Atypical acute myeloid leukemia-specific transcripts generate shared and immunogenic MHC class-I-associated epitopes.

Acute myeloid leukemia (AML) has not benefited from innovative immunotherapies, mainly because of the lack of actionable immune targets. Using an original proteogenomic approach, we analyzed the major histocompatibility complex class I (MHC class I)-associated immunopeptidome of 19 primary AML samples and identified 58 tumor-specific antigens (TSAs). These TSAs bore no mutations and derived mainly (86%) from supposedly non-coding genomic regions. Two AML-specific aberrations were instrumental in the biogenesis of TSAs, intron retention, and epigenetic changes. Indeed, 48% of TSAs resulted from intron retention and translation, and their RNA expression correlated with mutations of epigenetic modifiers (e.g., DNMT3A). AML TSA-coding transcripts were highly shared among patients and were expressed in both blasts and leukemic stem cells. In AML patients, the predicted number of TSAs correlated with spontaneous expansion of cognate T cell receptor clonotypes, accumulation of activated cytotoxic T cells, immunoediting, and improved survival. These TSAs represent attractive targets for AML immunotherapy.

Proteogenomics and Differential Ion Mobility Enable the Exploration of the Mutational Landscape in Colon Cancer Cells.

The sensitivity and depth of proteomic analyses are limited by isobaric ions and interferences that preclude the identification of low abundance peptides. Extensive sample fractionation is often required to extend proteome coverage when sample amount is not a limitation. Ion mobility devices provide a viable alternate approach to resolve confounding ions and improve peak capacity and mass spectrometry (MS) sensitivity. Here, we report the integration of differential ion mobility with segmented ion fractionation (SIFT) to enhance the comprehensiveness of proteomic analyses. The combination of differential ion mobility and SIFT, where narrow windows of ∼m/z 100 are acquired in turn, is found particularly advantageous in the analysis of protein digests and typically provided more than 60% gain in identification compared to conventional single-shot LC-MS/MS. The application of this approach is further demonstrated for the analysis of tryptic digests from different colorectal cancer cell lines where the enhanced sensitivity enabled the identification of single amino acid variants that were correlated with the corresponding transcriptomic data sets.

High frequency of germline RUNX1 mutations in patients with RUNX1-mutated AML.

RUNX1 is mutated in ∼10% of adult acute myeloid leukemia (AML). Although most RUNX1 mutations in this disease are believed to be acquired, they can also be germline. Indeed, germline RUNX1 mutations result in the well-described autosomal-dominant familial platelet disorder with predisposition to hematologic malignancies (RUNX1-FPD, FPD/AML, FPDMM); ∼44% of affected individuals progress to AML or myelodysplastic syndromes. Using the Leucegene RUNX1 AML patient group, we sought to investigate the proportion of germline vs acquired RUNX1 mutations in this cohort. Our results showed that 30% of RUNX1 mutations in our AML cohort are germline. Molecular profiling revealed higher frequencies of NRAS mutations and other mutations known to activate various signaling pathways in these patients with RUNX1 germline-mutated AML. Moreover, 2 patients (mother and son) had co-occurrence of RUNX1 and CEBPA germline mutations, with variable AML disease onset at 59 and 27 years, respectively. Together, these data suggest a higher than anticipated frequency of germline RUNX1 mutations in the Leucegene cohort and further highlight the importance of testing for RUNX1 mutations in instances in which allogeneic stem cell transplantation using a related donor is envisioned.

CAMAP: Artificial neural networks unveil the role of codon arrangement in modulating MHC-I peptides presentation.

MHC-I associated peptides (MAPs) play a central role in the elimination of virus-infected and neoplastic cells by CD8 T cells. However, accurately predicting the MAP repertoire remains difficult, because only a fraction of the transcriptome generates MAPs. In this study, we investigated whether codon arrangement (usage and placement) regulates MAP biogenesis. We developed an artificial neural network called Codon Arrangement MAP Predictor (CAMAP), predicting MAP presentation solely from mRNA sequences flanking the MAP-coding codons (MCCs), while excluding the MCC per se. CAMAP predictions were significantly more accurate when using original codon sequences than shuffled codon sequences which reflect amino acid usage. Furthermore, predictions were independent of mRNA expression and MAP binding affinity to MHC-I molecules and applied to several cell types and species. Combining MAP ligand scores, transcript expression level and CAMAP scores was particularly useful to increase MAP prediction accuracy. Using an in vitro assay, we showed that varying the synonymous codons in the regions flanking the MCCs (without changing the amino acid sequence) resulted in significant modulation of MAP presentation at the cell surface. Taken together, our results demonstrate the role of codon arrangement in the regulation of MAP presentation and support integration of both translational and post-translational events in predictive algorithms to ameliorate modeling of the immunopeptidome.

