Deoxyhypusine hydroxylase: A novel therapeutic target differentially expressed in short-term vs long-term survivors of glioblastoma.

Glioblastoma (GB) is the most aggressive neoplasm of the brain. Poor prognosis is mainly attributed to tumor heterogeneity, invasiveness and drug resistance. Only a small fraction of GB patients survives longer than 24 months from the time of diagnosis (ie, long-term survivors [LTS]). In our study, we aimed to identify molecular markers associated with favorable GB prognosis as a basis to develop therapeutic applications to improve patients' outcome. We have recently assembled a proteogenomic dataset of 87 GB clinical samples of varying survival rates. Following RNA-seq and mass spectrometry (MS)-based proteomics analysis, we identified several differentially expressed genes and proteins, including some known cancer-related pathways and some less established that showed higher expression in short-term (<6 months) survivors (STS) compared to LTS. One such target found was deoxyhypusine hydroxylase (DOHH), which is known to be involved in the biosynthesis of hypusine, an unusual amino acid essential for the function of the eukaryotic translation initiation factor 5A (eIF5A), which promotes tumor growth. We consequently validated DOHH overexpression in STS samples by quantitative polymerase chain reaction (qPCR) and immunohistochemistry. We further showed robust inhibition of proliferation, migration and invasion of GB cells following silencing of DOHH with short hairpin RNA (shRNA) or inhibition of its activity with small molecules, ciclopirox and deferiprone. Moreover, DOHH silencing led to significant inhibition of tumor progression and prolonged survival in GB mouse models. Searching for a potential mechanism by which DOHH promotes tumor aggressiveness, we found that it supports the transition of GB cells to a more invasive phenotype via epithelial-mesenchymal transition (EMT)-related pathways.

MYC Induces Immunotherapy and IFNγ Resistance Through Downregulation of JAK2.

Immunotherapy has revolutionized the treatment of advanced melanoma. Because the pathways mediating resistance to immunotherapy are largely unknown, we conducted transcriptome profiling of preimmunotherapy tumor biopsies from patients with melanoma that received PD-1 blockade or adoptive cell therapy with tumor-infiltrating lymphocytes. We identified two melanoma-intrinsic, mutually exclusive gene programs, which were controlled by IFNγ and MYC, and the association with immunotherapy outcome. MYC-overexpressing melanoma cells exhibited lower IFNγ responsiveness, which was linked with JAK2 downregulation. Luciferase activity assays, under the control of JAK2 promoter, demonstrated reduced activity in MYC-overexpressing cells, which was partly reversible upon mutagenesis of a MYC E-box binding site in the JAK2 promoter. Moreover, silencing of MYC or its cofactor MAX with siRNA increased JAK2 expression and IFNγ responsiveness of melanomas, while concomitantly enhancing the effector functions of T cells coincubated with MYC-overexpressing cells. Thus, we propose that MYC plays a pivotal role in immunotherapy resistance through downregulation of JAK2.

Adaptive sequencing using nanopores and deep learning of mitochondrial DNA.

Nanopore sequencing is an emerging technology that reads DNA by utilizing a unique method of detecting nucleic acid sequences and identifies the various chemical modifications they carry. Deep learning has increased in popularity as a useful technique to solve many complex computational tasks. 'Adaptive sequencing' is an implementation of selective sequencing, intended for use on the nanopore sequencing platform. In this study, we demonstrated an alternative method of software-based selective sequencing that is performed in real time by combining nanopore sequencing and deep learning. Our results showed the feasibility of using deep learning for classifying signals from only the first 200 nucleotides in a raw nanopore sequencing signal format. This was further demonstrated by comparing the accuracy of our deep learning classification model across data from several human cell lines and other eukaryotic organisms. We used custom deep learning models and a script that utilizes a 'Read Until' framework to target mitochondrial molecules in real time from a human cell line sample. This achieved a significant separation and enrichment ability of 2.3-fold. In a series of very short sequencing experiments (10, 30 and 120 min), we identified genomic and mitochondrial reads with accuracy above 90%, although mitochondrial DNA comprised only 0.1% of the total input material. The uniqueness of our method is the ability to distinguish two groups of DNA even without a labeled reference. This contrasts with studies that required a well-defined reference, whether of a DNA sequence or of another type of representation. Additionally, our method showed higher correlation to the theoretically possible enrichment factor, compared with other published methods. We believe that our results will lay the foundation for rapid and selective sequencing using nanopore technology and will pave the approach for clinical applications that use nanopore sequencing data.

LY6S, a New IFN-Inducible Human Member of the Ly6a Subfamily Expressed by Spleen Cells and Associated with Inflammation and Viral Resistance.

