S-nitrosylation of AMPKγ impairs coronary collateral circulation and disrupts VSMC reprogramming.

Collateral circulation is essential for blood resupply to the ischemic heart, which is dictated by the contractile phenotypic restoration of vascular smooth muscle cells (VSMC). Here we investigate whether S-nitrosylation of AMP-activated protein kinase (AMPK), a key regulator of the VSMC phenotype, impairs collateral circulation. In rats with collateral growth and development, nitroglycerin decreases coronary collateral blood flow (CCBF), inhibits vascular contractile phenotypic restoration, and increases myocardial infarct size, accompanied by reduced AMPK activity in the collateral zone. Nitric oxide (NO) S-nitrosylates human recombinant AMPKγ1 at cysteine 131 and decreases AMP sensitivity of AMPK. In VSMCs, exogenous expression of S-nitrosylation-resistant AMPKγ1 or deficient NO synthase (iNOS) prevents the disruption of VSMC reprogramming. Finally, hyperhomocysteinemia or hyperglycemia increases AMPKγ1 S-nitrosylation, prevents vascular contractile phenotypic restoration, reduces CCBF, and increases the infarct size of the heart in Apoe-/- mice, all of which is rescued in Apoe-/-/iNOSsm-/- mice or Apoe-/- mice with enforced expression of the AMPKγ1-C130A mutant following RI/MI. We conclude that nitrosative stress disrupts coronary collateral circulation during hyperhomocysteinemia or hyperglycemia through AMPK S-nitrosylation.

Establishment of machine learning-based tool for early detection of pulmonary embolism.

BACKGROUND AND OBJECTIVES: Pulmonary embolism (PE) is a complex disease with high mortality and morbidity rate, leading to increasing society burden. However, current diagnosis is solely based on symptoms and laboratory data despite its complex pathology, which easily leads to misdiagnosis and missed diagnosis by inexperienced doctors. Especially, CT pulmonary angiography, the gold standard method, is not widely available. In this study, we aim to establish a rapid and accurate screening model for pulmonary embolism using machine learning technology. Importantly, data required for disease prediction are easily accessed, including routine laboratory data and medical record information of patients. METHODS: We extracted features from patients' routine laboratory results and medical records, including blood routine, biochemical group, blood coagulation routine and other test results, as well as symptoms and medical history information. Samples with a feature loss rate greater than 0.8 were deleted from the original database. Data from 4723 cases were retained, 231 of which were positive for pulmonary embolism. 50 features were retained through the positive and negative statistical hypothesis testing which was used to build the predictive model. In order to avoid identification as majority-class samples caused by the imbalance of sample proportion, we used the method of Synthetic Minority Oversampling Technique (SMOTE) to increase the amount of information on minority samples. Five typical machine learning algorithms were used to model the screening of pulmonary embolism, including Support Vector Machines, Logistic Regression, Random Forest, XGBoost, and Back Propagation Neural Networks. To evaluate model performance, sensitivity, specificity and AUC curve were analyzed as the main evaluation indicators. Furthermore, a baseline model was established using the characteristics of the pulmonary embolism guidelines as a comparison model. RESULTS: We found that XGBoost showed better performance compared to other models, with the highest sensitivity and specificity (0.99 and 0.99, respectively). Moreover, it showed significant improvement in performance compared to the baseline model (sensitivity and specificity were 0.76 and 0.76 respectively). More important, our model showed low missed diagnosis rate (0.46) and high AUC value (0.992). Finally, the calculation time of our model is only about 0.05 s to obtain the possibility of pulmonary embolism. CONCLUSIONS: In this study, five machine learning classification models were established to assess the likelihood of patients suffering from pulmonary embolism, and the XGBoost model most significantly improved the precision, sensitivity, and AUC for pulmonary embolism screening. Collectively, we have established an AI-based model to accurately predict pulmonary embolism at early stage.

The prognosis of patients with postoperative hyperglycemia after Stanford type A aortic dissection surgery and construction of prediction model for postoperative hyperglycemia.

