In a recent article published in Nature Communications, researchers reported a proteogenomic approach to investigate fresh solid melanoma tumor material. The approach combines liquid chromatography-tandem mass spectrometry (LC-MS/MS) of immuno-precipitated human leukocyte antigen class I (pHLA-I) peptides with whole exome sequencing (WES) for the discovery and validation of neoantigen candidates at the protein level.

Study: Proteogenomic analysis reveals RNA as a source for tumor-agnostic neoantigen identification. Image Credit: Stock-Asso/Shutterstock.com

Background

Genetic mutations give rise to neoantigens that mount an anti-tumor immune response. The challenge is that most cancer patients do not share neoantigens. Thus, validating thousands of in silico-predicted neoantigen candidates is impractical in a clinical setting.

However, with the advancements in the proteogenomic-based discovery of neoantigens, it is becoming easier and facilitating the development of novel immunotherapies.

About the study

In the current study, researchers used tumor material and blood from 32 cancer patients of the ImmoNEO MASTER cohort with diverse tumors for the proteogenomic analysis to find new tumor-suppressing neoantigens.

Researchers phenotyped tumor microenvironments using single-cell suspensions from primary tumor tissues, which they also used for WES and whole-genome sequencing (WGS).

Further, they used bulk transcriptome analysis or ribonucleic acid sequencing (RNA-seq) for flow cytometric (FC) characterization of tumor-invading T cells and fluorescence-activated cell sorting (FACS)-sorting of CD8+ T cells. Subsequently, the researchers called mutant variants and filtered them for single-nucleotide polymorphisms (SNPs).

For the immunogenicity assessment of neoantigen candidates, they used the patient’s tumor-infiltrating lymphocytes (TILs) and peripheral blood mononuclear cells, and healthy donor-derived PBMCs using a co-cultured dendritic cell (acDC) assay.

Finally, the team validated all identified RNA variants and peptides (in-depth) for their tumor-specificity by analyzing their prevalence in tissue RNA-seq data from the genotype-tissue expression (GTEx) project.

Results

In this study, the authors used data sets with unfiltered deoxyribonucleic acid (DNA) and RNA variants to prevent the loss of potential neoantigen candidates. They found that somatic mutations on coding exons and non-coding transcripts, intronic regions, and splice sites represented a source of neoantigens.

On average, they detected 302 somatic mutations per tumor but a much higher number of variants at the RNA level, with an average of 4,024 variants per tumor.

Most neoantigen candidates originated from variants identified in the RNA data set, illustrating the relevance of RNA as a still understudied source of cancer antigens. Thus, RNA-centered variant detection has the potential to identify relevant neoantigen candidates.

Another intriguing observation was that tumors with low levels of somatic mutations also harbored a high amount of RNA variants. They detected nearly ten times more RNA variants compared to DNA variants.

Also, all patients shared a subset of RNA variants; however, only some patients shared DNA variants, and in insignificant frequency.

The authors detected a corresponding canonical sequence at the DNA level for most RNA variants, suggesting that RNA editing events gave rise to some of these variants. Furthermore, they noted that ~97% of variants were unique in this cohort at the DNA level but only 89% at the RNA level.

The team used a Prosit-based rescoring workflow to increase the number of identified neoantigen candidates by 13. It helped them identify 90 neoantigen candidates in 24 patients, with 1 to 13 neoantigen candidates per patient, highlighting that most cancer patients harbor potential targets for personalized immunotherapy.

The peptide length of all identified neoantigen candidates ranged from eight to 14 amino acids with a predominance of nonamers.

Strikingly, 79 of 90 identified neoantigen candidates were derived exclusively from RNA variants, three from DNA variants, and eight shared between sources. Finally, the researchers noted that immunogenic neoantigens were not limited to specific tumor entities and were present in patients with melanomas, sarcomas, and carcinomas.

Nearly 29% of the validated neoantigen candidates elicited an in vitro T-cell response, thus, reducing the need for practically unfeasible and tedious immunogenicity testing in a clinical environment.

Conclusions

The discovery of cancer-specific immunotherapies based on neoantigens is still critical; thus, this area of research might benefit from combinatorial approaches like proteogenomics.

In the current extensively characterized pan-cancer cohort, an improved proteogenomic pipeline established RNA as a significant source of neoantigen candidates and shared tumor antigens.

Furthermore, the authors combined proteogenomics with phenotypic and functional analyses to link the identified candidates with their immunological features. Additionally, they assessed their potential to drive T-cell immune responses.

Together, these advancements could guide the selection process of promising neoantigen candidates for preclinical testing, especially TCR-transgenic T cells represent a great immunotherapeutic target for cancer patients.