Surrogate Biomarker Prediction from Whole-Slide Images for Evaluating Overall Survival in Lung Adenocarcinoma.

BACKGROUND: Recent advances in computational pathology have shown potential in predicting biomarkers from haematoxylin and eosin (H&E) whole-slide images (WSI). However, predicting the outcome directly from WSIs remains a substantial challenge. In this study, we aimed to investigate how gene expression, predicted from WSIs, could be used to evaluate overall survival (OS) in patients with lung adenocarcinoma (LUAD). METHODS: Differentially expressed genes (DEGs) were identified from The Cancer Genome Atlas (TCGA)-LUAD cohort. Cox regression analysis was performed on DEGs to identify the gene prognostics of OS. Attention-based multiple instance learning (AMIL) models were trained to predict the expression of identified prognostic genes from WSIs using the TCGA-LUAD dataset. Models were externally validated in the Clinical Proteomic Tumour Analysis Consortium (CPTAC)-LUAD dataset. The prognostic value of predicted gene expression values was then compared to the true gene expression measurements. RESULTS: The expression of 239 prognostic genes could be predicted in TCGA-LUAD with cross-validated Pearson's R > 0.4. Predicted gene expression demonstrated prognostic performance, attaining a cross-validated concordance index of up to 0.615 in TCGA-LUAD through Cox regression. In total, 36 genes had predicted expression in the external validation cohort that was prognostic of OS. CONCLUSIONS: Gene expression predicted from WSIs is an effective method of evaluating OS in patients with LUAD. These results may open up new avenues of cost- and time-efficient prognosis assessment in LUAD treatment.

Cell-cell interactome of the hematopoietic niche and its changes in acute myeloid leukemia.

The bone marrow (BM) is a complex microenvironment, coordinating the production of billions of blood cells every day. Despite its essential role and its relevance to hematopoietic diseases, this environment remains poorly characterized. Here we present a high-resolution characterization of the niche in health and acute myeloid leukemia (AML) by establishing a single-cell gene expression database of 339,381 BM cells. We found significant changes in cell type proportions and gene expression in AML, indicating that the entire niche is disrupted. We then predicted interactions between hematopoietic stem and progenitor cells (HSPCs) and other BM cell types, revealing a remarkable expansion of predicted interactions in AML that promote HSPC-cell adhesion, immunosuppression, and cytokine signaling. In particular, predicted interactions involving transforming growth factor ß1 (TGFB1) become widespread, and we show that this can drive AML cell quiescence in vitro. Our results highlight potential mechanisms of enhanced AML-HSPC competitiveness and a skewed microenvironment, fostering AML growth.

Deep Learning of Histopathological Features for the Prediction of Tumour Molecular Genetics.

Advanced diagnostics are enabling cancer treatments to become increasingly tailored to the individual through developments in immunotherapies and targeted therapies. However, long turnaround times and high costs of molecular testing hinder the widespread implementation of targeted cancer treatments. Meanwhile, gold-standard histopathological assessment carried out by a trained pathologist is widely regarded as routine and mandatory in most cancers. Recently, methods have been developed to mine hidden information from histopathological slides using deep learning applied to scanned and digitized slides; deep learning comprises a collection of computational methods which learn patterns in data in order to make predictions. Such methods have been reported to be successful in a variety of cancers for predicting the presence of biomarkers such as driver mutations, tumour mutational burden, and microsatellite instability. This information could prove valuable to pathologists and oncologists in clinical decision making for cancer treatment and triage for in-depth sequencing. In addition to identifying molecular features, deep learning has been applied to predict prognosis and treatment response in certain cancers. Despite reported successes, many challenges remain before the clinical implementation of such diagnostic strategies in the clinical setting is possible. This review aims to outline recent developments in the field of deep learning for predicting molecular genetics from histopathological slides, as well as to highlight limitations and pitfalls of working with histopathology slides in deep learning.

Molecular signatures and clinical outcomes of transplant glomerulopathy stratified by microvascular inflammation and donor-specific antibody.

