SORBET: Automated cell-neighborhood analysis of spatial transcriptomics or proteomics for interpretable sample classification via GNN.

Spatially resolved transcriptomics or proteomics data have the potential to contribute fundamental insights into the mechanisms underlying physiologic and pathological processes. However, analysis of these data capable of relating spatial information, multiplexed markers, and their observed phenotypes remains technically challenging. To analyze these relationships, we developed SORBET, a deep learning framework that leverages recent advances in graph neural networks (GNN). We apply SORBET to predict tissue phenotypes, such as response to immunotherapy, across different disease processes and different technologies including both spatial proteomics and transcriptomics methods. Our results show that SORBET accurately learns biologically meaningful relationships across distinct tissue structures and data acquisition methods. Furthermore, we demonstrate that SORBET facilitates understanding of the spatially-resolved biological mechanisms underlying the inferred phenotypes. In sum, our method facilitates mapping between the rich spatial and marker information acquired from spatial 'omics technologies to emergent biological phenotypes. Moreover, we provide novel techniques for identifying the biological processes that comprise the predicted phenotypes.

Performance of TMRM and Mitotrackers in mitochondrial morphofunctional analysis of primary human skin fibroblasts.

Mitochondrial membrane potential (Δψ) and morphology are considered key readouts of mitochondrial functional state. This morphofunction can be studied using fluorescent dyes ("probes") like tetramethylrhodamine methyl ester (TMRM) and Mitotrackers (MTs). Although these dyes are broadly used, information comparing their performance in mitochondrial morphology quantification and Δψ-sensitivity in the same cell model is still scarce. Here we applied epifluorescence microscopy of primary human skin fibroblasts to evaluate TMRM, Mitotracker Red CMXros (CMXros), Mitotracker Red CMH2Xros (CMH2Xros), Mitotracker Green FM (MG) and Mitotracker Deep Red FM (MDR). All probes were suited for automated quantification of mitochondrial morphology parameters when Δψ was normal, although they did not deliver quantitatively identical results. The mitochondrial localization of TMRM and MTs was differentially sensitive to carbonyl cyanide-4-phenylhydrazone (FCCP)-induced Δψ depolarization, decreasing in the order: TMRM ≫ CHM2Xros = CMXros = MDR > MG. To study the effect of reversible Δψ changes, the impact of photo-induced Δψ "flickering" was studied in cells co-stained with TMRM and MG. During a flickering event, individual mitochondria displayed subsequent TMRM release and uptake, whereas this phenomenon was not observed for MG. Spatiotemporal and computational analysis of the flickering event provided evidence that TMRM redistributes between adjacent mitochondria by a mechanism dependent on Δψ and TMRM concentration. In summary, this study demonstrates that: (1) TMRM and MTs are suited for automated mitochondrial morphology quantification, (2) numerical data obtained with different probes is not identical, and (3) all probes are sensitive to FCCP-induced Δψ depolarization, with TMRM and MG displaying the highest and lowest sensitivity, respectively. We conclude that TMRM is better suited for integrated analysis of Δψ and mitochondrial morphology than the tested MTs under conditions that Δψ is not substantially depolarized.

Quantitative tissue analysis and role of myeloid cells in non-small cell lung cancer.

