Search

With the increasing amount of image data collected from biomedical experiments there is an urgent need for smarter and more effective analysis methods. Many scientific questions require analysis of image subregions related to some specific biology. Finding such regions of interest (ROIs) at low resolution and limiting the data subjected to final quantification at high resolution can reduce computational requirements and save time. In this paper we propose a three-step pipeline: First, bounding boxes for ROIs are located at low resolution. Next, ROIs are subjected to semantic segmentation into sub-regions at mid-resolution. We also estimate the confidence of the segmented sub-regions. Finally, quantitative measurements are extracted at high resolution. We use deep learning for the first two steps in the pipeline and conformal prediction for confidence assessment. We show that limiting final quantitative analysis to sub regions with high confidence reduces noise and increases separability of observed biological effects.

Nisha Kurian,1 Taylor S Cohen,2 Lisa Öberg,3 Erica De Zan,3 Gabriel Skogberg,4 Stefan Vollmer,3 Engin Baturcam,3 Petter Svanberg,5 Britta Bonn,5 Paul D Smith,6 Outi Vaarala,3 Danen M Cunoosamy3 1Respiratory Inflammation and Autoimmune (RIA) Precision Medicine Unit, Precision Medicine, Oncology R&D, AstraZeneca, Gothenburg, Sweden; 2Microbial Sciences, Medimmune, Gaithersburg, MD, USA; 3Translational Science and Experimental Medicine, Research and Early Development, RIA, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden; 4Bioscience, Research and Early Development, RIA, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden; 5Drug Metabolism and Pharmacokinetics, Research and Early Development, RIA, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden; 6Bioscience, Oncology R&D, AstraZeneca, Cambridge, UKCorrespondence: Nisha KurianRespiratory Inflammation and Autoimmune (RIA) Precision Medicine Unit, Precision Medicine, Oncology R&D, AstraZeneca, Gothenburg, SwedenTel +46 72 197 9662Email Nisha.e.kurian@astrazeneca.comBackground: Unlike p38 mitogen-activated protein Kinases (MAPK) that has been extensively studied in the context of lung-associated pathologies in COPD, the role of the dual-specificity mitogen-activated protein kinase kinase (MEK1/2) or its downstream signaling molecule extracellular signal-regulated kinases 1/2 (ERK1/2) in COPD is poorly understood.Objectives: The aim of this study was to address whether MEK1/2 pathway activation is linked to COPD and that targeting this pathway can improve lung inflammation through decreased immune-mediated inflammatory responses without compromising bacterial clearance.Methods: Association of MEK1/2 pathway activation to COPD was investigated by immunohistochemistry using lung tissue biopsies from COPD and healthy individuals and through analysis of sputum gene expression data from COPD patients. The anti-inflammatory effect of MEK1/2 inhibition was assessed on cytokine release from lipopolysaccharide-stimulated alveolar macrophages. The effect of MEK1/2 inhibition on bacterial clearance was assessed using Staphylococcus aureus killing assays with RAW 264.7 macrophage cell line and human neutrophils.Results: We report here MEK1/2 pathway activation demonstrated by increased pERK1/2 staining in bronchial epithelium and by the presence of MEK gene activation signature in sputum samples from COPD patients. Inhibition of MEK1/2 resulted in a superior anti-inflammatory effect in human alveolar macrophages in comparison to a p38 inhibitor. Furthermore, MEK1/2 inhibition led to an increase in bacterial killing in human neutrophils and RAW 264.7 cells that was not observed with the p38 inhibitor.Conclusion: Our data demonstrate the activation of MEK1/2 pathway in COPD and highlight a dual function of MEK1/2 inhibition in improving host defense responses whilst also controlling inflammation.Keywords: exacerbation, infection, alveolar macrophage, p38 MAPK, steroid, transcriptomics, sputum

Thymocytes develop in the thymus to become a functional pool of T-cells in the periphery. To achieve this, the thymocytes go through several steps of maturation and selection within the thymus. The goal is a cell population able to recognize foreign structures, in order to help in the defense against invading pathogens/infectious agents, which at the same time tolerate self-structures and thereby avoid autoimmunity. Exosomes are nano-sized vesicles released into the extracellular space by cells. They carry components such as proteins, micro-RNA and mRNA between cells and are able to participate in inter-cellular communication. We have isolated and characterized exosomes from human thymus and we demonstrate that exosomes are abundant in thymic tissue and that they share features with other exosomes but have their own characteristic niche. Tissue-restricted antigens (of which many are under the control of the autoimmune regulator gene) are antigens expressed by highly specialized medullary thymic epithelial cells. The role of these antigens is to mirror self antigens found in the periphery, in order to deplete self-reactive thymocytes in the negative selection process. Tissue-restricted antigens were found in exosomes from human thymic tissue and also in exosomes from primary cultures of human thymic epithelial cells. Also, several known autoantigens e.g. myelin basic protein and transglutaminase 2 were found in these exosomes. When studying the effects of exosomes on thymocyte maturation, with an in vitro model, we showed that thymic exosomes stimulated the final steps of single-positive CD4+ thymocyte maturation. Finally, since Down syndrome patients share autoimmunity and autoantibodies with patients carrying mutations in the autoimmune regulator, and this gene is located on chromosome 21, we investigated the expression of the autoimmune regulator in thymus from Down syndrome patients. The expression was up regulated and also accompanied by other thymic abnormalities such as accumulation of CD11c+ cells in the medulla and an altered protein composition of thymic exosomes.