Objective: The mortality of type A aortic dissection (TAAD) is extremely high. The effect of postoperative hyperglycemia (PHG) on the prognosis of TAAD surgery is unclear. This study aims to investigate the prognosis of patients with PHG after TAAD surgery and construct prediction model for PHG. Methods: Patients underwent TAAD surgery from January 2016 to December 2020 in Xiangya Hospital were collected. A total of 203 patients were included and patients were divided into non PHG group and PHG group. The occurrence of postoperative delirium, cardiac complications, spinal cord complication, cerebral complications, acute kidney injury (AKI), hepatic dysfunction, hypoxemia, and in-hospital mortality were compared between two groups. Data from MIMIC-IV database were further applied to validate the relationship between PHG and clinical outcomes. The prediction model for PHG was then constructed using Extreme Gradient Boosting (XGBoost) analysis. The predictive value of selected features was further validated using patient data from MIMIC-IV database. Finally, the 28-days survival rate of patient with PHG was analyzed using data from MIMIC-IV database. Results: There were 86 patients developed PHG. The incidences of postoperative AKI, hepatic dysfunction, and in-hospital mortality were significant higher in PHG group. The ventilation time after surgery was significant longer in PHG group. Data from MIMIC-IV database validated these results. Neutrophil, platelet, lactic acid, weight, and lymphocyte were selected as features for prediction model. The values of AUC in training and testing set were 0.8697 and 0.8286 respectively. Then, five features were applied to construct another prediction model using data from MIMIC-IV database and the value of AUC in the new model was 0.8185. Finally, 28-days survival rate of patients with PHG was significantly lower and PHG was an independent risk factor for 28-days mortality after TAAD surgery. Conclusion: PHG was significantly associated with the occurrence of AKI, hepatic dysfunction, increased ventilation time, and in-hospital mortality after TAAD surgery. The feature combination of neutrophil, platelet, lactic acid, weight, and lymphocyte could effectively predict PHG. The 28-days survival rate of patients with PHG was significantly lower. Moreover, PHG was an independent risk factor for 28-days mortality after TAAD surgery.

Analysis of infiltrated immune cells in left atriums from patients with atrial fibrillation and identification of circRNA biomarkers for postoperative atrial fibrillation.

Background: Atrial fibrillation (AF) increases the risk of stroke and heart failure. Postoperative AF (POAF) increases the risk of mortality after cardiac surgery. This study aims to explore mechanisms underlying AF, analyze infiltration of immune cells in left atrium (LA) from patients with AF, and identify potential circular RNA (circRNA) biomarkers for POAF. Methods: Raw data of GSE797689, GSE115574, and GSE97455 were downloaded and processed. AF-related gene co-expression network was constructed using weighted gene correlation network analysis and enrichment analysis of genes in relevant module was conducted. Gene set enrichment analysis (GSEA) and gene set variation analysis (GSVA) were applied to investigate pathways significantly enriched in AF group. Infiltration of immune cells was analyzed using single-sample GSEA. Differentially expressed genes (DEGs) between patients with or without AF were identified and competing endogenous RNA (ceRNA) networks of DEGs were constructed. To screen biomarkers for POAF, differentially expressed circRNAs (DEcircRNAs) between patients with or without POAF were identified. Intersection between DEcircRNAs and circRNAs in ceRNA networks of DEGs were extracted and circRNAs in the intersection were further screened using support vector machine, random forest, and neural network to identify biomarkers for POAF. Results: Three modules were found to be relevant with AF and enrichment analysis indicated that genes in these modules were enriched in synthesis of extracellular matrix and inflammatory response. The results of GSEA and GSVA suggested that inflammatory response-related pathways were significantly enriched in AF group. Immune cells like macrophages, mast cells, and neutrophils were significantly infiltrated in LA tissues from patients with AF. The expression levels of immune genes such as CHGB, HLA-DRA, LYZ, IGKV1-17 and TYROBP were significantly upregulated in patients with AF, which were correlated with infiltration of immune cells. ceRNA networks of DEGs were constructed and has_circ_0006314 and hsa_circ_0055387 were found to have potential predictive values for POAF. Conclusion: Synthesis of extracellular matrix and inflammatory response were main processes involved in development and progression of AF. Infiltration of immune cells was significantly different between patients with or without AF. Has_circ_0006314 and hsa_circ_0055387 were found to have potential predictive values for POAF.

Aortic Dissection Auxiliary Diagnosis Model and Applied Research Based on Ensemble Learning.

Aortic dissection (AD), a dangerous disease threatening to human beings, has a hidden onset and rapid progression and has few effective methods in its early diagnosis. At present, although CT angiography acts as the gold standard on AD diagnosis, it is so expensive and time-consuming that it can hardly offer practical help to patients. Meanwhile, the artificial intelligence technology may provide a cheap but effective approach to building an auxiliary diagnosis model for improving the early AD diagnosis rate by taking advantage of the data of the general conditions of AD patients, such as the data about the basic inspection information. Therefore, this study proposes to hybrid five types of machine learning operators into an integrated diagnosis model, as an auxiliary diagnostic approach, to cooperate with the AD-clinical analysis. To improve the diagnose accuracy, the participating rate of each operator in the proposed model may adjust adaptively according to the result of the data learning. After a set of experimental evaluations, the proposed model, acting as the preliminary AD-discriminant, has reached an accuracy of over 80%, which provides a promising instance for medical colleagues.

AIBP, Angiogenesis, Hematopoiesis, and Atherogenesis.