BACKGROUND: Despite the prominent role of innate immunity in the antitumor response, little is known about the myeloid composition of human non-small cell lung cancer (NSCLC) with respect to histology and molecular subtype. We used multiplexed quantitative immunofluorescence (QIF) to measure the distribution and clinical significance of major myeloid cell subsets in large retrospective NSCLC collections. METHODS: We established a QIF panel to map major myeloid cell subsets in fixed human NSCLC including 4',6-Diamidino-2-Phenylindole for all cells, pancytokeratin for tumor-epithelial cells, CD68 for M1-like macrophages; and CD11b plus HLA-DR to interrogate mature and immature myeloid cell populations such as myeloid derived suppressor cells (MDSCs). We interrogated 793 NSCLCs represented in four tissue microarray-based cohorts: #1 (Yale, n=379) and #2 (Greece, n=230) with diverse NSCLC subtypes; #3 (Yale, n=138) with molecularly annotated lung adenocarcinomas (ADC); and #4 (Yale, n=46) with patient-matched NSCLC and morphologically-normal lung tissue. We examined associations between marker levels, myeloid cell profiles, clinicopathologic/molecular variables and survival. RESULTS: The levels of CD68+ M1 like macrophages were significantly lower and the fraction of CD11b+/HLA-DR- MDSC-like cells was prominently higher in tumor than in matched non-tumor lung tissues. HLA-DR was consistently higher in myeloid cells from tumors with elevated CD68 expression. Stromal CD11b was significantly higher in squamous cell carcinomas (SCC) than in ADC across the cohorts and EGFR-mutated lung ADCs displayed lower CD11b levels than KRAS-mutant tumors. Increased stromal CD68- and HLA-DR-expressing cells was associated with better survival in ADCs from two independent NSCLC cohorts. In SCC, increased stromal CD11b or HLA-DR expression was associated with a trend towards shorter 5-year survival. CONCLUSIONS: NSCLCs display an unfavorable myeloid immune contexture relative to non-tumor lung and exhibit distinct myeloid-cell profiles across histologies and presence of major oncogenic driver-mutations. Elevated M1-like stromal proinflammatory myeloid cells are prognostic in lung ADC, but not in SCC.

Tumor cell SYK expression modulates the tumor immune microenvironment composition in human cancer via TNF-α dependent signaling.

BACKGROUND: The expression of SYK in cancer cells has been associated with both tumor promoting and tumor suppressive effects. Despite being proposed as anticancer therapeutic target, the possible role of SYK in modulating local adaptive antitumor immune responses remains uncertain. Using detailed analysis of primary human tumors and in vitro models, we reveal the immunomodulatory effect of SYK protein in human solid cancer. METHODS: We spatially mapped SYK kinase in tumor cells, stromal cells and tumor-infiltrating leukocytes (TILs) in 808 primary non-small cell lung carcinomas (NSCLCs) from two cohorts and in 374 breast carcinomas (BCs) from two independent cohorts. We established the associations of localized SYK with clinicopathologic variables and outcomes. The immunomodulatory role of SYK on tumor cells was assessed using in vitro cytokine stimulation, transcriptomic analysis and selective SYK blockade using a small molecule inhibitor. Functional responses were assessed using cocultures of tumor cells with peripheral blood lymphocytes. T cell responses in baseline and post-treatment biopsies from patients with BC treated with a SYK inhibitor in a phase I clinical trial were also studied. RESULTS: Elevated tumor cell or leukocyte SYK expression was associated with high CD4+ and CD8+ TILs and better outcome in both NSCLC and BC. Tumor cell SYK was associated with oncogenic driver mutations in EGFR or KRAS in lung adenocarcinomas and with triple negative phenotype in BC. In cultured tumor cells, SYK was upregulated by TNFα and required for the TNFα-induced proinflammatory responses and T cell activation. SYK blockade after nivolumab in a phase I clinical trial including three patients with advanced triple negative BC reduced TILs and T cell proliferation. Our work establishes the proinflammatory function of tumor cell SYK in lung and breast cancer. SYK signaling in cultured tumor cells is required for T cell activation and SYK blockade limits adaptive antitumor immune responses and tumor rejection in patients with cancer. CONCLUSIONS: Together, our results establish the immunomodulatory role of SYK expression in human solid tumors. This information could be used to develop novel biomarkers and/or therapeutic strategies.

A Burned-Out CD8+ T-cell Subset Expands in the Tumor Microenvironment and Curbs Cancer Immunotherapy.

