Integrating lncRNAs and mRNAs expression profiles in terminal hindgut of fetal rats with anorectal malformations.

Abstract

The detailed embryonic etiology and pathogenesis of anorectal malformations (ARMs) remains unclear. Recent studies have shown that gene expression abnormalities were the key factors that result in ARMs. Long non-coding RNAs (lncRNAs) were reported as the 'transcriptional noise' within the genome. The expression profiles of lncRNA and mRNA remain less characterized in the pathogenesis of ARMs. Furthermore, the function of lncRNAs in the regulation of this process has not been investigated so far. Therefore, this current study was aimed to integrate lncRNA and mRNA expression profiles in terminal hindgut of ethylenethiourea (ETU)-induced ARM rats using Agilents lncRNA and mRNA co-expression microarrays. ARM model was induced with ethylenethiourea (ETU) on gestational day 10. Cesarean deliveries were conducted to collect the embryos on gestational day 20. For the extraction of total RNA, 1-cm terminal hindgut tissues were collected from three fetal rats with similair weights. The microarrays and quantitative RT-PCR analysis were conducted to evaluate the lncRNA and mRNA expression profiles in normal fetal rats and ARM fetal rats. Compared with control group, 164 lncRNAs were observed to be aberrantly expressed (FC ≥ 2; P < 0.05) in ARM group, including 36 upregulated and 128 downregulated, while 772 mRNAs were observed to be aberrantly expressed (FC ≥ 2; P < 0.05) in the terminal hindgut, including 350 up-regulated and 422 down-regulated. The differential expression profiles between the ARM and the control group were used for gene ontology (GO) and pathway analysis. A subset of those RNAs was identified to be closely related to the development process of ARMs. The four RNAs that were differentially expressed between the two groups were selected for qPCR validation, and the results were in line with the microarray data. In addition, the lncRNAs and mRNA co-expression network was established according to the correlation analysis. We predicted the functions of transregulatory lncRNAs by the TFs (transcription factors) which might modulate their expression. In the core network of lncRNA-TF pairs, the lncRNAs can be classified into 5 categories of pathways governed by Jun, c-Myc, Usf1, Alf2, and Stat3. From the above results, it can be suggested that these aberrant lncRNAs might participate in the pathogenesis of ARM, and our present work may provide new research directions for future studies of ARMs.