Mollugin ameliorates murine allergic airway inflammation by inhibiting Th2 response and M2 macrophage activation.

Mollugin, isolated from Rubia cordifolia L, is a pharmacological compound with anti-inflammatory activity. This study aimed to investigate whether mollugin protects mice against shrimp tropomyosin (ST)-induced allergic airway inflammation. Mice were sensitized with ST combined with Al(OH)3 administered intraperitoneally (i.p.) once weekly for 3 wk followed by ST challenge for 5 d. Mice were i.p.-administered daily with mollugin for 7 d. Results showed that mollugin attenuated ST-induced infiltration of eosinophils and epithelial mucus secretion in the lung tissues and suppressed lung eosinophil peroxidase activity. Additionally, mollugin lowered the Th2 cytokine, IL-4 and IL-5, production and downregulated the mRNA levels of Il-4, Il-5, Il-13, eotaxin, Ccl-17, Muc5ac, arginase-1, Ym-1, and Fizz-1 in the lung tissues. Network pharmacology was employed to predict core targets, and the molecular docking approach was used to verify the compound targets. The results of the molecular docking study of mollugin into p38 MAPK or poly(ADP-ribose) polymerase 1 (PARP1) binding sites revealed that its mechanism was possibly similar to that of SB203580 (a p38 MAPK inhibitor) or olaparib (a PARP1 inhibitor). Immunohistochemistry analysis revealed that mollugin mitigated ST-induced elevation of arginase-1 expression and macrophage levels in the lungs and bronchoalveolar lavage fluid, respectively. Furthermore, arginase-1 mRNA level and phosphorylation of p38 MAPK were inhibited in IL-4-stimulated peritoneal macrophages. In ST-stimulated mouse primary splenocytes, mollugin notably inhibited IL-4 and IL-5 production and downregulated PARP1 and PAR protein expression. According to our findings, mollugin ameliorated allergic airway inflammation by inhibiting Th2 response and macrophage polarization.

A Universal Detection Method for Adversarial Examples and Fake Images.

Deep-learning technologies have shown impressive performance on many tasks in recent years. However, there are multiple serious security risks when using deep-learning technologies. For examples, state-of-the-art deep-learning technologies are vulnerable to adversarial examples that make the model's predictions wrong due to some specific subtle perturbation, and these technologies can be abused for the tampering with and forgery of multimedia, i.e., deep forgery. In this paper, we propose a universal detection framework for adversarial examples and fake images. We observe some differences in the distribution of model outputs for normal and adversarial examples (fake images) and train the detector to learn the differences. We perform extensive experiments on the CIFAR10 and CIFAR100 datasets. Experimental results show that the proposed framework has good feasibility and effectiveness in detecting adversarial examples or fake images. Moreover, the proposed framework has good generalizability for the different datasets and model structures.