Specific mechanisms by which tumor-infiltrating lymphocytes (TIL) become dysfunctional remain poorly understood. Here, we employed a two-pronged approach using single-cell mass cytometry and tissue imaging technologies to dissect TILs from 25 patients with resectable and 35 patients with advanced non-small cell lung cancer (NSCLC). We identified a burned-out CD8+ TIL subset (Ebo) that specifically accumulated within the tumor microenvironment (TME) but not in adjacent nontumoral tissues. Ebo showed the highest expression of proliferation and activation markers but produced the lowest amount of IFNγ and were the most apoptotic CD8+ TIL subset. Using a humanized patient-derived tumor xenograft model, we demonstrated that Ebo expansion occurred within the TME in a PD-1/B7-H1 pathway-dependent manner. Ebo abundance in baseline tumor tissues was associated with resistance to anti-PD therapy in patients with NSCLC. Our study identifies a dysfunctional TIL subset, with distinct features from previously described exhausted T cells, and implies strategies to overcome immunotherapy resistance. SIGNIFICANCE: We identified a highly proliferative, overactivated, and apoptotic dysfunctional CD8+ tumor-infiltrating subpopulation that is functionally distinct from previously described exhausted T cells. This population is expanded in lung cancer tissues in a PD-1/B7-H1-dependent manner, and its abundance is associated with resistance to cancer immunotherapy, thus becoming a potential tissue biomarker.This article is highlighted in the In This Issue feature, p. 1601.

Host-dependent molecular factors mediating SARS-CoV-2 infection to gain clinical insights for developing effective targeted therapy.

Coronavirus disease 2019 (COVID-19), a recent viral pandemic that first began in December 2019, in Hunan wildlife market, Wuhan, China. The infection is caused by a coronavirus, SARS-CoV-2 and clinically characterized by common symptoms including fever, dry cough, loss of taste/smell, myalgia and pneumonia in severe cases. With overwhelming spikes in infection and death, its pathogenesis yet remains elusive. Since the infection spread rapidly, its healthcare demands are overwhelming with uncontrollable emergencies. Although laboratory testing and analysis are developing at an enormous pace, the high momentum of severe cases demand more rapid strategies for initial screening and patient stratification. Several molecular biomarkers like C-reactive protein, interleukin-6 (IL6), eosinophils and cytokines, and artificial intelligence (AI) based screening approaches have been developed by various studies to assist this vast medical demand. This review is an attempt to collate the outcomes of such studies, thus highlighting the utility of AI in rapid screening of molecular markers along with chest X-rays and other COVID-19 symptoms to enable faster diagnosis and patient stratification. By doing so, we also found that molecular markers such as C-reactive protein, IL-6 eosinophils, etc. showed significant differences between severe and non-severe cases of COVID-19 patients. CT findings in the lungs also showed different patterns like lung consolidation significantly higher in patients with poor recovery and lung lesions and fibrosis being higher in patients with good recovery. Thus, from these evidences we perceive that an initial rapid screening using integrated AI approach could be a way forward in efficient patient stratification.

Graph of graphs analysis for multiplexed data with application to imaging mass cytometry.

Imaging Mass Cytometry (IMC) combines laser ablation and mass spectrometry to quantitate metal-conjugated primary antibodies incubated in intact tumor tissue slides. This strategy allows spatially-resolved multiplexing of dozens of simultaneous protein targets with 1µm resolution. Each slide is a spatial assay consisting of high-dimensional multivariate observations (m-dimensional feature space) collected at different spatial positions and capturing data from a single biological sample or even representative spots from multiple samples when using tissue microarrays. Often, each of these spatial assays could be characterized by several regions of interest (ROIs). To extract meaningful information from the multi-dimensional observations recorded at different ROIs across different assays, we propose to analyze such datasets using a two-step graph-based approach. We first construct for each ROI a graph representing the interactions between the m covariates and compute an m dimensional vector characterizing the steady state distribution among features. We then use all these m-dimensional vectors to construct a graph between the ROIs from all assays. This second graph is subjected to a nonlinear dimension reduction analysis, retrieving the intrinsic geometric representation of the ROIs. Such a representation provides the foundation for efficient and accurate organization of the different ROIs that correlates with their phenotypes. Theoretically, we show that when the ROIs have a particular bi-modal distribution, the new representation gives rise to a better distinction between the two modalities compared to the maximum a posteriori (MAP) estimator. We applied our method to predict the sensitivity to PD-1 axis blockers treatment of lung cancer subjects based on IMC data, achieving 97.3% average accuracy on two IMC datasets. This serves as empirical evidence that the graph of graphs approach enables us to integrate multiple ROIs and the intra-relationships between the features at each ROI, giving rise to an informative representation that is strongly associated with the phenotypic state of the entire image.

