1.Prediction of pathological type of early lung adenocarcinoma using machine learning based on SHOX2 and RASSF1A methylation levels
Runqi HUANG ; Guangliang QIANG ; Yifei LIU ; Jiahai SHI
Chinese Journal of Clinical Thoracic and Cardiovascular Surgery 2025;32(01):67-72
Objective To explore the accuracy of machine learning algorithms based on SHOX2 and RASSF1A methylation levels in predicting early-stage lung adenocarcinoma pathological types. Methods A retrospective analysis was conducted on formalin-fixed paraffin-embedded (FFPE) specimens from patients who underwent lung tumor resection surgery at Affiliated Hospital of Nantong University from January 2021 to January 2023. Based on the pathological classification of the tumors, patients were divided into three groups: a benign tumor/adenocarcinoma in situ (BT/AIS) group, a minimally invasive adenocarcinoma (MIA) group, and an invasive adenocarcinoma (IA) group. The methylation levels of SHOX2 and RASSF1A in FFPE specimens were measured using the LungMe kit through methylation-specific PCR (MS-PCR). Using the methylation levels of SHOX2 and RASSF1A as predictive variables, various machine learning algorithms (including logistic regression, XGBoost, random forest, and naive Bayes) were employed to predict different lung adenocarcinoma pathological types. Results A total of 272 patients were included. The average ages of patients in the BT/AIS, MIA, and IA groups were 57.97, 61.31, and 63.84 years, respectively. The proportions of female patients were 55.38%, 61.11%, and 61.36%, respectively. In the early-stage lung adenocarcinoma prediction model established based on SHOX2 and RASSF1A methylation levels, the random forest and XGBoost models performed well in predicting each pathological type. The C-statistics of the random forest model for the BT/AIS, MIA, and IA groups were 0.71, 0.72, and 0.78, respectively. The C-statistics of the XGBoost model for the BT/AIS, MIA, and IA groups were 0.70, 0.75, and 0.77, respectively. The naive Bayes model only showed robust performance in the IA group, with a C-statistic of 0.73, indicating some predictive ability. The logistic regression model performed the worst among all groups, showing no predictive ability for any group. Through decision curve analysis, the random forest model demonstrated higher net benefit in predicting BT/AIS and MIA pathological types, indicating its potential value in clinical application. Conclusion Machine learning algorithms based on SHOX2 and RASSF1A methylation levels have high accuracy in predicting early-stage lung adenocarcinoma pathological types.
2.Role and mechanism of microRNA-145-5p in hypoxia-induced pyroptosis of human alveolar epithelial cells.
Runqi YUAN ; Junmiao GUO ; Zhenting LIANG ; Yongxin ZHENG ; Yongbo HUANG ; Yonghao XU ; Pu MAO ; Jinglan SHAN
Chinese Critical Care Medicine 2025;37(4):354-360
OBJECTIVE:
To elucidate the role and mechanism of microRNA-145-5p (miR-145-5p) in hypoxia-induced pyroptosis of human alveolar epithelial cells.
METHODS:
In vitro, human alveolar epithelial cell line BEAS-2B was cultured. Cells in the logarithmic growth phase were cultured to 80% confluence and then used for the experiment. (1) BEAS-2B cells were cultured under 1% O2 hypoxic condition, with a normoxic control group. Western blotting was employed to detect the expressions of pyroptosis marker proteins [NOD-like receptor protein 3 (NLRP3), Gasdermin D N-terminal domain (GSDMD-N), and caspase-1] in cells cultured for 24 hours. Real-time fluorescent quantitative reverse transcription-polymerase chain reaction (RT-PCR) was used to detect the expression of miR-145-5p in cells cultured for 6 hours and 12 hours. (2) Cells were transfected with 30 nmol/L miR-145-5p mimic to overexpress miR-145-5p expression under normoxic condition or 30 nmol/L miR-145-5p inhibitor to suppress miR-145-5p expression under hypoxic condition. Control group and negative control group were respectively set up. After 24 hours of cell culture, Western blotting was used to detect the expressions of pyroptosis marker proteins and nuclear factor-E2-related factor 2 (Nrf2) in cells. Flow cytometry was applied to detect the level of reactive oxygen species (ROS) in cells. The target genes of miR-145-5p were predicted by miR target gene prediction software miRWalk and verified by Western blotting. (3) Under hypoxic condition, cells were transfected with 6.94 ng/μL silent information regulator 5 (Sirt5) overexpression plasmid or pretreated with 12.5 mmol/L N-acetyl-L-cysteine (NAC) as an ROS inhibitor. The empty plasmid group and control group were set up. After 24 hours of cell culture, Western blotting was used to detect the expressions of Sirt5, Nrf2, and pyroptosis marker proteins in cells. Flow cytometry was used to detect the level of ROS in cells.
