironment; and (b) the low, intermittent dosing possible due to the long half-life of antibodies and the long-term persistence of memory T cells after therapy is complete does not result in the sustained systemic exposure of kinase or metabolic inhibitors required for their toxic effects. Tox screens for candidate antibodies typically include screens for cross-reactivity with a wide panel of normal human tissues and any antibody showing any level of binding to essential organs is either redesigned or humanized to eliminate the off-target binding. This is reflected clinically by the relative rarity of grade-3/4 non-immune toxicities; relatively few patients experience cytopenias, hepatotoxicity, or cardiotoxicity, for example, compared to cytotoxic drugs such as platinum salts or topoisomerase inhibitors. Low-grade immune-related toxicities, when they occur, can often be treated with short-term immunosuppression without loss of antitumor activity.
Clinical success of anti–PD-1 or anti–CTLA-4 has created a platform to which other agents can be added. Vaccines, adoptive T cells and cytokines all act in a synergistic fashion with checkpoint blockade, as the latter counteracts the constitutive inhibitory signals that tumors exert on primed lymphocytes. A cancer vaccine, for instance, that expands a mutation-specific CD8⁺ T-cell population within lymph nodes will, by itself, have little antitumor activity. However, its combination with PD-1 blockade, by preserving the function of those newly primed cells in the tumor microenvironment, can turn immunologic flare- ups into long-term tumor regression. The complementarity of vaccines and checkpoint inhibitors is further enhanced by their non-overlapping toxicity profiles: whereas checkpoint-related adverse events mainly include colitis, dermatitis, or endocrinopathies mediated by the expression of the same antigen within normal tissues, peptide vaccines or transgenic T cells cause only a brief febrile response. This means that the dose-finding phase can escalate both agents to their respective biologically active doses without cumulative organ damage. Translational endpoints such as peripheral clonal expansion, intratumoral interferon- γ signatures, and epitope spreading can then provide immediate feedback that the combination regimen elicits deeper immune activation than either single agent. Regulatory considerations are also favorable: accelerated approval pathways can be used for the initial licensure of the checkpoint inhibitor, with companion vaccines or cellular therapies subsequently evaluated under the same investigational new drug application, cutting development times and avoiding repetitive safety studies. The end result is a modular therapeutic landscape in which vaccination, cellular engineering, and checkpoint modulation can be used sequentially or in parallel to maintain adaptive immunity for long enough to eradicate any residual disease and to establish memory surveillance.
Neoantigen peptides are reshaping cancer immunotherapy by targeting patient-specific mutations. These peptides trigger immune responses against tumor cells while sparing healthy tissue. Our services cover bioinformatics prediction, synthesis, and validation of neoantigen peptides for research use. By tailoring therapies to individual patients, neoantigen peptides increase efficacy and reduce toxicity compared to conventional treatments. With GMP-grade production and immune assay validation, we provide reliable solutions for translating personalized immunotherapy from bench to bedside.
Products & Services:
Advantages:
Neoantigen peptides bring precision and personalization to cancer immunotherapy. Partner with us to design custom peptides that advance individualized treatment strategies.
1. What are neoantigen peptides?
Tumor-specific peptides from patient mutations.
2. Why are they effective?
They are unique to cancer cells and avoid healthy tissues.
3. Do you provide GMP-grade peptides?
Yes, suitable for clinical development.
4. How are neoantigens identified?
Through bioinformatics and sequencing.
References
