Background: Long noncoding RNAs (lncRNAs) have been shown to play pivotal role in pathogenesis and prognosis of cancers. Identification of novel clinical biomarkers in advanced stage colorectal cancer (CRC) is warranted.

Aims: To identify potential lncRNAs associated with progression of stage III/IV CRC and illuminate regulatory mechanisms.

Methods: Differentially expressed lncRNAs, mRNAs and miRNAs (DElncRNAs, DEmRNAs, and DEmiRNAs) were extracted between stage III/IV CRC and normal tissues. We used DEGs to construct a ceRNA network and analyzed correlations between key lncRNAs and overall survivals (OS) of stage III/IV CRC patients. Weighted gene co-expression network analysis (WGCNA) was applied to a pivotal lncRNA. We conducted functional enrichment analysis on target genes and constructed lncRNA-TF-mRNA network by overlapping mRNAs co-expressed with the key lncRNA and target genes of transcriptional factors (TFs).

Results: A total of 26 DElncRNAs, 398 DEmiRNAs, 2155 DEmRNAs were identified. A ceRNA network was constructed with 16 lncRNAs, 20 miRNAs, and 59 mRNAs, in which MFI2-AS1 exhibited promising diagnostic efficiency. (AUC was 0.938.) MFI2-AS1 was negatively correlated to OS of stage III/IV CRC patients (P value < 0.05). KEGG analysis showed potential mRNA targets of MFI2-AS1 mainly involved in cell cycle and cytokine-cytokine receptor interaction. We identified 17 potential TFs of MFI2-AS1 and built a lncRNA-TF-mRNA network.

Conclusion: Our study provides novel insights into lncRNAs associated regulatory networks and reveals a promising lncRNA biomarker, MFI2-AS1, as an independent prognostic indicator and potential therapeutic target for CRC.