Biology subjects

Trautwein, C.

Publications and source records attributed to Trautwein, C..

16 recordsLinked to original sources

bioRxiv2026

Evidence for a rod-to-cone lactate shuttle in the mammalian retina

In his seminal 1920s studies, Otto Warburg found the retina to generate large amounts of lactate. However, it was unclear what retinal cells produced lactate and whether it was a metabolic waste product or used further. Here, we show that lactate produced by rod photoreceptors fuels the energy-intensive function and viability of cone photoreceptors. In an initial expression analysis, we found monocarboxylate transporter-1 (MCT1), lactate-producing lactate-dehydrogenase-A (LDHA), and pyruvate carboxykinase-1 (PCK1) localized to rod photoreceptors, while high-affinity MCT2, pyruvate-producing LDHB, and PCK2 were expressed in cones. We then exposed retina to defined media containing either glucose or lactate as caloric component, and applied specific MCT inhibitors. In glucose-containing medium, 1H-NMR metabolomics showed rod MCT1 inhibition to increase retinal lactate, suggesting rods as a major source of lactate. In lactate-only medium, functional recordings using micro-electroretinography showed decreased rod function, while cone function was maintained. In glucose-containing medium, blocking rod MCT1 abolished cone function. Long-term treatment with MCT inhibitors selectively decreased photoreceptor viability. Conversely, supplementing the defined medium with lactate preserved cone viability in the rd1 mouse model for Retinitis Pigmentosa. Together, our data suggest that lactate shuttling from rods is crucial for cone function and viability. This may explain cone degeneration seen in various retinal diseases and provides an entirely new avenue for metabolism-based treatment development. The discovery of a lactate-shuttle between two functionally similar, yet distinct types of neurons may have far-reaching implications for our understanding of the central nervous system in general.

neuroscience↗

bioRxiv2025

Methanobrevibacter smithii associates with colorectal cancer through trophic control of the cancer bacteriome

The human gut is colonized by trillions of microbes that influence the health of their human host. Whereas many bacterial species have now been linked to a variety of different diseases, the involvement of Archaea in human disease remains elusive. Here we searched for gut archaeal signatures of disease by screening 19 cross-sectional clinical studies comprising more than 1,800 individuals. We found that associations between Archaea and medical disorders are common but highly variable and are dominated by a significant increase of Methanobrevibacter smithii in colorectal cancer (CRC) patients. Metabolic modelling and in vitro co-culture identified distinct mutualistic interactions of M. smithii with CRC-causing bacteria such as Fusobacterium nucleatum, including metabolic enhancement. Metabolomics further revealed archaeal-derived compounds with tumor-modulating properties. This provides the first mechanistic link between human gut archaeome and CRC and highlights its role in modulating health in humans through trophic control of the resident bacteriome.

microbiology↗

bioRxiv2023

Beneficial effects of intermittent fasting in NASH and subsequent HCC development are executed by concerted PPAR alpha and PCK1 action in hepatocytes

The role and molecular mechanisms of intermittent fasting (IF) in non-alcoholic steatohepatitis (NASH) and its transition to hepatocellular carcinoma (HCC) are unknown. Here, we identified that an IF 5:2 regimen (two non-consecutive days of food deprivation per week), initiated in the active phase of mice, prevents/ameliorates NASH and fibrosis as well as reduces subsequent HCC development without affecting total calorie intake. The timing, length and number of fasting cycles as well as the type of NASH diet were all critical parameters determining the effectiveness of the fasting benefits. Combined proteomic, transcriptomic and metabolomic analyses identified that PPAR and glucocorticoid receptor (GR)-PCK1 act co-operatively as hepatic executors of the fasting response by promoting fatty acid catabolism and gluconeogenesis whilst suppressing anabolic lipogenesis. In line, PPAR targets and PCK1 were reduced in human NASH. Additionally, dynamic [18F]FDG-PET analysis in vivo revealed increased [18F]FDG uptake/retention and enhanced gluconeogenesis in the liver upon fasting (in accordance with PPAR and GR-PCK1 activation) when assessed by compartmental modelling. Hepatocyte-specific GR deletion only partially abrogated the hepatic fasting response. In contrast, the combined knockdown of Ppara and Pck1 in vivo abolished the beneficial outcomes of fasting against inflammation and fibrosis, confirming their causal relationship in integrating systemic signalling in hepatocytes. Notably, PPAR agonist pemafibrate recapitulated key aspects of hepatic fasting signalling at a molecular level. Therefore, IF or pharmacological mimetics of the PPAR and/or GR-PCK1 axis could be a viable intervention against NASH and subsequent liver cancer. One-Sentence SummaryIntermittent fasting protects against fatty liver disease and liver cancer through concerted PPAR and GR-PCK1 action in hepatocytes.

