PurposeBreast cancer (BC) is one of the biggest threats to women's health. LncRNA HOTAIR is related to the recurrence and metastasis of BC. Whether HOTAIR can serve as an effective biomarker to distinguish BC patients with different prognosis need to be further studied.MethodsThe miRNA and mRNA expression profile data of BC patients were downloaded from TCGA database. Univariate Cox regression was used to screen differential expression genes (DEGs). The miRcode database and miRWalk database were used to predict miRNA binding to HOTAIR and binding sites of miRNAs, respectively. Kaplan-Meier (KM) analysis was used to estimate the overall survival rate of BC patients. Finally, qRT-PCR and western blot were applied to evaluate the expression level of HOTAIR and mRNAs between BC cells and normal mammary cells.ResultsThe patients with high HOTAIR expression had poor prognosis in BC. Totally 10 genes correlated with BC prognosis were identified from 170 DEGs, among which PAX7, IYD, ZIC2, MS4A1, TPRXL, CD24, LHX1 were positively correlated with HOTAIR, while CHAD, NPY1R, TPRG1 were opposite. The levels of IYD, ZIC2, CD24 mRNA and protein were increased in BC tissues and BC cells. In BC cells, the levels of IYD, ZIC2 and CD24 mRNA and protein were significantly increased in HOTAIR overexpressed group. HOTAIR had the strongest interaction with hsa-miR-129-5p, followed by hsa-miR-107.ConclusionHOTAIR regulated the expression of downstream genes by interacting with 8 miRNAs and ultimately affected the prognosis of BC patients.