Factorized embeddings learns rich and biologically meaningful embedding spaces using factorized tensor decomposition.

MOTIVATION: The recent development of sequencing technologies revolutionized our understanding of the inner workings of the cell as well as the way disease is treated. A single RNA sequencing (RNA-Seq) experiment, however, measures tens of thousands of parameters simultaneously. While the results are information rich, data analysis provides a challenge. Dimensionality reduction methods help with this task by extracting patterns from the data by compressing it into compact vector representations. RESULTS: We present the factorized embeddings (FE) model, a self-supervised deep learning algorithm that learns simultaneously, by tensor factorization, gene and sample representation spaces. We ran the model on RNA-Seq data from two large-scale cohorts and observed that the sample representation captures information on single gene and global gene expression patterns. Moreover, we found that the gene representation space was organized such that tissue-specific genes, highly correlated genes as well as genes participating in the same GO terms were grouped. Finally, we compared the vector representation of samples learned by the FE model to other similar models on 49 regression tasks. We report that the representations trained with FE rank first or second in all of the tasks, surpassing, sometimes by a considerable margin, other representations. AVAILABILITY AND IMPLEMENTATION: A toy example in the form of a Jupyter Notebook as well as the code and trained embeddings for this project can be found at: https://github.com/TrofimovAssya/FactorizedEmbeddings. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Hepatic leukemia factor is a novel leukemic stem cell regulator in DNMT3A, NPM1, and FLT3-ITD triple-mutated AML.

FLT3, DNMT3A, and NPM1 are the most frequently mutated genes in cytogenetically normal acute myeloid leukemia (AML), but little is known about how these mutations synergize upon cooccurrence. Here we show that triple-mutated AML is characterized by high leukemia stem cell (LSC) frequency, an aberrant leukemia-specific GPR56 highCD34low immunophenotype, and synergistic upregulation of Hepatic Leukemia Factor (HLF). Cell sorting based on the LSC marker GPR56 allowed isolation of triple-mutated from DNMT3A/NPM1 double-mutated subclones. Moreover, in DNMT3A R882-mutated patients, CpG hypomethylation at the HLF transcription start site correlated with high HLF mRNA expression, which was itself associated with poor survival. Loss of HLF via CRISPR/Cas9 significantly reduced the CD34+GPR56+ LSC compartment of primary human triple-mutated AML cells in serial xenotransplantation assays. HLF knockout cells were more actively cycling when freshly harvested from mice, but rapidly exhausted when reintroduced in culture. RNA sequencing of primary human triple-mutated AML cells after shRNA-mediated HLF knockdown revealed the NOTCH target Hairy and Enhancer of Split 1 (HES1) and the cyclin-dependent kinase inhibitor CDKN1C/p57 as novel targets of HLF, potentially mediating these effects. Overall, our data establish HLF as a novel LSC regulator in this genetically defined high-risk AML subgroup.

Legal and Ethical Considerations for the Design and Use of Web Portals for Researchers, Clinicians, and Patients: Scoping Literature Review.