cancer biology↗

bioRxiv2023

Western diet increases brain metabolism and adaptive immune responses in a mouse model of amyloidosis

Diet-induced body weight gain is a growing health problem worldwide, leading to several serious systemic diseases such as diabetes. Because it is often accompanied by a low-grade metabolic inflammation that alters systemic function, dietary changes may also contribute to the progression of neurodegenerative diseases. Here we demonstrate disrupted glucose and fatty acid metabolism and a disrupted plasma metabolome in a mouse model of Alzheimers disease following a western diet using a multimodal imaging approach and NMR-based metabolomics. We did not detect glial-dependent neuroinflammation, however using flow cytometry we observed T cell recruitment in the brains of western diet-fed mice. Our study highlights the role of the brain-liver-fat-axis and the adaptive immune system in the disruption of brain homeostasis due to a Western diet.

immunology↗

bioRxiv2023

Maintained imbalance of triglycerides, apolipoproteins, energy metabolites and cytokines in long-term COVID-19 syndrome (LTCS) patients

Deep metabolomic, proteomic and immunologic phenotyping of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) patients have matched a wide diversity of clinical symptoms with potential biomarkers for coronavirus disease 2019 (COVID-19). Within here, several studies described the role of metabolites, lipoproteins and inflammation markers during infection and in recovered patients. In fact, after SARS-CoV-2 viral infection almost 20-30% of patients experience persistent symptoms even after 12 weeks of recovery which has been defined as long-term COVID-19 syndrome (LTCS). Emerging evidence revealed that a dysregulated immune system and persisting inflammation could be one of the key drivers of LTCS. However, how these small biomolecules such as metabolites, lipoprotein, cytokines and chemokines altogether govern pathophysiology is largely underexplored. Thus, a clear understanding how these parameters into an integrated fashion could predict the disease course may help to stratify LTCS patients from acute COVID-19 or recovered specimen and would help to elucidate a potential mechanistic role of these biomolecules during the disease course. Here, we report an integrated analysis of blood serum and plasma by in vitro diagnostics research NMR spectroscopy and flow cytometry-based cytokine quantification in a total of 125 individuals (healthy controls (HC; n=73), recovered (n=12), acute (n=7) and LTCS (n=33)). We identified that in LTCS patients lactate and pyruvate were significantly different from either healthy controls or acute COVID-19 patients. Further correlational analysis of cytokines and metabolites indicated that creatine, glutamine, and high-density lipoprotein (HDL) phospholipids were distributed differentially amongst patients or individuals. Of note, triglycerides and several lipoproteins (apolipoproteins Apo-A1 and A2) in LTCS patients demonstrate COVID-19-like alterations compared to HC. Interestingly, LTCS and acute COVID-19 samples were distinguished mostly by their creatinine, phenylalanine, succinate, 3-hydroxybutyrate (3-HB) and glucose concentrations, illustrating an imbalanced energy metabolism. Most of the cytokines and chemokines were present at low levels in LTCS patients compared with HC except IL-18 chemokine, which tended to be higher in LTCS patients and correlated positively with several amino acids (creatine, histidine, leucine, and valine), metabolites (lactate and 3-HB) and lipoproteins. The identification of these persisting plasma metabolites, lipoprotein and inflammation alterations will help to better stratify LTCS patients from other diseases and could help to predict ongoing severity of LTCS patients. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=127 SRC="FIGDIR/small/523998v1_ufig1.gif" ALT="Figure 1"> View larger version (50K): org.highwire.dtl.DTLVardef@ecdeacorg.highwire.dtl.DTLVardef@10f0f80org.highwire.dtl.DTLVardef@1c2ddaborg.highwire.dtl.DTLVardef@672120_HPS_FORMAT_FIGEXP M_FIG C_FIG Layman summary & significance of the researchAlmost 20-30% of individuals infected with the SARS-CoV-2 virus regardless of hospitalization status experience long-term COVID-19 syndrome (LTCS). It is devasting for millions of individuals worldwide and hardly anything is known about why some people experience these symptoms even after 3 to 12 months after the acute phase. In this, we attempted to understand whether dysregulated metabolism and inflammation could be contributing factors to the ongoing symptoms in LTCS patients. Total blood triglycerides and the Cory cycle metabolites (lactate and pyruvate) were significantly higher, lipoproteins (Apo-A1 and A2) were drastically lower in LTCS patients compared to healthy controls. Correlation analysis revealed that either age or gender are positively correlated with several metabolites (citrate, glutamate, 3-hydroxybutyrate, glucose) and lipoproteins (Apo-A1, HDL Apo-A1, LDL triglycerides) in LTCS patients. Several cytokines and chemokines were also positively correlated with metabolites and lipoproteins thus, dysregulation in metabolism and inflammation could be a potential contributory factor for LTCS symptoms.

