Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Oct 23;14(1):6731.
doi: 10.1038/s41467-023-42417-w.

DNMT and HDAC inhibition induces immunogenic neoantigens from human endogenous retroviral element-derived transcripts

Ashish Goyal #  1 , Jens Bauer #  2   3   4 , Joschka Hey #  1   5   6 , Dimitris N Papageorgiou  7   8 , Ekaterina Stepanova  9 , Michael Daskalakis  1   10 , Jonas Scheid  2   3   4   11 , Marissa Dubbelaar  2   3   4   11 , Boris Klimovich  4   12 , Dominic Schwarz  7 , Melanie Märklin  4   12 , Malte Roerden  2   4 , Yu-Yu Lin  1 , Tobias Ma  13 , Oliver Mücke  1 , Hans-Georg Rammensee  3   4   14 , Michael Lübbert  13 , Fabricio Loayza-Puch  9 , Jeroen Krijgsveld  7   8 , Juliane S Walz  15   16   17   18 , Christoph Plass  19   20   21
Affiliations

DNMT and HDAC inhibition induces immunogenic neoantigens from human endogenous retroviral element-derived transcripts

Ashish Goyal et al. Nat Commun. .

Abstract

Immunotherapies targeting cancer-specific neoantigens have revolutionized the treatment of cancer patients. Recent evidence suggests that epigenetic therapies synergize with immunotherapies, mediated by the de-repression of endogenous retroviral element (ERV)-encoded promoters, and the initiation of transcription. Here, we use deep RNA sequencing from cancer cell lines treated with DNA methyltransferase inhibitor (DNMTi) and/or Histone deacetylase inhibitor (HDACi), to assemble a de novo transcriptome and identify several thousand ERV-derived, treatment-induced novel polyadenylated transcripts (TINPATs). Using immunopeptidomics, we demonstrate the human leukocyte antigen (HLA) presentation of 45 spectra-validated treatment-induced neopeptides (t-neopeptides) arising from TINPATs. We illustrate the potential of the identified t-neopeptides to elicit a T-cell response to effectively target cancer cells. We further verify the presence of t-neopeptides in AML patient samples after in vivo treatment with the DNMT inhibitor Decitabine. Our findings highlight the potential of ERV-derived neoantigens in epigenetic and immune therapies.

PubMed Disclaimer

Conflict of interest statement

The University Hospital Tübingen and German Cancer Research Center (DKFZ, Heidelberg) is in the process of applying for a patent application (Patent number: EP23170469.3, Title: Neoepitope immunogenic peptides for cancer treatment and diagnosis) covering t-neopeptides that lists A.G., J.B., J.H., J.S.W, and C.P. as inventors. The other authors declare no competing interests.

