|Home | About | Journals | Submit | Contact Us | Français|
This article has been retractedRetraction in:
Down syndrome (DS), or Trisomy 21, is the most common genetic cause of cognitive impairment and congenital heart defects in the human population. To date, the contribution of microRNAs (miRNAs) in DS has not been investigated. Bioinformatic analyses demonstrate that human chromosome 21 (Hsa21) harbors five miRNA genes; miR-99a, let-7c, miR-125b-2, miR-155 and miR-802. MiRNA expression profiling, miRNA RT-PCR and miRNA in situ hybridization experiments demonstrate that these miRNAs are over-expressed in fetal brain and heart specimens from individuals with DS when compared with age- and sex-matched controls. We hypothesize that trisomic 21 gene dosage over-expression of Hsa21-derived miRNAs results in the decreased expression of specific target proteins and contribute, in part, to features of the neuronal and cardiac DS phenotype. Importantly, Hsa21-derived miRNAs may provide novel therapeutic targets in the treatment of individuals with DS.
Down syndrome or Trisomy 21 is a result of the presence of an extra copy of all or part of human chromosome 21 (Hsa21) . DS is the most frequent survivable congenital chromosomal abnormality occurring in approximately 750–1000 live births [2,3]. Although the phenotypes of DS are complex and variable they include congenital heart defects, craniofacial abnormalities, gastrointestinal anomalies, cognitive impairment, leukemia and Alzheimer’s disease [4,5]. Over-expression of the genes on Hsa21 by 50% in many tissues is thought to initiate the DS phenotypes, however, there is currently no explanation for how this relatively small increase in transcript levels results in any specific feature of DS [6,7].
MicroRNAs (miRNAs) are small, non-protein coding RNAs that base pair with specific mRNA targets and leads to translational repression or mRNA cleavage [8–10]. MiRNAs are expressed as long primary transcripts that are subsequently processed into mature miRNAs (~22 nucleotides) by several nuclear and cytoplasmic enzymatic steps [9–10]. MiRNAs have been shown to play a fundamental role in diverse biological and pathological processes, including cell proliferation, differentiation, apoptosis, carcinogenesis and cardiovascular disease [9–10]. In this study, we tested the hypothesis that Trisomy 21 results in the over-expression of Hsa21-derived miRNAs. Importantly, we demonstrate that all five Hsa21-derived miRNAs are over-expressed in fetal DS brain and heart specimens.
Human fetal hippocampus (HIPP) and heart samples, age- and sex-matched controls (n=3–5) and DS (n=3–5), were obtained from the Brain and Tissue Bank for Developmental Disorders, University of Maryland at Baltimore in contract with the National Institute of Child Health and Human Development.
Six hundred ng of fetal hippocampus control and DS total RNA was briefly treated with DNase I (n=3–5 per group). Five hundred ng of the DNase-treated RNA was converted to cDNA using gene-specific primers to 446 mature miRNAs (TaqMan® MicroRNA Assays, Applied Biosystems, Foster City, CA) per the manufacturer’s recommendation. Primers to the internal controls, snoRNAs, U38B and U43, as well as 18S rRNA, and U6 RNA were included in the mix of primers. Real-time PCR was performed in 5 µl reactions using standard conditions as described . The expression of 446 human mature miRNAs was profiled using an Applied Biosystems 7900HT real-time PCR instrument equipped with a 384-well reaction plate. Liquid handling robots and the Zymak Twister robot were used to increase throughput and reduce error. The relative expression of each miRNA was calculated from the equation 2−ΔCT, where Δ CT = CTmiRNA − CTinternal control . 18S rRNA was used as the internal control.
Total RNA was isolated from frozen human fetal hippocampus and heart control and DS specimens using Trizol (Invitrogen). The RNA was subsequently treated with RNase-free DNase I, and mature mir-99a, let-7c, miR-125b-2, and miR-155 was quantified utilizing specific TaqMan microRNA assay kits (373124, Applied Biosystems, Foster City, CA) as previously described [13,14]. Briefly, 100 ng of total RNA was heated for 5 min at 80°C with 2.5 µM of the 18S rRNA anti-sense primer followed by 5 min at 60°C then cooling to room temperature. The resulting solution was then added to a reverse transcriptase cocktail and transcription was performed in 20 µl according to the manufacturer’s recommendations (Catalog #4366596, Applied Biosystems). Quantitative real-time PCR (20 µl total reaction) was performed using 5 µl of a 1:50 dilution of cDNA. Gene expression was calculated relative to 18S rRNA and Ct values were normalized to “1” for normal control samples to simplify data presentation. Because Taqman primers for miR-802 are not available, this miRNA was quantitated utilizing fetal hippocampus total RNA which was reverse-transcribed and PCR amplified using the mirScript Reverse Transcription Kit and the miR-802 miScript Primer Assay system (Catalog # MS00010598, Qiagen).