immunology↗

bioRxiv2022

Retinal energy metabolism: Photoreceptors switch between Cori, Cahill, and mini-Krebs cycles to uncouple glycolysis from mitochondrial respiration

The retina consumes massive amounts of energy, yet its metabolism and substrate exploitation remain poorly understood. Here, we used a murine explant model to manipulate retinal energy metabolism under entirely controlled conditions and utilized 1H-NMR spectroscopy-based metabolomics, in situenzyme detection, and cell viability readouts to uncover the pathways of retinal energy production. Our experimental manipulations resulted in varying degrees of photoreceptor degeneration, while the inner retina and retinal pigment epithelium were essentially unaffected. This selective vulnerability of photoreceptors suggested very specific adaptations in their energy metabolism. Rod photoreceptors were found to rely strongly on oxidative phosphorylation, but only mildly on glycolysis. Conversely, cone photoreceptors were dependent on glycolysis but insensitive to electron transport chain decoupling. Importantly, photoreceptors appeared to uncouple glycolytic and Krebs-cycle metabolism via three different pathways: 1) the mini-Krebs-cycle, fueled by glutamine and branched-chain amino acids, generating N-acetylaspartate; 2) the alanine-generating Cahill-cycle; 3) the lactate-releasing Cori-cycle. Moreover, the metabolomic data indicated a shuttling of taurine and hypotaurine between the retinal pigment epithelium and photoreceptors, likely resulting in an additional net transfer of reducing power to photoreceptors. These findings expand our understanding of retinal physiology and pathology and shed new light on neuronal energy homeostasis and the pathogenesis of neurodegenerative diseases. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=187 SRC="FIGDIR/small/496788v3_figa1.gif" ALT="Figure 1"> View larger version (67K): org.highwire.dtl.DTLVardef@1ef3692org.highwire.dtl.DTLVardef@1e7e193org.highwire.dtl.DTLVardef@8aa6fborg.highwire.dtl.DTLVardef@d8c5bd_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOGraphical abstractC_FLOATNO Retinal photoreceptors employ both glucose and glutamate as fuels. While rod photoreceptors rely strongly on oxidative phosphorylation and the N-acetylaspartate producing mini-Krebs-cycle, cone photoreceptors rely on the lactate-producing Cori cycle and the oxidative, alanine producing Cahill cycle. C_FIG HighlightsO_LIThe retina utilizes a complex energy switchboard consisting of the Krebs cycle, mini-Krebs cycle, Cahill cycle, and Cori cycle. C_LIO_LIMini-Krebs cycle runs more efficiently than full Krebs cycle. C_LIO_LIAlanine transaminase decouples glycolysis from the Krebs cycle. C_LIO_LILactate, alanine, and N-acetylaspartate are distinctive energetic pathway signatures. C_LI

