Background Pancreatic beta cell dysfunction and activated macrophage infiltration are early features in type 1 diabetes pathogenesis. A tricarboxylic acid cycle metabolite that can strongly activate NF-E2-related factor 2 (Nrf2) in macrophages, itaconate is important in a series of inflammatory-associated diseases via anti-inflammatory and antioxidant properties. However, its role in type 1 diabetes is unclear. We used 4-octyl itaconate (OI), the cell-permeable itaconate derivate, to explore its preventative and therapeutic effects in mouse models of type 1 diabetes and the potential mechanism of macrophage phenotype reprogramming. Methods The mouse models of streptozotocin (STZ)-induced type 1 diabetes and spontaneous autoimmune diabetes were used to evaluate the preventative and therapeutic effects of OI, which were performed by measuring blood glucose, insulin level, pro- and anti-inflammatory cytokine secretion, histopathology examination, flow cytometry, and islet proteomics. The protective effect and mechanism of OI were examined via peritoneal macrophages isolated from STZ-induced diabetic mice and co-cultured MIN6 cells with OI-pre-treated inflammatory macrophages in vitro. Moreover, the inflammatory status of peripheral blood mononuclear cells (PBMCs) from type 1 diabetes patients was evaluated after OI treatment. Results OI ameliorated glycemic deterioration, increased systemic insulin level, and improved glucose metabolism in STZ-induced diabetic mice and non-obese diabetic (NOD) mice. OI intervention significantly restored the islet insulitis and beta cell function. OI did not alter the macrophage count but significantly downregulated the proportion of M1 macrophages. Additionally, OI significantly inhibited MAPK activation in macrophages to attenuate the macrophage inflammatory response, eventually improving beta cell dysfunction in vitro. Furthermore, we detected higher IL-1β production upon lipopolysaccharide stimulation in the PBMCs from type 1 diabetes patients, which was attenuated by OI treatment. Conclusions These results provided the first evidence to date that OI can prevent the progression of glycemic deterioration, excessive inflammation, and beta cell dysfunction predominantly mediated by restricting macrophage M1 polarization in mouse models of type 1 diabetes.

As mitochondrial metabolism is a major determinant of β-cell insulin secretion, mitochondrial dysfunction underlies β-cell failure and type 2 diabetes mellitus progression. An algal polysaccharide of Laminaria japonica, sulfated fucogalactan (SFG) displays various pharmacological effects in a variety of conditions, including metabolic disease. We investigated the protective effects of SFG against hydrogen peroxide (H2O2)-induced β-cell failure in MIN6 cells and islets. SFG significantly promoted the H2O2-inhibited proliferation in the cells and ameliorated their senescence, and potentiated β-cell function by regulating β-cell identity and the insulin exocytosis-related genes and proteins in H2O2-induced β-cells. SFG also attenuated mitochondrial dysfunction, including alterations in ATP content, mitochondrial respiratory chain genes and proteins expression, and reactive oxygen species and superoxide dismutase levels. Furthermore, SFG resulted in SIRT1–PGC1-α pathway activation and upregulated the downstream Nrf2 and Tfam. Taken together, the results show that SFG attenuates H2O2-induced β-cell failure by improving mitochondrial function via SIRT1–PGC1-α signaling pathway activation. Therefore, SFG is implicated as a potential agent for treating pancreatic β-cell failure.

Federated learning is currently a popular distributed machine learning solution that often experiences cumbersome communication processes and challenging model convergence in practical edge deployments due to the training nature of its model information interactions. The paper proposes a hierarchical federated learning algorithm called FedDyn to address these challenges. FedDyn uses dynamic weighting to limit the negative effects of local model parameters with high dispersion and speed-up convergence. Additionally, an efficient aggregation-based hierarchical federated learning algorithm is proposed to improve training efficiency. The waiting time is set at the edge layer, enabling edge aggregation within a specified time, while the central server waits for the arrival of all edge aggregation models before integrating them. Dynamic grouping weighted aggregation is implemented during aggregation based on the average obsolescence of local models in various batches. The proposed algorithm is tested on the MNIST and CIFAR-10 datasets and compared with the FedAVG algorithm. The results show that FedDyn can reduce the negative effects of non-independent and identically distributed (IID) data on the model and shorten the total training time by 30% under the same accuracy rate compared to FedAVG.