BACKGROUND: We investigated clinical outcomes and molecular signatures of transplant glomerulopathy (TG) stratified by microvascular inflammation (MVI) and donor-specific antibody (DSA) status. METHODS: We performed a retrospective review of 749 kidney transplant patients who received a for-cause kidney biopsy from 2009 to 2014. We classified TG as MVI positive (MVI+) or MVI negative (MVI-), and with or without DSA. We obtained gene expression profiles for 44 biopsies by Affymetrix HuGene 1.0 ST expression arrays. RESULTS: A total of 100 patients had TG; 49 were MVI+, and 51 were MVI-. After a median post-biopsy follow-up of 2.08 years (range 0.43-4.59), Kaplan-Meier survival analysis demonstrated worse allograft survival in MVI+ TG patients compared with MVI- TG patients (P = 0.01), and time to graft failure was significantly shorter in MVI+ patients (1.08 ± 1.01 years vs 2.3 ± 1.8 years; P = 0.002). DSA status did not affect graft survival within MVI+ or MVI- groups. Analysis of pathogenesis-based transcripts (PBT) showed that MVI+ TG biopsies had increased expression of gamma interferon and rejection (GRIT) and DSA-associated transcripts (DSAST), as observed in antibody-mediated rejection. MVI- TG biopsies had increased expression of cytotoxic and regulatory T cell- and B cell-associated transcripts but not GRIT or DSAST. DSA status had no effect on expression of any PBTs studied in MVI- TG biopsies. CONCLUSIONS: Graft survival in TG is significantly worse in the presence of MVI. Gene expression profiles of MVI+ TG resemble antibody-mediated rejection while gene expression profiles of MVI- TG resemble cell-mediated rejection regardless of DSA status.

Resveratrol Improves Vascular Function and Mitochondrial Number but Not Glucose Metabolism in Older Adults.

BACKGROUND: Resveratrol, a plant-derived polyphenol, has been reported to improve glucose metabolism and vascular function and to extend life span in animal models, but studies in humans have been inconclusive. METHODS: In a randomized, double-blind crossover study, we treated older glucose-intolerant adults (n = 30) with resveratrol (2-3 g/daily) or placebo, each for 6 weeks. A standard mixed-meal test was used to assess insulin sensitivity (Matsuda index) and secretion (C-peptide deconvolution) and vascular function by reactive hyperemia peripheral arterial tonometry. Skeletal muscle samples were obtained for gene expression using RNA-Seq analysis and to assess mitochondrial morphology. RESULTS: There were no changes in glucose tolerance, insulin sensitivity, weight, blood pressure, or lipid profile following resveratrol treatment. Fasting reactive hyperemia index improved with resveratrol (2.02 ± 0.2 vs 1.76 ± 0.02, p = .002). RNA-Seq analysis yielded 140 differentially expressed transcripts (corrected p-value ≤ .05), predominantly associated with mitochondrial genes and noncoding RNA. Ingenuity Pathway Analysis confirmed that mitochondrial dysfunction (p = 2.77 × 10-12) and oxidative phosphorylation (p = 1.41 × 10-11) were the most significantly perturbed pathways. Mitochondrial number, but not size, was increased. CONCLUSIONS: Resveratrol treatment of older adults with impaired glucose regulation may have beneficial effects on vascular function, but not glucose metabolism or insulin sensitivity. Changes in gene expression suggest effects similar to those observed with caloric restriction, which has been shown to increase life and health span in animal models, although its significance for humans is uncertain. Future human studies should address the appropriate dose range and low bioavailability of resveratrol.

Deregulation of the Ras-Erk Signaling Axis Modulates the Enhancer Landscape.

Unrestrained receptor tyrosine kinase (RTK) signaling and epigenetic deregulation are root causes of tumorigenesis. We establish linkage between these processes by demonstrating that aberrant RTK signaling unleashed by oncogenic HRas(G12V) or loss of negative feedback through Sprouty gene deletion remodels histone modifications associated with active typical and super-enhancers. However, although both lesions disrupt the Ras-Erk axis, the expression programs, enhancer signatures, and transcription factor networks modulated upon HRas(G12V) transformation or Sprouty deletion are largely distinct. Oncogenic HRas(G12V) elevates histone 3 lysine 27 acetylation (H3K27ac) levels at enhancers near the transcription factor Gata4 and the kinase Prkcb, as well as their expression levels. We show that Gata4 is necessary for the aberrant gene expression and H3K27ac marking at enhancers, and Prkcb is required for the oncogenic effects of HRas(G12V)-driven cells. Taken together, our findings demonstrate that dynamic reprogramming of the cellular enhancer landscape is a major effect of oncogenic RTK signaling.

Intestinal microbiota-derived metabolomic blood plasma markers for prior radiation injury.