BACKGROUND: This study aims to identify a novel potential use for web portals in health care and health research: their adoption for the purposes of rapidly sharing health research findings with clinicians, scientists, and patients. In the era of precision medicine and learning health systems, the translation of research findings into targeted therapies depends on the availability of big data and emerging research results. Web portals may work to promote the availability of novel research, working in tandem with traditional scientific publications and conference proceedings. OBJECTIVE: This study aims to assess the potential use of web portals, which facilitate the sharing of health research findings among researchers, clinicians, patients, and the public. It also summarizes the potential legal, ethical, and policy implications associated with such tools for public use and in the management of patient care for complex diseases. METHODS: This study broadly adopts the methods for scoping literature reviews outlined by Arskey and O'Malley in 2005. Raised by the integration of web portals into patient care for complex diseases, we systematically searched 3 databases, PubMed, Scopus, and WestLaw Next, for sources describing web portals for sharing health research findings among clinicians, researchers, and patients and their associated legal, ethical, and policy challenges. Of the 719 candidate source citations, 22 were retained for the review. RESULTS: We found varied and inconsistent treatment of web portals for sharing health research findings among clinicians, researchers, and patients. Although the literature supports the view that portals of this kind are potentially highly promising, they remain novel and are not yet widely adopted. We also found a wide range of discussions on the legal, ethical, and policy issues related to the use of web portals to share research data. CONCLUSIONS: We identified 5 important legal and ethical challenges: privacy and confidentiality, patient health literacy, equity, training, and decision-making. We contend that each of these has meaningful implications for the increased integration of web portals into clinical care.

MAPDP: A Cloud-Based Computational Platform for Immunopeptidomics Analyses.

The immunopeptidome corresponds to the repertoire of peptides presented at the cell surface by the major histocompatibility complex (MHC) molecules. Cytotoxic T cells scan this repertoire to identify nonself antigens that can arise from tumors or infected cells. The identification of actionable antigenic targets is key to the development of therapeutic cancer vaccines, T-cell therapy, and other T-cell receptor-based biologics. The growing clinical interest for immunopeptidomics has accelerated the development of high throughput proteogenomic platforms that provide a system-level analysis of MHC-associated peptides. Improvement in sensitivity and throughput of mass spectrometers now allows the detection of a few thousands of peptides from less than 100 million cells. To manage the amount of data generated by these instruments, we have developed the MHC-associated peptide discovery platform (MAPDP), a novel open-source cloud-based computational platform for immunopeptidomic analyses. It provides convenient access from a web portal to immunopeptidomes stored in the database, filtering tools, various visualizations, annotations (e.g., IEDB, dbSNP, gnomAD), peptide-binding affinity prediction (mhcflurry, NetMHC), HLA genotyping, and the generation of personalized proteome databases. MAPDP functionalities are demonstrated here by the discovery of MHC peptides featuring new genetic variants identified in two previously published ovarian carcinoma data sets.

Enhancing the drug discovery process: Bayesian inference for the analysis and comparison of dose-response experiments.

MOTIVATION: The efficacy of a chemical compound is often tested through dose-response experiments from which efficacy metrics, such as the IC50, can be derived. The Marquardt-Levenberg algorithm (non-linear regression) is commonly used to compute estimations for these metrics. The analysis are however limited and can lead to biased conclusions. The approach does not evaluate the certainty (or uncertainty) of the estimates nor does it allow for the statistical comparison of two datasets. To compensate for these shortcomings, intuition plays an important role in the interpretation of results and the formulations of conclusions. We here propose a Bayesian inference methodology for the analysis and comparison of dose-response experiments. RESULTS: Our results well demonstrate the informativeness gain of our Bayesian approach in comparison to the commonly used Marquardt-Levenberg algorithm. It is capable to characterize the noise of dataset while inferring probable values distributions for the efficacy metrics. It can also evaluate the difference between the metrics of two datasets and compute the probability that one value is greater than the other. The conclusions that can be drawn from such analyzes are more precise. AVAILABILITY AND IMPLEMENTATION: We implemented a simple web interface that allows the users to analyze a single dose-response dataset, as well as to statistically compare the metrics of two datasets.

A key role for EZH2 and associated genes in mouse and human adult T-cell acute leukemia.

In this study, we show the high frequency of spontaneous γδ T-cell leukemia (T-ALL) occurrence in mice with biallelic deletion of enhancer of zeste homolog 2 (Ezh2). Tumor cells show little residual H3K27 trimethylation marks compared with controls. EZH2 is a component of the PRC2 Polycomb group protein complex, which is associated with DNA methyltransferases. Using next-generation sequencing, we identify alteration in gene expression levels of EZH2 and acquired mutations in PRC2-associated genes (DNMT3A and JARID2) in human adult T-ALL. Together, these studies document that deregulation of EZH2 and associated genes leads to the development of mouse, and likely human, T-ALL.