Figures

Fig. 1
Fig. 1. De novo transcriptome assembly identifies treatment-induced novel poly-adenylated transcripts (TINPATs).
a Schematic overview of the experimental setup comprising the de novo transcriptome assembly of DMSO, Decitabine (DAC), SB939, or DAC + SB939-treated NCI-H1299 cells as well as the isolation of (human leukocyte antigen) HLA-presented ligands from DMSO or DAC + SB939-treated NCI-H1299 cells followed by (mass spectrometry) MS-based immunopeptidome analysis (created with BioRender). b RNA-seq performed on biological replicates (n = 3) of NCI-H1299 cells treated with DMSO, SB939, DAC, or DAC + SB939 was used to generate a de novo transcriptome assembly (see methods). c Volcano plot of differential transcript expression (defined by DESeq2: adjusted p-value < 0.01, absolute log2 fold change > 2) analyses for indicated comparisons. The color of the dots indicates the transcript classification (gray: known, purple: non-chimeric, light purple: chimeric). d Heatmap of z-scaled variance stabilized transformed (vst) transcript expression of DAC + SB939-induced transcripts stratified by their transcript classification. Box plots of vst transcript expression are plotted alongside. e–g Number (e), length (f), and the number of exons (g) of DAC + SB939-induced transcripts stratified by their transcript classification. The number of mono- and multi-exonic transcripts are plotted alongside. Box plots indicate the largest value within the 1.5 times interquartile range above 75th percentile, 75th percentile, median, 25th percentile, and smallest value within the 1.5 times interquartile range below 25th percentile (n = 3 biological replicates). h Number of transcriptional start sites (TSSs) of DAC + SB939-induced transcripts, overlapping with transposable element (TE) families and their (i) locus overlap and enrichment analysis with TE families, subfamilies, and classes stratified by transcript classification. The most significant enrichment results for TE families and subfamilies, and LTR12-derived classes, are visualized. TSSs of all identified transcripts served as background (p-values were calculated using a Fisher’s test as the statistical framework within R package LOLA). j, k Locus plots of selected chimeric (j) and non-chimeric (k) novel transcripts, showing (from top to bottom) LTR12 repeats, the de novo transcriptome assembly, normalized RNA-seq coverage, splicing sites of DAC + SB939 treated NCI-H1299 cells.
Fig. 2
Fig. 2. Immunopeptidomics unravels DAC + SB939-induced novel ORFs-derived HLA ligands.
a Number of ORFs predicted from DAC + SB939-induced transcripts stratified by their host transcript classification. b Frequency distribution of the number and (c) length of the predicted ORFs. d Volcano plot of differential protein abundance analyses, comparing DAC + SB939 vs DMSO treated NCI-H1299 cells measured by MS using predicted ORFs from DAC + SB939-induced transcripts as reference. e MS-identified HLA class I ligands from DAC + SB939 vs DMSO treated cells (n = 3 biological replicates; mean values ± SD; p-value was calculated using an unpaired two-tailed t test). f Overlap analysis of MS-identified HLA ligands from DAC + SB939 vs DMSO treated cells from all identified ligands (left) or novel ORF-exclusive ligands (right). g Peptide length distribution of HLA ligands exclusively assigned to novel ORFs compared to all identified canonical HLA ligands. h Ranked intensity values of MS acquired data derived from the combined immunopeptidomes of DAC + SB939 and DMSO treated cells (n = 3). Positions of novel ORF-derived HLA ligands are projected on the curve. i Comparative profiling of ORF-derived HLA ligands (n = 112) based on HLA-restricted presentation frequency in immunopeptidomes of DAC + SB939 or DMSO treated cells. Frequencies of positive immunopeptidomes for the respective HLA ligands (x-axis) are indicated on the y-axis. The box highlights treatment-induced neopeptides (t-neopeptides; n = 15) with presentation in 100% of the biological replicates. The corresponding calculated spectral correlation coefficients (R2) of the fragment spectra comparison are indicated below. j Fragment spectra comparison (m/z on the x-axis) of the experimentally eluted DAC + SB939-induced novel ORF-derived HLA class I-presented ligand LLLSYRYIY (PA*32) extracted from the DAC + SB939 treated cells (identification) to the synthetic peptide (validation, mirrored on the x-axis) with the calculated R2. Identified b- and y-ions are marked in red and blue, respectively. k Calculated R2 of the fragment spectra comparison of experimentally eluted and synthetic peptides plotted against the number of samples with ligand identification (n = 6). The band indicates the median, and the box indicates the first and third quartiles.
Fig. 3
Fig. 3. t-neopeptide encoding transcripts are mostly LTR12-derived.
a The number of DAC + SB939-induced transcripts giving rise to peptides identified in more than one biological replicate of DAC + SB939 treated samples and absent in DMSO treated samples. b Locus plot of a selected treatment-induced novel polyA+ transcript (TINPAT) giving rise to t-neopeptides outside repeat sequence, showing (from top to bottom) LTRs, the de novo transcriptome assembly, normalized RNA-seq coverage of DMSO, SB939, DAC, and DAC + SB939 treated NCI-H1299 cells, predicted ORF, and corresponding t-neopeptide LLLSYRYIY. The region containing the t-neopeptide is zoomed in and reversed for improved visualization. c Classification of transcripts giving rise to t-neopeptides in known, chimeric, and non-chimeric novel transcripts. d The number of transcriptional start sites of transcripts that give rise to t-neopeptides and overlap with transposable element families. e Locus overlap and enrichment analysis of transcripts that give rise to t-neopeptides with transposable element families, subfamilies, and classes. The most significant enrichment result for transposable element families and subfamilies, as well as all LTR12-derived classes, are visualized. Transcriptional start sites of all DAC + SB939 treatment-induced transcripts were used as a background for enrichment analysis. p-values were calculated using Fisher’s exact test. f Transcript expression fold changes of transcripts that give rise to t-neopeptides (red) and all transcripts (gray) in the DAC + SB939 vs DMSO, DAC vs DMSO, and SB939 vs DMSO comparison. Box plots indicate the largest value within the 1.5 times interquartile range above 75th percentile, 75th percentile, median, 25th percentile, and smallest value within the 1.5 times interquartile range below 25th percentile (n = 3 biological replicates).
