Periodic Reporting for period 3 - CodingHeart (Novel Coding Factors in Heart Disease)
Periodo di rendicontazione: 2022-01-01 al 2023-06-30
Heart failure has become a worldwide epidemic with to date more than 26 million people affected resulting in devastating consequences for patients and an enormous burden on health care systems. One in five heart failure patients dies within a year of diagnosis and survival estimates after diagnosis are 50% and 10% at 5 and 10 years, respectively. Despite intensive investigation, the molecular mechanisms leading to heart failure remain poorly understood. We propose to narrow this critical gap in knowledge with a previously unattainable, comprehensive approach to define the genomic architecture and functional consequences of newly identified so-called micropeptides from multiple classes of RNAs, previously classified as non-coding, in cardiac biology and heart failure. In fact, the human proteome is expected to contain many micropeptides translated from presumed long noncoding RNAs (lncRNAs). For a long time, these had remained undetected because of their small size (< 100 amino acids). For a handful of these micropeptides, physiological roles have been uncovered to date, including the novel heart-specific SERCA modulators DWORF and Myoregulin. Yet, a genome-wide catalogue of micropeptides produced in human cardiac tissue is missing.
Throughout this project, we want to discern fundamental causes of maladaptive responses in the heart as well as open strategies to monitor and limit them. To solve all these scientific questions we follow four main objectives: (i) we aim to detect short open reading frames (sORFs) in transcribed lncRNAs and circRNAs in human and rat hearts, (ii) we planned a computational and single cell sequencing based micropeptide candidate selection, (iii) we plan high-throughput in vitro and in vivo characterization of candidate micropeptides, and (iv) we finally aim to define disease mechanisms and pathophysiological consequences of our detected novel micropeptides.
Throughout this project, we want to discern fundamental causes of maladaptive responses in the heart as well as open strategies to monitor and limit them. To solve all these scientific questions we follow four main objectives: (i) we aim to detect short open reading frames (sORFs) in transcribed lncRNAs and circRNAs in human and rat hearts, (ii) we planned a computational and single cell sequencing based micropeptide candidate selection, (iii) we plan high-throughput in vitro and in vivo characterization of candidate micropeptides, and (iv) we finally aim to define disease mechanisms and pathophysiological consequences of our detected novel micropeptides.
As proposed, we have generated a genome-wide catalog of novel cardiac micropeptides in rat and human hearts. We were able to identify 74 lncRNA encoded micropeptides in rat hearts, and 339 micropeptides in human hearts. Interestingly, we identified dozens of human micropeptides that are translated from lncRNAs with well characterized functions, suggesting undiscovered roles of these micropeptides, or their involvement in biological functions assigned to the lncRNA. Originally, we proposed to only investigate coding functions of lncRNAs. However, since it became evident that 5’ untranslated regions (5’UTRs) of canonical mRNAs can also harbor micropeptides, we extended our micropeptide search to 5’UTRs. We identified 1,125 micropeptides in rat and 1090 micropeptides in human hearts to be encoded by 5’UTRs. Our human data were published in 2019 in Cell (van Heesch et al., 2019), our rat data were published in 2021 in Genome Biology (Witte et al., 2021).