Spatial Analysis and Clinical Significance of HLA Class-I and Class-II Subunit Expression in Non-Small Cell Lung Cancer.

PURPOSE: To analyze the distribution, associated immune contexture, and clinical significance of human leukocyte antigen (HLA) class-I and HLA class-II subunits in non-small cell lung cancer (NSCLC). EXPERIMENTAL DESIGN: Using spatially resolved and quantitative multiplexed immunofluorescence we studied the tumor/stromal tissue distribution, cancer cell-specific defects, and clinicopathologic/survival associations of ß2 microglobulin (ß2M), HLA-A, and HLA-B,-C heavy chains, as well as HLA class-II ß chain in >700 immunotherapy-naïve NSCLCs from four independent cohorts. Genomic analysis of HLA genes in NSCLC was performed using two publicly available cohorts. RESULTS: Cancer cell-specific downregulation of HLA markers was identified in 30.4% of cases. ß2M was downregulated in 9.8% (70/714), HLA-A in 9% (65/722), HLA-B,-C in 12.1% (87/719), and HLA class-II in 17.7% (127/717) of evaluable samples. Concurrent downregulation of ß2M, HLA-B,-C, and HLA class-II was commonly identified. Deleterious mutations in HLA genes were detected in <5% of lung malignancies. Tumors with cancer cell-specific ß2M downregulation displayed reduced T cells and increased natural killer (NK)-cell infiltration. Samples with cancer cell HLA-A downregulation displayed modest increase in CD8+ T cells and NK-cell infiltration. Samples with cancer cell-selective HLA-B,-C or HLA class-II downregulation displayed reduced T cells and NK-cell infiltration. There was limited association of the markers with clinicopathologic variables and KRAS/EGFR mutations. Cancer cell-selective downregulation of the HLA subunits was associated with shorter overall survival. CONCLUSIONS: Our results reveal frequent and differential defects in HLA class-I and HLA class-II protein subunit expression in immunotherapy-naïve NSCLCs associated with distinct tumor microenvironment composition and patient survival.

Alzheimer's genetic risk factor FERMT2 (Kindlin-2) controls axonal growth and synaptic plasticity in an APP-dependent manner.

Although APP metabolism is being intensively investigated, a large fraction of its modulators is yet to be characterized. In this context, we combined two genome-wide high-content screenings to assess the functional impact of miRNAs and genes on APP metabolism and the signaling pathways involved. This approach highlighted the involvement of FERMT2 (or Kindlin-2), a genetic risk factor of Alzheimer's disease (AD), as a potential key modulator of axon guidance, a neuronal process that depends on the regulation of APP metabolism. We found that FERMT2 directly interacts with APP to modulate its metabolism, and that FERMT2 underexpression impacts axonal growth, synaptic connectivity, and long-term potentiation in an APP-dependent manner. Last, the rs7143400-T allele, which is associated with an increased AD risk and localized within the 3'UTR of FERMT2, induced a downregulation of FERMT2 expression through binding of miR-4504 among others. This miRNA is mainly expressed in neurons and significantly overexpressed in AD brains compared to controls. Altogether, our data provide strong evidence for a detrimental effect of FERMT2 underexpression in neurons and insight into how this may influence AD pathogenesis.

BIN1 recovers tauopathy-induced long-term memory deficits in mice and interacts with Tau through Thr348 phosphorylation.