PURPOSE OF REVIEW: The goal of this manuscript is to summarize the current understanding of the secreted APOA1 binding protein (AIBP), encoded by NAXE, in angiogenesis, hematopoiesis, and inflammation. The studies on AIBP illustrate a critical connection between lipid metabolism and the aforementioned endothelial and immune cell biology. RECENT FINDINGS: AIBP dictates both developmental processes such as angiogenesis and hematopoiesis, and pathological events such as inflammation, tumorigenesis, and atherosclerosis. Although cholesterol efflux dictates AIBP-mediated lipid raft disruption in many of the cell types, recent studies document cholesterol efflux-independent mechanism involving Cdc42-mediated cytoskeleton remodeling in macrophages. AIBP disrupts lipid rafts and impairs raft-associated VEGFR2 but facilitates non-raft-associated NOTCH1 signaling. Furthermore, AIBP can induce cholesterol biosynthesis gene SREBP2 activation, which in turn transactivates NOTCH1 and supports specification of hematopoietic stem and progenitor cells (HSPCs). In addition, AIBP also binds TLR4 and represses TLR4-mediated inflammation. In this review, we summarize the latest research on AIBP, focusing on its role in cholesterol metabolism and the attendant effects on lipid raft-regulated VEGFR2 and non-raft-associated NOTCH1 activation in angiogenesis, SREBP2-upregulated NOTCH1 signaling in hematopoiesis, and TLR4 signaling in inflammation and atherogenesis. We will discuss its potential therapeutic applications in angiogenesis and inflammation due to selective targeting of activated cells.

A study of aortic dissection screening method based on multiple machine learning models.

BACKGROUND: The main purpose of the study was to develop an early screening method for aortic dissection (AD) based on machine learning. Due to the rarity of AD and the complexity of symptoms, many doctors have no clinical experience with it. Many patients are not suspected of having AD, which lead to a high rate of misdiagnosis. Here, we report the preliminary study and feasibility of rapid and accurate screening method of AD with machine learning methods. METHODS: The dataset analyzed was composed by examination data provided by the Xiangya Hospital Central South University of China which include a total of 60,000 samples, including aortic patients and non-aortic ones. Each sample has 76 features which is consist of routine examinations and other easily accessible information. Since the proportion of people who are affected is usually imbalanced compared to non-diseased people, multiple machine learning models were used, include AdaBoost, SmoteBagging, EasyEnsemble and CalibratedAdaMEC. They used different methods such as ensemble learning, undersampling, oversampling, and cost-sensitivity to solve data imbalance problems. RESULTS: AdaBoost performed poorly with an average recall of 16.1% and a specificity of 99.8%. SmoteBagging achieved a statistically significant better performance for this problem with an average recall of 78.1% and a specificity of 79.2%. EasyEnsemble reached the values of 77.8% and 79.3% for recall and specificity respectively. CalibratedAdaMEC's recall and specificity are 75.8% and 76%. CONCLUSIONS: It was found that the screening performance of the models evaluated in this paper had a misdiagnosis rate lower than 25% except AdaBoost. The data used in these methods are only routine inspection data. This means that machine learning methods can help us build a fast, cheap, worthwhile and effective early screening approach for AD.

An early aortic dissection screening model and applied research based on ensemble learning.

BACKGROUND: As a particularly dangerous and rare cardiovascular disease, aortic dissection (AD) is characterized by complex and diverse symptoms and signs. In the early stage, the rate of misdiagnosis and missed diagnosis is relatively high. This study aimed to use machine learning technology to establish a fast and accurate screening model that requires only patients' routine examination data as input to obtain predictive results. METHODS: A retrospective analysis of the examination data and diagnosis results of 53,213 patients with cardiovascular disease was conducted. Among these samples, 802 samples had AD. Forty-two features were extracted from the patients' routine examination data to establish a prediction model. There were five ensemble learning models applied to explore the possibility of using machine learning methods to build screening models for AD, including AdaBoost, XGBoost, SmoteBagging, EasyEnsemble and XGBF. Among these, XGBF is an ensemble learning model that we propose to deal with the imbalance of the positive and negative samples. The seven-fold cross validation method was used to analyze and verify the performance of each model. Due to the imbalance of the samples, the evaluation indicators were sensitivity and specificity. RESULTS: Comparative experiments showed that the sensitivity of XGBF was 80.5%, which was better than the 16.1% of AdaBoost, 15.7% of XGBoost, 78.0% of SmoteBagging and 77.8% of EasyEnsemble. Additionally, XGBF had relatively high specificity, and the training time consumption was short. Based on these three indicators, XGBF performed best, and met the application requirements, which means through careful design, we can use machine learning technology to achieve early AD screening. CONCLUSIONS: Through reasonable design, the ensemble learning method can be used to build an effective screening model. The XGBF has high practical application value for screening for AD.

