Hypoxia-inducible factor-1 (HIF-1) is an important transcription factor that induces adaptive responses upon low oxygen conditions in human cancers and triggers off a poor prognostic outcome of conventional treatments. In this study, we discovered for the first time that brusatol (BRU), a quassinoid extracted from Brucea Esters, has the capability to inhibit HIF-1 signaling pathway. We found that BRU concentration-dependently down-regulated HIF-1α protein levels under hypoxia or CoCl2-induced mimic hypoxia in HCT116 cells without causing significant cytotoxicity. Besides, the transactivation activity of HIF-1 was suppressed by BRU under hypoxic conditions, as well as the expression of HIF-1 target genes, including VEGF, GLUT1, HK2 and LDHA. In addition, BRU can also decrease glucose consumption under hypoxia through inhibition of HIF-1 signaling pathway. Further studies revealed that the inhibitory effect of BRU on HIF-1 signaling pathway might be attributed to promoting degradation of HIF-1α. Interestingly, intracellular reactive oxygen species (ROS) levels and mitochondrial ROS level were both decreased by BRU treatment, indicating the involvment of mitochondrial ROS regulation in the action of BRU. Taken together, these results provided clear evidence for BRU-mediated HIF-1α regulation and suggested its therapeutic potential in colon tumors.

As one of the mainstream methods for transfer learning, Correlation Alignment (CORAL) has been widely applied in the field of fault diagnosis and has achieved certain achievements. However, CORAL ignores the differences between domain expectations in the matching process, which makes it difficult to accurately measure the discrepancies between domains. To compensate for the shortcomings of the CORAL method, this paper proposes a new feature correlation matching (FCM) method, and further uses it as an objective function to propose a deep feature correlation matching network (DFCMN). The FCM method focuses on both first-order feature correlation and second-order feature correlation of the source and target domains, which can measure the discrepancies between different domains more comprehensively and accurately. With the powerful feature mapping capability of neural network, DFCMN can improve the feature similarity in different domain centers while reducing the discrepancies of feature distribution between different domains, so as to obtain more reliable shared features and improve the cross-work-conditions diagnosis accuracy. The effectiveness of the proposed method was verified under multiple transfer tasks using the public rolling bearing dataset.

No abstract

Benefitting from the rapid development of artificial intelligence, the end-to-end fault diagnosis mode based on deep learning has become one of the most potential research directions. Nevertheless, regardless of the outstanding diagnostic accuracy, this kind of diagnostic procedure still faces the problems of high-dimensional data redundancy and subjective feature extraction. To overcome the above limitations, a novel fault diagnosis scheme for rolling bearings based on symbolic aggregate approximation (SAX) and a convolutional neural network with attention mechanism is developed in this work. In the developed diagnosis procedure, the raw data are first symbolized by the SAX approach, which can effectively reduce the dimensions of the data with relatively low subjectivity according to the characteristics of the original waveform. In addition, to enhance the feature abstraction ability of the model, channel-based attention is introduced into the deep architecture. The character strings generated by the SAX are fed into the attention mechanism-enhanced deep model to implement the training procedure, which significantly improves the diagnosis efficiency. The proposed method is tested on the common bearing fault data set of Western Reserve University, and the fault classification accuracy reaches 98.4%, with good fault diagnosis performance. The effectiveness of the proposed method is verified by comparing it with existing fault classification methods.

No abstract