Role of SFTPD/miR-335-5p/lnc-HNRNPUL2 axis in colorectal cancer: in silico characterization and clinical validation

The lnc-RNA, miRNA, and mRNA networks have been extensively studied for cancer regulation. Furthermore, ceRNA network analysis has proved as an effective candidate for identifying potential diagnostic biomarkers in a variety of cancers. In this study, we described a novel axis of serum Surfactant protein D (SFTPD)/miR-335-5p/lnc-HNRNPUL2 proposed by bioinformatics analysis and validated in a pilot human study of CRC versus control & in vitro assay to explore its role in the CRC pathogenesis. Real-time PCR and ELISA were used to validate the expression of this axis panel in serum samples from 81 CRC patients, 41 patients with benign tumors, and 30 healthy controls. Its expression was also confirmed in the CRC cell line HT29. Lnc-HNRNPUL2 and TGF-β expression levels were significantly upregulted in the sera of CRC patients compared to controls, along with a concurrent decrease in the expression of miR-335-5p and SFTPD mRNA in the CRC group relative to both benign and healthy control groups Interestingly, lnc-HNRNPUL2 and hsa-miR-335-5p had significant predictive values for CRC diagnosis due to their high sensitivities and specificities (AUC = 0.901 and 0.877, respectively). In addition, we silenced lnc-HNRNPUL2 in HT29 cells, which resulted in a significant decrease in cell count and viability, as well as a restoration of normal panel expression. Together, these findings point to an oncogenic role for lnc-HNRNPUL2 in CRC, which acts via miR-335-5p suppression, followed by a decrease in SFTPD mRNA that is involved in immune surveillance against several cancers. The SFTPD/miR-335-5p/lnc-HNRNPUL2 axis shows high diagnostic accuracy for CRC detection. Low miR-335-5p expression was significantly associated with metastasis in Cox regression analysis, but longer follow-up is needed to establish prognostic value.