The bridging integrator 1 gene (BIN1) is a major genetic risk factor for Alzheimer's disease (AD). In this report, we investigated how BIN1-dependent pathophysiological processes might be associated with Tau. We first generated a cohort of control and transgenic mice either overexpressing human MAPT (TgMAPT) or both human MAPT and BIN1 (TgMAPT;TgBIN1), which we followed-up from 3 to 15 months. In TgMAPT;TgBIN1 mice short-term memory deficits appeared earlier than in TgMAPT mice; however-unlike TgMAPT mice-TgMAPT;TgBIN1 mice did not exhibit any long-term or spatial memory deficits for at least 15 months. After killing the cohort at 18 months, immunohistochemistry revealed that BIN1 overexpression prevents both Tau mislocalization and somatic inclusion in the hippocampus, where an increase in BIN1-Tau interaction was also observed. We then sought mechanisms controlling the BIN1-Tau interaction. We developed a high-content screening approach to characterize modulators of the BIN1-Tau interaction in an agnostic way (1,126 compounds targeting multiple pathways), and we identified-among others-an inhibitor of calcineurin, a Ser/Thr phosphatase. We determined that calcineurin dephosphorylates BIN1 on a cyclin-dependent kinase phosphorylation site at T348, promoting the open conformation of the neuronal BIN1 isoform. Phosphorylation of this site increases the availability of the BIN1 SH3 domain for Tau interaction, as demonstrated by nuclear magnetic resonance experiments and in primary neurons. Finally, we observed that although the levels of the neuronal BIN1 isoform were unchanged in AD brains, phospho-BIN1(T348):BIN1 ratio was increased, suggesting a compensatory mechanism. In conclusion, our data support the idea that BIN1 modulates the AD risk through an intricate regulation of its interaction with Tau. Alteration in BIN1 expression or activity may disrupt this regulatory balance with Tau and have direct effects on learning and memory.

AP2 α modulates cystic fibrosis transmembrane conductance regulator function in the human intestine.

BACKGROUND: AP2 is a clathrin-based endocytic adaptor complex comprising α, ß2, µ2 and σ2 subunits. µ2 regulates CFTR endocytosis. The α subunit interacts with CFTR in the intestine but its physiologic significance is unclear. METHODS: CFTR short circuit current was measured in intestinal T84 cells following shRNA knock down of AP2α (AP2αKD). Clathrin-coated structures (CCS) were immunolabeled and quantified in AP2αKD intestinal Caco2BBe (C2BBe) cells. GST tagged human AP2α appendage domain was cloned and its interaction with CFTR determined by GST pull down assay. RESULT: AP2αKD in T84 cells resulted in higher CFTR current (57%) compared to control, consistent with increased functional CFTR and delayed endocytosis. Depletion of AP2α reduced CCS in C2BBe cells. Pull down assays revealed an interaction between human AP2α appendage domain and CFTR. CONCLUSION: AP2 α interacts with and modulates CFTR function in the intestine by participating in clathrin assembly and recruitment of CFTR to CCS.

Genome-wide, high-content siRNA screening identifies the Alzheimer's genetic risk factor FERMT2 as a major modulator of APP metabolism.

Genome-wide association studies (GWASs) have identified 19 susceptibility loci for Alzheimer's disease (AD). However, understanding how these genes are involved in the pathophysiology of AD is one of the main challenges of the "post-GWAS" era. At least 123 genes are located within the 19 susceptibility loci; hence, a conventional approach (studying the genes one by one) would not be time- and cost-effective. We therefore developed a genome-wide, high-content siRNA screening approach and used it to assess the functional impact of gene under-expression on APP metabolism. We found that 832 genes modulated APP metabolism. Eight of these genes were located within AD susceptibility loci. Only FERMT2 (a ß3-integrin co-activator) was also significantly associated with a variation in cerebrospinal fluid Aß peptide levels in 2886 AD cases. Lastly, we showed that the under-expression of FERMT2 increases Aß peptide production by raising levels of mature APP at the cell surface and facilitating its recycling. Taken as a whole, our data suggest that FERMT2 modulates the AD risk by regulating APP metabolism and Aß peptide production.