cell biology↗

bioRxiv2021

DeepMed: A unified, modular pipeline for end-to-end deep learning in computational pathology

The interpretation of digitized histopathology images has been transformed thanks to artificial intelligence (AI). End-to-end AI algorithms can infer high-level features directly from raw image data, extending the capabilities of human experts. In particular, AI can predict tumor subtypes, genetic mutations and gene expression directly from hematoxylin and eosin (H&E) stained pathology slides. However, existing end-to-end AI workflows are poorly standardized and not easily adaptable to new tasks. Here, we introduce DeepMed, a Python library for predicting any high-level attribute directly from histopathological whole slide images alone, or from images coupled with additional meta-data (https://github.com/KatherLab/deepmed). Unlike earlier computational pipelines, DeepMed is highly developer-friendly: its structure is modular and separates preprocessing, training, deployment, statistics, and visualization in such a way that any one of these processes can be altered without affecting the others. Also, DeepMed scales easily from local use on laptop computers to multi-GPU clusters in cloud computing services and therefore can be used for teaching, prototyping and for large-scale applications. Finally, DeepMed is user-friendly and allows researchers to easily test multiple hypotheses in a single dataset (via cross-validation) or in multiple datasets (via external validation). Here, we demonstrate and document DeepMeds abilities to predict molecular alterations, histopathological subtypes and molecular features from routine histopathology images, using a large benchmark dataset which we release publicly. In summary, DeepMed is a fully integrated and broadly applicable end-to-end AI pipeline for the biomedical research community.

cancer biology↗

bioRxiv2021

Swarm learning for decentralized artificial intelligence in cancer histopathology

Artificial Intelligence (AI) can extract clinically actionable information from medical image data. In cancer histopathology, AI can be used to predict the presence of molecular alterations directly from routine histopathology slides. However, training robust AI systems requires large datasets whose collection faces practical, ethical and legal obstacles. These obstacles could be overcome with swarm learning (SL) where partners jointly train AI models, while avoiding data transfer and monopolistic data governance. Here, for the first time, we demonstrate the successful use of SL in large, multicentric datasets of gigapixel histopathology images comprising over 5000 patients. We show that AI models trained using Swarm Learning can predict BRAF mutational status and microsatellite instability (MSI) directly from hematoxylin and eosin (H&E)-stained pathology slides of colorectal cancer (CRC). We trained AI models on three patient cohorts from Northern Ireland, Germany and the United States of America and validated the prediction performance in two independent datasets from the United Kingdom using SL-based AI models. Our data show that SL enables us to train AI models which outperform most locally trained models and perform on par with models which are centrally trained on the merged datasets. In addition, we show that SL-based AI models are data efficient and maintain a robust performance even if only subsets of local datasets are used for training. In the future, SL can be used to train distributed AI models for any histopathology image analysis tasks, overcoming the need for data transfer and without requiring institutions to give up control of the final AI model.

bioinformatics↗

bioRxiv2021

Classical Mathematical Models for Prediction of Response to Chemotherapy and Immunotherapy