PURPOSE: Assessing whole-body radiation injury and absorbed dose is essential for remediation efforts following accidental or deliberate exposure in medical, industrial, military, or terrorist incidents. We hypothesize that variations in specific metabolite concentrations extracted from blood plasma would correlate with whole-body radiation injury and dose. METHODS AND MATERIALS: Groups of C57BL/6 mice (n=12 per group) were exposed to 0, 2, 4, 8, and 10.4 Gy of whole-body gamma radiation. At 24 hours after treatment, all animals were euthanized, and both plasma and liver biopsy samples were obtained, the latter being used to identify a distinct hepatic radiation injury response within plasma. A semiquantitative, untargeted metabolite/lipid profile was developed using gas chromatography-mass spectrometry and liquid chromatography-tandem mass spectrometry, which identified 354 biochemical compounds. A second set of C57BL/6 mice (n=6 per group) were used to assess a subset of identified plasma markers beyond 24 hours. RESULTS: We identified a cohort of 37 biochemical compounds in plasma that yielded the optimal separation of the irradiated sample groups, with the most correlated metabolites associated with pyrimidine (positively correlated) and tryptophan (negatively correlated) metabolism. The latter were predominantly associated with indole compounds, and there was evidence that these were also correlated between liver and plasma. No evidence of saturation as a function of dose was observed, as has been noted for studies involving metabolite analysis of urine. CONCLUSIONS: Plasma profiling of specific metabolites related to pyrimidine and tryptophan pathways can be used to differentiate whole-body radiation injury and dose response. As the tryptophan-associated indole compounds have their origin in the intestinal microbiome and subsequently the liver, these metabolites particularly represent an attractive marker for radiation injury within blood plasma.

Genomics of BK viremia in kidney transplant recipients.

BACKGROUND: This study aimed to investigate global gene expression profiles of BK viremia and nephropathy (BKVN) samples using microarrays to investigate the immunologic response to BK virus. METHODS: Patients were monitored for BK viremia in the blood monthly for 6 months, then at 9 and 12 months after kidney transplantation. BKVN and normal transplant kidney biopsy samples, and whole blood samples of patients with and without BK viremia were analyzed by Affymetrix Human Gene 1.0 ST Arrays. RESULTS: During a mean follow-up of 917±325 days, 61 of the 289 patients (21%) developed BK viremia at a median 149 (27, 1,113) days after transplantation with a median peak PCR titers of 35,900 (1,000, 2,677,000). The only significant risk factor for development of BK viremia was induction with anti-thymocyte globulin (P=0.03). Only four patients developed BKVN (1.3%). Pathogenesis-based transcript analysis revealed a significant increased expression of interferon-gamma and rejection induced (GRIT), quantitative cytotoxic T-cell (QCAT), quantitative constitutive and alternate macrophage, B-cell and natural killer cell-associated transcripts (NKAT), indicating an active inflammatory immune response in BKVN biopsies (n=3) compared to normal transplant kidney biopsies with (n=3) and without BK viremia (n=11). The whole blood gene expression profiles of 19 BK viremia patients revealed significant increased expression of GRIT, QCAT, and NKAT compared to 14 patients without viremia. CONCLUSIONS: The results showed increased activity of cytotoxic T cells and natural killer cells in BKVN and viremia samples resembling acute rejection and suggested the involvement of both adaptive and innate immunity.

An integrated genome screen identifies the Wnt signaling pathway as a major target of WT1.

WT1, a critical regulator of kidney development, is a tumor suppressor for nephroblastoma but in some contexts functions as an oncogene. A limited number of direct transcriptional targets of WT1 have been identified to explain its complex roles in tumorigenesis and organogenesis. In this study we performed genome-wide screening for direct WT1 targets, using a combination of ChIP-ChIP and expression arrays. Promoter regions bound by WT1 were highly G-rich and resembled the sites for a number of other widely expressed transcription factors such as SP1, MAZ, and ZNF219. Genes directly regulated by WT1 were implicated in MAPK signaling, axon guidance, and Wnt pathways. Among directly bound and regulated genes by WT1, nine were identified in the Wnt signaling pathway, suggesting that WT1 modulates a subset of Wnt components and responsive genes by direct binding. To prove the biological importance of the interplay between WT1 and Wnt signaling, we showed that WT1 blocked the ability of Wnt8 to induce a secondary body axis during Xenopus embryonic development. WT1 inhibited TCF-mediated transcription activated by Wnt ligand, wild type and mutant, stabilized beta-catenin by preventing TCF4 loading onto a promoter. This was neither due to direct binding of WT1 to the TCF binding site nor to interaction between WT1 and TCF4, but by competition of WT1 and TCF4 for CBP. WT1 interference with Wnt signaling represents an important mode of its action relevant to the suppression of tumor growth and guidance of development.