Screening of potential therapeutic compounds for osteoporosis based on L1000 database
LIANG Pengchen1,ZHOU Ziyan2,3, CHANG Qing2, YI Qingqing2, SUN Miaomiao2,3, TANG Yeling2,3, CAO Liou4,5, YANG Jie6
(1. School of Microelectronics,Shanghai University, Shanghai 201800; 2. Clinical Research Center, Jiading District Central Hospital Affiliated to Shanghai University of Medicine ＆ Health Sciences, Shanghai 201800; 3. Graduate School, Shanghai University of Traditional Chinese Medicine, Shanghai 201203; 4. Department of Nephrology, Jiading District Central Hospital Affiliated to Shanghai University of Medicine ＆ Health Sciences, Shanghai 201800; 5. Department of Nephrology, Renji Hospital Affiliated to Medical College of Shanghai Jiaotong University, Shanghai 200127; 6. GCP Office, Jiading District Central Hospital Affiliated to Shanghai University of Medicine ＆ Health Sciences, Shanghai 201800, China)
Abstract：Objective:The L1000 micro-interference data set was used to screen compounds for the treatment of osteoporosis, and the evaluation and cytological activity verification by multiple biological information methods were conducted, aiming to discover new osteoporosis treatment drugs.Methods:The data related to osteoporosis（OP）were retrieved in the GEO（Gene Expression Omnibus）database, and the difference in gene expression levels on the datasets was analyzed using limma package in R language. Pathway enrichment analysis of differential genes was performed using MetaScape data platform. The resulting differential OP gene expression profiles were analyzed to match potential OP therapeutic compounds in the L1000 dataset of the Connectivity Map database. The molecular docking study of the matching compounds and bone metabolism targets was conducted by Autodock_Vina software, and the ADMET（absorption, distribution, metabolism, excretion, and toxicity）characteristics of the matching compounds were calculated and analyzed by using pkCSM database. The activity of the candidate compound was studied by a cell proliferative activity experiment.Results:After analyzing the dataset of osteoporosis expression profile, 195 differential genes related to osteoporosis were obtained, including 127 genes with significantly up-regulated expression and 68 genes with significantly downregulated expression. There were 146 GO Biological Processes with differential gene enrichment;There were nine enriched KEGG signaling pathways and nine enriched Reactome Gene Sets. A total of 10 potential therapeutic compounds were matched in the L1000 dataset: diethylstilbestrol, PLX-4720, mezlocillin, KU-C103428N, rhapontin, maravir, quinine, MST-312, risperidone, and CO-102862. Through molecular docking studies, we found that the matching compounds have multi-target characteristics, and the main docking activities with the compounds are CAⅡ, PGR, ERβ, BMP2 and other OP targets. The ADMET properties of potential therapeutic compounds were calculated using the ADMET calculation module of the pkCSM database, which showed that mezlocillin had good ADME performance and low toxicity, with good comprehensive evaluation. Cell proliferation experiment showed that mezlocillin at 20 μg/mL had the most significant effect on the proliferation of osteoblasts.Conclusion:The candidate OP therapeutic compound, mezlocillin, was screened from the L1000 data set with good ADMET characteristics and bone cell proliferation promoting activity.
［1］Brum AM, van de Peppel J, van der Leije CS, et al. Connectivity Mapbased discovery of parbendazole reveals targetable human osteogenic pathway［J］. Proc Natl Acad Sci U S A, 2015, 112（41）:12711-12716.
［2］Brum AM, Van de PJ, Nguyen L, et al. Using the Connectivity Map to discover compounds influencing human osteoblast differentiation［J］.J Cell Physiol, 2018, 233（6）:4895-4906.
［4］Subramanian A, Narayan R, Corsello SM, et al. A next generation connectivity map: L1000 platform and the first 1000000 profiles［J］. Cell, 2017, 171（6）: 1437-1452.
［5］Zhou Y, Zhou B, Pache L, et al. Metascape provides a biologistoriented resource for the analysis of systemslevel datasets［J］. Nat Commun, 2019,10（1）:1523.
［6］Wang Y, Zhang S, Li F,et al. Therapeutic target database 2020: enriched resource for facilitating research and early development of targeted therapeutics［J］. Nucleic Acids Res, 2020, 48（D1）: D1031-D1041.
［9］Pires DE, Blundell TL, Ascher DB. pkCSM: predicting smallmolecule pharmacokinetic and toxicity properties using graphbased signatures［J］. J Med Chem, 2015,58（9）:4066-4072.
［14］Han HW, Hahn S, Jeong HY, et al. LINCS L1000 datasetbased repositioning of CGP60474 as a highly potent antiendotoxemic agent［J］. Sci Rep, 2018, 8（1）:14969.
［15］Sendama W. L1000 connectivity map interrogation identifies candidate drugs for repurposing as SARSCoV2 antiviral therapies［J］. Comput Struct Biotechnol J, 2020, 18: 3947-3949.
［16］Zhang L, Wang Z, Liu R, et al. Connectivity mapping identifies BI2536 as a potential drug to treat diabetic kidney disease［J］. Diabetes, 2021,70（2）:589-602.