Induction of microRNA-199 by Nitric Oxide in Endothelial Cells Is Required for Nitrovasodilator Resistance via Targeting of Prostaglandin I2 Synthase.

BACKGROUND: Nitrates are widely used to treat coronary artery disease, but their therapeutic value is compromised by nitrate tolerance, because of the dysfunction of prostaglandin I2 synthase (PTGIS). MicroRNAs repress target gene expression and are recognized as important epigenetic regulators of endothelial function. The aim of this study was to determine whether nitrates induce nitrovasodilator resistance via microRNA-dependent repression of PTGIS gene expression. METHODS: Nitrovasodilator resistance was induced by nitroglycerin (100 mg·kg-1·d-1, 3 days) infusion in Apoe-/- mice. The responses of aortic arteries to nitric oxide donors were assessed in an organ chamber. The expression levels of microRNA-199 (miR-199)a/b were assayed by quantitative reverse transcription polymerase chain reaction or fluorescent in situ hybridization. RESULTS: In cultured human umbilical vein endothelial cells, nitric oxide donors induced miR-199a/b endogenous expression and downregulated PTGIS gene expression, both of which were reversed by 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide potassium salt or silence of serum response factor. Evidence from computational and luciferase reporter gene analyses indicates that the seed sequence of 976 to 982 in the 3'-untranslated region of PTGIS mRNA is a target of miR-199a/b. Gain functions of miR-199a/b resulting from chemical mimics or adenovirus-mediated overexpression increased PTGIS mRNA degradation in HEK293 cells and human umbilical vein endothelial cells. Furthermore, nitroglycerin-decreased PTGIS gene expression was prevented by miR-199a/b antagomirs or was mirrored by the enforced expression of miR-199a/b in human umbilical vein endothelial cells. In Apoe-/- mice, nitroglycerin induced the ectopic expression of miR-199a/b in the carotid arterial endothelium, decreased PTGIS gene expression, and instigated nitrovasodilator resistance, all of which were abrogated by miR-199a/b antagomirs or LNA-anti-miR-199. It is important that the effects of miR-199a/b inhibitions were abolished by adenovirus-mediated PTGIS deficiency. Moreover, the enforced expression of miR-199a/b in vivo repressed PTGIS gene expression and impaired the responses of aortic arteries to nitroglycerin/sodium nitroprusside/acetylcholine/cinaciguat/riociguat, whereas the exogenous expression of the PTGIS gene prevented nitrovasodilator resistance in Apoe-/- mice subjected to nitroglycerin infusion or miR-199a/b overexpression. Finally, indomethacin, iloprost, and SQ29548 improved vasorelaxation in nitroglycerin-infused Apoe-/- mice, whereas U51605 induced nitrovasodilator resistance. In humans, the increased expressions of miR-199a/b were closely associated with nitrate tolerance. CONCLUSIONS: Nitric oxide-induced ectopic expression of miR-199a/b in endothelial cells is required for nitrovasodilator resistance via the repression of PTGIS gene expression. Clinically, miR-199a/b is a novel target for the treatment of nitrate tolerance.

Plasma Omentin-1 Level as a Predictor of Good Coronary Collateral Circulation.

AIM: Coronary collateral circulation (CCC) is crucial during an acute ischemic attack. Evidences showed that omentin-1 exhibited remarkable antiatherogenic effects and ischemia-induced revascularization. The aim of this study was to investigate the relationship between plasma omentin-1 levels and CCC in patients with ≥90% angiography-proven coronary occlusion. METHODS: 142 patients with ≥90% luminal diameter stenosis in at least one major epicardial coronary artery were recruited. Among them, 79 patients with Rentrop 0-1 grade were classified into the poor CCC group and 63 patients with Rentrop 2-3 grade were included into the good CCC group. The association between plasma omentin-1 levels and CCC status was assessed. RESULTS: Plasma omentin-1 level was significantly higher in patients with good CCC than those with poor CCC (566.57±26.90 vs. 492.38±19.70 ng/mL, p=0.024). Besides, omentin-1 was positively correlated with total cholesterol (TC), high-density lipoprotein, and gensini score but inversely with hyperlipidemia and body mass index (all p values＜0.05). Multivariate regression analysis indicated that omentin-1 [odds ratio (OR)=1.002, 95% confidence interval (CI): 1.000-1.004, p=0.041)], TC, the number of the diseased vessels, a higher frequency of left circumflex artery and right coronary artery, chronic total occlusion, and gensini score remained as the independent predictors of good CCC. CONCLUSION: Higher plasma omentin-1 level was associated with better CCC development. Our findings suggest that omentin-1 may be an alternative marker for adequate CCC in patients with ≥90% coronary occlusion.