Proteogenomic identification of Hepatitis B virus (HBV) genotype-specific HLA-I restricted peptides from HBV-positive patient liver tissues

Mayank Srivastava, Richard Copin, Augustine Choy, Anbo Zhou, Olav Olsen, Sarah E. Wolf, Darshit Shah, Anna Rye-Weller, Lisa Chen, Newton Chan, Angel Coppola, Kathryn Lanza, Nicole Negron, Min Ni, Gurinder S. Atwal, Christos A. Kyratsous, William C. Olson, Robert SalzlerFrontiers in Immunology 2022 · DOI

TL;DR

This paper introduces a novel proteogenomic approach to identify Hepatitis B virus (HBV) peptides presented by human leukocyte antigen (HLA) class I molecules in HBV-infected liver tissues. By reconstructing patient-specific HBV genomes and combining this with immunopeptidomics, they discovered new HBV-derived peptides that could serve as targets for immunotherapies against HBV-related liver cancer.

What Problem Does This Paper Address?

Chronic Hepatitis B virus (HBV) infection is a major global health issue, leading to serious liver complications like cirrhosis and hepatocellular carcinoma (HCC). While the immune system, particularly CD8+ T cells, plays a crucial role in controlling viral infections, chronic HBV often leads to immune evasion. Developing effective immunotherapies for HBV-related HCC is challenging due to the significant genetic variability of HBV strains and the diversity of human HLA alleles, which dictate how viral peptides are presented to T cells. Existing methods struggle to identify relevant viral peptides directly from patient tissues, especially when HBV DNA integrates into the host genome, leading to fragmented and diverse viral sequences. This paper aims to overcome these challenges by developing a method to precisely identify HBV-specific HLA-I restricted peptides directly from patient liver samples.

Methods

The researchers developed a multi-step proteogenomic approach:

  1. Patient-Specific HBV Genome Reconstruction: They used RNA sequencing (RNA-seq) data from patient liver samples to reconstruct individual HBV genomes. This involved assembling RNA reads into contigs, identifying HBV-specific sequences, and then reconstructing the complete or partial viral genome for each patient. This is a key innovation, moving beyond reliance on generic reference genomes.
  2. HBV Integration Site Identification: They also identified instances where HBV DNA fragments integrated into the human genome by looking for "hybrid contigs" that contained both HBV and human sequences.
  3. HLA Genotyping: DNA sequencing was performed to determine the specific HLA class I alleles present in each patient.
  4. Immunopeptidomics: HLA class I molecules were isolated from liver tissue using an anti-HLA antibody. The peptides bound to these HLA molecules were then purified and analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS).
  5. Peptide Identification and Validation: The LC-MS/MS data was searched against a combined database of human and patient-specific HBV protein sequences. Novel peptides were validated using synthetic heavy peptide analogues, comparing their retention times and fragmentation patterns.

Datasets

The study utilized liver tissue samples from 48 HBV-infected patients, primarily from Asia. These samples were used for RNA sequencing, DNA sequencing (for HLA genotyping), and immunopeptidomics. The RNA-seq data was used to generate patient-specific HBV genome databases.

Key Insights and Ideas

  • Patient-Specific Databases are Crucial: Relying on generic HBV reference genomes is insufficient due to viral variability. Reconstructing patient-specific HBV genomes from RNA-seq data is essential for accurate peptide identification.
  • HBV Integration is Common and Variable: HBV DNA integration into the human genome is a significant factor in HBV-related HCC. The study identified breakpoints of these integrations, showing they often occur in specific regions of the HBV genome but with variable host genome locations.
  • Novel HBV Peptides Discovered: The proteogenomic approach successfully identified previously unreported HBV peptides, particularly from the HBV polymerase protein, that are presented by HLA class I molecules.
  • Genotype-Specific Peptides: The study highlighted the importance of considering HBV genotypes, as some newly identified peptides showed variations across different genotypes.
  • Therapeutic Potential: The identified peptides, especially those found frequently and predicted to bind strongly to patient HLA alleles, represent promising targets for developing T cell-based immunotherapies (e.g., bispecific antibodies, CAR-T cells) for HBV-related HCC.