Myelin-specific Th17 cells induce severe relapsing optic neuritis with irreversible loss of retinal ganglion cells in C57BL/6 mice.

PURPOSE: Optic neuritis affects most patients with multiple sclerosis (MS), and current treatments are unreliable. The purpose of this study was to characterize the contribution of Th1 and Th17 cells to the development of optic neuritis. METHODS: Mice were passively transferred myelin-specific Th1 or Th17 cells to induce experimental autoimmune encephalomyelitis (EAE), a model of neuroautoimmunity. Visual acuity was assessed daily with optokinetic tracking, and 1, 2, and 3 weeks post-induction, optic nerves and retinas were harvested for immunohistochemical analyses. RESULTS: Passive transfer experimental autoimmune encephalomyelitis elicits acute episodes of asymmetric visual deficits and is exacerbated in Th17-EAE relative to Th1-EAE. The Th17-EAE optic nerves contained more inflammatory infiltrates and an increased neutrophil to macrophage ratio. Significant geographic degeneration of the retinal ganglion cells accompanied Th17-EAE but not Th1. CONCLUSIONS: Th17-induced transfer EAE recapitulates pathologies observed in MS-associated optic neuritis, namely, monocular episodes of vision loss, optic nerve inflammation, and geographic retinal ganglion cell (RGC) degeneration.

Loss of Nrf2 exacerbates the visual deficits and optic neuritis elicited by experimental autoimmune encephalomyelitis.

PURPOSE: Optic neuritis, inflammation of the optic nerve, is experienced by most patients with multiple sclerosis (MS) and is typically characterized by episodes of acute, monocular vision loss. These episodes of inflammation can lead to damage or degeneration of the retinal ganglion cells (RGCs), the axons of which comprise the optic nerve. Experimental autoimmune encephalomyelitis (EAE) is a well-established model of MS in which mice are immunized to produce a neuroautoimmunity that recapitulates the cardinal hallmarks of human disease, namely, inflammation, demyelination, and neurodegeneration of the brain, spinal cord, and optic nerve. Inflammation-associated oxidative stress plays a key role in promoting spinal cord damage in EAE. However, the role of oxidative stress in optic neuritis and the associated visual deficits has not been studied. To address this gap in research, we sought to determine how a deficiency in the master antioxidant transcription factor (using nuclear factor-E2-related factor [Nrf2]-deficient mice) affects visual pathology in the EAE model. METHODS: EAE was induced in 8-week-old wild-type (WT) and Nrf2 knockout (KO) mice by immunization against the myelin oligodendrocyte glycoprotein (MOG) peptide antigen. Motor deficits were monitored daily, as was visual acuity using the established functional optokinetic tracking (OKT) assay. Mice were euthanized 21 days post-immunization for histological analyses. The optic nerves were paraffin-embedded and stained with hematoxylin and eosin (H&E) or immune cell type-specific antibodies to analyze inflammatory infiltrates. The retinas were flatmounted and stained with an RGC-specific antibody, and the RGCs were counted to assess neurodegeneration. T-helper (Th) cell-associated cytokines were measured in spleens with enzyme-linked immunosorbent assay (ELISA). Immune analyses of healthy, non-EAE mice were characterized with flow cytometry to assess the baseline immune cell profiles. RESULTS: Female Nrf2 KO mice exhibited more severe EAE-induced motor deficits compared with female WT mice. In both genders, EAE elicited more severe visual acuity deficits, inflammation of the optic nerve, and RGC degeneration in KO mice compared with their strain- and age-matched WT counterparts. Visual acuity deficits were primarily present in (and only exacerbated in) one eye of each mouse. Excess inflammatory cells within the optic nerves of the KO mice were primarily comprised of T-cells, and greater RGC degeneration in the KO mice was most prevalent in the central retina compared with the peripheral retina. Nrf2 KO spleens exhibited an increased Th1- but not Th17-associated immune response. This enhanced pathology in the KO mice was not due to global differences in immune system development between the two genotypes. CONCLUSIONS: This is the first study to report that genetic ablation of Nrf2 exacerbates visual deficits, inflammation of the optic nerve, and RGC degeneration in a murine model of MS, suggesting that Nrf2 plays a neuro- and cytoprotective role in EAE-associated optic neuritis.