Locked nucleic acid (LNA) probes complementary to human mature let 7-c (5′-ACAAGGAUGAAUCUUUGUUACUG-3′), miR-99a (5′-CACAAGAUCGGAUCUACGGGUU-3′), miR-125b (5′-UCAGAAGUUAGGGUCUCAGGGA-3′), miR-155 (ACCCCUAUCACGAUUAGCAUUAA-3′), miR-802 (5′-ACAAGGAUGAAUCUUUGUUACUG-3′), and scrambled negative control (5′-UUCACAAUGCGUUAUCGGAUGU-3′) digoxigenin-labeled at the 5′ position were purchased from Exiqon (Vedbaek, Denmark). Detection of mature miRNAs by in situ hybridization utilizing oligonucleotide probes were performed as previously described . Briefly, frozen control and DS hippocampus specimens were formalin-fixed, embedded in wax and subsequently treated with pepsin at 2 mg/ml in RNase-free water for 2 hrs as previously described . Hybridization was performed at 37°C overnight followed by a low stringency wash in 0.2 × SSC and 2% bovine serum albumin at 4°C for 10 min. The probe-target complex was visualized utilizing a digoxigenin antibody conjugated to alkaline phosphatase acting on the chromogen nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate. Therefore, positive expressing miRNA cells stain dark blue. Nuclear fast red served as the counterstain. Identical control experiments were performed utilizing a digoxigenin-labeled scrambled miRNA and control experiments were conducted where the miRNA probe was omitted. All specimens were sequentially viewed in their entirety under a ×100 objective and number of positively stained cells per field was scored.
All data are reported as means ± SEM. When comparisons were made between two different groups, statistical significance was determined using Student’s t test. When multiple comparisons were made, statistical significance was determined using one-way ANOVA followed by Tukey’s post-test. All statistical analysis was performed using the software package Prism 4.0b (GraphPad Software, San Diego, CA).
It has been previously demonstrated that mature miR-155 is processed from the primary non-protein-coding transcript of the bic gene that is harbored on Hsa21 . To investigate whether additional miRNA genes are also located on Hsa21, bioinformatic analyses were performed (http://microrna.sanger.ac.uk). Importantly, this study identified five miRNA genes harbored on Hsa21 including miR-99a, let-7c, miR-125b-2, miR-155 and miR-802 (Table 1). Mir-99a, let-7c, and miR-125b-2 are located in the sense orientation within an intron of the C21orf34 gene (intron number designation changes because of alternative C21orf34 promoter utilization). The C21orf34 gene is located at the beginning of q21.1 and the protein encoded by this gene has not been characterized. Mir-99a and let-7c are only 659 bp apart, whereas miR- 125b-2 lies just over 50,000 base pairs (bp) downstream of let-7c. Because these miRNAs are located within a C21orf34 intron, it is speculated that their expression pattern will be regulated by the promoter region of the C21orf34 gene. The bic/miR-155 gene is located almost nine million bp downstream from the C21orf34 gene at Hsa21 genomic position q21.2. The bic/miR-155 gene is well characterized and the expression of this gene has been shown to be regulated by lipopolysaccharide (LPS) and cytokines [17,18]. Finally, miR-802 is located just over ten million bp downstream from the bic/miR-155 gene in the anti-sense orientation within intron 1 of the RUNX1 gene at position q22.11. Since miR-802 was just identified, it has not been characterized and very little information is known regarding the expression distribution of this miRNA or the mechanisms by which this gene is regulated. Although the Hsa21 miRNA genes are distributed throughout a large portion of the long arm of chromosome 21, there is currently no experimental evidence to exclude any significant segment of 21q from containing genes relevant to the DS phenotypes [7,19,20]. Therefore, all Hsa21 genes encoding miRNAs must be also considered as candidates.
It has been previously demonstrated that miR-155 is over-expressed in fibroblasts isolated from an individual with DS when compared with miR-155 levels in fibroblasts isolated from their unaffected monozygotic twin . To extend these results and to determine whether all five Hsa21-derived miRNAs were over-expressed in DS specimens and to establish whether the expression of non-Hsa21-derived miRNAs were also aberrantly expressed in DS samples, miRNA microarray experiments were performed utilizing control and DS fetal hippocampus samples. Of the 424 human mature miRNAs investigated, 10 were over-expressed and 22 were under-expressed by at least 50% in hippocampus specimens from individuals with DS when compared to controls (Fig. 1). These experiments also demonstrated that 181 mature miRNA expression levels were not different between the two experimental groups and an additional 243 mature miRNAs were not expressed in the human hippocampus (data not included in Fig. 1). Importantly, these experiments demonstrated that of the Hsa21-derived miRNAs, miR-99a, let-7c and miR-155 were over-expressed in the profiling experiments. Unfortunately, the values for miR-125b-2 were inconclusive and due to the lack of primers specific for miR-802 we were unable to quantitate the levels of this miRNA.
