Activating the dark genome to illuminate cancer vaccine targets

Abstract

Epigenetic therapy awakens myriad transposable elements to generate new antigens that could prime tumor cells for immunotherapy. A new study of glioblastoma discovers indiscriminate awakening in normal cells also and then presents a more selective strategy for potential therapeutic targeting.

New therapies are needed for glioblastoma (GBM), the most common and aggressive primary brain tumor in adults.¹ Although immunotherapy holds promise for improving outcomes, the complex molecular and micro-environmental architecture of GBMs has thus far rendered immunotherapies ineffective in extending survival. Mutations have been the traditional source of tumor-specific antigens (TSAs) for immunotherapy; however, GBM has relatively few of them, and they may be present in only one part of the tumor.^2–4 These issues highlight the need for expanding the repertoire of targetable antigens in heterogeneous and mutationally sparse cancers like GBM. Cancer scientists, therefore, have sought alternative sources, such as transposable elements (TEs) that can be activated by epigenetic drugs.^5,6 Families of TEs have invaded and evolved with genomes, cumulatively accounting for an astounding 50% of the human genome known as the dark genome or, pejoratively, junk DNA. Thousands of members of each TE family have similar DNA sequences, presenting a significant computational challenge, especially when relying on short-read sequencing. Jang et al.⁷ is a stand-out among the ongoing efforts for its comprehensive computational filtering and attention to the tumor-to-normal differential, narrowing thousands of candidates to a handful of attractive TE targets that have the potential for cancer cell selectivity.

Inhibitors of DNA methyltransferases (DNMT) and histone deacetylases (HDAC) reverse aberrant epigenetic modifications contributing to cancer progression⁸, and activate latent cancer-testis antigens and viral defense genes.^5,6,9 In this study, Jang et al. treated patient-derived GBM stem cells (GSCs), along with dividing and non-dividing normal cells, with a DNMT1 inhibitor (Decitabine) and HDAC inhibitor (Panobinostat) and observed cell type-enriched clusters of newly opened chromatin regions following treatment.¹⁰ Notably, GSC-enriched open chromatin regions proximal to immune-related genes were associated with increased expression of interferon response and antigen-presentation genes. These findings extend prior studies showing that epigenetic therapy primes immune pathways, potentially increasing sensitivity to immunotherapy.

The combination of inhibitors also reactivates normally silenced TEs and cryptic promoters via altering local DNA methylation and histone modifications¹¹, a major focus of this study. When a gene contains an intronic TE, the primary transcript includes both gene exons and TE sequences. Inappropriate splicing may produce a chimeric transcript in which the TE sequence is retained and incorporated into the mature mRNA. Alternatively, cryptic promoters in regulatory remnants of TE can be activated in a similar fashion, initiating transcription at novel sites within the genome to generate chimeric transcripts.¹² The immunogenicity of TSAs created by these activation events has been demonstrated in multiple cancers.^13,14 Here, Jang et al. employed a suite of aptly abbreviated algorithms (SQuIRE, MaxQuant, and pFind) to identify HLA-bound TE-derived antigens from treated GSCs. Filtering against normal tissue expression data (GTEx and BLAST) identified cancer-specific epitopes, and short- and long-read RNA sequencing confirmed or ruled out the expression of their corresponding transcripts. Altogether, this study elegantly demonstrates that epigenetic therapy-induced TEs generate HLA-bound peptides in GSCs.

Epigenetic therapies have the potential to remodel immunosuppressive tumor microenvironments, but they indiscriminately activate hundreds of TE-derived antigens. Many are not tumor-specific and could trigger an immune response against healthy tissue, potentially resulting in autoimmune reactions and toxicities. In this study, a sizeable proportion of the TE-chimeric transcripts were activated by treatment not only in GSCs, but also in dividing or quiescent normal cells at low levels. Jang et al., therefore, revisited HLA-pulldown and mass spectrometry analysis of treated GSCs and quiescent cells, honing in on the small number of candidates that were more cancer-specific. In contrast to indiscriminate activation, the transformative technology CRISPR offers single-locus selectivity. Jang et al. explore the potential of a CRISPR spin-off, CRISPR activation (CRISPRa) technology to precisely activate an individual TE promoter while minimizing off-target effects. Jang et al. used this approach to successfully demethylate individual target DNA regions to reactivate cryptic promoters. This approach best illustrates the direct relationship between DNA methylation and suppression of TE-related cryptic promoters and the feasibility of selectively activating one or a few TE-related transcripts from a sea of similar sequences.

Targetable neoepitopes derived from TEs and other multi-copy DNA sequences date back to the early discoveries of DNA demethylation inhibitors triggering the expression of cancer-testes antigens and human endogenous retrovirus (HERV) TEs.^5,6,15 The exciting results presented by Jang et al.⁷ further highlight the immunotherapeutic potential of inducing TE-derived antigens and present limitations to overcome. GBMs are notoriously heterogeneous and evolving, and even the best TSA candidates may fail if not expressed homogenously and stably over time, issues that are relevant to most cancer therapies. Although Jang’s study suggests that treatment-induced TEs can generate HLA-presented antigens, variations in TE splicing patterns across the tumor landscape may surface as a new challenge in identifying targets that are consistent over space and time. This would necessitate approaches to find TEs that are reliably expressed across the tumor landscape. Future research may evaluate these antigens in preclinical settings using animal models of cancer. It remains to be determined whether the antigen presentation described by Jang generates a durable immune response. Evaluating the long-term efficacy and safety of inducing and targeting TE-related antigens in preclinical settings will be essential to translating these discoveries into clinical trials. With the early success of personalized mRNA cancer vaccines, going to the dark side of the genome may be on the horizon.

Figure.

a. Epigenetic inhibitors activate chimeric transcripts. The TE-derived chimeric transcripts are translated, creating neoepitopes that are presented on the cell surface by HLA molecules. b. Targeted CRISPRa demethylation selectively activates individual TEs.