RESULTS:
(1) Compared with the normoxic control group, the expression levels of pyroptosis marker proteins in the 24-hour hypoxia group was significantly increased, indicating that hypoxia could induce pyroptosis in BEAS-2B cells. The expression level of miR-145-5p in cells gradually increased with the extension of hypoxia induction time, indicating that hypoxia could cause the increase of miR-145-5p expression level. (2) The expression levels of pyroptosis marker proteins in cells of miR-145-5p mimic group significantly increased under normoxic condition as compared with the control and negative control groups [NLRP3 protein (NLRP3/β-actin): 1.58±0.07 vs. 1.00±0.01, 0.98±0.07, GSDMD-N protein (GSDMD-N/β-actin): 1.71±0.03 vs. 1.01±0.01, 0.85±0.03, caspase-1 protein (caspase-1/β-actin): 2.33±0.04 vs. 1.01±0.01, 1.05±0.04, all P < 0.05], Nrf2 protein expression level was significantly decreased (Nrf2/β-actin: 0.79±0.03 vs. 1.00±0.01, 1.03±0.04, both P < 0.05), ROS level was significantly up-regulated (fluorescence intensity: 1.74±0.03 vs. 1.00±0.01, 0.92±0.03, both P < 0.05). Under hypoxia condition, compared with control group and negative control group, the expression levels of pyroptosis marker proteins in miR-145-5p inhibitor group were significantly decreased [NLRP3 protein (NLRP3/β-actin): 0.21±0.04 vs. 1.70±0.02, 1.63±0.04; GSDMD-N protein (GSDMD-N/β-actin): 1.32±0.02 vs. 2.51±0.02, 2.72±0.03; caspase-1 protein (caspase-1/β-actin): 0.56±0.01 vs. 2.77±0.02, 3.12±0.03; all P < 0.05], Nrf2 protein expression level was significantly increased (Nrf2/β-actin: 1.57±0.04 vs. 1.22±0.01, 1.28±0.04, both P < 0.05), ROS level was significantly down-regulated (fluorescence intensity: 0.64±0.05 vs. 1.87±0.04, 1.70±0.07, both P < 0.05). The results indicated that miR-145-5p could promote cell pyrodeath. The predictive result of miRWalk showed that the 3' untranslated region (3'UTR) of Sirt5 had complementary base binding sites with miR-145-5p. The expression level of Sirt5 protein in cells of miR-145-5p mimic group was significantly lower than that of control group and negative control group under normoxic condition (Sirt5/β-actin: 0.59±0.03 vs. 1.00±0.01, 1.01±0.03, both P < 0.05), which verified that Sirt5 was the target gene of miR-145-5p. (3) The occurrence of pyrodeath could be partially reversed by transfection with Sirt5 overexpression plasmid or adding ROS inhibitor NAC into cells, and Sirt5 overexpression could also up-regulate Nrf2 expression and eliminate intracellular ROS.
CONCLUSION
In human alveolar epithelial cells, miR-145-5p can down-regulate Nrf2 by targeting Sirt5, thereby increasing ROS expression and inducing pyrodeath.