MiSTIC, an integrated platform for the analysis of heterogeneity in large tumour transcriptome datasets.

Genome-wide transcriptome profiling has enabled non-supervised classification of tumours, revealing different sub-groups characterized by specific gene expression features. However, the biological significance of these subtypes remains for the most part unclear. We describe herein an interactive platform, Minimum Spanning Trees Inferred Clustering (MiSTIC), that integrates the direct visualization and comparison of the gene correlation structure between datasets, the analysis of the molecular causes underlying co-variations in gene expression in cancer samples, and the clinical annotation of tumour sets defined by the combined expression of selected biomarkers. We have used MiSTIC to highlight the roles of specific transcription factors in breast cancer subtype specification, to compare the aspects of tumour heterogeneity targeted by different prognostic signatures, and to highlight biomarker interactions in AML. A version of MiSTIC preloaded with datasets described herein can be accessed through a public web server (http://mistic.iric.ca); in addition, the MiSTIC software package can be obtained (github.com/iric-soft/MiSTIC) for local use with personalized datasets.

Chemo-genomic interrogation of CEBPA mutated AML reveals recurrent CSF3R mutations and subgroup sensitivity to JAK inhibitors.

In this study, we analyzed RNA-sequencing data of 14 samples characterized by biallelic CEBPA (CEBPA(bi)) mutations included in the Leucegene collection of 415 primary acute myeloid leukemia (AML) specimens, and describe for the first time high frequency recurrent mutations in the granulocyte colony-stimulating factor receptor gene CSF3R, which signals through JAK-STAT proteins. Chemical interrogation of these primary human specimens revealed a uniform and specific sensitivity to all JAK inhibitors tested irrespective of their CSF3R mutation status, indicating a general sensitization of JAK-STAT signaling in this leukemia subset. Altogether, these results identified the co-occurrence of mutations in CSF3R and CEBPA in a well-defined AML subset, which uniformly responds to JAK inhibitors and paves the way to personalized clinical trials for this disease.

GPR56 identifies primary human acute myeloid leukemia cells with high repopulating potential in vivo.

Acute myeloid leukemia (AML) is a genetically heterogeneous hematologic malignancy, which is initiated and driven by a rare fraction of leukemia stem cells (LSCs). Despite the difficulties of identifying a common LSC phenotype, there is increasing evidence that high expression of stem cell gene signatures is associated with poor clinical outcome. Identification of functionally distinct subpopulations in this disease is therefore crucial to dissecting the molecular machinery underlying LSC self-renewal. Here, we combined next-generation sequencing technology with in vivo assessment of LSC frequencies and identified the adhesion G protein-coupled receptor 56 (GPR56) as a novel and stable marker for human LSCs for the majority of AML samples. High GPR56 expression was significantly associated with high-risk genetic subgroups and poor outcome. Analysis of GPR56 in combination with CD34 expression revealed engraftment potential of GPR56(+)cells in both the CD34(-)and CD34(+)fractions, thus defining a novel LSC compartment independent of the CD34(+)CD38(-)LSC phenotype.

The phagosomal proteome in interferon-gamma-activated macrophages.

The ability of macrophages to clear pathogens and elicit a sustained immune response is regulated by various cytokines, including interferon-gamma (IFN-gamma). To investigate the molecular mechanisms by which IFN-gamma modulates phagosome functions, we profiled the changes in composition, abundance, and phosphorylation of phagosome proteins in resting and activated macrophages by using quantitative proteomics and bioinformatics approaches. We identified 2415 phagosome proteins together with 2975 unique phosphorylation sites with a high level of sensitivity. Using network analyses, we determined that IFN-gamma delays phagosomal acquisition of lysosomal hydrolases and peptidases for the gain of antigen presentation. Furthermore, this gain in antigen presentation is dependent on phagosomal networks of the actin cytoskeleton and vesicle-trafficking proteins, as well as Src kinases and calpain proteases. Major histocompatibility complex class I antigen-presentation assays validated the molecular participation of these networks in the enhanced capacity of IFN-gamma-activated macrophages to crosspresent exogenous antigens to CD8(+) T cells.

Identification of small molecules that support human leukemia stem cell activity ex vivo.