Syntenic genomic loci on human chromosome 8 and mouse chromosome 15 (mChr15) code for LY6/Ly6 (lymphocyte Ag 6) family proteins. The 23 murine Ly6 family genes include eight genes that are flanked by the murine Ly6e and Ly6l genes and form an Ly6 subgroup referred to in this article as the Ly6a subfamily gene cluster. Ly6a, also known as Stem Cell Ag-1 and T cell-activating protein, is a member of the Ly6a subfamily gene cluster. No LY6 genes have been annotated within the syntenic LY6E to LY6L human locus. We report in this article on LY6S, a solitary human LY6 gene that is syntenic with the murine Ly6a subfamily gene cluster, and with which it shares a common ancestry. LY6S codes for the IFN-inducible GPI-linked LY6S-iso1 protein that contains only 9 of the 10 consensus LY6 cysteine residues and is most highly expressed in a nonclassical spleen cell population. Its expression leads to distinct shifts in patterns of gene expression, particularly of genes coding for inflammatory and immune response proteins, and LY6S-iso1-expressing cells show increased resistance to viral infection. Our findings reveal the presence of a previously unannotated human IFN-stimulated gene, LY6S, which has a 1:8 ortholog relationship with the genes of the Ly6a subfamily gene cluster, is most highly expressed in spleen cells of a nonclassical cell lineage, and whose expression induces viral resistance and is associated with an inflammatory phenotype and with the activation of genes that regulate immune responses.

Proteogenomics of glioblastoma associates molecular patterns with survival.

Glioblastoma (GBM) is the most aggressive form of glioma, with poor prognosis exhibited by most patients, and a median survival time of less than 2 years. We assemble a cohort of 87 GBM patients whose survival ranges from less than 3 months and up to 10 years and perform both high-resolution mass spectrometry proteomics and RNA sequencing (RNA-seq). Integrative analysis of protein expression, RNA expression, and patient clinical information enables us to identify specific immune, metabolic, and developmental processes associated with survival as well as determine whether they are shared between expression layers or are layer specific. Our analyses reveal a stronger association between proteomic profiles and survival and identify unique protein-based classification, distinct from the established RNA-based classification. By integrating published single-cell RNA-seq data, we find a connection between subpopulations of GBM tumors and survival. Overall, our findings establish proteomic heterogeneity in GBM as a gateway to understanding poor survival.

Deep Learning Applied on Next Generation Sequencing Data Analysis.

Deep learning is defined as the group of computational techniques allowing for the discovery of latent information within large amounts of data. Recently, many fields have seen the immense potential of deep learning to solve various tasks in ways which outperformed many other traditional methods. Genomic research could be the next frontier to take advantage of deep learning, as it has the perfect combination of vast amounts of data and diverse tasks. Here we present the platform we generated to combine deep learning and genomic sequencing data. We tested the platform on publicly available sequencing data from the gut microbiome of cancer patients. We showed that our platform is capable of classifying patients with higher accuracy than other methods, with some caveats. Overall, we believe genomic research is the next frontline for deep learning as there are exciting avenues waiting to be explored. We think that our platform, presented here, could serve as the basis for such future research.

Overcoming Interpretability in Deep Learning Cancer Classification.

Since its inception, deep learning has revolutionized the field of machine learning and data-driven science. One such data-driven science to be transformed by deep learning is genomics. In the past decade, numerous genomics studies have adopted deep learning and its applications range from predicting regulatory elements to cancer classification. Despite its dominating efficacy in these applications, deep learning is not without drawbacks. A prominent shortcoming of deep learning is the lack of interpretability. Hence, the main objective of this study is to address this obstacle in the deep learning cancer classification. Here we adopt a feature importance scoring methodology (Gradient-based class activation mapping or Grad-CAM) on a quasi-recurrent neural network model that classify cancer based on FASTA sequencing data. In this study, we managed to formulate a nucleotide-to-genomic-region Grad-CAM scoring methodology, as well as, validate the use this methodology for the chosen model. Consequently, this allows for the utilization of the Grad-CAM scoring methodology for feature importance in deep learning cancer classification. The results from our study identify potential novel candidate genes, genomic elements, and mechanisms for future cancer research.

Bayesian-based noninvasive prenatal diagnosis of single-gene disorders.

In the last decade, noninvasive prenatal diagnosis (NIPD) has emerged as an effective procedure for early detection of inherited diseases during pregnancy. This technique is based on using cell-free DNA (cfDNA) and fetal cfDNA (cffDNA) in maternal blood, and hence, has minimal risk for the mother and fetus compared with invasive techniques. NIPD is currently used for identifying chromosomal abnormalities (in some instances) and for single-gene disorders (SGDs) of paternal origin. However, for SGDs of maternal origin, sensitivity poses a challenge that limits the testing to one genetic disorder at a time. Here, we present a Bayesian method for the NIPD of monogenic diseases that is independent of the mode of inheritance and parental origin. Furthermore, we show that accounting for differences in the length distribution of fetal- and maternal-derived cfDNA fragments results in increased accuracy. Our model is the first to predict inherited insertions-deletions (indels). The method described can serve as a general framework for the NIPD of SGDs; this will facilitate easy integration of further improvements. One such improvement that is presented in the current study is a machine learning model that corrects errors based on patterns found in previously processed data. Overall, we show that next-generation sequencing (NGS) can be used for the NIPD of a wide range of monogenic diseases, simultaneously. We believe that our study will lead to the achievement of a comprehensive NIPD for monogenic diseases.