Humans
;
MicroRNAs
;
Pyroptosis
;
Cell Hypoxia
;
Alveolar Epithelial Cells/cytology*
;
Cell Line
;
NLR Family, Pyrin Domain-Containing 3 Protein
;
Caspase 1/metabolism*
;
Epithelial Cells/metabolism*
;
Gasdermins
;
Phosphate-Binding Proteins
3.Correlation between blood biochemical indexes and milk composition in early lactation cows in dairy cow
Runqi LIU ; Wei YANG ; Cheng XIA ; Hongyou ZHANG ; Yuanyuan CHEN ; Hongjiang YU ; Taiyu SHEN ; Sansi GAO ; Baoyin HUANG ; Chuang XU
Chinese Journal of Veterinary Science 2017;37(8):1566-1570
Biochemical indexes and blood composition in early lactation health,ketosis and hypocalcemia dairy cows were analyzed to make sure the milk composition characteristics with related diseases,the correlation analysis between early lactating dairy milk composition and blood biochemical were also make in order to provide support for the cattle health assessment.According to theblood index,72 Holstein cows 7-21 d postpartum,were divided into group subclinical hypocalcemia,ketosis test group and control group,24 heads each group.The blood and milk of cows were collected and used to analyze the correlation between blood biochemical indexes and milk composition.The results showed that ketosis and hypocalcemia induced the level of milk protein and non fat milk solids decreased,while the content of citric acid in milk increased.The correlation equation between citric acid in milk and serum NEFA,BHBA and GLU was y=3.192x-0.802,(R2 =0.363),y=4.594x-0.793,(R2 =0.320),y=1.228x+0.775,(R2 =0.261),in which x was the content of citric acid in milk.The results showed that the content of citric acid in milk had positively related to blood NEFA,which could be used as an early marker for the diagnosis of negative energy balance.The levels of BUN and ALB in blood can be used to evaluate the levels of milk protein and urea.
4.Analysis of the concentrations of calcium, phosphorus and magnesium in beef cattle which in seven different areas of Heilongjiang province
Baoyin HUANG ; Ping HE ; Hongjiang YU ; Taiyu SHEN ; Sansi GAO ; Runqi LIU ; Zhihao DONG ; Ruirui LI ; Yuanyuan CHEN ; Wei YANG ; Chuang XU
Chinese Journal of Veterinary Science 2017;37(8):1562-1565
The objective of this study was to determine the concentrations of calcium,phosphorus and magnesium in blood and compare the differences in seven areas of Heilongjiang province,and then estimate calcuim-phosphorus metabolism of beef cattle in seven beef cattle farms to provi detheoretical foundation for the prevention of calcium phosphorus metabolism diseases of beef cattle.Seven beef cattle farms of Daqing,Shuangyashan,Jiusan and Mudanjiang in Heilongjiang province were selected as the survey sites,which were recorded as group A (both grazing and stall-feeding in Shuangyashan),group B (stall-feeding mode in Shuangyashan),group C (stall-feeding mode in Daqing),group D (grazing mode in Jiusan),group E (both grazing and stall-feeding in Daqing),group F (both grazing and stall-feeding in Daqing) and group G (stall-feeding mode in Mudanjiang).Then the concentrations of Ca,Mg,P,free fatty acid (NEFA),glucose (Glc) and β-hydroxybutyric acid (BHBA) in the blood were compared to estimate the calcuim-phosphorus metabolic states.Results showed that the concentrations of calcium,magnesium and phosphorus in the plasma of 65 beef cattle in seven survey sites were within the normal range,and there was no significant difference in calcium concentration among seven sites.The P contents in group C and G were significantly higher those that in group A and B(P<0.01),which in group C was significantly higher than those in group D,E and F (P<0.01),which in group G was significantly higher than those in group D,E and F(P<0.01).NEFA content in group B was significantly higher than that in group D (P<0.05),and there was no difference among other groups.The concentration of Glc in group A was significantly higher than that in group E (P<0.05),which in group B was significantly higher than those in group A and D,and was very significantly higher than those in group E,F and G (P<0.01),which in group C was very significantly higher than those in group A,D,E,F and G (P<0.01),which in group C was significantly higher than those in group A,D,E,F and G (P<0.01),which in group D was significantly higher than that in group E (P<0.05),and which in group F was significantly higher than that in group E (P<0.05).The concentration of BHBA in group C,D and E were significantly higher than that in group A (P<0.01),which in group C was significantly higher than that in group B (P<0.05),which in group D and E was significantly higher than that in group B (P<0.01),which in group C was significantly higher than that in group G (P<0.01),which in group D was significantly higher than those in group F and G (P<0.01),and in group E was highly significantly higher than that in group F and G (P<0.01).Overall,there were not calcuim-phosphorus metabolic disorders within the seven beef cattle farms which were selected,but it is also necessary to strengthen feeding management and health care to prevent the occurrence of nutrition and metabolic diseases.

Result Analysis
Print
Save
E-mail