Predicting and elucidating the etiology of fatty liver disease: A machine learning modeling and validation study in the IMI DIRECT cohorts.

BACKGROUND: Non-alcoholic fatty liver disease (NAFLD) is highly prevalent and causes serious health complications in individuals with and without type 2 diabetes (T2D). Early diagnosis of NAFLD is important, as this can help prevent irreversible damage to the liver and, ultimately, hepatocellular carcinomas. We sought to expand etiological understanding and develop a diagnostic tool for NAFLD using machine learning. METHODS AND FINDINGS: We utilized the baseline data from IMI DIRECT, a multicenter prospective cohort study of 3,029 European-ancestry adults recently diagnosed with T2D (n = 795) or at high risk of developing the disease (n = 2,234). Multi-omics (genetic, transcriptomic, proteomic, and metabolomic) and clinical (liver enzymes and other serological biomarkers, anthropometry, measures of beta-cell function, insulin sensitivity, and lifestyle) data comprised the key input variables. The models were trained on MRI-image-derived liver fat content (<5% or ≥5%) available for 1,514 participants. We applied LASSO (least absolute shrinkage and selection operator) to select features from the different layers of omics data and random forest analysis to develop the models. The prediction models included clinical and omics variables separately or in combination. A model including all omics and clinical variables yielded a cross-validated receiver operating characteristic area under the curve (ROCAUC) of 0.84 (95% CI 0.82, 0.86; p < 0.001), which compared with a ROCAUC of 0.82 (95% CI 0.81, 0.83; p < 0.001) for a model including 9 clinically accessible variables. The IMI DIRECT prediction models outperformed existing noninvasive NAFLD prediction tools. One limitation is that these analyses were performed in adults of European ancestry residing in northern Europe, and it is unknown how well these findings will translate to people of other ancestries and exposed to environmental risk factors that differ from those of the present cohort. Another key limitation of this study is that the prediction was done on a binary outcome of liver fat quantity (<5% or ≥5%) rather than a continuous one. CONCLUSIONS: In this study, we developed several models with different combinations of clinical and omics data and identified biological features that appear to be associated with liver fat accumulation. In general, the clinical variables showed better prediction ability than the complex omics variables. However, the combination of omics and clinical variables yielded the highest accuracy. We have incorporated the developed clinical models into a web interface (see: https://www.predictliverfat.org/) and made it available to the community. TRIAL REGISTRATION: ClinicalTrials.gov NCT03814915.

Systematic Development of Sandwich Immunoassays for the Plasma Secretome.

The plasma proteome offers a clinically useful window into human health. Recent advances from highly multiplexed assays now call for appropriate pipelines to validate individual candidates. Here, a workflow is developed to build dual binder sandwich immunoassays (SIA) and for proteins predicted to be secreted into plasma. Utilizing suspension bead arrays, ≈1800 unique antibody pairs are first screened against 209 proteins with recombinant proteins as well as EDTA plasma. Employing 624 unique antibodies, dilution-dependent curves in plasma and concentration-dependent curves of full-length proteins for 102 (49%) of the targets are obtained. For 22 protein assays, the longitudinal, interindividual, and technical performance is determined in a set of plasma samples collected from 18 healthy subjects every third month over 1 year. Finally, 14 of these assays are compared with with SIAs composed of other binders, proximity extension assays, and affinity-free targeted mass spectrometry. The workflow provides a multiplexed approach to screen for SIA pairs that suggests using at least three antibodies per target. This design is applicable for a wider range of targets of the plasma proteome, and the assays can be applied for discovery but also to validate emerging candidates derived from other platforms.

Facets of individual-specific health signatures determined from longitudinal plasma proteome profiling.

BACKGROUND: Precision medicine approaches aim to tackle diseases on an individual level through molecular profiling. Despite the growing knowledge about diseases and the reported diversity of molecular phenotypes, the descriptions of human health on an individual level have been far less elaborate. METHODS: To provide insights into the longitudinal protein signatures of well-being, we profiled blood plasma collected over one year from 101 clinically healthy individuals using multiplexed antibody assays. After applying an antibody validation scheme, we utilized > 700 protein profiles for in-depth analyses of the individuals' short-term health trajectories. FINDINGS: We found signatures of circulating proteomes to be highly individual-specific. Considering technical and longitudinal variability, we observed that 49% of the protein profiles were stable over one year. We also identified eight networks of proteins in which 11-242 proteins covaried over time. For each participant, there were unique protein profiles of which some could be explained by associations to genetic variants. INTERPRETATION: This observational and non-interventional study identifyed noticeable diversity among clinically healthy subjects, and facets of individual-specific signatures emerged by monitoring the variability of the circulating proteomes over time. To enable more personal hence precise assessments of health states, longitudinal profiling of circulating proteomes can provide a valuable component for precision medicine approaches. FUNDING: This work was supported by the Erling Persson Foundation, the Swedish Heart and Lung Foundation, the Knut and Alice Wallenberg Foundation, Science for Life Laboratory, and the Swedish Research Council.

MEK inhibitors activate Wnt signalling and induce stem cell plasticity in colorectal cancer.

In colorectal cancer (CRC), aberrant Wnt signalling is essential for tumorigenesis and maintenance of cancer stem cells. However, how other oncogenic pathways converge on Wnt signalling to modulate stem cell homeostasis in CRC currently remains poorly understood. Using large-scale compound screens in CRC, we identify MEK1/2 inhibitors as potent activators of Wnt/ß-catenin signalling. Targeting MEK increases Wnt activity in different CRC cell lines and murine intestine in vivo. Truncating mutations of APC generated by CRISPR/Cas9 strongly synergize with MEK inhibitors in enhancing Wnt responses in isogenic CRC models. Mechanistically, we demonstrate that MEK inhibition induces a rapid downregulation of AXIN1. Using patient-derived CRC organoids, we show that MEK inhibition leads to increased Wnt activity, elevated LGR5 levels and enrichment of gene signatures associated with stemness and cancer relapse. Our study demonstrates that clinically used MEK inhibitors inadvertently induce stem cell plasticity, revealing an unknown side effect of RAS pathway inhibition.

The human secretome.

The proteins secreted by human cells (collectively referred to as the secretome) are important not only for the basic understanding of human biology but also for the identification of potential targets for future diagnostics and therapies. Here, we present a comprehensive analysis of proteins predicted to be secreted in human cells, which provides information about their final localization in the human body, including the proteins actively secreted to peripheral blood. The analysis suggests that a large number of the proteins of the secretome are not secreted out of the cell, but instead are retained intracellularly, whereas another large group of proteins were identified that are predicted to be retained locally at the tissue of expression and not secreted into the blood. Proteins detected in the human blood by mass spectrometry-based proteomics and antibody-based immunoassays are also presented with estimates of their concentrations in the blood. The results are presented in an updated version 19 of the Human Protein Atlas in which each gene encoding a secretome protein is annotated to provide an open-access knowledge resource of the human secretome, including body-wide expression data, spatial localization data down to the single-cell and subcellular levels, and data about the presence of proteins that are detectable in the blood.