Leukemic stem cells (LSCs) are considered a major cause of relapse in acute myeloid leukemia (AML). Defining pathways that control LSC self-renewal is crucial for a better understanding of underlying mechanisms and for the development of targeted therapies. However, currently available culture conditions do not prevent spontaneous differentiation of LSCs, which greatly limits the feasibility of cell-based assays. To overcome these constraints we conducted a high-throughput chemical screen and identified small molecules that inhibit differentiation and support LSC activity in vitro. Similar to reports with cord blood stem cells, several of these compounds suppressed the aryl-hydrocarbon receptor (AhR) pathway, which we show to be inactive in vivo and rapidly activated ex vivo in AML cells. We also identified a compound, UM729, that collaborates with AhR suppressors in preventing AML cell differentiation. Together, these findings provide newly defined culture conditions for improved ex vivo culture of primary human AML cells.

The NMR structure of the II-III-VI three-way junction from the Neurospora VS ribozyme reveals a critical tertiary interaction and provides new insights into the global ribozyme structure.

As part of an effort to structurally characterize the complete Neurospora VS ribozyme, NMR solution structures of several subdomains have been previously determined, including the internal loops of domains I and VI, the I/V kissing-loop interaction and the III-IV-V junction. Here, we expand this work by determining the NMR structure of a 62-nucleotide RNA (J236) that encompasses the VS ribozyme II-III-VI three-way junction and its adjoining stems. In addition, we localize Mg(2+)-binding sites within this structure using Mn(2+)-induced paramagnetic relaxation enhancement. The NMR structure of the J236 RNA displays a family C topology with a compact core stabilized by continuous stacking of stems II and III, a cis WC/WC G•A base pair, two base triples and two Mg(2+) ions. Moreover, it reveals a remote tertiary interaction between the adenine bulges of stems II and VI. Additional NMR studies demonstrate that both this bulge-bulge interaction and Mg(2+) ions are critical for the stable folding of the II-III-VI junction. The NMR structure of the J236 RNA is consistent with biochemical studies on the complete VS ribozyme, but not with biophysical studies performed with a minimal II-III-VI junction that does not contain the II-VI bulge-bulge interaction. Together with previous NMR studies, our findings provide important new insights into the three-dimensional architecture of this unique ribozyme.

EVI1-rearranged acute myeloid leukemias are characterized by distinct molecular alterations.

The genetic and transcriptional signature of EVI1 (ecotropic viral integration site 1)-rearranged (EVI1-r) acute myeloid leukemias (AMLs) remains poorly defined. We performed RNA sequencing of 12 EVI1-r AMLs and compared the results with those of other AML subtypes (n = 139) and normal CD34(+) cells (n = 17). Results confirm high frequencies of RAS and other activated signaling mutations (10/12 AMLs) and identify new recurrent mutations in splicing factors (5/12 AMLs in SF3B1 and 2/12 AMLs in U2AF1), IKZF1 (3/12 AMLs), and TP53 (3/12 AMLs). Mutations in IKZF1, a gene located on chromosome 7, and monosomy 7 are mutually exclusive in this disease. Moreover IKZF1 expression is halved in monosomy 7 leukemias. EVI-r AMLs are also characterized by a unique transcriptional signature with high expression levels of MECOM, PREX2, VIP, MYCT1, and PAWR. Our results suggest that EVI1-r AMLs could be molecularly defined by specific transcriptomic anomalies and a hitherto unseen mutational pattern. Larger patient cohorts will better determine the frequency of these events.

A functional screen reveals an extensive layer of transcriptional and splicing control underlying RAS/MAPK signaling in Drosophila.

