My current research
identification of non-canonical neoantigens in tumors
Transposable elements in GLIOBLASTOMA
Transposable elements (TE) populate 45% of the genome and they are generally overexpressed in cancer due to epigenetic disregulations. We identified that 2 subpopulations of TEs are enriched in glioblastoma immunopeptidome. First, young TEs (mainly LINEs) that share homology with many other transposed copies in the genome, being then redundant and less tumor-specific. A second class of evolutionary older TEs, with mutated and degenerated sequences, was more tumor-enriched and potentially more immunogenic.
You can find the paper in here.
splicing between exons and transposable elements
Aberrant splicing is known to be upregulated in tumors. We then investigated if there are tumor-specific splicing junctions between exons and transposable elements (JETs) in LUAD tumors. We found a population of shared JETs that was translated and presented in MHC-I molecules of tumor cells. These MHC peptide were recognised by Tumor-infiltrating T cells, suggesting an endogenous immune response. In addition, specific T cell clones were capable to kill cell lines endogenously expressing the JET.
Transposable elements & Protein evolution
You can find the paper in here.
mRNA splicing enlarges the diversity of the protein repertoire, mainly through alternative exon retention and skipping. A small proportion of annotated exons in the human proteome is derived from transposable elements (TEs) that were exonized during evolution. Recent developments in transcript analyses further identified unannotated splicing variants that exonize intronic TEs and are largely characterized at the mRNA level. Their contribution to the cellular proteome and their possible biological relevance, however, remain unclear. Here, we use transcriptome assembly, ribosome profiling, and mass spectrometry to describe a population of translated isoforms from known genes that are generated by non-canonical splicing between exons and TEs. Despite being shorter and expressed at relatively low levels, these isoforms are shared between individuals, efficiently translated and functional. We show that TE exonization occurs preferentially in evolutionary ancient genes (which have more introns), but the splice sites involved are relatively recent. Isoforms expressed more frequently in human populations use more evolutionarily conserved splice sites, and exonized TE sequences can adopt secondary structures, enriched in α-helices. We conclude that non-canonical, lowly expressed TE-containing splicing variants represent a diversity reservoir of functional protein isoforms on which natural selection can act.