Bioinformatic analysis and experimental identification of blood biomarkers for chronic nonunion

Background Incomplete fracture healing may lead to chronic nonunion; thus, determining fracture healing is the primary issue in the clinical treatment. However, there are no validated early diagnostic biomarkers for assessing chronic nonunion. In this study, bioinformatics analysis combined with an experimental verification strategy was used to identify blood biomarkers for chronic nonunion. Methods First, differentially expressed genes in chronic nonunion were identified by microarray data analysis. Second, Dipsaci Radix (DR), a traditional Chinese medicine for fracture treatment, was used to screen the drug target genes. Third, the drug-disease network was determined, and biomarker genes were obtained. Finally, the potential blood biomarkers were verified by ELISA and qPCR methods. Results Fifty-five patients with open long bone fractures (39 healed and 16 nonunion) were enrolled in this study, and urgent surgical debridement and the severity of soft tissue injury had a significant effect on the prognosis of fracture. After the systems pharmacology analysis, six genes, including QPCT, CA1, LDHB, MMP9, UGCG, and HCAR2, were chosen for experimental validation. We found that all six genes in peripheral blood mononuclear cells (PBMCs) and serum were differentially expressed after injury, and five genes (QPCT, CA1, MMP9, UGCG, and HCAR2) were significantly lower in nonunion patients. Further, CA1, MMP9, and QPCT were markedly increased after DR treatment. Conclusion CA1, MMP9, and QPCT are biomarkers of nonunion patients and DR treatment targets. However, HCAR2 and UGCG are biomarkers of nonunion patients but not DR treatment targets. Therefore, our findings may provide valuable information for nonunion diagnosis and DR treatment.