Classical mathematical models of tumor growth have shaped our understanding of cancer and have broad practical implications for treatment scheduling and dosage. However, even the simplest textbook models have been barely validated in real world-data of human patients. In this study, we fitted a range of differential equation models to tumor volume measurements of patients undergoing chemotherapy or cancer immunotherapy for solid tumors. We used a large dataset of 1472 patients with three or more measurements per target lesion, of which 652 patients had six or more data points. We show that the early treatment response shows only moderate correlation with the final treatment response, demonstrating the need for nuanced models. We then perform a head-to-head comparison of six classical models which are widely used in the field: the Exponential, Logistic, Classic Bertalanffy, General Bertalanffy, Classic Gompertz and General Gompertz model. Several models provide a good fit to tumor volume measurements, with the Gompertz model providing the best balance between goodness of fit and number of parameters. Similarly, when fitting to early treatment data, the general Bertalanffy and Gompertz models yield the lowest mean absolute error to forecasted data, indicating that these models could potentially be effective at predicting treatment outcome. In summary, we provide a quantitative benchmark for classical textbook models and state-of-the art models of human tumor growth. We publicly release an anonymized version of our original data, providing the first benchmark set of human tumor growth data for evaluation of mathematical models. Author SummaryMathematical oncology uses quantitative models for prediction of tumor growth and treatment response. The theoretical foundation of mathematical oncology is provided by six classical mathematical models: the Exponential, Logistic, Classic Bertalanffy, General Bertalanffy, Classic Gompertz and General Gompertz model. These models have been introduced decades ago, have been used in thousands of scientific articles and are part of textbooks and curricula in mathematical oncology. However, these models have not been systematically tested in clinical data from actual patients. In this study, we have collected quantitative tumor volume measurements from thousands of patients in five large clinical trials of cancer immunotherapy. We use this dataset to systematically investigate how accurately mathematical models can describe tumor growth, showing that there are pronounced differences between models. In addition, we show that two of these models can predict tumor response to immunotherapy and chemotherapy at later time points when trained on early tumor growth dynamics. Thus, our article closes a conceptual gap in the literature and at the same time provides a simple tool to predict response to chemotherapy and immunotherapy on the level of individual patients.

cancer biology↗

bioRxiv2021

Intercellular crosstalk regulating ARRB2/RARRES1 is involved in transition from fibrosis to cancer

Progressive fibrogenesis in chronic liver injury is often associated with cancer development. Beta-arrestin-2 (ARRB2) is a regulator of the profibrotic Angiotensin II type 1 receptor (AGTR1). The role of ARRB2 in liver fibrosis and in the transition from fibrosis to cancer is not fully understood and was investigated in this study. This study demonstrates that upregulation of the retinoic acid receptor responder 1 (RARRES1) in HSC mediated by ARRB2 leads to fibrosis. This process is driven by exosomal ARRB2 transfer to HSC, major fibrosis contributors, from injured hepatocytes, which highly express ARRB2. By contrast, downregulation of RARRES1 in hepatocytes induces malignant transformation and hepatocellular carcinoma (HCC) development. Consequently, Arrb2-deficient mice show higher number and size of liver tumors than wild-type mice in a hepatocellular carcinoma model with fibrosis. The identified relationship between ARRB2 and RARRES1 was observed in at least two species, including human cells and tissues in fibrosis and HCC and has a predictive value for survival in cancer patients. This study describes the discovery of a novel molecular pathway mediating the transition from fibrosis to cancer offering potential diagnostics and therapeutics.

molecular biology↗

bioRxiv2021

Benchmarking artificial intelligence methods for end-to-end computational pathology

Artificial intelligence (AI) can extract subtle visual information from digitized histopathology slides and yield scientific insight on genotype-phenotype interactions as well as clinically actionable recommendations. Classical weakly supervised pipelines use an end-to-end approach with residual neural networks (ResNets), modern convolutional neural networks such as EfficientNet, or non-convolutional architectures such as vision transformers (ViT). In addition, multiple-instance learning (MIL) and clustering-constrained attention MIL (CLAM) are being used for pathology image analysis. However, it is unclear how these different approaches perform relative to each other. Here, we implement and systematically compare all five methods in six clinically relevant end-to-end prediction tasks using data from N=4848 patients with rigorous external validation. We show that histological tumor subtyping of renal cell carcinoma is an easy task which approaches successfully solved with an area under the receiver operating curve (AUROC) of above 0.9 without any significant differences between approaches. In contrast, we report significant performance differences for mutation prediction in colorectal, gastric and bladder cancer. Weakly supervised ResNet-and ViT-based workflows significantly outperformed other methods, in particular MIL and CLAM for mutation prediction. As a reason for this higher performance we identify the ability of ResNet and ViT to assign high prediction scores to highly informative image regions with plausible histopathological image features. We make all source codes publicly available at https://github.com/KatherLab/HIA, allowing easy application of all methods on any end-to-end problem in computational pathology.