Pro-oncogenic Roles of HLXB9 Protein in Insulinoma Cells through Interaction with Nono Protein and Down-regulation of the c-Met Inhibitor Cblb (Casitas B-lineage Lymphoma b).

Pancreatic islet ß-cells that lack the MEN1-encoded protein menin develop into tumors. Such tumors express the phosphorylated isoform of the ß-cell differentiation transcription factor HLXB9. It is not known how phospho-HLXB9 acts as an oncogenic factor in insulin-secreting ß-cell tumors (insulinomas). In this study we investigated the binding partners and target genes of phospho-HLXB9 in mouse insulinoma MIN6 ß-cells. Co-immunoprecipitation coupled with mass spectrometry showed a significant association of phospho-HLXB9 with the survival factor p54nrb/Nono (54-kDa nuclear RNA-binding protein, non-POU-domain-containing octamer). Endogenous phospho-HLXB9 co-localized with endogenous Nono in the nucleus. Overexpression of HLXB9 decreased the level of overexpressed Nono but not endogenous Nono. Anti-phospho-HLXB9 chromatin immunoprecipitation followed by deep sequencing (ChIP-Seq) identified the c-Met inhibitor, Cblb, as a direct phospho-HLXB9 target gene. Phospho-HLXB9 occupied the promoter of Cblb and reduced the expression of Cblb mRNA. Cblb overexpression or HLXB9 knockdown decreased c-Met protein and reduced cell migration. Also, increased phospho-HLXB9 coincided with reduced Cblb and increased c-Met in insulinomas of two mouse models of menin loss. These data provide mechanistic insights into the role of phospho-HLXB9 as a pro-oncogenic factor by interacting with a survival factor and by promoting the oncogenic c-Met pathway. These mechanisms have therapeutic implications for reducing ß-cell proliferation in insulinomas by inhibiting phospho-HLXB9 or its interaction with Nono and modulating the expression of its direct (Cblb) or indirect (c-Met) targets. Our data also implicate the use of pro-oncogenic activities of phospho-HLXB9 in ß-cell expansion strategies to alleviate ß-cell loss in diabetes.

Consequence of Menin Deficiency in Mouse Adipocytes Derived by In Vitro Differentiation.

Lipoma in patients with the multiple endocrine neoplasia type 1 (MEN1) syndrome is a type of benign fat-cell tumor that has biallelic inactivation of MEN1 that encodes menin and could serve as a model to investigate normal and pathologic fat-cell (adipocyte) proliferation and function. The role of menin and its target genes in adipocytes is not known. We used in vitro differentiation to derive matched normal and menin-deficient adipocytes from wild type (WT) and menin-null (Men1-KO) mouse embryonic stem cells (mESCs), respectively, or 3T3-L1 cells without or with menin knockdown to investigate cell size, lipid content, and gene expression changes. Adipocytes derived from Men1-KO mESCs or after menin knockdown in 3T3-L1 cells showed a 1.5-1.7-fold increase in fat-cell size. Global gene expression analysis of mESC-derived adipocytes showed that lack of menin downregulated the expression of many differentially methylated genes including the tumor suppressor long noncoding RNA Meg3 but upregulated gene expression from the prolactin gene family locus. Our results show that menin deficiency leads to fat-cell hypertrophy and provide model systems that could be used to study the regulation of fat-cell size.

GSK-3ß protein phosphorylates and stabilizes HLXB9 protein in insulinoma cells to form a targetable mechanism of controlling insulinoma cell proliferation.