Fig. 4
Fig. 4. t-neopeptides activate T cells and induce cytotoxic responses.
a Representative example of flow cytometry-based characterization of PA*32-specific CD8+ T cells of a healthy volunteer (HV) after in vitro artificial antigen-presenting cell (aAPC)-priming with HLA-A*32-PA*32-monomer. b Frequencies of PA*32-specific CD8+ T cells compared to CD8+ T cells primed with an HLA-matched negative peptide in HVs (n = 3, frequency of T cells is indicated per well). All data points are shown, the band indicates the median, and the box indicates the first and third quartiles. c Representative example of IFN-γ and TNF production of PA*32-specific CD8+ T cells stimulated with PA*32 (upper panel) or an HLA-matched negative peptide (lower panel) after aAPC-priming. d Specific cell lysis by PA*32-specific CD8+ T cells of DAC + SB-treated NCI-H1299 cells at different effector-to-target cell ratios compared to PA*32-unspecific CD8+ T cells. Shown is the area of NCI-H1299 cells normalized to time point 0 h (t0) over time until 190 h. Results are shown as mean ± SEM for three independent technical replicates.
Fig. 5
Fig. 5. Induction of treatment-induced transcripts is conserved across different cancer entities.
a LTR12C expression analysis performed via quantitative PCR (qPCR) on biological replicates across a panel of cancer cell lines treated with DMSO, SB939, DAC, or DAC + SB939, respectively. Expression is normalized to GAPDH and the corresponding DMSO control (mean values ± SD; n = the number of biological replicates; p-values indicated above each bar, are calculated using a two-tailed unpaired t-test and corrected for multiple comparisons using the Holm-Sidak method). b Principal component (PC) analysis of the 5000 most variably expressed transcripts of DAC + SB939 (n = 3) and DMSO-treated (n = 3) cancer cell lines, assessed via RNA-seq. Transcript expression was quantified using the de novo transcriptome assembly of all cell lines. PC1 and PC3 are visualized. c The number of differentially induced transcripts, defined by DESeq2 (adjusted p-value < 0.01, log2 fold change >2,) in the DAC + SB939 vs DMSO comparisons of NCI-H1299 cells and the cell line panel de novo transcriptome assembly. d Differentially induced transcripts, in the DAC + SB939 vs DMSO comparisons of the cell line panel de novo transcriptome assembly. Colors are only mapped from the 1st to the 99th quantile of the gene expression matrix. The cell line identity was included in the design formula to adjust for cell line-specific differences. Box plots of vst transcript expression are plotted alongside and indicate the largest value within the 1.5 times interquartile range above 75th percentile, 75th percentile, median, 25th percentile, and smallest value within the 1.5 times interquartile range below 25th percentile (n = 3 biological replicates). e Locus plot of selected chimeric transcript DNAH3, identified in the cancer cell line panel post DAC + SB939 treatment, showing (from top to bottom) LTR12 repeats, the de novo cell line panel transcriptome assembly, normalized RNA-seq coverage of DMSO, and DAC + SB939 treated cancer cell lines.
Fig. 6
Fig. 6. Identification of t-neopeptides in AML patients treated with DAC.
a Schematic overview of the blood sampling from AML patients treated with DAC. Indicated are the blood collections pretreatment as well as 48 h and 96 h post-DAC treatment (created with BioRender). b, c The number of MS-identified HLA ligands from AML patients’ material. Available samples were from uniform patient number (UPN) 1 and UPN2, both pre- and post-DAC treatment of (b) HLA class I ligands and (c) HLA class II ligands, respectively. d, e Overlap analysis of MS-identified novel ORF-exclusive HLA ligands from AML patients UPN1 and UPN2 after DAC treatment (48 h and 96 h time point combined) compared to the pretreatment of (d) HLA class I ligands and (e), HLA class II ligands. (f) t-neopeptide-specific T cell responses of AML patient UPN9 after DNMTi and valproic acid (VPA) treatment assessed by IFN-γ ELISPOT assay after in vitro stimulation with the HLA class II peptide pool compared to the negative control. g Locus plot of the non-chimeric transcript MSTRG.21956.5 giving rise to a t-neopeptide in NCI-H1299 cells treated with DAC + SB939 and UPN2 treated with DAC. The locus plot shows (from top to bottom) LTR12 repeats, the de novo cell line panel transcriptome assembly, normalized RNA-seq coverage of DMSO, and DAC + SB939 treated cancer cell lines and Ribo-seq coverage (from the + and − strand) of DMSO and DAC + SB939 treated NCI-H1299 cells, as well as the predicted ORF and location of the t-neopeptides identified in NCI-H1299 cells (red) and UPN2 (blue). The region containing the t-neopeptides is zoomed in and reversed for improved visualization.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Maslak PG, et al. Phase 2 trial of a multivalent WT1 peptide vaccine (galinpepimut-S) in acute myeloid leukemia. Blood Adv. 2018;2:224–234. - PMC - PubMed
    1. Nelde A, et al. Upstream open reading frames regulate translation of cancer-associated transcripts and encode HLA-presented immunogenic tumor antigens. Cell Mol. Life Sci. 2022;79:171. - PMC - PubMed
    1. Sahin U, et al. Personalized RNA mutanome vaccines mobilize poly-specific therapeutic immunity against cancer. Nature. 2017;547:222–226. - PubMed
    1. Bauer, J., Nelde, A., Bilich, T. & Walz, J. S. Antigen Targets for the Development of Immunotherapies in Leukemia. Int. J. Mol. Sci.2010.3390/ijms20061397 (2019). - PMC - PubMed
    1. Nelde A, et al. Immunopeptidomics-guided warehouse design for peptide-based immunotherapy in chronic lymphocytic leukemia. Front. Immunol. 2021;12:705974. - PMC - PubMed

Publication types

Substances