To validate the microarray studies, RT-PCR studies were performed using primer sets specific for the five Hsa21-derived miRNAs. Importantly, these experiments demonstrated that mature miR-99a, let-7c, miR-125b-2, miR-155 and miR-802 (primers were available for this type of assay) were over-expressed by at least 50% in fetal hippocampus (n=5) and heart (n=3) samples obtained from DS individuals when compared with age- and sex-matched control specimens (Fig. 2A, 2B).
To support the RT-PCR data, Hsa21-derived miRNA expression was also investigated in human fetal hippocampus samples by in situ hybridization experiments utilizing anti-sense miRNA locked nucleic acid (LNA) probes specific for each Hsa21-derived miRNA. The representative photomicrographs of fixed control fetal hippocampus specimens, that were generated from the frozen brain samples utilized for Fig. 2A, demonstrated that very few miR-99a, let-7c, miR-125b-2, miR-155 and miR-802 positive stained cells (~1%) were detectable (lack of blue stained cells; Fig. 2A, 2C, 2E, 2G, 2I); whereas DS fetal hippocampus specimens showed an increase in each of the Hsa21-derived miRNA positive neurons (blue-stained cells denoted by arrow; Fig. 2B, 2D, 2F, 2H, 2J). Quantitative analyses of age- and sex-matched DS (n=5) and control (n=5) fetal hippocampus specimens demonstrated that let-7c, mir-125b-2, miR-155 and miR-802 expressing neurons were increased by 10–15-fold (Fig. 2B, 2F, 2H and 2J) in DS samples. Interestingly, miR-99a-expressing neurons were increased by ~50-fold (Fig. 2D). The determination that Hsa21-derived miRNA over-expression was observed in neurons and not microglial or astrocyte cells was based on cytologic features and neuron-specific enolase staining in adjacent serial sections (data not shown).
DS is a complex syndrome of genetic origin with multiple and variable clinical features [3–5]. These phenotypic outcomes seem to be the product of dosage effects of multiple genes affecting developmental processes and functions and not a simple one gene, one phenotype explanation [7,19,20]. To date, the contribution of miRNAs in DS has not been investigated. For the first time we demonstrate that mature miR99a, let-7c, miR-125b-2, miR-155 and miR- 802, which are encoded by genes harbored on Hsa21, are over-expressed in human fetal hippocampus and heart samples from individuals with DS. As previously described, the end result of miRNA-mediated gene regulation is a reduction in the total amount of target protein that is produced [8–10], we speculate that Trisomy 21-induced over-expression of Hsa21-derived miRNAs, results in the decreased expression of specific target proteins and contributes, in part, to features of the DS phenotype.
It is estimated that 30% to 90% of human genes are regulated by miRNAs [22,23]. Recently, several general principles regarding miRNAs and target-gene regulation have emerged. First, each miRNA can potentially regulate a large number of protein-coding genes. Second, many miRNAs potentially act in combination to regulate the same target genes. This allows for combinatorial regulation, where synergy between individual target sites has been demonstrated [9,10]. Finally, predicted miRNA target genes are not restricted to a particular functional category or biological pathway, but rather are involved in a wide variety of biological processes. Therefore, based on these parameters the over-expression of the five Hsa21-derived miRNAs in DS individuals may result in the aberrant expression of a myriad of critical proteins in a variety of tissues.
MiRNA-mediated regulation of gene expression results when a miRNA binds to its recognition element within the 3′-UTR of a target mRNA and suppresses its translation or initiates its degradation (8–10). The algorithms used to predict miRNA/mRNA targets typically develop scoring schemes based on sequence complementarity, free energy calculations of RNA:duplex formation, and phylogenetic conservation. For example, if the bioinformatic algorithm “TargetScan”  is utilized (http://www.targetscan.org) to predict Hsa21-derived miRNA/mRNA target sites, the following information is obtained: miR-99a, 36 putative mRNA targets; let-7c, 616 putative mRNA targets; miR-125b-2, 488 putative mRNA targets; miR-155, 240 putative mRNA targets, and miR-802, 190 putative mRNA targets. Current algorithms utilize conservation of miRNA/mRNA target sites across species as an important parameter; importantly however, the conservation of a miRNA target site is not a requirement for a functional miRNA. Therefore, the actual number of Hsa21-derived miRNA/mRNA targets is much greater than the number of in silico predicted recognition elements denoted above. Although Hsa21-derived miRNAs may regulate large numbers of mRNA targets, therapeutically this is not a disadvantage since inhibition or knock-down of these over-expressed miRNAs should normalize the expression level of all miRNA/mRNA targets back to non-trisomic 21 levels. Importantly, recent studies utilizing locked nucleic acid (LNA)-modified anti-sense oligonucleotides (antimiRs), which are complementary to specific miRNAs, have demonstrated that LNA antagonists have high metabolic stability, specificity, efficacy, and potency [24–26] and are therefore well suited for the development of novel therapeutics for individuals with DS.
This work was supported by National Institutes of Health grants #HL48848 (TSE) and #HL084498 (DSF), #ES012241 (AVT), AG21912 (EH).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.