bioinformatics↗

bioRxiv2020

Overexpression of human alpha-synuclein leads to dysregulated microbiome and metabolites with ageing in a rat model of Parkinson disease

Since Braaks hypothesis stating that sporadic Parkinsons disease follows a specific progression of the pathology from the peripheral to the central nervous system and can be monitored by detecting accumulation of the alpha-Synuclein protein. There is growing interest in understanding how the gut (commensal) microbiome can regulate alpha-Synuclein accumulation which can lead to PD. We studied a transgenic rat model overexpressing the human alpha-Synuclein and found that the protein overexpression resulted in gut alpha-Synuclein expression and aggregation in the gut neurons with advancing age. A progressive gut microbial composition alteration characterized by the reduction of Firmicutes to Bacteroidetes ratio could be detected in the young transgenic rat model and interestingly this ratio was then increased with aging. This observation was accompanied in older animals by intestinal inflammation, increase gut permeability and a robust alteration in metabolites production characterized by the increase of succinate level in the feces and serum. Manipulation of the gut bacteria by short-term antibiotics treatment revealed a complete loss of short-chain fatty acids (SCFAs) and reduction in succinate levels. Although antibiotics treatment did not change alpha-synuclein expression in the enteric nervous system of the colon, it can reduce alpha-synuclein expression in the olfactory bulb of the transgenic rats. In summary, synchronous with ageing, our data emphasize that the gut microbiome dysbiosis leads to a specific alteration of gut metabolites which are reflected in the serum and can be modulated by the environment.

neuroscience↗

bioRxiv2020

SARS-CoV-2 infection paralyzes cytotoxic and metabolic functions of immune cells

The SARS-CoV-2 virus is the causative agent of the global COVID-19 infectious disease outbreak, which can lead to acute respiratory distress syndrome (ARDS). However, it is still unclear how the virus interferes with immune cell and metabolic functions in the human body. In this study, we investigated the immune response in acute or convalescent COVID19 patients. We characterized the peripheral blood mononuclear cells (PBMCs) using flow cytometry and found that CD8+ T cells were significantly subsided in moderate COVID-19 and convalescent patients. Furthermore, characterization of CD8+ T cells suggested that patients with a mild and moderate course of the COVID-19 disease and convalescent patients have significantly diminished expression of both perforin and granzyme A in CD8+ T cells. Using 1H-NMR spectroscopy, we characterized the metabolic status of their autologous PBMCs. We found that fructose, lactate and taurine levels were elevated in infected (mild and moderate) patients compared with control and convalescent patients. Glucose, glutamate, formate and acetate levels were attenuated in COVID-19 (mild and moderate) patients. In summary, our report suggests that SARS-CoV-2 infection leads to disrupted CD8+ T cytotoxic functions and changes the overall metabolic functions of immune cells.

immunology↗

bioRxiv2019

Pan-cancer image-based detection of clinically actionable genetic alterations

Precision treatment of cancer relies on genetic alterations which are diagnosed by molecular biology assays.1 These tests can be a bottleneck in oncology workflows because of high turnaround time, tissue usage and costs.2 Here, we show that deep learning can predict point mutations, molecular tumor subtypes and immune-related gene expression signatures3,4 directly from routine histological images of tumor tissue. We developed and systematically optimized a one-stop-shop workflow and applied it to more than 4000 patients with breast5, colon and rectal6, head and neck7, lung8,9, pancreatic10, prostate11 cancer, melanoma12 and gastric13 cancer. Together, our findings show that a single deep learning algorithm can predict clinically actionable alterations from routine histology data. Our method can be implemented on mobile hardware14, potentially enabling point-of-care diagnostics for personalized cancer treatment in individual patients.

