Artificial intelligence (AI) is expected to support clinical judgement in medicine. We constructed a new predictive model for diabetic kidney diseases (DKD) using AI, processing natural language and longitudinal data with big data machine learning, based on the electronic medical records (EMR) of 64,059 diabetes patients. AI extracted raw features from the previous 6 months as the reference period and selected 24 factors to find time series patterns relating to 6-month DKD aggravation, using a convolutional autoencoder. AI constructed the predictive model with 3,073 features, including time series data using logistic regression analysis. AI could predict DKD aggravation with 71% accuracy. Furthermore, the group with DKD aggravation had a significantly higher incidence of hemodialysis than the non-aggravation group, over 10 years (N = 2,900). The new predictive model by AI could detect progression of DKD and may contribute to more effective and accurate intervention to reduce hemodialysis.

TAR DNA-binding protein 43 kDa (TDP-43), encoded by TARDBP, is an RNA-binding protein, the nuclear depletion of which is the histopathological hallmark of amyotrophic lateral sclerosis (ALS), a fatal neurodegenerative disorder affecting both upper and lower motor neurons. Besides motor symptoms, patients with ALS often develop nonneuronal signs including glucose intolerance, but the underlying pathomechanism is still controversial, i.e., whether it is impaired insulin secretion and/or insulin resistance. Here, we showed that ALS subjects reduced early-phase insulin secretion and that the nuclear localization of TDP-43 was lost in the islets of autopsied ALS pancreas. Loss of TDP-43 inhibited exocytosis by downregulating CaV1.2 calcium channels, thereby reducing early-phase insulin secretion in a cultured β cell line (MIN6) and β cell–specific Tardbp–knockout mice. Overexpression of CaV1.2 restored early-phase insulin secretion in Tardbp–knocked-down MIN6 cells. Our findings suggest that TDP-43 regulates cellular exocytosis mediated by L-type voltage–dependent calcium channels and, thus, plays an important role in the early phase of insulin secretion by pancreatic islets. Thus, nuclear loss of TDP-43 is implicated in not only the selective loss of motor neurons, but also in glucose intolerance due to impaired insulin secretion at an early stage of ALS.

Uptake of P_iat the cellular membrane is essential for the maintenance of cell viability. However, phosphate overload is also stressful for cells and can result in cellular damage. In the present study, we investigated the effects of the transgenic overexpression of type III P_itransporter Pit-1 to explore the role of extracellular P_iin glomerular sclerosis during chronic renal disease. Pit-1 transgenic (TG) rats showed progressive proteinuria associated with hypoalbuminemia and dyslipidemia. Ultrastructural analysis of TG rat kidney by transmission electron microscopy showed a diffuse effacement of the foot processes of podocytes and a thickening of the glomerular basement membrane, which were progressively exhibited since 8 wk after birth. TG rats died at 32 wk of age due to cachexia. At this time, more thickening of the glomerular basement membrane and segmental sclerosis were observed in glomeruli of the TG rats. Immunohistochemical examination using anti-connexin 43 and anti-desmin antibodies suggested the progressive injury of podocytes in TG rats. TG rats showed higher P_iuptake in podocytes than wild-type rats, especially under low P_iconcentration. When 8-wk-old wild-type and TG rats were fed a 0.6% normal phosphate (NP) or 1.2% phosphate (HP) diet for 12 wk, HP diet-treated TG rats showed more progressive proteinuria and higher serum creatinine levels than NP diet-treated TG rats. In conclusion, our findings suggest that overexpression of Pit-1 in rats induces phosphate-dependent podocyte injury and damage to the glomerular barrier, which result in the progression of glomerular sclerosis in the kidney.