Date Published: October 20, 2011
Publisher: Public Library of Science
Author(s): Francesca Megiorni, Samantha Cialfi, Carlo Dominici, Serena Quattrucci, Antonio Pizzuti, Iris Schrijver. http://doi.org/10.1371/journal.pone.0026601
Abstract: microRNAs (miRNAs) are a class of regulatory small non-coding molecules that control gene expression at post-transcriptional level. Deregulation of miRNA functions affects a variety of biological processes also involved in the etiology of several human mendelian and complex diseases. Recently, aberrant miRNA expression has been observed in Cystic Fibrosis (CF), an autosomal-recessive genetic disorder caused by mutations in the CFTR gene, in which a genotype-phenotype correlation is not always found. In order to determine miRNA role in CFTR post-transcriptional regulation, we searched for miR-responsive elements in the CFTR 3′-UTR. In silico analysis, performed using different computational on-line programs, identified some putative miRNAs. Both miR-101 and miR-494 synthetic mimics significantly inhibited the expression of a reporter construct containing the 3′-UTR of CFTR in luciferase assays. Interestingly, miR-101/miR-494 combination was able to markedly suppress CFTR activity by approximately 80% (p<0.001). This is one of the first in vitro studies implicating microRNAs as negative regulators of the CFTR gene expression. miRNA aberrant expression and function might explain the wide phenotypic variability observed among CF patients.
Partial Text: Cystic Fibrosis (CF) is a common monogenic disease caused by mutations in the Cystic Fibrosis Transmembrane conductance Regulator (CFTR) gene coding for an ATP-dependent anion channel-protein required to regulate sweat, digestive juices and mucus components . Even if CF is inherited as a recessive trait, with more than 1500 causative mutations, a wide range of inter- and intra-familial variability is reported so that individuals with the same genotype may show different phenotypes. Therefore, the clinical phenotype is not entirely due to the CFTR mutant homozygous or compound heterozygous status but genetic modifiers and environmental factors are likely to modulate the severity of the disease , . Recent papers focused on the possible role of microRNAs in the inter- and intra-familial variability of the CF clinical manifestations . MicroRNAs are evolutionarily conserved, small non-coding RNAs that negatively regulate gene expression at the post-transcriptional level by either repressing translation or decreasing mRNA stability , . Mature miRNA, a single-stranded RNA with a length of about 22 nucleotides, is incorporated into a RNA-induced silencing complex (RISC) and is able to bind target transcripts through base pairing with their 3′-untranslated regions (UTRs). The main sequence for mRNA recognition encompasses bases 2 to 8 at the 5′-end of the mature miRNA and it is known as the “seed sequence” , . To date, more than 800 miRNAs have been computationally identified in the human genome , , each one having the potential capacity to bind to hundreds of transcripts, and the biological functions of most miRNAs are rapidly emerging. As estimated by in silico prediction algorithms , , miRNAs regulate at least 30% of the protein-encoding genes and are involved in a broad range of cellular processes such as proliferation, differentiation, homeostasis and apoptosis. It is therefore predictable that the dysregulation of the miRNA pathway might contribute to human diseases, including mendelian disorders as demonstrated for Atassia telangectasia, Fragile X syndrome and Huntington’s disease . Emerging evidence suggests that changes in expression of microRNAs are associated with Cystic Fibrosis , –. In a recent work, Oglesby et al.  performed a miRNA profile screening in bronchial brushings from CF individuals versus non-CF controls and showed that miR-126 was specifically down-regulated in CF airway epithelial cells and targeted TOM1 gene transcripts. In the present work, we address the role of miRNAs in the post-transcriptional regulation of the CFTR gene. In particular, after the in silico identification of a list of putative miRNAs able to target CFTR mRNA, we demonstrated by in vitro analysis that miR-101 and miR-494 were able to markedly suppress CFTR expression either alone and in combination.
Computational prediction of microRNA responsive elements (MREs) within the 3′-UTR of CFTR transcript was performed using commonly prediction programs such as TargetScan, PicTar, and miRanda. These web servers use complicated algorithms searching for target sequences with perfect or nearly perfect pairing to the 3′-UTR sequence, evaluating the thermodynamic stability of miRNA-mRNA hybrids and performing comparative sequence analysis to check evolutionary conservation. Also, the 3′ pairing contribution, the local AU content and the distance to the nearest end of the annotated UTR were scored. Since various programs might predict different microRNAs for a given coding gene, we also used the “Predicted Targets” section of miRecords which gave a global view of our analyses as this meta-server integrates predicted miRNA targets produced by 11 established miRNA target prediction programs. Overall, we registered 496 putative CFTR targeting miRNAs. Three hits were predicted by five single programs and 20 hits by four target prediction softwares. The most likely candidate miRNAs targeting the CFTR 3’UTR which overlapped in at least three prediction programs resulted miR-101, miR-144, miR-199-3p, miR-345, miR-376b, miR-377, miR-380, miR-494, miR-509-3p, miR-600 and miR-645. We combined all these information with thermodynamic and on-line available expression data and, finally, we selected miR-101 and miR-494 microRNAs as the more likely regulators of the CFTR mRNA. Notably, these miRNAs looked up-regulated in the miRNA expression profile analysis performed in the CF bronchial brushings in comparison to healthy controls . Both miR-101 and miR-494 are conserved in Homo sapiens, Macaca mulatta and Mus musculus and each miRNA sequence has a single putative target site within the CFTR 3’UTR at position 1508–1514 and 1140–1147, respectively (Figure 1A). In particular, miR-101 responsive element is a 7mer-A1 having an exact match to positions 2–7 of the miRNA (the seed) and an immediate downstream ‘A’ which is across from microRNA nucleotide 1 while miR-494 target site is a 8-mer element that shares an exact complementarity to position 2–8 of the mature microRNA followed by an ‘A’. The miR-101:CFTR mRNA heteroduplex displayed a mfe value of −14.2 kcal/mol whereas miR-494:CFTR hybrid structure showed a −13.5 kcal/mol value (Figure 1B).