Insulinomas (pancreatic islet ß cell tumors) are the most common type of functioning pancreatic neuroendocrine tumors that occur sporadically or as a part of the MEN1 syndrome that is caused by germ line mutations in MEN1. Tissue-specific tumor predisposition from germ line mutations in ubiquitously expressed genes such as MEN1 could occur because of functional consequences on tissue-specific factors. We previously reported the proapoptotic ß cell differentiation factor HLXB9 as a downstream target of menin (encoded by MEN1). Here we show that GSK-3ß inactivates the proapoptotic activity of HLXB9 by phosphorylating HLXB9 at Ser-78/Ser-80 (pHLXB9). Although HLXB9 is found in the nucleus and cytoplasm, pHLXB9 is predominantly nuclear. Both pHLXB9 and active GSK-3ß are elevated in ß cells with menin knockdown, in MEN1-associated ß cell tumors (insulinomas), and also in human sporadic insulinomas. Pharmacologic inhibition of GSK-3ß blocked cell proliferation in three different rodent insulinoma cell lines by arresting the cells in G2/M phase and caused apoptosis. Taken together, these data suggest that the combination of GSK-3ß and pHLXB9 forms a therapeutically targetable mechanism of insulinoma pathogenesis. Our results reveal that GSK-3ß and pHLXB9 can serve as novel targets for insulinoma treatment and have implications for understanding the pathways associated with ß cell proliferation.

Tissue-specific isoform switch and DNA hypomethylation of the pyruvate kinase PKM gene in human cancers.

The M2 isoform of pyruvate kinase (PKM2) plays an important role in aerobic glycolysis and is a mediator of the Warburg effect in tumors. It was previously thought that tumor cells switch expression of PKM from normal tissue-expressed PKM1 to tumor-specific PKM2 via an alternative splicing mechanism. This view was challenged by a recent report demonstrating that PKM2 is already the major PKM isoform expressed in many differentiated normal tissues. Here, through analyses on sixteen tumor types using the cancer genome atlas RNA-Seq and exon array datasets, we confirmed that isoform switch from PKM1 to PKM2 occurred in glioblastomas but not in other tumor types examined. Despite lacking of isoform switches, PKM2 expression was found to be increased in all cancer types examined, and correlated strongly to poor prognosis in head and neck cancers. We further demonstrated that elevated PKM2 expression correlated well with the hypomethylation status of intron 1 of the PKM gene in multiple cancer types, suggesting epigenetic regulation by DNA methylation as a major mechanism in controlling PKM transcription in tumors. Our study suggests that isoform switch of PKM1 to PKM2 in cancers is tissue-specific and targeting PKM2 activity in tumors remains a promising approach for clinical intervention of multiple cancer types.

The embryonic transcription factor Hlxb9 is a menin interacting partner that controls pancreatic ß-cell proliferation and the expression of insulin regulators.

The multiple endocrine neoplasia type 1 (MEN1) syndrome is caused by germline mutations in the MEN1 gene encoding menin, with tissue-specific tumors of the parathyroids, anterior pituitary, and enteropancreatic endocrine tissues. Also, 30-40% of sporadic pancreatic endocrine tumors show somatic MEN1 gene inactivation. Although menin is expressed in all cell types of the pancreas, mouse models with loss of menin in either pancreatic α-cells, or ß-cells, or total pancreas develop ß-cell-specific endocrine tumors (insulinomas). Loss of widely expressed tumor suppressor genes may produce tissue-specific tumors by reactivating one or more embryonic-specific differentiation factors. Therefore, we determined the effect of menin overexpression or knockdown on the expression of ß-cell differentiation factors in a mouse ß-cell line (MIN6). We show that the ß-cell differentiation factor Hlxb9 is posttranscriptionally upregulated upon menin knockdown, and it interacts with menin. Hlxb9 reduces cell proliferation and causes apoptosis in the presence of menin, and it regulates genes that modulate insulin level. Thus, upon menin loss or from other causes, dysregulation of Hlxb9 predicts a possible combined mechanism for ß-cell proliferation and insulin production in insulinomas. These observations help to understand how a ubiquitously expressed protein such as menin might control tissue-specific tumorigenesis. Also, our findings identify Hlxb9 as an important factor for ß-cell proliferation and insulin regulation.