Key Results

  • The study identified a total of 54 unique HBV peptides of lengths 9-12 residues.
  • Of these, 45 were previously unreported.
  • A novel peptide, GSLPQEHIVQK (from the HBV polymerase protein, position P606-616), and its variants were identified in approximately 74% of HLA-A11+ patient samples. This peptide and its variants were found to be frequently detected.
  • Two previously reported HBV-derived peptides were also detected but at very low frequencies.
  • The study demonstrated that the majority of identified HBV peptides were predicted to bind to the patient's specific HLA alleles.
  • Validation using synthetic heavy peptide analogues confirmed the identity and fragmentation patterns of key identified peptides.

Limitations and Future Work

  • HBV Genotype Conservation: While novel peptides were identified, their conservation across all HBV genotypes was not ideal, with some amino acid changes observed across variants. The authors suggest that T cell therapeutics could potentially be designed around conserved regions of these peptides.
  • Limited Patient Sample Diversity: The study primarily focused on patients from Asia, and the diversity of HLA alleles and HBV genotypes represented might not cover all global populations.
  • Future Work: The authors propose using this proteogenomic approach to analyze more patient samples with diverse HLA alleles to identify peptides that better fit all desired criteria for therapeutic targets. They also suggest that understanding which regions of the HBV genome are frequently integrated into HCC cells is crucial for selecting optimal peptide targets.

Summary updated

Abstract

The presentation of virus-derived peptides by HLA class I molecules on the surface of an infected cell and the recognition of these HLA-peptide complexes by, and subsequent activation of, CD8<sup>+</sup> cytotoxic T cells provides an important mechanism for immune protection against viruses. Recent advances in proteogenomics have allowed researchers to discover a growing number of unique HLA-restricted viral peptides, resulting in a rapidly expanding repertoire of targets for immunotherapeutics (i.e. bispecific antibodies, engineered T-cell receptors (TCRs), chimeric antigen receptor T-cells (CAR-Ts)) to infected tissues. However, genomic variability between viral strains, such as Hepatitis-B virus (HBV), in combination with differences in patient HLA alleles, make it difficult to develop therapeutics against these targets. To address this challenge, we developed a novel proteogenomics approach for generating patient-specific databases that enable the identification of viral peptides based on the viral transcriptomes sequenced from individual patient liver samples. We also utilized DNA sequencing of patient samples to identify HLA genotypes and assist in target selection. Liver samples from 48 HBV infected patients, primarily from Asia, were examined to reconstruct patient-specific HBV genomes, identify regions within the human chromosomes targeted by HBV integrations and obtain a comprehensive view of HBV peptide epitopes using our HLA class-I (HLA-I) immunopeptidomics discovery platform. Two previously reported HLA associated HBV-derived peptides, HLA-A02 binder FLLTRILTI (S<sub>194-202</sub>) from the large surface antigen and HLA-A11 binder STLPETTVVRR (C<sub>141-151</sub>) from the capsid protein were validated by our discovery platform, but both were detected at very low frequencies. In addition, we identified and validated, using heavy peptide analogues, novel strain-specific HBV-HLA associated peptides, such as GSLPQEHIVQK (P<sub>606-616</sub>) and variants. Overall, our novel approach can guide the development of bispecific antibody, TCR-T, or CAR-T based therapeutics for the treatment of HBV-related HCC and inform vaccine development.
BibTeX
@inproceedings{srivastava2022proteogenomic,
  author = {Mayank Srivastava and Richard Copin and Augustine Choy and Anbo Zhou and Olav Olsen and Sarah E. Wolf and Darshit Shah and Anna Rye-Weller and Lisa Chen and Newton Chan and Angel Coppola and Kathryn Lanza and Nicole Negron and Min Ni and Gurinder S. Atwal and Christos A. Kyratsous and William C. Olson and Robert Salzler},
  title = {{Proteogenomic identification of Hepatitis B virus (HBV) genotype-specific HLA-I restricted peptides from HBV-positive patient liver tissues}},
  booktitle = {Frontiers in Immunology},
  year = {2022},
  doi = {10.3389/fimmu.2022.1032716},
}
Figure