To elucidate the biological role of detected micropeptides, we investigated several micropeptide features. These features include their expression pattern across tissues and cardiac cell types, secondary structure, differential expression in diseased hearts, and conservation. We found seventeen of the 169 translated lncRNAs to be specifically expressed in the heart based on GTEx expression data. Using the latest single cell technologies, we were able to generate a comprehensive single-nucleus and non-myocyte single-cell dataset, derived from 6 different atrial and ventricular locations of 14 different donor hearts (Litvinukova et al, Nature, 2020). We applied downstream statistical analysis on our single-cell atlas of the adult human heart to define cell-type specificity of identified translated lncRNAs. We found several lncRNAs to be cell-type specifically expressed, with the majority of cardiac-specific translated lncRNAs assigned to cardiomyocytes. In silico prediction of cleavage sites and secondary structures revealed 3 micropeptides that are potentially secreted, indicating a potential role as signaling molecules, and 7 micropeptides that might contain transmembrane domains, a feature often observed across the few known and biologically characterized micropeptides. Moreover, we find 34 and 7 micropeptides that are up- and downregulated in dilated cardioymyopathy (DCM) hearts, respectively. We only detect 17 micropeptides with strong sequence conservation to vertebrates based on PhyloCSF scores, but can align many lncRNAs to the genomes of other hominid species (chimp, gorilla, and orangutan; n = 79) or to the genomes of other primates or mammals (n = 31 or 43, respectively), with 16 being completely specific to humans (van Heesch et al., 2019). We found 46 micropeptides were localized to mitochondria, highlighting a potential involvement in cardiac energy metabolism. Additionally, we have designed and performed a high-throughput, peptide array based interactome screen termed PRISMA (protein interaction screen on peptide matrix). This screen is currently being analyzed, we have first results that are concordant with the detected localization of these micropeptides.
To elucidate the biological role of detected micropeptides, we investigated several micropeptide features. These features include their expression pattern across tissues and cardiac cell types, secondary structure, differential expression in diseased hearts, and conservation. We found seventeen of the 169 translated lncRNAs to be specifically expressed in the heart based on GTEx expression data. Using the latest single cell technologies, we were able to generate a comprehensive single-nucleus and non-myocyte single-cell dataset, derived from 6 different atrial and ventricular locations of 14 different donor hearts (Litvinukova et al, Nature, 2020). We applied downstream statistical analysis on our single-cell atlas of the adult human heart to define cell-type specificity of identified translated lncRNAs. We found several lncRNAs to be cell-type specifically expressed, with the majority of cardiac-specific translated lncRNAs assigned to cardiomyocytes. In silico prediction of cleavage sites and secondary structures revealed 3 micropeptides that are potentially secreted, indicating a potential role as signaling molecules, and 7 micropeptides that might contain transmembrane domains, a feature often observed across the few known and biologically characterized micropeptides. Moreover, we find 34 and 7 micropeptides that are up- and downregulated in dilated cardioymyopathy (DCM) hearts, respectively. We only detect 17 micropeptides with strong sequence conservation to vertebrates based on PhyloCSF scores, but can align many lncRNAs to the genomes of other hominid species (chimp, gorilla, and orangutan; n = 79) or to the genomes of other primates or mammals (n = 31 or 43, respectively), with 16 being completely specific to humans (van Heesch et al., 2019). We found 46 micropeptides were localized to mitochondria, highlighting a potential involvement in cardiac energy metabolism. Additionally, we have designed and performed a high-throughput, peptide array based interactome screen termed PRISMA (protein interaction screen on peptide matrix). This screen is currently being analyzed, we have first results that are concordant with the detected localization of these micropeptides.
To more precisely determine the phylogenetic age of the micropeptide itself rather than its encoding lncRNA, we developed a new pipeline based on frame sequences across 120 mammalian genomes revealing that half of the microproteins only emerged in primate species, whereas 30% of the microproteins recently evolved de novo in humans and chimpanzees. Lastly and in addition to our proposed work in work package 2, we found that the expression of many translated lncRNAs correlates with canonical, nuclear-encoded mitochondrial genes suggesting a potential involvement of encoded micropeptides in mitochondrial processes (van Heesch et al., 2019). To study the role of micropeptides in cardiac disease in-vivo, we are establishing a pooled CRISPR screen targeting all the 169 translated lncRNAs. Perturbation of the gene expression will be followed by single cell RNA-sequencing to check for possible effects on the transcriptome level when comparing with wild type cells. First, induced stem cell (iPSC)-derived cardiomyocytes will be targeted, in further steps we also aim to target other cardiac cell types such as myofibroblasts, but also cardiac organoids to approach a holistic overall image. Additionally, and alternatively to generating monoclonal knockout cell lines, the CRISPRi system can be used to generate stable knockdown cell lines, a process working quicker and requiring less testing than the first mentioned option. Several stable knockdown cell lines could be already generated and more will follow for genes of interest.