Search

ObjectiveObesity-induced metabolic dysfunction, tissue remodelling, and chronic inflammation in visceral white adipose-tissue (WAT) are correlated with insulin resistance, type II diabetes, and metabolic disease pathogenesis [1]. In this work, we sought to establish spatio-temporal context of adipose tissue macrophage (ATM) reprogramming during obesity.MethodsWe captured single-cell RNA-sequencing, spatial transcriptomics, and histological imagining of murine WAT over the course of diet-induced obesity to study macrophage phenotype dynamics. We developed a straightforward mathematical approach to integrating multi-modal data to quantify obesityinduced changes to WAT organization. We aligned ATM phenotypes with crown-like structures (CLS) in early obesity and used spatial network analysis to uncover signalling mechanisms implicated in CLS formation.ResultsWe identified novel diversity of the lipid-associated macrophage (LAM) phenotype, whose transcriptional profile, signaling mechanisms, and spatial context serve as indicators of CLS formation in early obesity. We demonstrated that dysregulation of lipid-metabolic signalling is a critical turning point in the monocyte-LAM lineage and identified novel ligand-receptor mechanisms includingApoe, Lrp1, LplandAppthat serve as hallmarks of nascent CLS in WAT.ConclusionsMulti-modal spatio-temporal profiling demonstrates that LAMs disproportionately accumulate in CLS and are preceded by a transition-state macrophage phenotype with monocytic origins. We identified novel ligand-receptor interactions implicated in nascent CLS regions which may guide future cellular-reprogramming interventions for obesity-related sequelae.Graphical AbstractHighlights-We characterize a novel lipid-associated macrophage (LAM) phenotype along the monocyte-LAM lineage-Integrated imaging, single-cell sequencing and spatial transcriptomics data show that LAMs accumulate at nascent CLS-Analysis of spatial transcriptomics data reveals a novel set of ligands and receptors that implicate immature LAMs in shaping the CLS microenvironment in early obesity-We present a simple mathematical framework for studying dynamics of tissue-structure over time

Adversarial attacks against deep neural networks (DNNs) are continuously evolving, requiring increasingly powerful defense strategies. We develop a novel adversarial defense framework inspired by the adaptive immune system: the Robust Adversarial Immune-inspired Learning System (RAILS). Initializing a population of exemplars that is balanced across classes, RAILS starts from a uniform label distribution that encourages diversity and uses an evolutionary optimization process to adaptively adjust the predictive label distribution in a manner that emulates the way the natural immune system recognizes novel pathogens. RAILS' evolutionary optimization process explicitly captures the tradeoff between robustness (diversity) and accuracy (specificity) of the network, and represents a new immune-inspired perspective on adversarial learning. The benefits of RAILS are empirically demonstrated under eight types of adversarial attacks on a DNN adversarial image classifier for several benchmark datasets, including: MNIST; SVHN; CIFAR-10; and CIFAR-10. We find that PGD is the most damaging attack strategy and that for this attack RAILS is significantly more robust than other methods, achieving improvements in adversarial robustness by $\geq 5.62\%, 12.5\%$, $10.32\%$, and $8.39\%$, on these respective datasets, without appreciable loss of classification accuracy. Codes for the results in this paper are available at https://github.com/wangren09/RAILS., arXiv admin note: text overlap with arXiv:2012.10485