The small GTPase RAS is among the most prevalent oncogenes. The evolutionarily conserved RAF-MEK-MAPK module that lies downstream of RAS is one of the main conduits through which RAS transmits proliferative signals in normal and cancer cells. Genetic and biochemical studies conducted over the last two decades uncovered a small set of factors regulating RAS/MAPK signaling. Interestingly, most of these were found to control RAF activation, thus suggesting a central regulatory role for this event. Whether additional factors are required at this level or further downstream remains an open question. To obtain a comprehensive view of the elements functionally linked to the RAS/MAPK cascade, we used a quantitative assay in Drosophila S2 cells to conduct a genome-wide RNAi screen for factors impacting RAS-mediated MAPK activation. The screen led to the identification of 101 validated hits, including most of the previously known factors associated to this pathway. Epistasis experiments were then carried out on individual candidates to determine their position relative to core pathway components. While this revealed several new factors acting at different steps along the pathway--including a new protein complex modulating RAF activation--we found that most hits unexpectedly work downstream of MEK and specifically influence MAPK expression. These hits mainly consist of constitutive splicing factors and thereby suggest that splicing plays a specific role in establishing MAPK levels. We further characterized two representative members of this group and surprisingly found that they act by regulating mapk alternative splicing. This study provides an unprecedented assessment of the factors modulating RAS/MAPK signaling in Drosophila. In addition, it suggests that pathway output does not solely rely on classical signaling events, such as those controlling RAF activation, but also on the regulation of MAPK levels. Finally, it indicates that core splicing components can also specifically impact alternative splicing.

Differential Features of AIRE-Induced and AIRE-Independent Promiscuous Gene Expression in Thymic Epithelial Cells.

Establishment of self-tolerance in the thymus depends on promiscuous expression of tissue-restricted Ags (TRA) by thymic epithelial cells (TEC). This promiscuous gene expression (pGE) is regulated in part by the autoimmune regulator (AIRE). To evaluate the commonalities and discrepancies between AIRE-dependent and -independent pGE, we analyzed the transcriptome of the three main TEC subsets in wild-type and Aire knockout mice. We found that the impact of AIRE-dependent pGE is not limited to generation of TRA. AIRE decreases, via non-cell autonomous mechanisms, the expression of genes coding for positive regulators of cell proliferation, and it thereby reduces the number of cortical TEC. In mature medullary TEC, AIRE-driven pGE upregulates non-TRA coding genes that enhance cell-cell interactions (e.g., claudins, integrins, and selectins) and are probably of prime relevance to tolerance induction. We also found that AIRE-dependent and -independent TRA present several distinctive features. In particular, relative to AIRE-induced TRA, AIRE-independent TRA are more numerous and show greater splicing complexity. Furthermore, we report that AIRE-dependent versus -independent TRA project nonredundant representations of peripheral tissues in the thymus.

SCL, LMO1 and Notch1 reprogram thymocytes into self-renewing cells.

The molecular determinants that render specific populations of normal cells susceptible to oncogenic reprogramming into self-renewing cancer stem cells are poorly understood. Here, we exploit T-cell acute lymphoblastic leukemia (T-ALL) as a model to define the critical initiating events in this disease. First, thymocytes that are reprogrammed by the SCL and LMO1 oncogenic transcription factors into self-renewing pre-leukemic stem cells (pre-LSCs) remain non-malignant, as evidenced by their capacities to generate functional T cells. Second, we provide strong genetic evidence that SCL directly interacts with LMO1 to activate the transcription of a self-renewal program coordinated by LYL1. Moreover, LYL1 can substitute for SCL to reprogram thymocytes in concert with LMO1. In contrast, inhibition of E2A was not sufficient to substitute for SCL, indicating that thymocyte reprogramming requires transcription activation by SCL-LMO1. Third, only a specific subset of normal thymic cells, known as DN3 thymocytes, is susceptible to reprogramming. This is because physiological NOTCH1 signals are highest in DN3 cells compared to other thymocyte subsets. Consistent with this, overexpression of a ligand-independent hyperactive NOTCH1 allele in all immature thymocytes is sufficient to sensitize them to SCL-LMO1, thereby increasing the pool of self-renewing cells. Surprisingly, hyperactive NOTCH1 cannot reprogram thymocytes on its own, despite the fact that NOTCH1 is activated by gain of function mutations in more than 55% of T-ALL cases. Rather, elevating NOTCH1 triggers a parallel pathway involving Hes1 and Myc that dramatically enhances the activity of SCL-LMO1 We conclude that the acquisition of self-renewal and the genesis of pre-LSCs from thymocytes with a finite lifespan represent a critical first event in T-ALL. Finally, LYL1 and LMO1 or LMO2 are co-expressed in most human T-ALL samples, except the cortical T subtype. We therefore anticipate that the self-renewal network described here may be relevant to a majority of human T-ALL.