Since aberrant level and function of microRNAs have been recently observed in Cystic Fibrosis tissues -, , the possibility they may act as phenotype modifiers was raised. Our data, together with the very recent publication by Gillen et al.,  support a role of some microRNAs in the post-transcriptional regulation of the CFTR channel synthesis. The CFTR mRNA 3’-UTR, the putative site of miRNA annealing, is highly conserved in different species and contains several regulatory motifs important for the mRNA stability and translation control . We identified a few miRNA-3’-UTR possible pairing with similar expression profiles and selected miR-101 and miR-494 for further analysis, since they showed the best possibility of positive interaction. Luciferase in vitro experiments confirmed that both miR-101 and miR-494 targeted and functionally suppressed CFTR construct mRNA translation with a moderate and strong action, respectively, and this activity was almost lost after mutation of the putative 3′ UTR target-sites. The different strength of the luciferase repression could be due to the secondary structure MRE accessibility, the relative position of the specific target sites in the CFTR 3’-UTR as well as the different seed match types. Furthermore, miR-101 and miR-494 seem to act synergistically on CFTR-reporter inhibition with a more than additive effect on the post-translational control and this could have a physiological relevance in the complex disease phenotypes observed in CF. Interestingly, our results are corroborated by expression profiling experiments of different human miRNAs from CF patients airway epithelial samples in which both miR-101 and miR-494 looked up-regulated (Relative Quantification ≥1.5) compared to non-CF individuals confirming an inverse correlation with CFTR levels . In particular, miR-494 levels were altered in four of the five CF samples showing an up to 6-fold increase respect to normal airway brushings. Up-regulation of miR-494 in asthmatic samples  and direct influence in cell cycle progression of primary murine bronchial epithelial cells exposed to carcinogens ,  suggest that miR-494 altered levels have effects mainly in the lung compartment. As far as mir-101 concerns, recent evidences underline its role in the biosynthesis of pro-inflammatory cytokines  and its function in central nervous system  providing a possible explanation for both lung excessive inflammation and neural symptoms observed in CF patients . Our study shows the possible relevance of miRNA pathways as genetic modifiers that may contribute to the varied CF phenotypes seen even among individuals carrying the same CFTR gene mutations. We speculate that microRNAs are largely responsible for how badly the illness affects patients since even minimal alterations in miR-101 and/or miR-494 levels could negatively influence the stability of the CFTR transcript which in turn might have an effect on the amount and maturation of CFTR protein and so on the degree of CF severity. This is well proved by the miRNA dose-dependent luciferase decreased activity and the concerted inhibitory interaction with CFTR 3′-UTR sequence showed in our in vitro results. Recent evidences also suggest that RNA binding proteins and microRNAs can regulate transcripts involved in a common pathway . In particular, Spence  identified conserved regulatory motifs in the CFTR 3’-UTR that had a partial identity with 3’-UTR of SEC24, a gene coding for a protein implicated in the CFTR translocation out of the endoplasmic reticulum. Our preliminary in silico analysis found that both miR-101 and miR-494 may target SEC24 3’-UTR and two other genes, TGFB1 and MBL2, which are known to modify the development and/or the severity of lung disease in CF , suggesting the existence of a coordinated network of gene expression control by microRNAs. Therefore, alteration of miR-101 and/or miR-494 levels in CF patients could influence the disease clinical expression with particular implications in CF lung function, such as increased susceptibility to infections, chronic airways inflammation and response to specific therapies. These miRNAs, as CF genetic modifiers, might act either by directly influence CFTR gene expression or/and other functionally related genes. Future investigations will be addressed to confirm this hypothesis such as the characterization of the entire CFTR 3’-UTR in a panel of CF subjects carrying the same mutations and different clinical phenotype looking for sequence variants in the miR-responsive elements that could dramatically alter CFTR regulation as well as mutations in the miR-101 and miR-494 genes which could explain their altered expression.