1 June 2023 in Journal of Clinical Investigation
Germline or somatic loss-of-function mutations of fumarate hydratase (FH) predispose patients to an aggressive form of renal cell carcinoma (RCC). Since other than tumor resection there is no effective therapy for metastatic FH-deficient RCC, an accurate method for early diagnosis is needed. Although MRI or CT scans are offered, they cannot differentiate FH-deficient tumors from other RCCs. Therefore, finding noninvasive plasma biomarkers suitable for rapid diagnosis, screening, and surveillance would improve clinical outcomes. Taking advantage of the robust metabolic rewiring that occurs in FH-deficient cells, we performed plasma metabolomics analysis and identified 2 tumor-derived metabolites, succinyl-adenosine and succinic-cysteine, as excellent plasma biomarkers for early diagnosis. These 2 molecules reliably reflected the FH mutation status and tumor mass. We further identified the enzymatic cooperativity by which these biomarkers are produced within the tumor microenvironment. Longitudinal monitoring of patients demonstrated that these circulating biomarkers can be used for reporting on treatment efficacy and identifying recurrent or metastatic tumors.
12 January 2024 in Science Immunology
Defective FAS (CD95/Apo-1/TNFRSF6) signaling causes autoimmune lymphoproliferative syndrome (ALPS). Hypergammaglobulinemia is a common feature in ALPS with FAS mutations (ALPS-FAS), but paradoxically, fewer conventional memory cells differentiate from FAS-expressing germinal center (GC) B cells. Resistance to FAS-induced apoptosis does not explain this phenotype. We tested the hypothesis that defective non-apoptotic FAS signaling may contribute to impaired B cell differentiation in ALPS. We analyzed secondary lymphoid organs of patients with ALPS-FAS and found low numbers of memory B cells, fewer GC B cells, and an expanded extrafollicular (EF) B cell response. Enhanced mTOR activity has been shown to favor EF versus GC fate decision, and we found enhanced PI3K/mTOR and BCR signaling in ALPS-FAS splenic B cells. Modeling initial T-dependent B cell activation with CD40L in vitro, we showed that FAS competent cells with transient FAS ligation showed specifically decreased mTOR axis activation without apoptosis. Mechanistically, transient FAS engagement with involvement of caspase-8 induced nuclear exclusion of PTEN, leading to mTOR inhibition. In addition, FASL-dependent PTEN nuclear exclusion and mTOR modulation were defective in patients with ALPS-FAS. In the early phase of activation, FAS stimulation promoted expression of genes related to GC initiation at the expense of processes related to the EF response. Hence, our data suggest that non-apoptotic FAS signaling acts as molecular switch between EF versus GC fate decisions via regulation of the mTOR axis and transcription. The defect of this modulatory circuit may explain the observed hypergammaglobulinemia and low memory B cell numbers in ALPS.
1 July 2019 in JAMIA Open
Objectives: Identifying new relations between medical entities, such as drugs, diseases, and side effects, is typically a resource-intensive task, involving experimentation and clinical trials. The increased availability of related data and curated knowledge enables a computational approach to this task, notably by training models to predict likely relations. Such models rely on meaningful representations of the medical entities being studied. We propose a generic features vector representation that leverages co-occurrences of medical terms, linked with PubMed citations. Materials and Methods: We demonstrate the usefulness of the proposed representation by inferring two types of relations: a drug causes a side effect and a drug treats an indication. To predict these relations and assess their effectiveness, we applied 2 modeling approaches: multi-task modeling using neural networks and single-task modeling based on gradient boosting machines and logistic regression. Results: These trained models, which predict either side effects or indications, obtained significantly better results than baseline models that use a single direct co-occurrence feature. The results demonstrate the advantage of a comprehensive representation. Discussion: Selecting the appropriate representation has an immense impact on the predictive performance of machine learning models. Our proposed representation is powerful, as it spans multiple medical domains and can be used to predict a wide range of relation types. Conclusion: The discovery of new relations between various medical entities can be translated into meaningful insights, for example, related to drug development or disease understanding. Our representation of medical entities can be used to train models that predict such relations, thus accelerating healthcare-related discoveries.