genetics↗

bioRxiv2019

DJ-1 (Park7) affects the gut microbiome, metabolites, and development of Innate Lymphoid Cells (ILCs)

The proper communication between gut and brain is pivotal for maintenance of health and dysregulation of the gut-brain axis can lead to several clinical disorders. Also, in Parkinsons disease (PD) 85% of all patients experienced constipation long before showing any signs of motor phenotypes. For differential diagnosis and when it comes to preventive treatment there is an urgent need for the identification of biomarkers indicating early disease stages long before the disease phenotype manifests. DJ-1 is a chaperon protein involved in the protection against PD and genetic mutations in this protein have been shown to cause familial PD. However, how the deficiency of DJ-1 modifies the PD risk remains incompletely understood. In the present study we provide evidence that DJ-1 is implicated in shaping the gut microbiome including their metabolite production or innate immune cells (ILCs) development. We revealed that in 4 months old mice genetic deficiency of DJ-1 leads to significantly decrease in several bacterial genera and significantly increase in two specific genera, namely Alistipes and Rikenella. DJ-1 deficient mice have a higher production of calprotectin/MCP-1 inflammatory protein - a known protein involved in colonic inflammation - and significantly higher expression of glial fibrillary acidic protein (GFAP) than control littermates. Expression of a-Synuclein, a key protein in Lewy bodies, in the colon was not significantly different between genotypes. Metabolic profiles of feces extracts analysed by H1-NMR spectroscopy showed increased short chain fatty acids (SCFAs) and decreased amino acid levels, suggesting a general switch from protein towards fibre degrading strains in DJ-1 deficient mice. We observed that Malonate - which is known to influence the immune system - has significantly higher concentrations in DJ-1 deficient mice. Moreover, DJ-1 deficient mice have high levels of the phenol derivate 3-(3-Hydroxyphenyl) propanoic acid (3-HPPA) which is a breakdown product of aromatic substrates like tyrosine, phenylalanine and polyphenols. DJ-1 deficient mice also showed significantly reduced percentage of ILCs. Thus, our data suggests that absence of DJ-1 leads to increase in gut inflammatory bacteria composition, deregulated metabolites and dysregulated innate immunity which could be a key factor in the initiation of PD disease in the gut, and potentially also in brain during disease progression.

immunology↗

bioRxiv2019

Deep learning detects virus presence in cancer histology

Oncogenic viruses like human papilloma virus (HPV) or Epstein Barr virus (EBV) are a major cause of human cancer. Viral oncogenesis has a direct impact on treatment decisions because virus-associated tumors can demand a lower intensity of chemotherapy and radiation or can be more susceptible to immune check-point inhibition. However, molecular tests for HPV and EBV are not ubiquitously available.\n\nWe hypothesized that the histopathological features of virus-driven and non-virus driven cancers are sufficiently different to be detectable by artificial intelligence (AI) through deep learning-based analysis of images from routine hematoxylin and eosin (HE) stained slides. We show that deep transfer learning can predict presence of HPV in head and neck cancer with a patient-level 3-fold cross validated area-under-the-curve (AUC) of 0.89 [0.82; 0.94]. The same workflow was used for Epstein-Barr virus (EBV) driven gastric cancer achieving a cross-validated AUC of 0.80 [0.70; 0.92] and a similar performance in external validation sets. Reverse-engineering our deep neural networks, we show that the key morphological features can be made understandable to humans.\n\nThis workflow could enable a fast and low-cost method to identify virus-induced cancer in clinical trials or clinical routine. At the same time, our approach for feature visualization allows pathologists to look into the black box of deep learning, enabling them to check the plausibility of computer-based image classification.

cancer biology↗